From 41b7beb7404c134820870ab21ae0242dfc749b2c Mon Sep 17 00:00:00 2001 From: Luke Shumaker Date: Wed, 26 Oct 2016 22:49:41 -0400 Subject: ./tools/notsd-move --- src/grp-system/90-systemd.preset | 24 + src/grp-system/Makefile | 32 + src/grp-system/bootup.xml | 305 + src/grp-system/grp-utils/Makefile | 32 + .../grp-utils/systemd-analyze/.gitignore | 1 + src/grp-system/grp-utils/systemd-analyze/Makefile | 39 + .../grp-utils/systemd-analyze/analyze-verify.c | 305 + .../grp-utils/systemd-analyze/analyze-verify.h | 26 + src/grp-system/grp-utils/systemd-analyze/analyze.c | 1486 ++++ .../systemd-analyze.completion.bash | 117 + .../systemd-analyze/systemd-analyze.completion.zsh | 58 + .../grp-utils/systemd-analyze/systemd-analyze.xml | 388 + src/grp-system/grp-utils/systemd-delta/Makefile | 33 + src/grp-system/grp-utils/systemd-delta/delta.c | 635 ++ .../systemd-delta/systemd-delta.completion.bash | 61 + .../systemd-delta/systemd-delta.completion.zsh | 15 + .../grp-utils/systemd-delta/systemd-delta.xml | 205 + .../grp-utils/systemd-fstab-generator/Makefile | 34 + .../systemd-fstab-generator/fstab-generator.c | 723 ++ .../systemd-fstab-generator.xml | 183 + src/grp-system/grp-utils/systemd-run/Makefile | 33 + src/grp-system/grp-utils/systemd-run/run.c | 1261 +++ .../systemd-run/systemd-run.completion.bash | 117 + .../systemd-run/systemd-run.completion.zsh | 60 + .../grp-utils/systemd-run/systemd-run.xml | 459 ++ .../grp-utils/systemd-sysv-generator/.gitignore | 1 + .../grp-utils/systemd-sysv-generator/Makefile | 46 + .../grp-utils/systemd-sysv-generator/README.in | 27 + .../systemd-sysv-generator.xml | 97 + .../systemd-sysv-generator/sysv-generator.c | 982 +++ src/grp-system/kernel-command-line.xml | 384 + src/grp-system/libcore/.gitignore | 3 + src/grp-system/libcore/Makefile | 170 + src/grp-system/libcore/include/core/automount.h | 59 + src/grp-system/libcore/include/core/bus-policy.h | 64 + src/grp-system/libcore/include/core/busname.h | 69 + src/grp-system/libcore/include/core/cgroup.h | 183 + src/grp-system/libcore/include/core/dbus-manager.h | 28 + src/grp-system/libcore/include/core/device.h | 47 + src/grp-system/libcore/include/core/execute.h | 292 + .../libcore/include/core/failure-action.h | 42 + .../libcore/include/core/hostname-setup.h | 22 + src/grp-system/libcore/include/core/ima-setup.h | 24 + src/grp-system/libcore/include/core/job.h | 242 + src/grp-system/libcore/include/core/kill.h | 65 + src/grp-system/libcore/include/core/killall.h | 22 + src/grp-system/libcore/include/core/kmod-setup.h | 22 + .../libcore/include/core/load-fragment.h | 123 + .../libcore/include/core/loopback-setup.h | 22 + .../libcore/include/core/machine-id-setup.h | 23 + src/grp-system/libcore/include/core/manager.h | 380 + src/grp-system/libcore/include/core/mount-setup.h | 30 + src/grp-system/libcore/include/core/mount.h | 108 + src/grp-system/libcore/include/core/namespace.h | 63 + src/grp-system/libcore/include/core/path.h | 93 + src/grp-system/libcore/include/core/scope.h | 58 + .../libcore/include/core/selinux-setup.h | 24 + src/grp-system/libcore/include/core/service.h | 223 + src/grp-system/libcore/include/core/show-status.h | 39 + src/grp-system/libcore/include/core/slice.h | 32 + src/grp-system/libcore/include/core/smack-setup.h | 24 + src/grp-system/libcore/include/core/socket.h | 186 + src/grp-system/libcore/include/core/swap.h | 110 + src/grp-system/libcore/include/core/target.h | 30 + src/grp-system/libcore/include/core/timer.h | 89 + src/grp-system/libcore/include/core/unit.h | 644 ++ src/grp-system/libcore/src/Makefile | 26 + src/grp-system/libcore/src/audit-fd.c | 73 + src/grp-system/libcore/src/audit-fd.h | 23 + src/grp-system/libcore/src/automount.c | 1124 +++ src/grp-system/libcore/src/bus-policy.c | 180 + src/grp-system/libcore/src/busname.c | 1078 +++ src/grp-system/libcore/src/cgroup.c | 1991 +++++ src/grp-system/libcore/src/dbus-automount.c | 35 + src/grp-system/libcore/src/dbus-automount.h | 23 + src/grp-system/libcore/src/dbus-busname.c | 38 + src/grp-system/libcore/src/dbus-busname.h | 23 + src/grp-system/libcore/src/dbus-cgroup.c | 1110 +++ src/grp-system/libcore/src/dbus-cgroup.h | 28 + src/grp-system/libcore/src/dbus-device.c | 29 + src/grp-system/libcore/src/dbus-device.h | 24 + src/grp-system/libcore/src/dbus-execute.c | 1555 ++++ src/grp-system/libcore/src/dbus-execute.h | 45 + src/grp-system/libcore/src/dbus-job.c | 194 + src/grp-system/libcore/src/dbus-job.h | 31 + src/grp-system/libcore/src/dbus-kill.c | 123 + src/grp-system/libcore/src/dbus-kill.h | 29 + src/grp-system/libcore/src/dbus-manager.c | 2307 ++++++ src/grp-system/libcore/src/dbus-mount.c | 212 + src/grp-system/libcore/src/dbus-mount.h | 29 + src/grp-system/libcore/src/dbus-path.c | 87 + src/grp-system/libcore/src/dbus-path.h | 24 + src/grp-system/libcore/src/dbus-scope.c | 230 + src/grp-system/libcore/src/dbus-scope.h | 31 + src/grp-system/libcore/src/dbus-service.c | 323 + src/grp-system/libcore/src/dbus-service.h | 29 + src/grp-system/libcore/src/dbus-slice.c | 53 + src/grp-system/libcore/src/dbus-slice.h | 29 + src/grp-system/libcore/src/dbus-socket.c | 185 + src/grp-system/libcore/src/dbus-socket.h | 29 + src/grp-system/libcore/src/dbus-swap.c | 116 + src/grp-system/libcore/src/dbus-swap.h | 30 + src/grp-system/libcore/src/dbus-target.c | 27 + src/grp-system/libcore/src/dbus-target.h | 24 + src/grp-system/libcore/src/dbus-timer.c | 353 + src/grp-system/libcore/src/dbus-timer.h | 28 + src/grp-system/libcore/src/dbus-unit.c | 1424 ++++ src/grp-system/libcore/src/dbus-unit.h | 41 + src/grp-system/libcore/src/dbus.c | 1244 +++ src/grp-system/libcore/src/dbus.h | 44 + src/grp-system/libcore/src/device.c | 877 +++ src/grp-system/libcore/src/execute.c | 3284 ++++++++ src/grp-system/libcore/src/failure-action.c | 128 + src/grp-system/libcore/src/hostname-setup.c | 68 + src/grp-system/libcore/src/ima-setup.c | 80 + src/grp-system/libcore/src/job.c | 1260 +++ src/grp-system/libcore/src/kill.c | 68 + src/grp-system/libcore/src/killall.c | 248 + src/grp-system/libcore/src/kmod-setup.c | 128 + src/grp-system/libcore/src/linux/auto_dev-ioctl.h | 228 + src/grp-system/libcore/src/load-dropin.c | 91 + src/grp-system/libcore/src/load-dropin.h | 34 + .../libcore/src/load-fragment-gperf.gperf.m4 | 402 + src/grp-system/libcore/src/load-fragment.c | 4132 ++++++++++ src/grp-system/libcore/src/locale-setup.c | 125 + src/grp-system/libcore/src/locale-setup.h | 22 + src/grp-system/libcore/src/loopback-setup.c | 89 + src/grp-system/libcore/src/machine-id-setup.c | 258 + src/grp-system/libcore/src/manager.c | 3149 ++++++++ src/grp-system/libcore/src/mount-setup.c | 419 + src/grp-system/libcore/src/mount.c | 1883 +++++ src/grp-system/libcore/src/namespace.c | 664 ++ src/grp-system/libcore/src/path.c | 785 ++ src/grp-system/libcore/src/scope.c | 622 ++ src/grp-system/libcore/src/selinux-access.c | 284 + src/grp-system/libcore/src/selinux-access.h | 45 + src/grp-system/libcore/src/selinux-setup.c | 121 + src/grp-system/libcore/src/service.c | 3392 ++++++++ src/grp-system/libcore/src/show-status.c | 124 + src/grp-system/libcore/src/slice.c | 347 + src/grp-system/libcore/src/smack-setup.c | 346 + src/grp-system/libcore/src/socket.c | 2994 ++++++++ src/grp-system/libcore/src/swap.c | 1533 ++++ src/grp-system/libcore/src/target.c | 224 + src/grp-system/libcore/src/timer.c | 860 +++ src/grp-system/libcore/src/transaction.c | 1103 +++ src/grp-system/libcore/src/transaction.h | 51 + src/grp-system/libcore/src/unit-printf.c | 305 + src/grp-system/libcore/src/unit-printf.h | 26 + src/grp-system/libcore/src/unit.c | 3872 ++++++++++ src/grp-system/systemctl/.gitignore | 2 + src/grp-system/systemctl/Makefile | 33 + src/grp-system/systemctl/halt.xml | 176 + src/grp-system/systemctl/runlevel.xml | 192 + src/grp-system/systemctl/shutdown.xml | 175 + src/grp-system/systemctl/systemctl.c | 8079 ++++++++++++++++++++ .../systemctl/systemctl.completion.bash.in | 284 + .../systemctl/systemctl.completion.zsh.in | 391 + src/grp-system/systemctl/systemctl.xml | 1826 +++++ .../systemctl/systemd-sysv-install.SKELETON | 47 + src/grp-system/systemctl/systemd.preset.xml | 189 + src/grp-system/systemctl/telinit.xml | 179 + src/grp-system/systemd-shutdown/Makefile | 39 + src/grp-system/systemd-shutdown/halt.target | 17 + src/grp-system/systemd-shutdown/kexec.target | 17 + src/grp-system/systemd-shutdown/poweroff.target | 19 + src/grp-system/systemd-shutdown/reboot.target | 19 + src/grp-system/systemd-shutdown/shutdown.c | 446 ++ .../systemd-shutdown/systemd-halt.service.in | 17 + .../systemd-shutdown/systemd-kexec.service.in | 17 + .../systemd-shutdown/systemd-poweroff.service.in | 17 + .../systemd-shutdown/systemd-reboot.service.in | 17 + .../systemd-shutdown/systemd-shutdown.xml | 119 + src/grp-system/systemd-shutdown/umount.c | 616 ++ src/grp-system/systemd-shutdown/umount.h | 28 + src/grp-system/systemd/50-systemd-user.xorg | 7 + src/grp-system/systemd/Makefile | 73 + src/grp-system/systemd/macros.systemd.in | 113 + src/grp-system/systemd/main.c | 2195 ++++++ .../systemd/org.freedesktop.systemd1.conf | 232 + .../systemd/org.freedesktop.systemd1.policy.in.in | 70 + .../systemd/org.freedesktop.systemd1.service | 11 + src/grp-system/systemd/system.conf | 61 + src/grp-system/systemd/systemd-system.conf.xml | 393 + src/grp-system/systemd/systemd-tmpfs.tmpfiles | 14 + src/grp-system/systemd/systemd.automount.xml | 173 + src/grp-system/systemd/systemd.device.xml | 182 + src/grp-system/systemd/systemd.exec.xml | 1612 ++++ src/grp-system/systemd/systemd.generator.xml | 348 + src/grp-system/systemd/systemd.journal-fields.xml | 525 ++ src/grp-system/systemd/systemd.kill.xml | 189 + src/grp-system/systemd/systemd.link.xml | 477 ++ src/grp-system/systemd/systemd.mount.xml | 406 + src/grp-system/systemd/systemd.netdev.xml | 1164 +++ src/grp-system/systemd/systemd.network.xml | 1312 ++++ src/grp-system/systemd/systemd.nspawn.xml | 463 ++ src/grp-system/systemd/systemd.offline-updates.xml | 169 + src/grp-system/systemd/systemd.path.xml | 202 + src/grp-system/systemd/systemd.pc.in | 34 + .../systemd/systemd.resource-control.xml | 696 ++ src/grp-system/systemd/systemd.scope.xml | 108 + src/grp-system/systemd/systemd.service.xml | 1350 ++++ src/grp-system/systemd/systemd.slice.xml | 132 + src/grp-system/systemd/systemd.socket.xml | 860 +++ src/grp-system/systemd/systemd.special.xml | 941 +++ src/grp-system/systemd/systemd.swap.xml | 250 + src/grp-system/systemd/systemd.target.xml | 112 + src/grp-system/systemd/systemd.time.xml | 316 + src/grp-system/systemd/systemd.timer.xml | 313 + src/grp-system/systemd/systemd.tmpfiles | 20 + src/grp-system/systemd/systemd.unit.xml | 1484 ++++ src/grp-system/systemd/systemd.xml | 1155 +++ src/grp-system/systemd/triggers.systemd.in | 66 + src/grp-system/systemd/user.conf | 44 + 214 files changed, 92409 insertions(+) create mode 100644 src/grp-system/90-systemd.preset create mode 100644 src/grp-system/Makefile create mode 100644 src/grp-system/bootup.xml create mode 100644 src/grp-system/grp-utils/Makefile create mode 100644 src/grp-system/grp-utils/systemd-analyze/.gitignore create mode 100644 src/grp-system/grp-utils/systemd-analyze/Makefile create mode 100644 src/grp-system/grp-utils/systemd-analyze/analyze-verify.c create mode 100644 src/grp-system/grp-utils/systemd-analyze/analyze-verify.h create mode 100644 src/grp-system/grp-utils/systemd-analyze/analyze.c create mode 100644 src/grp-system/grp-utils/systemd-analyze/systemd-analyze.completion.bash create mode 100644 src/grp-system/grp-utils/systemd-analyze/systemd-analyze.completion.zsh create mode 100644 src/grp-system/grp-utils/systemd-analyze/systemd-analyze.xml create mode 100644 src/grp-system/grp-utils/systemd-delta/Makefile create mode 100644 src/grp-system/grp-utils/systemd-delta/delta.c create mode 100644 src/grp-system/grp-utils/systemd-delta/systemd-delta.completion.bash create mode 100644 src/grp-system/grp-utils/systemd-delta/systemd-delta.completion.zsh create mode 100644 src/grp-system/grp-utils/systemd-delta/systemd-delta.xml create mode 100644 src/grp-system/grp-utils/systemd-fstab-generator/Makefile create mode 100644 src/grp-system/grp-utils/systemd-fstab-generator/fstab-generator.c create mode 100644 src/grp-system/grp-utils/systemd-fstab-generator/systemd-fstab-generator.xml create mode 100644 src/grp-system/grp-utils/systemd-run/Makefile create mode 100644 src/grp-system/grp-utils/systemd-run/run.c create mode 100644 src/grp-system/grp-utils/systemd-run/systemd-run.completion.bash create mode 100644 src/grp-system/grp-utils/systemd-run/systemd-run.completion.zsh create mode 100644 src/grp-system/grp-utils/systemd-run/systemd-run.xml create mode 100644 src/grp-system/grp-utils/systemd-sysv-generator/.gitignore create mode 100644 src/grp-system/grp-utils/systemd-sysv-generator/Makefile create mode 100644 src/grp-system/grp-utils/systemd-sysv-generator/README.in create mode 100644 src/grp-system/grp-utils/systemd-sysv-generator/systemd-sysv-generator.xml create mode 100644 src/grp-system/grp-utils/systemd-sysv-generator/sysv-generator.c create mode 100644 src/grp-system/kernel-command-line.xml create mode 100644 src/grp-system/libcore/.gitignore create mode 100644 src/grp-system/libcore/Makefile create mode 100644 src/grp-system/libcore/include/core/automount.h create mode 100644 src/grp-system/libcore/include/core/bus-policy.h create mode 100644 src/grp-system/libcore/include/core/busname.h create mode 100644 src/grp-system/libcore/include/core/cgroup.h create mode 100644 src/grp-system/libcore/include/core/dbus-manager.h create mode 100644 src/grp-system/libcore/include/core/device.h create mode 100644 src/grp-system/libcore/include/core/execute.h create mode 100644 src/grp-system/libcore/include/core/failure-action.h create mode 100644 src/grp-system/libcore/include/core/hostname-setup.h create mode 100644 src/grp-system/libcore/include/core/ima-setup.h create mode 100644 src/grp-system/libcore/include/core/job.h create mode 100644 src/grp-system/libcore/include/core/kill.h create mode 100644 src/grp-system/libcore/include/core/killall.h create mode 100644 src/grp-system/libcore/include/core/kmod-setup.h create mode 100644 src/grp-system/libcore/include/core/load-fragment.h create mode 100644 src/grp-system/libcore/include/core/loopback-setup.h create mode 100644 src/grp-system/libcore/include/core/machine-id-setup.h create mode 100644 src/grp-system/libcore/include/core/manager.h create mode 100644 src/grp-system/libcore/include/core/mount-setup.h create mode 100644 src/grp-system/libcore/include/core/mount.h create mode 100644 src/grp-system/libcore/include/core/namespace.h create mode 100644 src/grp-system/libcore/include/core/path.h create mode 100644 src/grp-system/libcore/include/core/scope.h create mode 100644 src/grp-system/libcore/include/core/selinux-setup.h create mode 100644 src/grp-system/libcore/include/core/service.h create mode 100644 src/grp-system/libcore/include/core/show-status.h create mode 100644 src/grp-system/libcore/include/core/slice.h create mode 100644 src/grp-system/libcore/include/core/smack-setup.h create mode 100644 src/grp-system/libcore/include/core/socket.h create mode 100644 src/grp-system/libcore/include/core/swap.h create mode 100644 src/grp-system/libcore/include/core/target.h create mode 100644 src/grp-system/libcore/include/core/timer.h create mode 100644 src/grp-system/libcore/include/core/unit.h create mode 100644 src/grp-system/libcore/src/Makefile create mode 100644 src/grp-system/libcore/src/audit-fd.c create mode 100644 src/grp-system/libcore/src/audit-fd.h create mode 100644 src/grp-system/libcore/src/automount.c create mode 100644 src/grp-system/libcore/src/bus-policy.c create mode 100644 src/grp-system/libcore/src/busname.c create mode 100644 src/grp-system/libcore/src/cgroup.c create mode 100644 src/grp-system/libcore/src/dbus-automount.c create mode 100644 src/grp-system/libcore/src/dbus-automount.h create mode 100644 src/grp-system/libcore/src/dbus-busname.c create mode 100644 src/grp-system/libcore/src/dbus-busname.h create mode 100644 src/grp-system/libcore/src/dbus-cgroup.c create mode 100644 src/grp-system/libcore/src/dbus-cgroup.h create mode 100644 src/grp-system/libcore/src/dbus-device.c create mode 100644 src/grp-system/libcore/src/dbus-device.h create mode 100644 src/grp-system/libcore/src/dbus-execute.c create mode 100644 src/grp-system/libcore/src/dbus-execute.h create mode 100644 src/grp-system/libcore/src/dbus-job.c create mode 100644 src/grp-system/libcore/src/dbus-job.h create mode 100644 src/grp-system/libcore/src/dbus-kill.c create mode 100644 src/grp-system/libcore/src/dbus-kill.h create mode 100644 src/grp-system/libcore/src/dbus-manager.c create mode 100644 src/grp-system/libcore/src/dbus-mount.c create mode 100644 src/grp-system/libcore/src/dbus-mount.h create mode 100644 src/grp-system/libcore/src/dbus-path.c create mode 100644 src/grp-system/libcore/src/dbus-path.h create mode 100644 src/grp-system/libcore/src/dbus-scope.c create mode 100644 src/grp-system/libcore/src/dbus-scope.h create mode 100644 src/grp-system/libcore/src/dbus-service.c create mode 100644 src/grp-system/libcore/src/dbus-service.h create mode 100644 src/grp-system/libcore/src/dbus-slice.c create mode 100644 src/grp-system/libcore/src/dbus-slice.h create mode 100644 src/grp-system/libcore/src/dbus-socket.c create mode 100644 src/grp-system/libcore/src/dbus-socket.h create mode 100644 src/grp-system/libcore/src/dbus-swap.c create mode 100644 src/grp-system/libcore/src/dbus-swap.h create mode 100644 src/grp-system/libcore/src/dbus-target.c create mode 100644 src/grp-system/libcore/src/dbus-target.h create mode 100644 src/grp-system/libcore/src/dbus-timer.c create mode 100644 src/grp-system/libcore/src/dbus-timer.h create mode 100644 src/grp-system/libcore/src/dbus-unit.c create mode 100644 src/grp-system/libcore/src/dbus-unit.h create mode 100644 src/grp-system/libcore/src/dbus.c create mode 100644 src/grp-system/libcore/src/dbus.h create mode 100644 src/grp-system/libcore/src/device.c create mode 100644 src/grp-system/libcore/src/execute.c create mode 100644 src/grp-system/libcore/src/failure-action.c create mode 100644 src/grp-system/libcore/src/hostname-setup.c create mode 100644 src/grp-system/libcore/src/ima-setup.c create mode 100644 src/grp-system/libcore/src/job.c create mode 100644 src/grp-system/libcore/src/kill.c create mode 100644 src/grp-system/libcore/src/killall.c create mode 100644 src/grp-system/libcore/src/kmod-setup.c create mode 100644 src/grp-system/libcore/src/linux/auto_dev-ioctl.h create mode 100644 src/grp-system/libcore/src/load-dropin.c create mode 100644 src/grp-system/libcore/src/load-dropin.h create mode 100644 src/grp-system/libcore/src/load-fragment-gperf.gperf.m4 create mode 100644 src/grp-system/libcore/src/load-fragment.c create mode 100644 src/grp-system/libcore/src/locale-setup.c create mode 100644 src/grp-system/libcore/src/locale-setup.h create mode 100644 src/grp-system/libcore/src/loopback-setup.c create mode 100644 src/grp-system/libcore/src/machine-id-setup.c create mode 100644 src/grp-system/libcore/src/manager.c create mode 100644 src/grp-system/libcore/src/mount-setup.c create mode 100644 src/grp-system/libcore/src/mount.c create mode 100644 src/grp-system/libcore/src/namespace.c create mode 100644 src/grp-system/libcore/src/path.c create mode 100644 src/grp-system/libcore/src/scope.c create mode 100644 src/grp-system/libcore/src/selinux-access.c create mode 100644 src/grp-system/libcore/src/selinux-access.h create mode 100644 src/grp-system/libcore/src/selinux-setup.c create mode 100644 src/grp-system/libcore/src/service.c create mode 100644 src/grp-system/libcore/src/show-status.c create mode 100644 src/grp-system/libcore/src/slice.c create mode 100644 src/grp-system/libcore/src/smack-setup.c create mode 100644 src/grp-system/libcore/src/socket.c create mode 100644 src/grp-system/libcore/src/swap.c create mode 100644 src/grp-system/libcore/src/target.c create mode 100644 src/grp-system/libcore/src/timer.c create mode 100644 src/grp-system/libcore/src/transaction.c create mode 100644 src/grp-system/libcore/src/transaction.h create mode 100644 src/grp-system/libcore/src/unit-printf.c create mode 100644 src/grp-system/libcore/src/unit-printf.h create mode 100644 src/grp-system/libcore/src/unit.c create mode 100644 src/grp-system/systemctl/.gitignore create mode 100644 src/grp-system/systemctl/Makefile create mode 100644 src/grp-system/systemctl/halt.xml create mode 100644 src/grp-system/systemctl/runlevel.xml create mode 100644 src/grp-system/systemctl/shutdown.xml create mode 100644 src/grp-system/systemctl/systemctl.c create mode 100644 src/grp-system/systemctl/systemctl.completion.bash.in create mode 100644 src/grp-system/systemctl/systemctl.completion.zsh.in create mode 100644 src/grp-system/systemctl/systemctl.xml create mode 100755 src/grp-system/systemctl/systemd-sysv-install.SKELETON create mode 100644 src/grp-system/systemctl/systemd.preset.xml create mode 100644 src/grp-system/systemctl/telinit.xml create mode 100644 src/grp-system/systemd-shutdown/Makefile create mode 100644 src/grp-system/systemd-shutdown/halt.target create mode 100644 src/grp-system/systemd-shutdown/kexec.target create mode 100644 src/grp-system/systemd-shutdown/poweroff.target create mode 100644 src/grp-system/systemd-shutdown/reboot.target create mode 100644 src/grp-system/systemd-shutdown/shutdown.c create mode 100644 src/grp-system/systemd-shutdown/systemd-halt.service.in create mode 100644 src/grp-system/systemd-shutdown/systemd-kexec.service.in create mode 100644 src/grp-system/systemd-shutdown/systemd-poweroff.service.in create mode 100644 src/grp-system/systemd-shutdown/systemd-reboot.service.in create mode 100644 src/grp-system/systemd-shutdown/systemd-shutdown.xml create mode 100644 src/grp-system/systemd-shutdown/umount.c create mode 100644 src/grp-system/systemd-shutdown/umount.h create mode 100755 src/grp-system/systemd/50-systemd-user.xorg create mode 100644 src/grp-system/systemd/Makefile create mode 100644 src/grp-system/systemd/macros.systemd.in create mode 100644 src/grp-system/systemd/main.c create mode 100644 src/grp-system/systemd/org.freedesktop.systemd1.conf create mode 100644 src/grp-system/systemd/org.freedesktop.systemd1.policy.in.in create mode 100644 src/grp-system/systemd/org.freedesktop.systemd1.service create mode 100644 src/grp-system/systemd/system.conf create mode 100644 src/grp-system/systemd/systemd-system.conf.xml create mode 100644 src/grp-system/systemd/systemd-tmpfs.tmpfiles create mode 100644 src/grp-system/systemd/systemd.automount.xml create mode 100644 src/grp-system/systemd/systemd.device.xml create mode 100644 src/grp-system/systemd/systemd.exec.xml create mode 100644 src/grp-system/systemd/systemd.generator.xml create mode 100644 src/grp-system/systemd/systemd.journal-fields.xml create mode 100644 src/grp-system/systemd/systemd.kill.xml create mode 100644 src/grp-system/systemd/systemd.link.xml create mode 100644 src/grp-system/systemd/systemd.mount.xml create mode 100644 src/grp-system/systemd/systemd.netdev.xml create mode 100644 src/grp-system/systemd/systemd.network.xml create mode 100644 src/grp-system/systemd/systemd.nspawn.xml create mode 100644 src/grp-system/systemd/systemd.offline-updates.xml create mode 100644 src/grp-system/systemd/systemd.path.xml create mode 100644 src/grp-system/systemd/systemd.pc.in create mode 100644 src/grp-system/systemd/systemd.resource-control.xml create mode 100644 src/grp-system/systemd/systemd.scope.xml create mode 100644 src/grp-system/systemd/systemd.service.xml create mode 100644 src/grp-system/systemd/systemd.slice.xml create mode 100644 src/grp-system/systemd/systemd.socket.xml create mode 100644 src/grp-system/systemd/systemd.special.xml create mode 100644 src/grp-system/systemd/systemd.swap.xml create mode 100644 src/grp-system/systemd/systemd.target.xml create mode 100644 src/grp-system/systemd/systemd.time.xml create mode 100644 src/grp-system/systemd/systemd.timer.xml create mode 100644 src/grp-system/systemd/systemd.tmpfiles create mode 100644 src/grp-system/systemd/systemd.unit.xml create mode 100644 src/grp-system/systemd/systemd.xml create mode 100644 src/grp-system/systemd/triggers.systemd.in create mode 100644 src/grp-system/systemd/user.conf (limited to 'src/grp-system') diff --git a/src/grp-system/90-systemd.preset b/src/grp-system/90-systemd.preset new file mode 100644 index 0000000000..138937011c --- /dev/null +++ b/src/grp-system/90-systemd.preset @@ -0,0 +1,24 @@ +# This file is part of systemd. +# +# systemd is free software; you can redistribute it and/or modify it +# under the terms of the GNU Lesser General Public License as published by +# the Free Software Foundation; either version 2.1 of the License, or +# (at your option) any later version. + +# These ones should be enabled by default, even if distributions +# generally follow a default-off policy. + +enable remote-fs.target +enable machines.target + +enable getty@.service + +disable console-getty.service +disable console-shell.service +disable debug-shell.service + +disable halt.target +disable kexec.target +disable poweroff.target +disable reboot.target +disable rescue.target diff --git a/src/grp-system/Makefile b/src/grp-system/Makefile new file mode 100644 index 0000000000..482fbb5670 --- /dev/null +++ b/src/grp-system/Makefile @@ -0,0 +1,32 @@ +# -*- Mode: makefile; indent-tabs-mode: t -*- +# +# This file is part of systemd. +# +# Copyright 2010-2012 Lennart Poettering +# Copyright 2010-2012 Kay Sievers +# Copyright 2013 Zbigniew Jędrzejewski-Szmek +# Copyright 2013 David Strauss +# Copyright 2016 Luke Shumaker +# +# systemd is free software; you can redistribute it and/or modify it +# under the terms of the GNU Lesser General Public License as published by +# the Free Software Foundation; either version 2.1 of the License, or +# (at your option) any later version. +# +# systemd is distributed in the hope that it will be useful, but +# WITHOUT ANY WARRANTY; without even the implied warranty of +# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU +# Lesser General Public License for more details. +# +# You should have received a copy of the GNU Lesser General Public License +# along with systemd; If not, see . +include $(dir $(lastword $(MAKEFILE_LIST)))/../../config.mk +include $(topsrcdir)/build-aux/Makefile.head.mk + +nested.subdirs += grp-utils +nested.subdirs += libcore +nested.subdirs += systemctl +nested.subdirs += systemd +nested.subdirs += systemd-shutdown + +include $(topsrcdir)/build-aux/Makefile.tail.mk diff --git a/src/grp-system/bootup.xml b/src/grp-system/bootup.xml new file mode 100644 index 0000000000..986996398c --- /dev/null +++ b/src/grp-system/bootup.xml @@ -0,0 +1,305 @@ + + + + + + + + + bootup + systemd + + + + Developer + Lennart + Poettering + lennart@poettering.net + + + + + + bootup + 7 + + + + bootup + System bootup process + + + + Description + + A number of different components are involved in the system + boot. Immediately after power-up, the system BIOS will do minimal + hardware initialization, and hand control over to a boot loader + stored on a persistent storage device. This boot loader will then + invoke an OS kernel from disk (or the network). In the Linux case, + this kernel (optionally) extracts and executes an initial RAM disk + image (initrd), such as generated by + dracut8, + which looks for the root file system (possibly using + systemd1 + for this). After the root file system is found and mounted, the + initrd hands over control to the host's system manager (such as + systemd1) + stored on the OS image, which is then responsible for probing all + remaining hardware, mounting all necessary file systems and + spawning all configured services. + + On shutdown, the system manager stops all services, unmounts + all file systems (detaching the storage technologies backing + them), and then (optionally) jumps back into the initrd code which + unmounts/detaches the root file system and the storage it resides + on. As a last step, the system is powered down. + + Additional information about the system boot process may be + found in + boot7. + + + + System Manager Bootup + + At boot, the system manager on the OS image is responsible + for initializing the required file systems, services and drivers + that are necessary for operation of the system. On + systemd1 + systems, this process is split up in various discrete steps which + are exposed as target units. (See + systemd.target5 + for detailed information about target units.) The boot-up process + is highly parallelized so that the order in which specific target + units are reached is not deterministic, but still adheres to a + limited amount of ordering structure. + + When systemd starts up the system, it will activate all + units that are dependencies of default.target + (as well as recursively all dependencies of these dependencies). + Usually, default.target is simply an alias of + graphical.target or + multi-user.target, depending on whether the + system is configured for a graphical UI or only for a text + console. To enforce minimal ordering between the units pulled in, + a number of well-known target units are available, as listed on + systemd.special7. + + The following chart is a structural overview of these + well-known units and their position in the boot-up logic. The + arrows describe which units are pulled in and ordered before which + other units. Units near the top are started before units nearer to + the bottom of the chart. + +local-fs-pre.target + | + v +(various mounts and (various swap (various cryptsetup + fsck services...) devices...) devices...) (various low-level (various low-level + | | | services: udevd, API VFS mounts: + v v v tmpfiles, random mqueue, configfs, + local-fs.target swap.target cryptsetup.target seed, sysctl, ...) debugfs, ...) + | | | | | + \__________________|_________________ | ___________________|____________________/ + \|/ + v + sysinit.target + | + ____________________________________/|\________________________________________ + / | | | \ + | | | | | + v v | v v + (various (various | (various rescue.service + timers...) paths...) | sockets...) | + | | | | v + v v | v rescue.target + timers.target paths.target | sockets.target + | | | | + v \_________________ | ___________________/ + \|/ + v + basic.target + | + ____________________________________/| emergency.service + / | | | + | | | v + v v v emergency.target + display- (various system (various system + manager.service services services) + | required for | + | graphical UIs) v + | | multi-user.target + | | | + \_________________ | _________________/ + \|/ + v + graphical.target + + Target units that are commonly used as boot targets are + emphasized. These units are good choices as + goal targets, for example by passing them to the + systemd.unit= kernel command line option (see + systemd1) + or by symlinking default.target to them. + + + timers.target is pulled-in by + basic.target asynchronously. This allows + timers units to depend on services which become only available + later in boot. + + + + Bootup in the Initial RAM Disk (initrd) + The initial RAM disk implementation (initrd) can be set up + using systemd as well. In this case, boot up inside the initrd + follows the following structure. + + The default target in the initrd is + initrd.target. The bootup process begins + identical to the system manager bootup (see above) until it + reaches basic.target. From there, systemd + approaches the special target initrd.target. + When the root device becomes available, + initd-root-device.target is reached. + If the root device can be mounted at + /sysroot, the + sysroot.mount unit becomes active and + initrd-root-fs.target is reached. The service + initrd-parse-etc.service scans + /sysroot/etc/fstab for a possible + /usr mount point and additional entries + marked with the x-initrd.mount option. All + entries found are mounted below /sysroot, and + initrd-fs.target is reached. The service + initrd-cleanup.service isolates to the + initrd-switch-root.target, where cleanup + services can run. As the very last step, the + initrd-switch-root.service is activated, + which will cause the system to switch its root to + /sysroot. + + + : (beginning identical to above) + : + v + basic.target + | emergency.service + ______________________/| | + / | v + | initrd-root-device.target emergency.target + | | + | v + | sysroot.mount + | | + | v + | initrd-root-fs.target + | | + | v + v initrd-parse-etc.service + (custom initrd | + services...) v + | (sysroot-usr.mount and + | various mounts marked + | with fstab option + | x-initrd.mount...) + | | + | v + | initrd-fs.target + \______________________ | + \| + v + initrd.target + | + v + initrd-cleanup.service + isolates to + initrd-switch-root.target + | + v + ______________________/| + / v + | initrd-udevadm-cleanup-db.service + v | + (custom initrd | + services...) | + \______________________ | + \| + v + initrd-switch-root.target + | + v + initrd-switch-root.service + | + v + Transition to Host OS + + + + System Manager Shutdown + + System shutdown with systemd also consists of various target + units with some minimal ordering structure applied: + + (conflicts with (conflicts with + all system all file system + services) mounts, swaps, + | cryptsetup + | devices, ...) + | | + v v + shutdown.target umount.target + | | + \_______ ______/ + \ / + v + (various low-level + services) + | + v + final.target + | + _____________________________________/ \_________________________________ + / | | \ + | | | | + v v v v +systemd-reboot.service systemd-poweroff.service systemd-halt.service systemd-kexec.service + | | | | + v v v v + reboot.target poweroff.target halt.target kexec.target + + Commonly used system shutdown targets are + emphasized. + + + + See Also + + systemd1, + boot7, + systemd.special7, + systemd.target5, + dracut8 + + + + diff --git a/src/grp-system/grp-utils/Makefile b/src/grp-system/grp-utils/Makefile new file mode 100644 index 0000000000..2c8cc75a7c --- /dev/null +++ b/src/grp-system/grp-utils/Makefile @@ -0,0 +1,32 @@ +# -*- Mode: makefile; indent-tabs-mode: t -*- +# +# This file is part of systemd. +# +# Copyright 2010-2012 Lennart Poettering +# Copyright 2010-2012 Kay Sievers +# Copyright 2013 Zbigniew Jędrzejewski-Szmek +# Copyright 2013 David Strauss +# Copyright 2016 Luke Shumaker +# +# systemd is free software; you can redistribute it and/or modify it +# under the terms of the GNU Lesser General Public License as published by +# the Free Software Foundation; either version 2.1 of the License, or +# (at your option) any later version. +# +# systemd is distributed in the hope that it will be useful, but +# WITHOUT ANY WARRANTY; without even the implied warranty of +# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU +# Lesser General Public License for more details. +# +# You should have received a copy of the GNU Lesser General Public License +# along with systemd; If not, see . +include $(dir $(lastword $(MAKEFILE_LIST)))/../../../config.mk +include $(topsrcdir)/build-aux/Makefile.head.mk + +nested.subdirs += systemd-analyze +nested.subdirs += systemd-delta +nested.subdirs += systemd-fstab-generator +nested.subdirs += systemd-run +nested.subdirs += systemd-sysv-generator + +include $(topsrcdir)/build-aux/Makefile.tail.mk diff --git a/src/grp-system/grp-utils/systemd-analyze/.gitignore b/src/grp-system/grp-utils/systemd-analyze/.gitignore new file mode 100644 index 0000000000..752ea236c8 --- /dev/null +++ b/src/grp-system/grp-utils/systemd-analyze/.gitignore @@ -0,0 +1 @@ +/systemd-analyze diff --git a/src/grp-system/grp-utils/systemd-analyze/Makefile b/src/grp-system/grp-utils/systemd-analyze/Makefile new file mode 100644 index 0000000000..cb1b72e77d --- /dev/null +++ b/src/grp-system/grp-utils/systemd-analyze/Makefile @@ -0,0 +1,39 @@ +# -*- Mode: makefile; indent-tabs-mode: t -*- +# +# This file is part of systemd. +# +# Copyright 2010-2012 Lennart Poettering +# Copyright 2010-2012 Kay Sievers +# Copyright 2013 Zbigniew Jędrzejewski-Szmek +# Copyright 2013 David Strauss +# Copyright 2016 Luke Shumaker +# +# systemd is free software; you can redistribute it and/or modify it +# under the terms of the GNU Lesser General Public License as published by +# the Free Software Foundation; either version 2.1 of the License, or +# (at your option) any later version. +# +# systemd is distributed in the hope that it will be useful, but +# WITHOUT ANY WARRANTY; without even the implied warranty of +# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU +# Lesser General Public License for more details. +# +# You should have received a copy of the GNU Lesser General Public License +# along with systemd; If not, see . +include $(dir $(lastword $(MAKEFILE_LIST)))/../../../../config.mk +include $(topsrcdir)/build-aux/Makefile.head.mk + +bin_PROGRAMS += systemd-analyze +systemd_analyze_SOURCES = \ + src/analyze/analyze.c \ + src/analyze/analyze-verify.c \ + src/analyze/analyze-verify.h + +systemd_analyze_CFLAGS = \ + $(SECCOMP_CFLAGS) \ + $(MOUNT_CFLAGS) + +systemd_analyze_LDADD = \ + libcore.la + +include $(topsrcdir)/build-aux/Makefile.tail.mk diff --git a/src/grp-system/grp-utils/systemd-analyze/analyze-verify.c b/src/grp-system/grp-utils/systemd-analyze/analyze-verify.c new file mode 100644 index 0000000000..620ae7ee35 --- /dev/null +++ b/src/grp-system/grp-utils/systemd-analyze/analyze-verify.c @@ -0,0 +1,305 @@ +/*** + This file is part of systemd. + + Copyright 2014 Zbigniew Jędrzejewski-Szmek + + systemd is free software; you can redistribute it and/or modify it + under the terms of the GNU Lesser General Public License as published by + the Free Software Foundation; either version 2.1 of the License, or + (at your option) any later version. + + systemd is distributed in the hope that it will be useful, but + WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + Lesser General Public License for more details. + + You should have received a copy of the GNU Lesser General Public License + along with systemd; If not, see . +***/ + +#include + +#include "core/manager.h" +#include "sd-bus/bus-error.h" +#include "sd-bus/bus-util.h" +#include "systemd-basic/alloc-util.h" +#include "systemd-basic/log.h" +#include "systemd-basic/path-util.h" +#include "systemd-basic/strv.h" +#include "systemd-basic/unit-name.h" +#include "systemd-shared/pager.h" + +#include "analyze-verify.h" + +static int prepare_filename(const char *filename, char **ret) { + int r; + const char *name; + _cleanup_free_ char *abspath = NULL; + _cleanup_free_ char *dir = NULL; + _cleanup_free_ char *with_instance = NULL; + char *c; + + assert(filename); + assert(ret); + + r = path_make_absolute_cwd(filename, &abspath); + if (r < 0) + return r; + + name = basename(abspath); + if (!unit_name_is_valid(name, UNIT_NAME_ANY)) + return -EINVAL; + + if (unit_name_is_valid(name, UNIT_NAME_TEMPLATE)) { + r = unit_name_replace_instance(name, "i", &with_instance); + if (r < 0) + return r; + } + + dir = dirname_malloc(abspath); + if (!dir) + return -ENOMEM; + + if (with_instance) + c = path_join(NULL, dir, with_instance); + else + c = path_join(NULL, dir, name); + if (!c) + return -ENOMEM; + + *ret = c; + return 0; +} + +static int generate_path(char **var, char **filenames) { + char **filename; + + _cleanup_strv_free_ char **ans = NULL; + int r; + + STRV_FOREACH(filename, filenames) { + char *t; + + t = dirname_malloc(*filename); + if (!t) + return -ENOMEM; + + r = strv_consume(&ans, t); + if (r < 0) + return r; + } + + assert_se(strv_uniq(ans)); + + r = strv_extend(&ans, ""); + if (r < 0) + return r; + + *var = strv_join(ans, ":"); + if (!*var) + return -ENOMEM; + + return 0; +} + +static int verify_socket(Unit *u) { + int r; + + assert(u); + + if (u->type != UNIT_SOCKET) + return 0; + + /* Cannot run this without the service being around */ + + /* This makes sure instance is created if necessary. */ + r = socket_instantiate_service(SOCKET(u)); + if (r < 0) { + log_unit_error_errno(u, r, "Socket cannot be started, failed to create instance: %m"); + return r; + } + + /* This checks both type of sockets */ + if (UNIT_ISSET(SOCKET(u)->service)) { + Service *service; + + service = SERVICE(UNIT_DEREF(SOCKET(u)->service)); + log_unit_debug(u, "Using %s", UNIT(service)->id); + + if (UNIT(service)->load_state != UNIT_LOADED) { + log_unit_error(u, "Service %s not loaded, %s cannot be started.", UNIT(service)->id, u->id); + return -ENOENT; + } + } + + return 0; +} + +static int verify_executable(Unit *u, ExecCommand *exec) { + if (exec == NULL) + return 0; + + if (access(exec->path, X_OK) < 0) + return log_unit_error_errno(u, errno, "Command %s is not executable: %m", exec->path); + + return 0; +} + +static int verify_executables(Unit *u) { + ExecCommand *exec; + int r = 0, k; + unsigned i; + + assert(u); + + exec = u->type == UNIT_SOCKET ? SOCKET(u)->control_command : + u->type == UNIT_MOUNT ? MOUNT(u)->control_command : + u->type == UNIT_SWAP ? SWAP(u)->control_command : NULL; + k = verify_executable(u, exec); + if (k < 0 && r == 0) + r = k; + + if (u->type == UNIT_SERVICE) + for (i = 0; i < ELEMENTSOF(SERVICE(u)->exec_command); i++) { + k = verify_executable(u, SERVICE(u)->exec_command[i]); + if (k < 0 && r == 0) + r = k; + } + + if (u->type == UNIT_SOCKET) + for (i = 0; i < ELEMENTSOF(SOCKET(u)->exec_command); i++) { + k = verify_executable(u, SOCKET(u)->exec_command[i]); + if (k < 0 && r == 0) + r = k; + } + + return r; +} + +static int verify_documentation(Unit *u, bool check_man) { + char **p; + int r = 0, k; + + STRV_FOREACH(p, u->documentation) { + log_unit_debug(u, "Found documentation item: %s", *p); + + if (check_man && startswith(*p, "man:")) { + k = show_man_page(*p + 4, true); + if (k != 0) { + if (k < 0) + log_unit_error_errno(u, r, "Can't show %s: %m", *p); + else { + log_unit_error_errno(u, r, "man %s command failed with code %d", *p + 4, k); + k = -ENOEXEC; + } + if (r == 0) + r = k; + } + } + } + + /* Check remote URLs? */ + + return r; +} + +static int verify_unit(Unit *u, bool check_man) { + _cleanup_(sd_bus_error_free) sd_bus_error err = SD_BUS_ERROR_NULL; + int r, k; + + assert(u); + + if (log_get_max_level() >= LOG_DEBUG) + unit_dump(u, stdout, "\t"); + + log_unit_debug(u, "Creating %s/start job", u->id); + r = manager_add_job(u->manager, JOB_START, u, JOB_REPLACE, &err, NULL); + if (r < 0) + log_unit_error_errno(u, r, "Failed to create %s/start: %s", u->id, bus_error_message(&err, r)); + + k = verify_socket(u); + if (k < 0 && r == 0) + r = k; + + k = verify_executables(u); + if (k < 0 && r == 0) + r = k; + + k = verify_documentation(u, check_man); + if (k < 0 && r == 0) + r = k; + + return r; +} + +int verify_units(char **filenames, UnitFileScope scope, bool check_man) { + _cleanup_(sd_bus_error_free) sd_bus_error err = SD_BUS_ERROR_NULL; + _cleanup_free_ char *var = NULL; + Manager *m = NULL; + FILE *serial = NULL; + FDSet *fdset = NULL; + char **filename; + int r = 0, k; + + Unit *units[strv_length(filenames)]; + int i, count = 0; + + if (strv_isempty(filenames)) + return 0; + + /* set the path */ + r = generate_path(&var, filenames); + if (r < 0) + return log_error_errno(r, "Failed to generate unit load path: %m"); + + assert_se(set_unit_path(var) >= 0); + + r = manager_new(scope, true, &m); + if (r < 0) + return log_error_errno(r, "Failed to initialize manager: %m"); + + log_debug("Starting manager..."); + + r = manager_startup(m, serial, fdset); + if (r < 0) { + log_error_errno(r, "Failed to start manager: %m"); + goto finish; + } + + manager_clear_jobs(m); + + log_debug("Loading remaining units from the command line..."); + + STRV_FOREACH(filename, filenames) { + _cleanup_free_ char *prepared = NULL; + + log_debug("Handling %s...", *filename); + + k = prepare_filename(*filename, &prepared); + if (k < 0) { + log_error_errno(k, "Failed to prepare filename %s: %m", *filename); + if (r == 0) + r = k; + continue; + } + + k = manager_load_unit(m, NULL, prepared, &err, &units[count]); + if (k < 0) { + log_error_errno(k, "Failed to load %s: %m", *filename); + if (r == 0) + r = k; + } else + count++; + } + + for (i = 0; i < count; i++) { + k = verify_unit(units[i], check_man); + if (k < 0 && r == 0) + r = k; + } + +finish: + manager_free(m); + + return r; +} diff --git a/src/grp-system/grp-utils/systemd-analyze/analyze-verify.h b/src/grp-system/grp-utils/systemd-analyze/analyze-verify.h new file mode 100644 index 0000000000..33b809997a --- /dev/null +++ b/src/grp-system/grp-utils/systemd-analyze/analyze-verify.h @@ -0,0 +1,26 @@ +#pragma once + +/*** + This file is part of systemd. + + Copyright 2014 Zbigniew Jędrzejewski-Szmek + + systemd is free software; you can redistribute it and/or modify it + under the terms of the GNU Lesser General Public License as published by + the Free Software Foundation; either version 2.1 of the License, or + (at your option) any later version. + + systemd is distributed in the hope that it will be useful, but + WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + Lesser General Public License for more details. + + You should have received a copy of the GNU Lesser General Public License + along with systemd; If not, see . +***/ + +#include + +#include "systemd-shared/path-lookup.h" + +int verify_units(char **filenames, UnitFileScope scope, bool check_man); diff --git a/src/grp-system/grp-utils/systemd-analyze/analyze.c b/src/grp-system/grp-utils/systemd-analyze/analyze.c new file mode 100644 index 0000000000..ea775c8dd1 --- /dev/null +++ b/src/grp-system/grp-utils/systemd-analyze/analyze.c @@ -0,0 +1,1486 @@ +/*** + This file is part of systemd. + + Copyright 2010-2013 Lennart Poettering + Copyright 2013 Simon Peeters + + systemd is free software; you can redistribute it and/or modify it + under the terms of the GNU Lesser General Public License as published by + the Free Software Foundation; either version 2.1 of the License, or + (at your option) any later version. + + systemd is distributed in the hope that it will be useful, but + WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + Lesser General Public License for more details. + + You should have received a copy of the GNU Lesser General Public License + along with systemd; If not, see . +***/ + +#include +#include +#include +#include + +#include + +#include "sd-bus/bus-error.h" +#include "sd-bus/bus-util.h" +#include "systemd-basic/alloc-util.h" +#include "systemd-basic/glob-util.h" +#include "systemd-basic/hashmap.h" +#include "systemd-basic/locale-util.h" +#include "systemd-basic/log.h" +#include "systemd-basic/parse-util.h" +#include "systemd-basic/special.h" +#include "systemd-basic/strv.h" +#include "systemd-basic/strxcpyx.h" +#include "systemd-basic/terminal-util.h" +#include "systemd-basic/unit-name.h" +#include "systemd-basic/util.h" +#include "systemd-shared/bus-unit-util.h" +#include "systemd-shared/pager.h" + +#include "analyze-verify.h" + +#define SCALE_X (0.1 / 1000.0) /* pixels per us */ +#define SCALE_Y (20.0) + +#define compare(a, b) (((a) > (b))? 1 : (((b) > (a))? -1 : 0)) + +#define svg(...) printf(__VA_ARGS__) + +#define svg_bar(class, x1, x2, y) \ + svg(" \n", \ + (class), \ + SCALE_X * (x1), SCALE_Y * (y), \ + SCALE_X * ((x2) - (x1)), SCALE_Y - 1.0) + +#define svg_text(b, x, y, format, ...) \ + do { \ + svg(" ", (b) ? "left" : "right", SCALE_X * (x) + (b ? 5.0 : -5.0), SCALE_Y * (y) + 14.0); \ + svg(format, ## __VA_ARGS__); \ + svg("\n"); \ + } while (false) + +static enum dot { + DEP_ALL, + DEP_ORDER, + DEP_REQUIRE +} arg_dot = DEP_ALL; +static char** arg_dot_from_patterns = NULL; +static char** arg_dot_to_patterns = NULL; +static usec_t arg_fuzz = 0; +static bool arg_no_pager = false; +static BusTransport arg_transport = BUS_TRANSPORT_LOCAL; +static char *arg_host = NULL; +static bool arg_user = false; +static bool arg_man = true; + +struct boot_times { + usec_t firmware_time; + usec_t loader_time; + usec_t kernel_time; + usec_t kernel_done_time; + usec_t initrd_time; + usec_t userspace_time; + usec_t finish_time; + usec_t security_start_time; + usec_t security_finish_time; + usec_t generators_start_time; + usec_t generators_finish_time; + usec_t unitsload_start_time; + usec_t unitsload_finish_time; + + /* + * If we're analyzing the user instance, all timestamps will be offset + * by its own start-up timestamp, which may be arbitrarily big. + * With "plot", this causes arbitrarily wide output SVG files which almost + * completely consist of empty space. Thus we cancel out this offset. + * + * This offset is subtracted from times above by acquire_boot_times(), + * but it still needs to be subtracted from unit-specific timestamps + * (so it is stored here for reference). + */ + usec_t reverse_offset; +}; + +struct unit_times { + char *name; + usec_t activating; + usec_t activated; + usec_t deactivated; + usec_t deactivating; + usec_t time; +}; + +struct host_info { + char *hostname; + char *kernel_name; + char *kernel_release; + char *kernel_version; + char *os_pretty_name; + char *virtualization; + char *architecture; +}; + +static int bus_get_uint64_property(sd_bus *bus, const char *path, const char *interface, const char *property, uint64_t *val) { + _cleanup_(sd_bus_error_free) sd_bus_error error = SD_BUS_ERROR_NULL; + int r; + + assert(bus); + assert(path); + assert(interface); + assert(property); + assert(val); + + r = sd_bus_get_property_trivial( + bus, + "org.freedesktop.systemd1", + path, + interface, + property, + &error, + 't', val); + + if (r < 0) { + log_error("Failed to parse reply: %s", bus_error_message(&error, -r)); + return r; + } + + return 0; +} + +static int bus_get_unit_property_strv(sd_bus *bus, const char *path, const char *property, char ***strv) { + _cleanup_(sd_bus_error_free) sd_bus_error error = SD_BUS_ERROR_NULL; + int r; + + assert(bus); + assert(path); + assert(property); + assert(strv); + + r = sd_bus_get_property_strv( + bus, + "org.freedesktop.systemd1", + path, + "org.freedesktop.systemd1.Unit", + property, + &error, + strv); + if (r < 0) { + log_error("Failed to get unit property %s: %s", property, bus_error_message(&error, -r)); + return r; + } + + return 0; +} + +static int compare_unit_time(const void *a, const void *b) { + return compare(((struct unit_times *)b)->time, + ((struct unit_times *)a)->time); +} + +static int compare_unit_start(const void *a, const void *b) { + return compare(((struct unit_times *)a)->activating, + ((struct unit_times *)b)->activating); +} + +static void free_unit_times(struct unit_times *t, unsigned n) { + struct unit_times *p; + + for (p = t; p < t + n; p++) + free(p->name); + + free(t); +} + +static void subtract_timestamp(usec_t *a, usec_t b) { + assert(a); + + if (*a > 0) { + assert(*a >= b); + *a -= b; + } +} + +static int acquire_boot_times(sd_bus *bus, struct boot_times **bt) { + static struct boot_times times; + static bool cached = false; + + if (cached) + goto finish; + + assert_cc(sizeof(usec_t) == sizeof(uint64_t)); + + if (bus_get_uint64_property(bus, + "/org/freedesktop/systemd1", + "org.freedesktop.systemd1.Manager", + "FirmwareTimestampMonotonic", + ×.firmware_time) < 0 || + bus_get_uint64_property(bus, + "/org/freedesktop/systemd1", + "org.freedesktop.systemd1.Manager", + "LoaderTimestampMonotonic", + ×.loader_time) < 0 || + bus_get_uint64_property(bus, + "/org/freedesktop/systemd1", + "org.freedesktop.systemd1.Manager", + "KernelTimestamp", + ×.kernel_time) < 0 || + bus_get_uint64_property(bus, + "/org/freedesktop/systemd1", + "org.freedesktop.systemd1.Manager", + "InitRDTimestampMonotonic", + ×.initrd_time) < 0 || + bus_get_uint64_property(bus, + "/org/freedesktop/systemd1", + "org.freedesktop.systemd1.Manager", + "UserspaceTimestampMonotonic", + ×.userspace_time) < 0 || + bus_get_uint64_property(bus, + "/org/freedesktop/systemd1", + "org.freedesktop.systemd1.Manager", + "FinishTimestampMonotonic", + ×.finish_time) < 0 || + bus_get_uint64_property(bus, + "/org/freedesktop/systemd1", + "org.freedesktop.systemd1.Manager", + "SecurityStartTimestampMonotonic", + ×.security_start_time) < 0 || + bus_get_uint64_property(bus, + "/org/freedesktop/systemd1", + "org.freedesktop.systemd1.Manager", + "SecurityFinishTimestampMonotonic", + ×.security_finish_time) < 0 || + bus_get_uint64_property(bus, + "/org/freedesktop/systemd1", + "org.freedesktop.systemd1.Manager", + "GeneratorsStartTimestampMonotonic", + ×.generators_start_time) < 0 || + bus_get_uint64_property(bus, + "/org/freedesktop/systemd1", + "org.freedesktop.systemd1.Manager", + "GeneratorsFinishTimestampMonotonic", + ×.generators_finish_time) < 0 || + bus_get_uint64_property(bus, + "/org/freedesktop/systemd1", + "org.freedesktop.systemd1.Manager", + "UnitsLoadStartTimestampMonotonic", + ×.unitsload_start_time) < 0 || + bus_get_uint64_property(bus, + "/org/freedesktop/systemd1", + "org.freedesktop.systemd1.Manager", + "UnitsLoadFinishTimestampMonotonic", + ×.unitsload_finish_time) < 0) + return -EIO; + + if (times.finish_time <= 0) { + log_error("Bootup is not yet finished. Please try again later."); + return -EINPROGRESS; + } + + if (arg_user) { + /* + * User-instance-specific timestamps processing + * (see comment to reverse_offset in struct boot_times). + */ + times.reverse_offset = times.userspace_time; + + times.firmware_time = times.loader_time = times.kernel_time = times.initrd_time = times.userspace_time = 0; + subtract_timestamp(×.finish_time, times.reverse_offset); + + subtract_timestamp(×.security_start_time, times.reverse_offset); + subtract_timestamp(×.security_finish_time, times.reverse_offset); + + subtract_timestamp(×.generators_start_time, times.reverse_offset); + subtract_timestamp(×.generators_finish_time, times.reverse_offset); + + subtract_timestamp(×.unitsload_start_time, times.reverse_offset); + subtract_timestamp(×.unitsload_finish_time, times.reverse_offset); + } else { + if (times.initrd_time) + times.kernel_done_time = times.initrd_time; + else + times.kernel_done_time = times.userspace_time; + } + + cached = true; + +finish: + *bt = × + return 0; +} + +static void free_host_info(struct host_info *hi) { + + if (!hi) + return; + + free(hi->hostname); + free(hi->kernel_name); + free(hi->kernel_release); + free(hi->kernel_version); + free(hi->os_pretty_name); + free(hi->virtualization); + free(hi->architecture); + free(hi); +} + +DEFINE_TRIVIAL_CLEANUP_FUNC(struct host_info*, free_host_info); + +static int acquire_time_data(sd_bus *bus, struct unit_times **out) { + _cleanup_(sd_bus_message_unrefp) sd_bus_message *reply = NULL; + _cleanup_(sd_bus_error_free) sd_bus_error error = SD_BUS_ERROR_NULL; + int r, c = 0; + struct boot_times *boot_times = NULL; + struct unit_times *unit_times = NULL; + size_t size = 0; + UnitInfo u; + + r = acquire_boot_times(bus, &boot_times); + if (r < 0) + goto fail; + + r = sd_bus_call_method( + bus, + "org.freedesktop.systemd1", + "/org/freedesktop/systemd1", + "org.freedesktop.systemd1.Manager", + "ListUnits", + &error, &reply, + NULL); + if (r < 0) { + log_error("Failed to list units: %s", bus_error_message(&error, -r)); + goto fail; + } + + r = sd_bus_message_enter_container(reply, SD_BUS_TYPE_ARRAY, "(ssssssouso)"); + if (r < 0) { + bus_log_parse_error(r); + goto fail; + } + + while ((r = bus_parse_unit_info(reply, &u)) > 0) { + struct unit_times *t; + + if (!GREEDY_REALLOC(unit_times, size, c+1)) { + r = log_oom(); + goto fail; + } + + t = unit_times+c; + t->name = NULL; + + assert_cc(sizeof(usec_t) == sizeof(uint64_t)); + + if (bus_get_uint64_property(bus, u.unit_path, + "org.freedesktop.systemd1.Unit", + "InactiveExitTimestampMonotonic", + &t->activating) < 0 || + bus_get_uint64_property(bus, u.unit_path, + "org.freedesktop.systemd1.Unit", + "ActiveEnterTimestampMonotonic", + &t->activated) < 0 || + bus_get_uint64_property(bus, u.unit_path, + "org.freedesktop.systemd1.Unit", + "ActiveExitTimestampMonotonic", + &t->deactivating) < 0 || + bus_get_uint64_property(bus, u.unit_path, + "org.freedesktop.systemd1.Unit", + "InactiveEnterTimestampMonotonic", + &t->deactivated) < 0) { + r = -EIO; + goto fail; + } + + subtract_timestamp(&t->activating, boot_times->reverse_offset); + subtract_timestamp(&t->activated, boot_times->reverse_offset); + subtract_timestamp(&t->deactivating, boot_times->reverse_offset); + subtract_timestamp(&t->deactivated, boot_times->reverse_offset); + + if (t->activated >= t->activating) + t->time = t->activated - t->activating; + else if (t->deactivated >= t->activating) + t->time = t->deactivated - t->activating; + else + t->time = 0; + + if (t->activating == 0) + continue; + + t->name = strdup(u.id); + if (t->name == NULL) { + r = log_oom(); + goto fail; + } + c++; + } + if (r < 0) { + bus_log_parse_error(r); + goto fail; + } + + *out = unit_times; + return c; + +fail: + if (unit_times) + free_unit_times(unit_times, (unsigned) c); + return r; +} + +static int acquire_host_info(sd_bus *bus, struct host_info **hi) { + static const struct bus_properties_map hostname_map[] = { + { "Hostname", "s", NULL, offsetof(struct host_info, hostname) }, + { "KernelName", "s", NULL, offsetof(struct host_info, kernel_name) }, + { "KernelRelease", "s", NULL, offsetof(struct host_info, kernel_release) }, + { "KernelVersion", "s", NULL, offsetof(struct host_info, kernel_version) }, + { "OperatingSystemPrettyName", "s", NULL, offsetof(struct host_info, os_pretty_name) }, + {} + }; + + static const struct bus_properties_map manager_map[] = { + { "Virtualization", "s", NULL, offsetof(struct host_info, virtualization) }, + { "Architecture", "s", NULL, offsetof(struct host_info, architecture) }, + {} + }; + + _cleanup_(sd_bus_error_free) sd_bus_error error = SD_BUS_ERROR_NULL; + _cleanup_(free_host_infop) struct host_info *host; + int r; + + host = new0(struct host_info, 1); + if (!host) + return log_oom(); + + r = bus_map_all_properties(bus, + "org.freedesktop.hostname1", + "/org/freedesktop/hostname1", + hostname_map, + host); + if (r < 0) + log_debug_errno(r, "Failed to get host information from systemd-hostnamed: %s", bus_error_message(&error, r)); + + r = bus_map_all_properties(bus, + "org.freedesktop.systemd1", + "/org/freedesktop/systemd1", + manager_map, + host); + if (r < 0) + return log_error_errno(r, "Failed to get host information from systemd: %s", bus_error_message(&error, r)); + + *hi = host; + host = NULL; + + return 0; +} + +static int pretty_boot_time(sd_bus *bus, char **_buf) { + char ts[FORMAT_TIMESPAN_MAX]; + struct boot_times *t; + static char buf[4096]; + size_t size; + char *ptr; + int r; + + r = acquire_boot_times(bus, &t); + if (r < 0) + return r; + + ptr = buf; + size = sizeof(buf); + + size = strpcpyf(&ptr, size, "Startup finished in "); + if (t->firmware_time) + size = strpcpyf(&ptr, size, "%s (firmware) + ", format_timespan(ts, sizeof(ts), t->firmware_time - t->loader_time, USEC_PER_MSEC)); + if (t->loader_time) + size = strpcpyf(&ptr, size, "%s (loader) + ", format_timespan(ts, sizeof(ts), t->loader_time, USEC_PER_MSEC)); + if (t->kernel_time) + size = strpcpyf(&ptr, size, "%s (kernel) + ", format_timespan(ts, sizeof(ts), t->kernel_done_time, USEC_PER_MSEC)); + if (t->initrd_time > 0) + size = strpcpyf(&ptr, size, "%s (initrd) + ", format_timespan(ts, sizeof(ts), t->userspace_time - t->initrd_time, USEC_PER_MSEC)); + + size = strpcpyf(&ptr, size, "%s (userspace) ", format_timespan(ts, sizeof(ts), t->finish_time - t->userspace_time, USEC_PER_MSEC)); + strpcpyf(&ptr, size, "= %s", format_timespan(ts, sizeof(ts), t->firmware_time + t->finish_time, USEC_PER_MSEC)); + + ptr = strdup(buf); + if (!ptr) + return log_oom(); + + *_buf = ptr; + return 0; +} + +static void svg_graph_box(double height, double begin, double end) { + long long i; + + /* outside box, fill */ + svg("\n", + SCALE_X * (end - begin), SCALE_Y * height); + + for (i = ((long long) (begin / 100000)) * 100000; i <= end; i+=100000) { + /* lines for each second */ + if (i % 5000000 == 0) + svg(" \n" + " %.01fs\n", + SCALE_X * i, SCALE_X * i, SCALE_Y * height, SCALE_X * i, -5.0, 0.000001 * i); + else if (i % 1000000 == 0) + svg(" \n" + " %.01fs\n", + SCALE_X * i, SCALE_X * i, SCALE_Y * height, SCALE_X * i, -5.0, 0.000001 * i); + else + svg(" \n", + SCALE_X * i, SCALE_X * i, SCALE_Y * height); + } +} + +static int analyze_plot(sd_bus *bus) { + _cleanup_(free_host_infop) struct host_info *host = NULL; + struct unit_times *times; + struct boot_times *boot; + int n, m = 1, y=0; + double width; + _cleanup_free_ char *pretty_times = NULL; + struct unit_times *u; + + n = acquire_boot_times(bus, &boot); + if (n < 0) + return n; + + n = pretty_boot_time(bus, &pretty_times); + if (n < 0) + return n; + + n = acquire_host_info(bus, &host); + if (n < 0) + return n; + + n = acquire_time_data(bus, ×); + if (n <= 0) + return n; + + qsort(times, n, sizeof(struct unit_times), compare_unit_start); + + width = SCALE_X * (boot->firmware_time + boot->finish_time); + if (width < 800.0) + width = 800.0; + + if (boot->firmware_time > boot->loader_time) + m++; + if (boot->loader_time) { + m++; + if (width < 1000.0) + width = 1000.0; + } + if (boot->initrd_time) + m++; + if (boot->kernel_time) + m++; + + for (u = times; u < times + n; u++) { + double text_start, text_width; + + if (u->activating < boot->userspace_time || + u->activating > boot->finish_time) { + u->name = mfree(u->name); + continue; + } + + /* If the text cannot fit on the left side then + * increase the svg width so it fits on the right. + * TODO: calculate the text width more accurately */ + text_width = 8.0 * strlen(u->name); + text_start = (boot->firmware_time + u->activating) * SCALE_X; + if (text_width > text_start && text_width + text_start > width) + width = text_width + text_start; + + if (u->deactivated > u->activating && u->deactivated <= boot->finish_time + && u->activated == 0 && u->deactivating == 0) + u->activated = u->deactivating = u->deactivated; + if (u->activated < u->activating || u->activated > boot->finish_time) + u->activated = boot->finish_time; + if (u->deactivating < u->activated || u->activated > boot->finish_time) + u->deactivating = boot->finish_time; + if (u->deactivated < u->deactivating || u->deactivated > boot->finish_time) + u->deactivated = boot->finish_time; + m++; + } + + svg("\n" + "\n"); + + svg("\n\n", + 80.0 + width, 150.0 + (m * SCALE_Y) + + 5 * SCALE_Y /* legend */); + + /* write some basic info as a comment, including some help */ + svg("\n" + "\n" + "\n" + "\n" + "\n\n" + "\n\n", VERSION); + + /* style sheet */ + svg("\n \n\n\n"); + + svg("\n"); + svg("%s", pretty_times); + svg("%s %s (%s %s %s) %s %s", + isempty(host->os_pretty_name) ? "GNU/Linux" : host->os_pretty_name, + strempty(host->hostname), + strempty(host->kernel_name), + strempty(host->kernel_release), + strempty(host->kernel_version), + strempty(host->architecture), + strempty(host->virtualization)); + + svg("\n", 20.0 + (SCALE_X * boot->firmware_time)); + svg_graph_box(m, -(double) boot->firmware_time, boot->finish_time); + + if (boot->firmware_time) { + svg_bar("firmware", -(double) boot->firmware_time, -(double) boot->loader_time, y); + svg_text(true, -(double) boot->firmware_time, y, "firmware"); + y++; + } + if (boot->loader_time) { + svg_bar("loader", -(double) boot->loader_time, 0, y); + svg_text(true, -(double) boot->loader_time, y, "loader"); + y++; + } + if (boot->kernel_time) { + svg_bar("kernel", 0, boot->kernel_done_time, y); + svg_text(true, 0, y, "kernel"); + y++; + } + if (boot->initrd_time) { + svg_bar("initrd", boot->initrd_time, boot->userspace_time, y); + svg_text(true, boot->initrd_time, y, "initrd"); + y++; + } + svg_bar("active", boot->userspace_time, boot->finish_time, y); + svg_bar("security", boot->security_start_time, boot->security_finish_time, y); + svg_bar("generators", boot->generators_start_time, boot->generators_finish_time, y); + svg_bar("unitsload", boot->unitsload_start_time, boot->unitsload_finish_time, y); + svg_text(true, boot->userspace_time, y, "systemd"); + y++; + + for (u = times; u < times + n; u++) { + char ts[FORMAT_TIMESPAN_MAX]; + bool b; + + if (!u->name) + continue; + + svg_bar("activating", u->activating, u->activated, y); + svg_bar("active", u->activated, u->deactivating, y); + svg_bar("deactivating", u->deactivating, u->deactivated, y); + + /* place the text on the left if we have passed the half of the svg width */ + b = u->activating * SCALE_X < width / 2; + if (u->time) + svg_text(b, u->activating, y, "%s (%s)", + u->name, format_timespan(ts, sizeof(ts), u->time, USEC_PER_MSEC)); + else + svg_text(b, u->activating, y, "%s", u->name); + y++; + } + + svg("\n"); + + /* Legend */ + svg("\n"); + y++; + svg_bar("activating", 0, 300000, y); + svg_text(true, 400000, y, "Activating"); + y++; + svg_bar("active", 0, 300000, y); + svg_text(true, 400000, y, "Active"); + y++; + svg_bar("deactivating", 0, 300000, y); + svg_text(true, 400000, y, "Deactivating"); + y++; + svg_bar("security", 0, 300000, y); + svg_text(true, 400000, y, "Setting up security module"); + y++; + svg_bar("generators", 0, 300000, y); + svg_text(true, 400000, y, "Generators"); + y++; + svg_bar("unitsload", 0, 300000, y); + svg_text(true, 400000, y, "Loading unit files"); + y++; + + svg("\n\n"); + + svg("\n"); + + free_unit_times(times, (unsigned) n); + + n = 0; + return n; +} + +static int list_dependencies_print(const char *name, unsigned int level, unsigned int branches, + bool last, struct unit_times *times, struct boot_times *boot) { + unsigned int i; + char ts[FORMAT_TIMESPAN_MAX], ts2[FORMAT_TIMESPAN_MAX]; + + for (i = level; i != 0; i--) + printf("%s", special_glyph(branches & (1 << (i-1)) ? TREE_VERTICAL : TREE_SPACE)); + + printf("%s", special_glyph(last ? TREE_RIGHT : TREE_BRANCH)); + + if (times) { + if (times->time) + printf("%s%s @%s +%s%s", ansi_highlight_red(), name, + format_timespan(ts, sizeof(ts), times->activating - boot->userspace_time, USEC_PER_MSEC), + format_timespan(ts2, sizeof(ts2), times->time, USEC_PER_MSEC), ansi_normal()); + else if (times->activated > boot->userspace_time) + printf("%s @%s", name, format_timespan(ts, sizeof(ts), times->activated - boot->userspace_time, USEC_PER_MSEC)); + else + printf("%s", name); + } else + printf("%s", name); + printf("\n"); + + return 0; +} + +static int list_dependencies_get_dependencies(sd_bus *bus, const char *name, char ***deps) { + _cleanup_free_ char *path = NULL; + + assert(bus); + assert(name); + assert(deps); + + path = unit_dbus_path_from_name(name); + if (path == NULL) + return -ENOMEM; + + return bus_get_unit_property_strv(bus, path, "After", deps); +} + +static Hashmap *unit_times_hashmap; + +static int list_dependencies_compare(const void *_a, const void *_b) { + const char **a = (const char**) _a, **b = (const char**) _b; + usec_t usa = 0, usb = 0; + struct unit_times *times; + + times = hashmap_get(unit_times_hashmap, *a); + if (times) + usa = times->activated; + times = hashmap_get(unit_times_hashmap, *b); + if (times) + usb = times->activated; + + return usb - usa; +} + +static int list_dependencies_one(sd_bus *bus, const char *name, unsigned int level, char ***units, + unsigned int branches) { + _cleanup_strv_free_ char **deps = NULL; + char **c; + int r = 0; + usec_t service_longest = 0; + int to_print = 0; + struct unit_times *times; + struct boot_times *boot; + + if (strv_extend(units, name)) + return log_oom(); + + r = list_dependencies_get_dependencies(bus, name, &deps); + if (r < 0) + return r; + + qsort_safe(deps, strv_length(deps), sizeof (char*), list_dependencies_compare); + + r = acquire_boot_times(bus, &boot); + if (r < 0) + return r; + + STRV_FOREACH(c, deps) { + times = hashmap_get(unit_times_hashmap, *c); + if (times + && times->activated + && times->activated <= boot->finish_time + && (times->activated >= service_longest + || service_longest == 0)) { + service_longest = times->activated; + break; + } + } + + if (service_longest == 0 ) + return r; + + STRV_FOREACH(c, deps) { + times = hashmap_get(unit_times_hashmap, *c); + if (times && times->activated && times->activated <= boot->finish_time && (service_longest - times->activated) <= arg_fuzz) + to_print++; + } + + if (!to_print) + return r; + + STRV_FOREACH(c, deps) { + times = hashmap_get(unit_times_hashmap, *c); + if (!times + || !times->activated + || times->activated > boot->finish_time + || service_longest - times->activated > arg_fuzz) + continue; + + to_print--; + + r = list_dependencies_print(*c, level, branches, to_print == 0, times, boot); + if (r < 0) + return r; + + if (strv_contains(*units, *c)) { + r = list_dependencies_print("...", level + 1, (branches << 1) | (to_print ? 1 : 0), + true, NULL, boot); + if (r < 0) + return r; + continue; + } + + r = list_dependencies_one(bus, *c, level + 1, units, + (branches << 1) | (to_print ? 1 : 0)); + if (r < 0) + return r; + + if (!to_print) + break; + } + return 0; +} + +static int list_dependencies(sd_bus *bus, const char *name) { + _cleanup_strv_free_ char **units = NULL; + char ts[FORMAT_TIMESPAN_MAX]; + struct unit_times *times; + int r; + const char *id; + _cleanup_free_ char *path = NULL; + _cleanup_(sd_bus_message_unrefp) sd_bus_message *reply = NULL; + _cleanup_(sd_bus_error_free) sd_bus_error error = SD_BUS_ERROR_NULL; + struct boot_times *boot; + + assert(bus); + + path = unit_dbus_path_from_name(name); + if (path == NULL) + return -ENOMEM; + + r = sd_bus_get_property( + bus, + "org.freedesktop.systemd1", + path, + "org.freedesktop.systemd1.Unit", + "Id", + &error, + &reply, + "s"); + if (r < 0) { + log_error("Failed to get ID: %s", bus_error_message(&error, -r)); + return r; + } + + r = sd_bus_message_read(reply, "s", &id); + if (r < 0) + return bus_log_parse_error(r); + + times = hashmap_get(unit_times_hashmap, id); + + r = acquire_boot_times(bus, &boot); + if (r < 0) + return r; + + if (times) { + if (times->time) + printf("%s%s +%s%s\n", ansi_highlight_red(), id, + format_timespan(ts, sizeof(ts), times->time, USEC_PER_MSEC), ansi_normal()); + else if (times->activated > boot->userspace_time) + printf("%s @%s\n", id, format_timespan(ts, sizeof(ts), times->activated - boot->userspace_time, USEC_PER_MSEC)); + else + printf("%s\n", id); + } + + return list_dependencies_one(bus, name, 0, &units, 0); +} + +static int analyze_critical_chain(sd_bus *bus, char *names[]) { + struct unit_times *times; + unsigned int i; + Hashmap *h; + int n, r; + + n = acquire_time_data(bus, ×); + if (n <= 0) + return n; + + h = hashmap_new(&string_hash_ops); + if (!h) + return -ENOMEM; + + for (i = 0; i < (unsigned)n; i++) { + r = hashmap_put(h, times[i].name, ×[i]); + if (r < 0) + return r; + } + unit_times_hashmap = h; + + pager_open(arg_no_pager, false); + + puts("The time after the unit is active or started is printed after the \"@\" character.\n" + "The time the unit takes to start is printed after the \"+\" character.\n"); + + if (!strv_isempty(names)) { + char **name; + STRV_FOREACH(name, names) + list_dependencies(bus, *name); + } else + list_dependencies(bus, SPECIAL_DEFAULT_TARGET); + + hashmap_free(h); + free_unit_times(times, (unsigned) n); + return 0; +} + +static int analyze_blame(sd_bus *bus) { + struct unit_times *times; + unsigned i; + int n; + + n = acquire_time_data(bus, ×); + if (n <= 0) + return n; + + qsort(times, n, sizeof(struct unit_times), compare_unit_time); + + pager_open(arg_no_pager, false); + + for (i = 0; i < (unsigned) n; i++) { + char ts[FORMAT_TIMESPAN_MAX]; + + if (times[i].time > 0) + printf("%16s %s\n", format_timespan(ts, sizeof(ts), times[i].time, USEC_PER_MSEC), times[i].name); + } + + free_unit_times(times, (unsigned) n); + return 0; +} + +static int analyze_time(sd_bus *bus) { + _cleanup_free_ char *buf = NULL; + int r; + + r = pretty_boot_time(bus, &buf); + if (r < 0) + return r; + + puts(buf); + return 0; +} + +static int graph_one_property(sd_bus *bus, const UnitInfo *u, const char* prop, const char *color, char* patterns[], char* from_patterns[], char* to_patterns[]) { + _cleanup_strv_free_ char **units = NULL; + char **unit; + int r; + bool match_patterns; + + assert(u); + assert(prop); + assert(color); + + match_patterns = strv_fnmatch(patterns, u->id, 0); + + if (!strv_isempty(from_patterns) && + !match_patterns && + !strv_fnmatch(from_patterns, u->id, 0)) + return 0; + + r = bus_get_unit_property_strv(bus, u->unit_path, prop, &units); + if (r < 0) + return r; + + STRV_FOREACH(unit, units) { + bool match_patterns2; + + match_patterns2 = strv_fnmatch(patterns, *unit, 0); + + if (!strv_isempty(to_patterns) && + !match_patterns2 && + !strv_fnmatch(to_patterns, *unit, 0)) + continue; + + if (!strv_isempty(patterns) && !match_patterns && !match_patterns2) + continue; + + printf("\t\"%s\"->\"%s\" [color=\"%s\"];\n", u->id, *unit, color); + } + + return 0; +} + +static int graph_one(sd_bus *bus, const UnitInfo *u, char *patterns[], char *from_patterns[], char *to_patterns[]) { + int r; + + assert(bus); + assert(u); + + if (IN_SET(arg_dot, DEP_ORDER, DEP_ALL)) { + r = graph_one_property(bus, u, "After", "green", patterns, from_patterns, to_patterns); + if (r < 0) + return r; + } + + if (IN_SET(arg_dot, DEP_REQUIRE, DEP_ALL)) { + r = graph_one_property(bus, u, "Requires", "black", patterns, from_patterns, to_patterns); + if (r < 0) + return r; + r = graph_one_property(bus, u, "Requisite", "darkblue", patterns, from_patterns, to_patterns); + if (r < 0) + return r; + r = graph_one_property(bus, u, "Wants", "grey66", patterns, from_patterns, to_patterns); + if (r < 0) + return r; + r = graph_one_property(bus, u, "Conflicts", "red", patterns, from_patterns, to_patterns); + if (r < 0) + return r; + } + + return 0; +} + +static int expand_patterns(sd_bus *bus, char **patterns, char ***ret) { + _cleanup_strv_free_ char **expanded_patterns = NULL; + char **pattern; + int r; + + STRV_FOREACH(pattern, patterns) { + _cleanup_(sd_bus_error_free) sd_bus_error error = SD_BUS_ERROR_NULL; + _cleanup_free_ char *unit = NULL, *unit_id = NULL; + + if (strv_extend(&expanded_patterns, *pattern) < 0) + return log_oom(); + + if (string_is_glob(*pattern)) + continue; + + unit = unit_dbus_path_from_name(*pattern); + if (!unit) + return log_oom(); + + r = sd_bus_get_property_string( + bus, + "org.freedesktop.systemd1", + unit, + "org.freedesktop.systemd1.Unit", + "Id", + &error, + &unit_id); + if (r < 0) + return log_error_errno(r, "Failed to get ID: %s", bus_error_message(&error, r)); + + if (!streq(*pattern, unit_id)) { + if (strv_extend(&expanded_patterns, unit_id) < 0) + return log_oom(); + } + } + + *ret = expanded_patterns; + expanded_patterns = NULL; /* do not free */ + + return 0; +} + +static int dot(sd_bus *bus, char* patterns[]) { + _cleanup_(sd_bus_message_unrefp) sd_bus_message *reply = NULL; + _cleanup_(sd_bus_error_free) sd_bus_error error = SD_BUS_ERROR_NULL; + _cleanup_strv_free_ char **expanded_patterns = NULL; + _cleanup_strv_free_ char **expanded_from_patterns = NULL; + _cleanup_strv_free_ char **expanded_to_patterns = NULL; + int r; + UnitInfo u; + + r = expand_patterns(bus, patterns, &expanded_patterns); + if (r < 0) + return r; + + r = expand_patterns(bus, arg_dot_from_patterns, &expanded_from_patterns); + if (r < 0) + return r; + + r = expand_patterns(bus, arg_dot_to_patterns, &expanded_to_patterns); + if (r < 0) + return r; + + r = sd_bus_call_method( + bus, + "org.freedesktop.systemd1", + "/org/freedesktop/systemd1", + "org.freedesktop.systemd1.Manager", + "ListUnits", + &error, + &reply, + ""); + if (r < 0) { + log_error("Failed to list units: %s", bus_error_message(&error, -r)); + return r; + } + + r = sd_bus_message_enter_container(reply, SD_BUS_TYPE_ARRAY, "(ssssssouso)"); + if (r < 0) + return bus_log_parse_error(r); + + printf("digraph systemd {\n"); + + while ((r = bus_parse_unit_info(reply, &u)) > 0) { + + r = graph_one(bus, &u, expanded_patterns, expanded_from_patterns, expanded_to_patterns); + if (r < 0) + return r; + } + if (r < 0) + return bus_log_parse_error(r); + + printf("}\n"); + + log_info(" Color legend: black = Requires\n" + " dark blue = Requisite\n" + " dark grey = Wants\n" + " red = Conflicts\n" + " green = After\n"); + + if (on_tty()) + log_notice("-- You probably want to process this output with graphviz' dot tool.\n" + "-- Try a shell pipeline like 'systemd-analyze dot | dot -Tsvg > systemd.svg'!\n"); + + return 0; +} + +static int dump(sd_bus *bus, char **args) { + _cleanup_(sd_bus_message_unrefp) sd_bus_message *reply = NULL; + _cleanup_(sd_bus_error_free) sd_bus_error error = SD_BUS_ERROR_NULL; + const char *text = NULL; + int r; + + if (!strv_isempty(args)) { + log_error("Too many arguments."); + return -E2BIG; + } + + pager_open(arg_no_pager, false); + + r = sd_bus_call_method( + bus, + "org.freedesktop.systemd1", + "/org/freedesktop/systemd1", + "org.freedesktop.systemd1.Manager", + "Dump", + &error, + &reply, + ""); + if (r < 0) + return log_error_errno(r, "Failed issue method call: %s", bus_error_message(&error, r)); + + r = sd_bus_message_read(reply, "s", &text); + if (r < 0) + return bus_log_parse_error(r); + + fputs(text, stdout); + return 0; +} + +static int set_log_level(sd_bus *bus, char **args) { + _cleanup_(sd_bus_error_free) sd_bus_error error = SD_BUS_ERROR_NULL; + int r; + + assert(bus); + assert(args); + + if (strv_length(args) != 1) { + log_error("This command expects one argument only."); + return -E2BIG; + } + + r = sd_bus_set_property( + bus, + "org.freedesktop.systemd1", + "/org/freedesktop/systemd1", + "org.freedesktop.systemd1.Manager", + "LogLevel", + &error, + "s", + args[0]); + if (r < 0) + return log_error_errno(r, "Failed to issue method call: %s", bus_error_message(&error, r)); + + return 0; +} + +static int set_log_target(sd_bus *bus, char **args) { + _cleanup_(sd_bus_error_free) sd_bus_error error = SD_BUS_ERROR_NULL; + int r; + + assert(bus); + assert(args); + + if (strv_length(args) != 1) { + log_error("This command expects one argument only."); + return -E2BIG; + } + + r = sd_bus_set_property( + bus, + "org.freedesktop.systemd1", + "/org/freedesktop/systemd1", + "org.freedesktop.systemd1.Manager", + "LogTarget", + &error, + "s", + args[0]); + if (r < 0) + return log_error_errno(r, "Failed to issue method call: %s", bus_error_message(&error, r)); + + return 0; +} + +static void help(void) { + + pager_open(arg_no_pager, false); + + printf("%s [OPTIONS...] {COMMAND} ...\n\n" + "Profile systemd, show unit dependencies, check unit files.\n\n" + " -h --help Show this help\n" + " --version Show package version\n" + " --no-pager Do not pipe output into a pager\n" + " --system Operate on system systemd instance\n" + " --user Operate on user systemd instance\n" + " -H --host=[USER@]HOST Operate on remote host\n" + " -M --machine=CONTAINER Operate on local container\n" + " --order Show only order in the graph\n" + " --require Show only requirement in the graph\n" + " --from-pattern=GLOB Show only origins in the graph\n" + " --to-pattern=GLOB Show only destinations in the graph\n" + " --fuzz=SECONDS Also print also services which finished SECONDS\n" + " earlier than the latest in the branch\n" + " --man[=BOOL] Do [not] check for existence of man pages\n\n" + "Commands:\n" + " time Print time spent in the kernel\n" + " blame Print list of running units ordered by time to init\n" + " critical-chain Print a tree of the time critical chain of units\n" + " plot Output SVG graphic showing service initialization\n" + " dot Output dependency graph in dot(1) format\n" + " set-log-level LEVEL Set logging threshold for manager\n" + " set-log-target TARGET Set logging target for manager\n" + " dump Output state serialization of service manager\n" + " verify FILE... Check unit files for correctness\n" + , program_invocation_short_name); + + /* When updating this list, including descriptions, apply + * changes to shell-completion/bash/systemd-analyze and + * shell-completion/zsh/_systemd-analyze too. */ +} + +static int parse_argv(int argc, char *argv[]) { + enum { + ARG_VERSION = 0x100, + ARG_ORDER, + ARG_REQUIRE, + ARG_USER, + ARG_SYSTEM, + ARG_DOT_FROM_PATTERN, + ARG_DOT_TO_PATTERN, + ARG_FUZZ, + ARG_NO_PAGER, + ARG_MAN, + }; + + static const struct option options[] = { + { "help", no_argument, NULL, 'h' }, + { "version", no_argument, NULL, ARG_VERSION }, + { "order", no_argument, NULL, ARG_ORDER }, + { "require", no_argument, NULL, ARG_REQUIRE }, + { "user", no_argument, NULL, ARG_USER }, + { "system", no_argument, NULL, ARG_SYSTEM }, + { "from-pattern", required_argument, NULL, ARG_DOT_FROM_PATTERN }, + { "to-pattern", required_argument, NULL, ARG_DOT_TO_PATTERN }, + { "fuzz", required_argument, NULL, ARG_FUZZ }, + { "no-pager", no_argument, NULL, ARG_NO_PAGER }, + { "man", optional_argument, NULL, ARG_MAN }, + { "host", required_argument, NULL, 'H' }, + { "machine", required_argument, NULL, 'M' }, + {} + }; + + int r, c; + + assert(argc >= 0); + assert(argv); + + while ((c = getopt_long(argc, argv, "hH:M:", options, NULL)) >= 0) + switch (c) { + + case 'h': + help(); + return 0; + + case ARG_VERSION: + return version(); + + case ARG_USER: + arg_user = true; + break; + + case ARG_SYSTEM: + arg_user = false; + break; + + case ARG_ORDER: + arg_dot = DEP_ORDER; + break; + + case ARG_REQUIRE: + arg_dot = DEP_REQUIRE; + break; + + case ARG_DOT_FROM_PATTERN: + if (strv_extend(&arg_dot_from_patterns, optarg) < 0) + return log_oom(); + + break; + + case ARG_DOT_TO_PATTERN: + if (strv_extend(&arg_dot_to_patterns, optarg) < 0) + return log_oom(); + + break; + + case ARG_FUZZ: + r = parse_sec(optarg, &arg_fuzz); + if (r < 0) + return r; + break; + + case ARG_NO_PAGER: + arg_no_pager = true; + break; + + case 'H': + arg_transport = BUS_TRANSPORT_REMOTE; + arg_host = optarg; + break; + + case 'M': + arg_transport = BUS_TRANSPORT_MACHINE; + arg_host = optarg; + break; + + case ARG_MAN: + if (optarg) { + r = parse_boolean(optarg); + if (r < 0) { + log_error("Failed to parse --man= argument."); + return -EINVAL; + } + + arg_man = !!r; + } else + arg_man = true; + + break; + + case '?': + return -EINVAL; + + default: + assert_not_reached("Unhandled option code."); + } + + return 1; /* work to do */ +} + +int main(int argc, char *argv[]) { + int r; + + setlocale(LC_ALL, ""); + setlocale(LC_NUMERIC, "C"); /* we want to format/parse floats in C style */ + log_parse_environment(); + log_open(); + + r = parse_argv(argc, argv); + if (r <= 0) + goto finish; + + if (streq_ptr(argv[optind], "verify")) + r = verify_units(argv+optind+1, + arg_user ? UNIT_FILE_USER : UNIT_FILE_SYSTEM, + arg_man); + else { + _cleanup_(sd_bus_flush_close_unrefp) sd_bus *bus = NULL; + + r = bus_connect_transport_systemd(arg_transport, arg_host, arg_user, &bus); + if (r < 0) { + log_error_errno(r, "Failed to create bus connection: %m"); + goto finish; + } + + if (!argv[optind] || streq(argv[optind], "time")) + r = analyze_time(bus); + else if (streq(argv[optind], "blame")) + r = analyze_blame(bus); + else if (streq(argv[optind], "critical-chain")) + r = analyze_critical_chain(bus, argv+optind+1); + else if (streq(argv[optind], "plot")) + r = analyze_plot(bus); + else if (streq(argv[optind], "dot")) + r = dot(bus, argv+optind+1); + else if (streq(argv[optind], "dump")) + r = dump(bus, argv+optind+1); + else if (streq(argv[optind], "set-log-level")) + r = set_log_level(bus, argv+optind+1); + else if (streq(argv[optind], "set-log-target")) + r = set_log_target(bus, argv+optind+1); + else + log_error("Unknown operation '%s'.", argv[optind]); + } + +finish: + pager_close(); + + strv_free(arg_dot_from_patterns); + strv_free(arg_dot_to_patterns); + + return r < 0 ? EXIT_FAILURE : EXIT_SUCCESS; +} diff --git a/src/grp-system/grp-utils/systemd-analyze/systemd-analyze.completion.bash b/src/grp-system/grp-utils/systemd-analyze/systemd-analyze.completion.bash new file mode 100644 index 0000000000..7a5f46ba1d --- /dev/null +++ b/src/grp-system/grp-utils/systemd-analyze/systemd-analyze.completion.bash @@ -0,0 +1,117 @@ +# systemd-analyze(1) completion -*- shell-script -*- +# +# This file is part of systemd. +# +# Copyright 2010 Ran Benita +# Copyright 2013 Harald Hoyer +# +# systemd is free software; you can redistribute it and/or modify it +# under the terms of the GNU Lesser General Public License as published by +# the Free Software Foundation; either version 2.1 of the License, or +# (at your option) any later version. +# +# systemd is distributed in the hope that it will be useful, but +# WITHOUT ANY WARRANTY; without even the implied warranty of +# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU +# General Public License for more details. +# +# You should have received a copy of the GNU Lesser General Public License +# along with systemd; If not, see . + +__contains_word () { + local w word=$1; shift + for w in "$@"; do + [[ $w = "$word" ]] && return + done +} + +__get_machines() { + local a b + machinectl list --no-legend --no-pager | { while read a b; do echo " $a"; done; }; +} + +_systemd_analyze() { + local i verb comps + local cur=${COMP_WORDS[COMP_CWORD]} prev=${COMP_WORDS[COMP_CWORD-1]} + + local -A OPTS=( + [STANDALONE]='--help --version --system --user --order --require --no-pager --man' + [ARG]='-H --host -M --machine --fuzz --from-pattern --to-pattern ' + ) + + local -A VERBS=( + [STANDALONE]='time blame plot dump' + [CRITICAL_CHAIN]='critical-chain' + [DOT]='dot' + [LOG_LEVEL]='set-log-level' + [VERIFY]='verify' + ) + + _init_completion || return + + for ((i=0; i < COMP_CWORD; i++)); do + if __contains_word "${COMP_WORDS[i]}" ${VERBS[*]} && + ! __contains_word "${COMP_WORDS[i-1]}" ${OPTS[ARG]}; then + verb=${COMP_WORDS[i]} + break + fi + done + + if __contains_word "$prev" ${OPTS[ARG]}; then + case $prev in + --host|-H) + comps=$(compgen -A hostname) + ;; + --machine|-M) + comps=$( __get_machines ) + ;; + esac + COMPREPLY=( $(compgen -W '$comps' -- "$cur") ) + return 0 + fi + + if [[ -z $verb && $cur = -* ]]; then + COMPREPLY=( $(compgen -W '${OPTS[*]}' -- "$cur") ) + return 0 + fi + + if [[ -z $verb ]]; then + comps=${VERBS[*]} + + elif __contains_word "$verb" ${VERBS[STANDALONE]}; then + if [[ $cur = -* ]]; then + comps='--help --version --system --user' + fi + + elif __contains_word "$verb" ${VERBS[CRITICAL_CHAIN]}; then + if [[ $cur = -* ]]; then + comps='--help --version --system --user --fuzz' + fi + + elif __contains_word "$verb" ${VERBS[DOT]}; then + if [[ $cur = -* ]]; then + comps='--help --version --system --user --from-pattern --to-pattern --order --require' + fi + + elif __contains_word "$verb" ${VERBS[LOG_LEVEL]}; then + if [[ $cur = -* ]]; then + comps='--help --version --system --user' + else + comps='debug info notice warning err crit alert emerg' + fi + + elif __contains_word "$verb" ${VERBS[VERIFY]}; then + if [[ $cur = -* ]]; then + comps='--help --version --system --user --man' + else + comps=$( compgen -A file -- "$cur" ) + compopt -o filenames + fi + + fi + + COMPREPLY=( $(compgen -W '$comps' -- "$cur") ) + return 0 +} + +complete -F _systemd_analyze systemd-analyze diff --git a/src/grp-system/grp-utils/systemd-analyze/systemd-analyze.completion.zsh b/src/grp-system/grp-utils/systemd-analyze/systemd-analyze.completion.zsh new file mode 100644 index 0000000000..efafddc686 --- /dev/null +++ b/src/grp-system/grp-utils/systemd-analyze/systemd-analyze.completion.zsh @@ -0,0 +1,58 @@ +#compdef systemd-analyze + +_systemd_analyze_set-log-level() { + local -a _levels + _levels=(debug info notice warning err crit alert emerg) + _describe -t level 'logging level' _levels || compadd "$@" +} + +_systemd_analyze_verify() { + _sd_unit_files +} + +_systemd_analyze_command(){ + local -a _systemd_analyze_cmds + # Descriptions taken from systemd-analyze --help. + _systemd_analyze_cmds=( + 'time:Print time spent in the kernel before reaching userspace' + 'blame:Print list of running units ordered by time to init' + 'critical-chain:Print a tree of the time critical chain of units' + 'plot:Output SVG graphic showing service initialization' + 'dot:Dump dependency graph (in dot(1) format)' + 'dump:Dump server status' + 'set-log-level:Set systemd log threshold' + 'verify:Check unit files for correctness' + ) + + if (( CURRENT == 1 )); then + _describe "options" _systemd_analyze_cmds + else + local curcontext="$curcontext" + cmd="${${_systemd_analyze_cmds[(r)$words[1]:*]%%:*}}" + if (( $#cmd )); then + if (( $+functions[_systemd_analyze_$cmd] )) && (( CURRENT == 2 )); then + _systemd_analyze_$cmd + else + _message "no more options" + fi + else + _message "unknown systemd-analyze command: $words[1]" + fi + fi +} + +_arguments \ + {-h,--help}'[Show help text]' \ + '--version[Show package version]' \ + '--system[Operate on system systemd instance]' \ + '--user[Operate on user systemd instance]' \ + '--no-pager[Do not pipe output into a pager]' \ + '--man=[Do (not) check for existence of man pages]:boolean:(1 0)' \ + '--order[When generating graph for dot, show only order]' \ + '--require[When generating graph for dot, show only requirement]' \ + '--fuzz=[When printing the tree of the critical chain, print also services, which finished TIMESPAN earlier, than the latest in the branch]:TIMESPAN' \ + '--from-pattern=[When generating a dependency graph, filter only origins]:GLOB' \ + '--to-pattern=[When generating a dependency graph, filter only destinations]:GLOB' \ + {-H+,--host=}'[Operate on remote host]:userathost:_sd_hosts_or_user_at_host' \ + {-M+,--machine=}'[Operate on local container]:machine:_sd_machines' \ + '*::systemd-analyze commands:_systemd_analyze_command' diff --git a/src/grp-system/grp-utils/systemd-analyze/systemd-analyze.xml b/src/grp-system/grp-utils/systemd-analyze/systemd-analyze.xml new file mode 100644 index 0000000000..bc37765dff --- /dev/null +++ b/src/grp-system/grp-utils/systemd-analyze/systemd-analyze.xml @@ -0,0 +1,388 @@ + + + + + + + + + systemd-analyze + systemd + + + + Developer + Lennart + Poettering + lennart@poettering.net + + + Developer + Harald + Hoyer + harald@redhat.com + + + + + + systemd-analyze + 1 + + + + systemd-analyze + Analyze system boot-up performance + + + + + systemd-analyze + OPTIONS + time + + + systemd-analyze + OPTIONS + blame + + + systemd-analyze + OPTIONS + critical-chain + UNIT + + + systemd-analyze + OPTIONS + plot + > file.svg + + + systemd-analyze + OPTIONS + dot + PATTERN + > file.dot + + + systemd-analyze + OPTIONS + dump + + + systemd-analyze + OPTIONS + set-log-level + LEVEL + + + systemd-analyze + OPTIONS + set-log-target + TARGET + + + systemd-analyze + OPTIONS + verify + FILES + + + + + Description + + systemd-analyze may be used to determine + system boot-up performance statistics and retrieve other state and + tracing information from the system and service manager, and to + verify the correctness of unit files. + + systemd-analyze time prints the time + spent in the kernel before userspace has been reached, the time + spent in the initial RAM disk (initrd) before normal system + userspace has been reached, and the time normal system userspace + took to initialize. Note that these measurements simply measure + the time passed up to the point where all system services have + been spawned, but not necessarily until they fully finished + initialization or the disk is idle. + + systemd-analyze blame prints a list of + all running units, ordered by the time they took to initialize. + This information may be used to optimize boot-up times. Note that + the output might be misleading as the initialization of one + service might be slow simply because it waits for the + initialization of another service to complete. + + systemd-analyze critical-chain + [UNIT...] prints a tree of + the time-critical chain of units (for each of the specified + UNITs or for the default target + otherwise). The time after the unit is active or started is + printed after the "@" character. The time the unit takes to start + is printed after the "+" character. Note that the output might be + misleading as the initialization of one service might depend on + socket activation and because of the parallel execution of + units. + + systemd-analyze plot prints an SVG + graphic detailing which system services have been started at what + time, highlighting the time they spent on initialization. + + systemd-analyze dot generates textual + dependency graph description in dot format for further processing + with the GraphViz + dot1 + tool. Use a command line like systemd-analyze dot | dot + -Tsvg > systemd.svg to generate a graphical dependency + tree. Unless or + is passed, the generated graph will + show both ordering and requirement dependencies. Optional pattern + globbing style specifications (e.g. *.target) + may be given at the end. A unit dependency is included in the + graph if any of these patterns match either the origin or + destination node. + + systemd-analyze dump outputs a (usually + very long) human-readable serialization of the complete server + state. Its format is subject to change without notice and should + not be parsed by applications. + + systemd-analyze set-log-level + LEVEL changes the current log + level of the systemd daemon to + LEVEL (accepts the same values as + described in + systemd1). + + systemd-analyze set-log-target + TARGET changes the current log + target of the systemd daemon to + TARGET (accepts the same values as + , described in + systemd1). + + systemd-analyze verify will load unit + files and print warnings if any errors are detected. Files + specified on the command line will be loaded, but also any other + units referenced by them. This command works by prepending the + directories for all command line arguments at the beginning of the + unit load path, which means that all units files found in those + directories will be used in preference to the unit files found in + the standard locations, even if not listed explicitly. + + If no command is passed, systemd-analyze + time is implied. + + + + + Options + + The following options are understood: + + + + + + Operates on the user systemd + instance. + + + + + + Operates on the system systemd instance. This + is the implied default. + + + + + + + When used in conjunction with the + dot command (see above), selects which + dependencies are shown in the dependency graph. If + is passed, only dependencies of type + After= or Before= are + shown. If is passed, only + dependencies of type Requires=, + Requisite=, + Wants= and Conflicts= + are shown. If neither is passed, this shows dependencies of + all these types. + + + + + + + When used in conjunction with the + dot command (see above), this selects which + relationships are shown in the dependency graph. Both options + require a + glob7 + pattern as an argument, which will be matched against the + left-hand and the right-hand, respectively, nodes of a + relationship. + + Each of these can be used more than once, in which case + the unit name must match one of the values. When tests for + both sides of the relation are present, a relation must pass + both tests to be shown. When patterns are also specified as + positional arguments, they must match at least one side of the + relation. In other words, patterns specified with those two + options will trim the list of edges matched by the positional + arguments, if any are given, and fully determine the list of + edges shown otherwise. + + + + timespan + + When used in conjunction with the + critical-chain command (see above), also + show units, which finished timespan + earlier, than the latest unit in the same level. The unit of + timespan is seconds unless + specified with a different unit, e.g. + "50ms". + + + + + + Do not invoke man to verify the existence of + man pages listed in Documentation=. + + + + + + + + + + + + + + + Exit status + + On success, 0 is returned, a non-zero failure code + otherwise. + + + + Examples for <command>dot</command> + + + Plots all dependencies of any unit whose name starts with + <literal>avahi-daemon</literal> + + $ systemd-analyze dot 'avahi-daemon.*' | dot -Tsvg > avahi.svg + $ eog avahi.svg + + + + Plots the dependencies between all known target units + + systemd-analyze dot --to-pattern='*.target' --from-pattern='*.target' | dot -Tsvg > targets.svg +$ eog targets.svg + + + + + Examples for <command>verify</command> + + The following errors are currently detected: + + unknown sections and directives, + + + missing dependencies which are required to start + the given unit, + + man pages listed in + Documentation= which are not found in the + system, + + commands listed in ExecStart= + and similar which are not found in the system or not + executable. + + + + Misspelt directives + + $ cat ./user.slice +[Unit] +WhatIsThis=11 +Documentation=man:nosuchfile(1) +Requires=different.service + +[Service] +Desription=x + +$ systemd-analyze verify ./user.slice +[./user.slice:9] Unknown lvalue 'WhatIsThis' in section 'Unit' +[./user.slice:13] Unknown section 'Service'. Ignoring. +Error: org.freedesktop.systemd1.LoadFailed: + Unit different.service failed to load: + No such file or directory. +Failed to create user.slice/start: Invalid argument +user.slice: man nosuchfile(1) command failed with code 16 + + + + + Missing service units + + $ tail ./a.socket ./b.socket +==> ./a.socket <== +[Socket] +ListenStream=100 + +==> ./b.socket <== +[Socket] +ListenStream=100 +Accept=yes + +$ systemd-analyze verify ./a.socket ./b.socket +Service a.service not loaded, a.socket cannot be started. +Service b@0.service not loaded, b.socket cannot be started. + + + + + + + + See Also + + systemd1, + systemctl1 + + + + diff --git a/src/grp-system/grp-utils/systemd-delta/Makefile b/src/grp-system/grp-utils/systemd-delta/Makefile new file mode 100644 index 0000000000..7273647c52 --- /dev/null +++ b/src/grp-system/grp-utils/systemd-delta/Makefile @@ -0,0 +1,33 @@ +# -*- Mode: makefile; indent-tabs-mode: t -*- +# +# This file is part of systemd. +# +# Copyright 2010-2012 Lennart Poettering +# Copyright 2010-2012 Kay Sievers +# Copyright 2013 Zbigniew Jędrzejewski-Szmek +# Copyright 2013 David Strauss +# Copyright 2016 Luke Shumaker +# +# systemd is free software; you can redistribute it and/or modify it +# under the terms of the GNU Lesser General Public License as published by +# the Free Software Foundation; either version 2.1 of the License, or +# (at your option) any later version. +# +# systemd is distributed in the hope that it will be useful, but +# WITHOUT ANY WARRANTY; without even the implied warranty of +# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU +# Lesser General Public License for more details. +# +# You should have received a copy of the GNU Lesser General Public License +# along with systemd; If not, see . +include $(dir $(lastword $(MAKEFILE_LIST)))/../../../../config.mk +include $(topsrcdir)/build-aux/Makefile.head.mk + +bin_PROGRAMS += systemd-delta +systemd_delta_SOURCES = \ + src/delta/delta.c + +systemd_delta_LDADD = \ + libsystemd-shared.la + +include $(topsrcdir)/build-aux/Makefile.tail.mk diff --git a/src/grp-system/grp-utils/systemd-delta/delta.c b/src/grp-system/grp-utils/systemd-delta/delta.c new file mode 100644 index 0000000000..0142803b46 --- /dev/null +++ b/src/grp-system/grp-utils/systemd-delta/delta.c @@ -0,0 +1,635 @@ +/*** + This file is part of systemd. + + Copyright 2012 Lennart Poettering + Copyright 2013 Zbigniew Jędrzejewski-Szmek + + systemd is free software; you can redistribute it and/or modify it + under the terms of the GNU Lesser General Public License as published by + the Free Software Foundation; either version 2.1 of the License, or + (at your option) any later version. + + systemd is distributed in the hope that it will be useful, but + WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + Lesser General Public License for more details. + + You should have received a copy of the GNU Lesser General Public License + along with systemd; If not, see . +***/ + +#include +#include +#include +#include +#include + +#include "systemd-basic/alloc-util.h" +#include "systemd-basic/dirent-util.h" +#include "systemd-basic/fd-util.h" +#include "systemd-basic/fs-util.h" +#include "systemd-basic/hashmap.h" +#include "systemd-basic/locale-util.h" +#include "systemd-basic/log.h" +#include "systemd-basic/parse-util.h" +#include "systemd-basic/path-util.h" +#include "systemd-basic/process-util.h" +#include "systemd-basic/signal-util.h" +#include "systemd-basic/stat-util.h" +#include "systemd-basic/string-util.h" +#include "systemd-basic/strv.h" +#include "systemd-basic/terminal-util.h" +#include "systemd-basic/util.h" +#include "systemd-shared/pager.h" + +static const char prefixes[] = + "/etc\0" + "/run\0" + "/usr/local/lib\0" + "/usr/local/share\0" + "/usr/lib\0" + "/usr/share\0" +#ifdef HAVE_SPLIT_USR + "/lib\0" +#endif + ; + +static const char suffixes[] = + "sysctl.d\0" + "tmpfiles.d\0" + "modules-load.d\0" + "binfmt.d\0" + "systemd/system\0" + "systemd/user\0" + "systemd/system-preset\0" + "systemd/user-preset\0" + "udev/rules.d\0" + "modprobe.d\0"; + +static const char have_dropins[] = + "systemd/system\0" + "systemd/user\0"; + +static bool arg_no_pager = false; +static int arg_diff = -1; + +static enum { + SHOW_MASKED = 1 << 0, + SHOW_EQUIVALENT = 1 << 1, + SHOW_REDIRECTED = 1 << 2, + SHOW_OVERRIDDEN = 1 << 3, + SHOW_UNCHANGED = 1 << 4, + SHOW_EXTENDED = 1 << 5, + + SHOW_DEFAULTS = + (SHOW_MASKED | SHOW_EQUIVALENT | SHOW_REDIRECTED | SHOW_OVERRIDDEN | SHOW_EXTENDED) +} arg_flags = 0; + +static int equivalent(const char *a, const char *b) { + _cleanup_free_ char *x = NULL, *y = NULL; + + x = canonicalize_file_name(a); + if (!x) + return -errno; + + y = canonicalize_file_name(b); + if (!y) + return -errno; + + return path_equal(x, y); +} + +static int notify_override_masked(const char *top, const char *bottom) { + if (!(arg_flags & SHOW_MASKED)) + return 0; + + printf("%s%s%s %s %s %s\n", + ansi_highlight_red(), "[MASKED]", ansi_normal(), + top, special_glyph(ARROW), bottom); + return 1; +} + +static int notify_override_equivalent(const char *top, const char *bottom) { + if (!(arg_flags & SHOW_EQUIVALENT)) + return 0; + + printf("%s%s%s %s %s %s\n", + ansi_highlight_green(), "[EQUIVALENT]", ansi_normal(), + top, special_glyph(ARROW), bottom); + return 1; +} + +static int notify_override_redirected(const char *top, const char *bottom) { + if (!(arg_flags & SHOW_REDIRECTED)) + return 0; + + printf("%s%s%s %s %s %s\n", + ansi_highlight(), "[REDIRECTED]", ansi_normal(), + top, special_glyph(ARROW), bottom); + return 1; +} + +static int notify_override_overridden(const char *top, const char *bottom) { + if (!(arg_flags & SHOW_OVERRIDDEN)) + return 0; + + printf("%s%s%s %s %s %s\n", + ansi_highlight(), "[OVERRIDDEN]", ansi_normal(), + top, special_glyph(ARROW), bottom); + return 1; +} + +static int notify_override_extended(const char *top, const char *bottom) { + if (!(arg_flags & SHOW_EXTENDED)) + return 0; + + printf("%s%s%s %s %s %s\n", + ansi_highlight(), "[EXTENDED]", ansi_normal(), + top, special_glyph(ARROW), bottom); + return 1; +} + +static int notify_override_unchanged(const char *f) { + if (!(arg_flags & SHOW_UNCHANGED)) + return 0; + + printf("[UNCHANGED] %s\n", f); + return 1; +} + +static int found_override(const char *top, const char *bottom) { + _cleanup_free_ char *dest = NULL; + int k; + pid_t pid; + + assert(top); + assert(bottom); + + if (null_or_empty_path(top) > 0) + return notify_override_masked(top, bottom); + + k = readlink_malloc(top, &dest); + if (k >= 0) { + if (equivalent(dest, bottom) > 0) + return notify_override_equivalent(top, bottom); + else + return notify_override_redirected(top, bottom); + } + + k = notify_override_overridden(top, bottom); + if (!arg_diff) + return k; + + putchar('\n'); + + fflush(stdout); + + pid = fork(); + if (pid < 0) + return log_error_errno(errno, "Failed to fork off diff: %m"); + else if (pid == 0) { + + (void) reset_all_signal_handlers(); + (void) reset_signal_mask(); + assert_se(prctl(PR_SET_PDEATHSIG, SIGTERM) == 0); + + execlp("diff", "diff", "-us", "--", bottom, top, NULL); + log_error_errno(errno, "Failed to execute diff: %m"); + _exit(EXIT_FAILURE); + } + + wait_for_terminate_and_warn("diff", pid, false); + putchar('\n'); + + return k; +} + +static int enumerate_dir_d(Hashmap *top, Hashmap *bottom, Hashmap *drops, const char *toppath, const char *drop) { + _cleanup_free_ char *unit = NULL; + _cleanup_free_ char *path = NULL; + _cleanup_strv_free_ char **list = NULL; + char **file; + char *c; + int r; + + assert(!endswith(drop, "/")); + + path = strjoin(toppath, "/", drop, NULL); + if (!path) + return -ENOMEM; + + log_debug("Looking at %s", path); + + unit = strdup(drop); + if (!unit) + return -ENOMEM; + + c = strrchr(unit, '.'); + if (!c) + return -EINVAL; + *c = 0; + + r = get_files_in_directory(path, &list); + if (r < 0) + return log_error_errno(r, "Failed to enumerate %s: %m", path); + + STRV_FOREACH(file, list) { + Hashmap *h; + int k; + char *p; + char *d; + + if (!endswith(*file, ".conf")) + continue; + + p = strjoin(path, "/", *file, NULL); + if (!p) + return -ENOMEM; + d = p + strlen(toppath) + 1; + + log_debug("Adding at top: %s %s %s", d, special_glyph(ARROW), p); + k = hashmap_put(top, d, p); + if (k >= 0) { + p = strdup(p); + if (!p) + return -ENOMEM; + d = p + strlen(toppath) + 1; + } else if (k != -EEXIST) { + free(p); + return k; + } + + log_debug("Adding at bottom: %s %s %s", d, special_glyph(ARROW), p); + free(hashmap_remove(bottom, d)); + k = hashmap_put(bottom, d, p); + if (k < 0) { + free(p); + return k; + } + + h = hashmap_get(drops, unit); + if (!h) { + h = hashmap_new(&string_hash_ops); + if (!h) + return -ENOMEM; + hashmap_put(drops, unit, h); + unit = strdup(unit); + if (!unit) + return -ENOMEM; + } + + p = strdup(p); + if (!p) + return -ENOMEM; + + log_debug("Adding to drops: %s %s %s %s %s", + unit, special_glyph(ARROW), basename(p), special_glyph(ARROW), p); + k = hashmap_put(h, basename(p), p); + if (k < 0) { + free(p); + if (k != -EEXIST) + return k; + } + } + return 0; +} + +static int enumerate_dir(Hashmap *top, Hashmap *bottom, Hashmap *drops, const char *path, bool dropins) { + _cleanup_closedir_ DIR *d; + + assert(top); + assert(bottom); + assert(drops); + assert(path); + + log_debug("Looking at %s", path); + + d = opendir(path); + if (!d) { + if (errno == ENOENT) + return 0; + + return log_error_errno(errno, "Failed to open %s: %m", path); + } + + for (;;) { + struct dirent *de; + int k; + char *p; + + errno = 0; + de = readdir(d); + if (!de) + return -errno; + + dirent_ensure_type(d, de); + + if (dropins && de->d_type == DT_DIR && endswith(de->d_name, ".d")) + enumerate_dir_d(top, bottom, drops, path, de->d_name); + + if (!dirent_is_file(de)) + continue; + + p = strjoin(path, "/", de->d_name, NULL); + if (!p) + return -ENOMEM; + + log_debug("Adding at top: %s %s %s", basename(p), special_glyph(ARROW), p); + k = hashmap_put(top, basename(p), p); + if (k >= 0) { + p = strdup(p); + if (!p) + return -ENOMEM; + } else if (k != -EEXIST) { + free(p); + return k; + } + + log_debug("Adding at bottom: %s %s %s", basename(p), special_glyph(ARROW), p); + free(hashmap_remove(bottom, basename(p))); + k = hashmap_put(bottom, basename(p), p); + if (k < 0) { + free(p); + return k; + } + } +} + +static int process_suffix(const char *suffix, const char *onlyprefix) { + const char *p; + char *f; + Hashmap *top, *bottom, *drops; + Hashmap *h; + char *key; + int r = 0, k; + Iterator i, j; + int n_found = 0; + bool dropins; + + assert(suffix); + assert(!startswith(suffix, "/")); + assert(!strstr(suffix, "//")); + + dropins = nulstr_contains(have_dropins, suffix); + + top = hashmap_new(&string_hash_ops); + bottom = hashmap_new(&string_hash_ops); + drops = hashmap_new(&string_hash_ops); + if (!top || !bottom || !drops) { + r = -ENOMEM; + goto finish; + } + + NULSTR_FOREACH(p, prefixes) { + _cleanup_free_ char *t = NULL; + + t = strjoin(p, "/", suffix, NULL); + if (!t) { + r = -ENOMEM; + goto finish; + } + + k = enumerate_dir(top, bottom, drops, t, dropins); + if (r == 0) + r = k; + } + + HASHMAP_FOREACH_KEY(f, key, top, i) { + char *o; + + o = hashmap_get(bottom, key); + assert(o); + + if (!onlyprefix || startswith(o, onlyprefix)) { + if (path_equal(o, f)) { + notify_override_unchanged(f); + } else { + k = found_override(f, o); + if (k < 0) + r = k; + else + n_found += k; + } + } + + h = hashmap_get(drops, key); + if (h) + HASHMAP_FOREACH(o, h, j) + if (!onlyprefix || startswith(o, onlyprefix)) + n_found += notify_override_extended(f, o); + } + +finish: + hashmap_free_free(top); + hashmap_free_free(bottom); + + HASHMAP_FOREACH_KEY(h, key, drops, i) { + hashmap_free_free(hashmap_remove(drops, key)); + hashmap_remove(drops, key); + free(key); + } + hashmap_free(drops); + + return r < 0 ? r : n_found; +} + +static int process_suffixes(const char *onlyprefix) { + const char *n; + int n_found = 0, r; + + NULSTR_FOREACH(n, suffixes) { + r = process_suffix(n, onlyprefix); + if (r < 0) + return r; + + n_found += r; + } + + return n_found; +} + +static int process_suffix_chop(const char *arg) { + const char *p; + + assert(arg); + + if (!path_is_absolute(arg)) + return process_suffix(arg, NULL); + + /* Strip prefix from the suffix */ + NULSTR_FOREACH(p, prefixes) { + const char *suffix; + + suffix = startswith(arg, p); + if (suffix) { + suffix += strspn(suffix, "/"); + if (*suffix) + return process_suffix(suffix, NULL); + else + return process_suffixes(arg); + } + } + + log_error("Invalid suffix specification %s.", arg); + return -EINVAL; +} + +static void help(void) { + printf("%s [OPTIONS...] [SUFFIX...]\n\n" + "Find overridden configuration files.\n\n" + " -h --help Show this help\n" + " --version Show package version\n" + " --no-pager Do not pipe output into a pager\n" + " --diff[=1|0] Show a diff when overridden files differ\n" + " -t --type=LIST... Only display a selected set of override types\n" + , program_invocation_short_name); +} + +static int parse_flags(const char *flag_str, int flags) { + const char *word, *state; + size_t l; + + FOREACH_WORD_SEPARATOR(word, l, flag_str, ",", state) { + if (strneq("masked", word, l)) + flags |= SHOW_MASKED; + else if (strneq ("equivalent", word, l)) + flags |= SHOW_EQUIVALENT; + else if (strneq("redirected", word, l)) + flags |= SHOW_REDIRECTED; + else if (strneq("overridden", word, l)) + flags |= SHOW_OVERRIDDEN; + else if (strneq("unchanged", word, l)) + flags |= SHOW_UNCHANGED; + else if (strneq("extended", word, l)) + flags |= SHOW_EXTENDED; + else if (strneq("default", word, l)) + flags |= SHOW_DEFAULTS; + else + return -EINVAL; + } + return flags; +} + +static int parse_argv(int argc, char *argv[]) { + + enum { + ARG_NO_PAGER = 0x100, + ARG_DIFF, + ARG_VERSION + }; + + static const struct option options[] = { + { "help", no_argument, NULL, 'h' }, + { "version", no_argument, NULL, ARG_VERSION }, + { "no-pager", no_argument, NULL, ARG_NO_PAGER }, + { "diff", optional_argument, NULL, ARG_DIFF }, + { "type", required_argument, NULL, 't' }, + {} + }; + + int c; + + assert(argc >= 1); + assert(argv); + + while ((c = getopt_long(argc, argv, "ht:", options, NULL)) >= 0) + + switch (c) { + + case 'h': + help(); + return 0; + + case ARG_VERSION: + return version(); + + case ARG_NO_PAGER: + arg_no_pager = true; + break; + + case 't': { + int f; + f = parse_flags(optarg, arg_flags); + if (f < 0) { + log_error("Failed to parse flags field."); + return -EINVAL; + } + arg_flags = f; + break; + } + + case ARG_DIFF: + if (!optarg) + arg_diff = 1; + else { + int b; + + b = parse_boolean(optarg); + if (b < 0) { + log_error("Failed to parse diff boolean."); + return -EINVAL; + } + + arg_diff = b; + } + break; + + case '?': + return -EINVAL; + + default: + assert_not_reached("Unhandled option"); + } + + return 1; +} + +int main(int argc, char *argv[]) { + int r, k, n_found = 0; + + log_parse_environment(); + log_open(); + + r = parse_argv(argc, argv); + if (r <= 0) + goto finish; + + if (arg_flags == 0) + arg_flags = SHOW_DEFAULTS; + + if (arg_diff < 0) + arg_diff = !!(arg_flags & SHOW_OVERRIDDEN); + else if (arg_diff) + arg_flags |= SHOW_OVERRIDDEN; + + pager_open(arg_no_pager, false); + + if (optind < argc) { + int i; + + for (i = optind; i < argc; i++) { + path_kill_slashes(argv[i]); + + k = process_suffix_chop(argv[i]); + if (k < 0) + r = k; + else + n_found += k; + } + + } else { + k = process_suffixes(NULL); + if (k < 0) + r = k; + else + n_found += k; + } + + if (r >= 0) + printf("%s%i overridden configuration files found.\n", n_found ? "\n" : "", n_found); + +finish: + pager_close(); + + return r < 0 ? EXIT_FAILURE : EXIT_SUCCESS; +} diff --git a/src/grp-system/grp-utils/systemd-delta/systemd-delta.completion.bash b/src/grp-system/grp-utils/systemd-delta/systemd-delta.completion.bash new file mode 100644 index 0000000000..cb1732895f --- /dev/null +++ b/src/grp-system/grp-utils/systemd-delta/systemd-delta.completion.bash @@ -0,0 +1,61 @@ +# systemd-delta(1) completion -*- shell-script -*- +# +# This file is part of systemd. +# +# Copyright 2014 Thomas H.P. Andersen +# +# systemd is free software; you can redistribute it and/or modify it +# under the terms of the GNU Lesser General Public License as published by +# the Free Software Foundation; either version 2.1 of the License, or +# (at your option) any later version. +# +# systemd is distributed in the hope that it will be useful, but +# WITHOUT ANY WARRANTY; without even the implied warranty of +# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU +# General Public License for more details. +# +# You should have received a copy of the GNU Lesser General Public License +# along with systemd; If not, see . + +__contains_word() { + local w word=$1; shift + for w in "$@"; do + [[ $w = "$word" ]] && return + done +} + +_systemd-delta() { + local cur=${COMP_WORDS[COMP_CWORD]} prev=${COMP_WORDS[COMP_CWORD-1]} + local comps + + local -A OPTS=( + [STANDALONE]='--help -h --no-pager --version' + [ARG]='--diff --type -t' + ) + + _init_completion || return + + + if __contains_word "$prev" ${OPTS[ARG]}; then + case $prev in + --diff) + comps='yes no' + ;; + --type|-t) + comps='masked equivalent redirected overridden unchanged extended default' + ;; + esac + COMPREPLY=( $(compgen -W '$comps' -- "$cur") ) + return 0 + fi + + if [[ "$cur" = -* ]]; then + COMPREPLY=( $(compgen -W '${OPTS[*]}' -- "$cur") ) + return 0 + fi + + COMPREPLY=( $(compgen -W '$comps' -- "$cur") ) + return 0 +} + +complete -F _systemd-delta systemd-delta diff --git a/src/grp-system/grp-utils/systemd-delta/systemd-delta.completion.zsh b/src/grp-system/grp-utils/systemd-delta/systemd-delta.completion.zsh new file mode 100644 index 0000000000..757f1b66fb --- /dev/null +++ b/src/grp-system/grp-utils/systemd-delta/systemd-delta.completion.zsh @@ -0,0 +1,15 @@ +#compdef systemd-delta + +_delta_type() { + local -a _delta_types + _delta_types=(masked equivalent redirected overridden unchanged) + _values -s , "${_delta_types[@]}" +} + +_arguments \ + {-h,--help}'[Show this help]' \ + '--version[Show package version]' \ + '--no-pager[Do not pipe output into a pager]' \ + '--diff=[Show a diff when overridden files differ]:boolean:(1 0)' \ + {-t+,--type=}'[Only display a selected set of override types]:types:_delta_type' \ + ':SUFFIX:(tmpfiles.d sysctl.d systemd/system)' diff --git a/src/grp-system/grp-utils/systemd-delta/systemd-delta.xml b/src/grp-system/grp-utils/systemd-delta/systemd-delta.xml new file mode 100644 index 0000000000..99709604aa --- /dev/null +++ b/src/grp-system/grp-utils/systemd-delta/systemd-delta.xml @@ -0,0 +1,205 @@ + + + + + + + + + systemd-delta + systemd + + + + Developer + Lennart + Poettering + lennart@poettering.net + + + + + + systemd-delta + 1 + + + + systemd-delta + Find overridden configuration files + + + + + systemd-delta + OPTIONS + PREFIX/SUFFIX|SUFFIX + + + + + Description + + systemd-delta may be used to identify and + compare configuration files that override other configuration + files. Files in /etc have highest priority, + files in /run have the second highest + priority, ..., files in /lib have lowest + priority. Files in a directory with higher priority override files + with the same name in directories of lower priority. In addition, + certain configuration files can have .d + directories which contain "drop-in" files with configuration + snippets which augment the main configuration file. "Drop-in" + files can be overridden in the same way by placing files with the + same name in a directory of higher priority (except that, in case + of "drop-in" files, both the "drop-in" file name and the name of + the containing directory, which corresponds to the name of the + main configuration file, must match). For a fuller explanation, + see + systemd.unit5. + + + The command line argument will be split into a prefix and a + suffix. Either is optional. The prefix must be one of the + directories containing configuration files + (/etc, /run, + /usr/lib, ...). If it is given, only + overriding files contained in this directory will be shown. + Otherwise, all overriding files will be shown. The suffix must be + a name of a subdirectory containing configuration files like + tmpfiles.d, sysctl.d or + systemd/system. If it is given, only + configuration files in this subdirectory (across all configuration + paths) will be analyzed. Otherwise, all configuration files will + be analyzed. If the command line argument is not given at all, all + configuration files will be analyzed. See below for some + examples. + + + + Options + + The following options are understood: + + + + + + + When listing the differences, only list those + that are asked for. The list itself is a comma-separated list + of desired difference types. + + Recognized types are: + + + + masked + + Show masked files + + + + equivalent + + Show overridden files that while + overridden, do not differ in content. + + + + redirected + + Show files that are redirected to + another. + + + + overridden + + Show overridden, and changed + files. + + + + extended + + Show *.conf files + in drop-in directories for units. + + + + unchanged + + Show unmodified files + too. + + + + + + + + + When showing modified files, when a file is + overridden show a diff as well. This option takes a boolean + argument. If omitted, it defaults to + . + + + + + + + + + + Examples + + To see all local configuration: + systemd-delta + + To see all runtime configuration: + systemd-delta /run + + To see all system unit configuration changes: + systemd-delta systemd/system + + To see all runtime "drop-in" changes for system units: + systemd-delta --type=extended /run/systemd/system + + + + Exit status + + On success, 0 is returned, a non-zero failure code + otherwise. + + + + See Also + + systemd1, + systemd.unit5 + + + + diff --git a/src/grp-system/grp-utils/systemd-fstab-generator/Makefile b/src/grp-system/grp-utils/systemd-fstab-generator/Makefile new file mode 100644 index 0000000000..43475f69e1 --- /dev/null +++ b/src/grp-system/grp-utils/systemd-fstab-generator/Makefile @@ -0,0 +1,34 @@ +# -*- Mode: makefile; indent-tabs-mode: t -*- +# +# This file is part of systemd. +# +# Copyright 2010-2012 Lennart Poettering +# Copyright 2010-2012 Kay Sievers +# Copyright 2013 Zbigniew Jędrzejewski-Szmek +# Copyright 2013 David Strauss +# Copyright 2016 Luke Shumaker +# +# systemd is free software; you can redistribute it and/or modify it +# under the terms of the GNU Lesser General Public License as published by +# the Free Software Foundation; either version 2.1 of the License, or +# (at your option) any later version. +# +# systemd is distributed in the hope that it will be useful, but +# WITHOUT ANY WARRANTY; without even the implied warranty of +# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU +# Lesser General Public License for more details. +# +# You should have received a copy of the GNU Lesser General Public License +# along with systemd; If not, see . +include $(dir $(lastword $(MAKEFILE_LIST)))/../../../../config.mk +include $(topsrcdir)/build-aux/Makefile.head.mk + +systemgenerator_PROGRAMS += systemd-fstab-generator +systemd_fstab_generator_SOURCES = \ + src/fstab-generator/fstab-generator.c \ + src/core/mount-setup.c + +systemd_fstab_generator_LDADD = \ + libsystemd-shared.la + +include $(topsrcdir)/build-aux/Makefile.tail.mk diff --git a/src/grp-system/grp-utils/systemd-fstab-generator/fstab-generator.c b/src/grp-system/grp-utils/systemd-fstab-generator/fstab-generator.c new file mode 100644 index 0000000000..f4b3a048e9 --- /dev/null +++ b/src/grp-system/grp-utils/systemd-fstab-generator/fstab-generator.c @@ -0,0 +1,723 @@ +/*** + This file is part of systemd. + + Copyright 2012 Lennart Poettering + + systemd is free software; you can redistribute it and/or modify it + under the terms of the GNU Lesser General Public License as published by + the Free Software Foundation; either version 2.1 of the License, or + (at your option) any later version. + + systemd is distributed in the hope that it will be useful, but + WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + Lesser General Public License for more details. + + You should have received a copy of the GNU Lesser General Public License + along with systemd; If not, see . +***/ + +#include +#include +#include +#include +#include + +#include "core/mount-setup.h" +#include "systemd-basic/alloc-util.h" +#include "systemd-basic/fd-util.h" +#include "systemd-basic/fileio.h" +#include "systemd-basic/log.h" +#include "systemd-basic/mkdir.h" +#include "systemd-basic/mount-util.h" +#include "systemd-basic/parse-util.h" +#include "systemd-basic/path-util.h" +#include "systemd-basic/proc-cmdline.h" +#include "systemd-basic/special.h" +#include "systemd-basic/stat-util.h" +#include "systemd-basic/string-util.h" +#include "systemd-basic/strv.h" +#include "systemd-basic/unit-name.h" +#include "systemd-basic/util.h" +#include "systemd-basic/virt.h" +#include "systemd-shared/fstab-util.h" +#include "systemd-shared/generator.h" + +static const char *arg_dest = "/tmp"; +static bool arg_fstab_enabled = true; +static char *arg_root_what = NULL; +static char *arg_root_fstype = NULL; +static char *arg_root_options = NULL; +static int arg_root_rw = -1; +static char *arg_usr_what = NULL; +static char *arg_usr_fstype = NULL; +static char *arg_usr_options = NULL; + +static int add_swap( + const char *what, + struct mntent *me, + bool noauto, + bool nofail) { + + _cleanup_free_ char *name = NULL, *unit = NULL, *lnk = NULL; + _cleanup_fclose_ FILE *f = NULL; + int r; + + assert(what); + assert(me); + + if (access("/proc/swaps", F_OK) < 0) { + log_info("Swap not supported, ignoring fstab swap entry for %s.", what); + return 0; + } + + if (detect_container() > 0) { + log_info("Running in a container, ignoring fstab swap entry for %s.", what); + return 0; + } + + r = unit_name_from_path(what, ".swap", &name); + if (r < 0) + return log_error_errno(r, "Failed to generate unit name: %m"); + + unit = strjoin(arg_dest, "/", name, NULL); + if (!unit) + return log_oom(); + + f = fopen(unit, "wxe"); + if (!f) + return log_error_errno(errno, + errno == EEXIST ? + "Failed to create swap unit file %s, as it already exists. Duplicate entry in /etc/fstab?" : + "Failed to create unit file %s: %m", + unit); + + fprintf(f, + "# Automatically generated by systemd-fstab-generator\n\n" + "[Unit]\n" + "SourcePath=/etc/fstab\n" + "Documentation=man:fstab(5) man:systemd-fstab-generator(8)\n\n" + "[Swap]\n" + "What=%s\n", + what); + + if (!isempty(me->mnt_opts) && !streq(me->mnt_opts, "defaults")) + fprintf(f, "Options=%s\n", me->mnt_opts); + + r = fflush_and_check(f); + if (r < 0) + return log_error_errno(r, "Failed to write unit file %s: %m", unit); + + /* use what as where, to have a nicer error message */ + r = generator_write_timeouts(arg_dest, what, what, me->mnt_opts, NULL); + if (r < 0) + return r; + + if (!noauto) { + lnk = strjoin(arg_dest, "/" SPECIAL_SWAP_TARGET, + nofail ? ".wants/" : ".requires/", name, NULL); + if (!lnk) + return log_oom(); + + mkdir_parents_label(lnk, 0755); + if (symlink(unit, lnk) < 0) + return log_error_errno(errno, "Failed to create symlink %s: %m", lnk); + } + + return 0; +} + +static bool mount_is_network(struct mntent *me) { + assert(me); + + return fstab_test_option(me->mnt_opts, "_netdev\0") || + fstype_is_network(me->mnt_type); +} + +static bool mount_in_initrd(struct mntent *me) { + assert(me); + + return fstab_test_option(me->mnt_opts, "x-initrd.mount\0") || + streq(me->mnt_dir, "/usr"); +} + +static int write_idle_timeout(FILE *f, const char *where, const char *opts) { + _cleanup_free_ char *timeout = NULL; + char timespan[FORMAT_TIMESPAN_MAX]; + usec_t u; + int r; + + r = fstab_filter_options(opts, "x-systemd.idle-timeout\0", NULL, &timeout, NULL); + if (r < 0) + return log_warning_errno(r, "Failed to parse options: %m"); + if (r == 0) + return 0; + + r = parse_sec(timeout, &u); + if (r < 0) { + log_warning("Failed to parse timeout for %s, ignoring: %s", where, timeout); + return 0; + } + + fprintf(f, "TimeoutIdleSec=%s\n", format_timespan(timespan, sizeof(timespan), u, 0)); + + return 0; +} + +static int write_requires_after(FILE *f, const char *opts) { + _cleanup_strv_free_ char **names = NULL, **units = NULL; + _cleanup_free_ char *res = NULL; + char **s; + int r; + + assert(f); + assert(opts); + + r = fstab_extract_values(opts, "x-systemd.requires", &names); + if (r < 0) + return log_warning_errno(r, "Failed to parse options: %m"); + if (r == 0) + return 0; + + STRV_FOREACH(s, names) { + char *x; + + r = unit_name_mangle_with_suffix(*s, UNIT_NAME_NOGLOB, ".mount", &x); + if (r < 0) + return log_error_errno(r, "Failed to generate unit name: %m"); + r = strv_consume(&units, x); + if (r < 0) + return log_oom(); + } + + if (units) { + res = strv_join(units, " "); + if (!res) + return log_oom(); + fprintf(f, "After=%1$s\nRequires=%1$s\n", res); + } + + return 0; +} + +static int write_requires_mounts_for(FILE *f, const char *opts) { + _cleanup_strv_free_ char **paths = NULL; + _cleanup_free_ char *res = NULL; + int r; + + assert(f); + assert(opts); + + r = fstab_extract_values(opts, "x-systemd.requires-mounts-for", &paths); + if (r < 0) + return log_warning_errno(r, "Failed to parse options: %m"); + if (r == 0) + return 0; + + res = strv_join(paths, " "); + if (!res) + return log_oom(); + + fprintf(f, "RequiresMountsFor=%s\n", res); + + return 0; +} + +static int add_mount( + const char *what, + const char *where, + const char *fstype, + const char *opts, + int passno, + bool noauto, + bool nofail, + bool automount, + const char *post, + const char *source) { + + _cleanup_free_ char + *name = NULL, *unit = NULL, *lnk = NULL, + *automount_name = NULL, *automount_unit = NULL, + *filtered = NULL; + _cleanup_fclose_ FILE *f = NULL; + int r; + + assert(what); + assert(where); + assert(opts); + assert(post); + assert(source); + + if (streq_ptr(fstype, "autofs")) + return 0; + + if (!is_path(where)) { + log_warning("Mount point %s is not a valid path, ignoring.", where); + return 0; + } + + if (mount_point_is_api(where) || + mount_point_ignore(where)) + return 0; + + if (path_equal(where, "/")) { + if (noauto) + log_warning("Ignoring \"noauto\" for root device"); + if (nofail) + log_warning("Ignoring \"nofail\" for root device"); + if (automount) + log_warning("Ignoring automount option for root device"); + + noauto = nofail = automount = false; + } + + r = unit_name_from_path(where, ".mount", &name); + if (r < 0) + return log_error_errno(r, "Failed to generate unit name: %m"); + + unit = strjoin(arg_dest, "/", name, NULL); + if (!unit) + return log_oom(); + + f = fopen(unit, "wxe"); + if (!f) + return log_error_errno(errno, + errno == EEXIST ? + "Failed to create mount unit file %s, as it already exists. Duplicate entry in /etc/fstab?" : + "Failed to create unit file %s: %m", + unit); + + fprintf(f, + "# Automatically generated by systemd-fstab-generator\n\n" + "[Unit]\n" + "SourcePath=%s\n" + "Documentation=man:fstab(5) man:systemd-fstab-generator(8)\n", + source); + + if (!noauto && !nofail && !automount) + fprintf(f, "Before=%s\n", post); + + if (!automount && opts) { + r = write_requires_after(f, opts); + if (r < 0) + return r; + r = write_requires_mounts_for(f, opts); + if (r < 0) + return r; + } + + if (passno != 0) { + r = generator_write_fsck_deps(f, arg_dest, what, where, fstype); + if (r < 0) + return r; + } + + fprintf(f, + "\n" + "[Mount]\n" + "What=%s\n" + "Where=%s\n", + what, + where); + + if (!isempty(fstype) && !streq(fstype, "auto")) + fprintf(f, "Type=%s\n", fstype); + + r = generator_write_timeouts(arg_dest, what, where, opts, &filtered); + if (r < 0) + return r; + + if (!isempty(filtered) && !streq(filtered, "defaults")) + fprintf(f, "Options=%s\n", filtered); + + r = fflush_and_check(f); + if (r < 0) + return log_error_errno(r, "Failed to write unit file %s: %m", unit); + + if (!noauto && !automount) { + lnk = strjoin(arg_dest, "/", post, nofail ? ".wants/" : ".requires/", name, NULL); + if (!lnk) + return log_oom(); + + mkdir_parents_label(lnk, 0755); + if (symlink(unit, lnk) < 0) + return log_error_errno(errno, "Failed to create symlink %s: %m", lnk); + } + + if (automount) { + r = unit_name_from_path(where, ".automount", &automount_name); + if (r < 0) + return log_error_errno(r, "Failed to generate unit name: %m"); + + automount_unit = strjoin(arg_dest, "/", automount_name, NULL); + if (!automount_unit) + return log_oom(); + + fclose(f); + f = fopen(automount_unit, "wxe"); + if (!f) + return log_error_errno(errno, "Failed to create unit file %s: %m", automount_unit); + + fprintf(f, + "# Automatically generated by systemd-fstab-generator\n\n" + "[Unit]\n" + "SourcePath=%s\n" + "Documentation=man:fstab(5) man:systemd-fstab-generator(8)\n", + source); + + fprintf(f, "Before=%s\n", post); + + if (opts) { + r = write_requires_after(f, opts); + if (r < 0) + return r; + r = write_requires_mounts_for(f, opts); + if (r < 0) + return r; + } + + fprintf(f, + "\n" + "[Automount]\n" + "Where=%s\n", + where); + + r = write_idle_timeout(f, where, opts); + if (r < 0) + return r; + + r = fflush_and_check(f); + if (r < 0) + return log_error_errno(r, "Failed to write unit file %s: %m", automount_unit); + + free(lnk); + lnk = strjoin(arg_dest, "/", post, nofail ? ".wants/" : ".requires/", automount_name, NULL); + if (!lnk) + return log_oom(); + + mkdir_parents_label(lnk, 0755); + if (symlink(automount_unit, lnk) < 0) + return log_error_errno(errno, "Failed to create symlink %s: %m", lnk); + } + + return 0; +} + +static int parse_fstab(bool initrd) { + _cleanup_endmntent_ FILE *f = NULL; + const char *fstab_path; + struct mntent *me; + int r = 0; + + fstab_path = initrd ? "/sysroot/etc/fstab" : "/etc/fstab"; + f = setmntent(fstab_path, "re"); + if (!f) { + if (errno == ENOENT) + return 0; + + return log_error_errno(errno, "Failed to open %s: %m", fstab_path); + } + + while ((me = getmntent(f))) { + _cleanup_free_ char *where = NULL, *what = NULL; + bool noauto, nofail; + int k; + + if (initrd && !mount_in_initrd(me)) + continue; + + what = fstab_node_to_udev_node(me->mnt_fsname); + if (!what) + return log_oom(); + + if (is_device_path(what) && path_is_read_only_fs("sys") > 0) { + log_info("Running in a container, ignoring fstab device entry for %s.", what); + continue; + } + + where = initrd ? strappend("/sysroot/", me->mnt_dir) : strdup(me->mnt_dir); + if (!where) + return log_oom(); + + if (is_path(where)) + path_kill_slashes(where); + + noauto = fstab_test_yes_no_option(me->mnt_opts, "noauto\0" "auto\0"); + nofail = fstab_test_yes_no_option(me->mnt_opts, "nofail\0" "fail\0"); + log_debug("Found entry what=%s where=%s type=%s nofail=%s noauto=%s", + what, where, me->mnt_type, + yes_no(noauto), yes_no(nofail)); + + if (streq(me->mnt_type, "swap")) + k = add_swap(what, me, noauto, nofail); + else { + bool automount; + const char *post; + + automount = fstab_test_option(me->mnt_opts, + "comment=systemd.automount\0" + "x-systemd.automount\0"); + if (initrd) + post = SPECIAL_INITRD_FS_TARGET; + else if (mount_is_network(me)) + post = SPECIAL_REMOTE_FS_TARGET; + else + post = SPECIAL_LOCAL_FS_TARGET; + + k = add_mount(what, + where, + me->mnt_type, + me->mnt_opts, + me->mnt_passno, + noauto, + nofail, + automount, + post, + fstab_path); + } + + if (k < 0) + r = k; + } + + return r; +} + +static int add_sysroot_mount(void) { + _cleanup_free_ char *what = NULL; + const char *opts; + int r; + + if (isempty(arg_root_what)) { + log_debug("Could not find a root= entry on the kernel command line."); + return 0; + } + + if (streq(arg_root_what, "gpt-auto")) { + /* This is handled by the gpt-auto generator */ + log_debug("Skipping root directory handling, as gpt-auto was requested."); + return 0; + } + + if (path_equal(arg_root_what, "/dev/nfs")) { + /* This is handled by the kernel or the initrd */ + log_debug("Skipping root directory handling, as /dev/nfs was requested."); + return 0; + } + + what = fstab_node_to_udev_node(arg_root_what); + if (!what) + return log_oom(); + + if (!arg_root_options) + opts = arg_root_rw > 0 ? "rw" : "ro"; + else if (arg_root_rw >= 0 || + !fstab_test_option(arg_root_options, "ro\0" "rw\0")) + opts = strjoina(arg_root_options, ",", arg_root_rw > 0 ? "rw" : "ro"); + else + opts = arg_root_options; + + log_debug("Found entry what=%s where=/sysroot type=%s", what, strna(arg_root_fstype)); + + if (is_device_path(what)) { + r = generator_write_initrd_root_device_deps(arg_dest, what); + if (r < 0) + return r; + } + + return add_mount(what, + "/sysroot", + arg_root_fstype, + opts, + is_device_path(what) ? 1 : 0, /* passno */ + false, /* noauto off */ + false, /* nofail off */ + false, /* automount off */ + SPECIAL_INITRD_ROOT_FS_TARGET, + "/proc/cmdline"); +} + +static int add_sysroot_usr_mount(void) { + _cleanup_free_ char *what = NULL; + const char *opts; + + if (!arg_usr_what && !arg_usr_fstype && !arg_usr_options) + return 0; + + if (arg_root_what && !arg_usr_what) { + /* Copy over the root device, in case the /usr mount just differs in a mount option (consider btrfs subvolumes) */ + arg_usr_what = strdup(arg_root_what); + if (!arg_usr_what) + return log_oom(); + } + + if (arg_root_fstype && !arg_usr_fstype) { + arg_usr_fstype = strdup(arg_root_fstype); + if (!arg_usr_fstype) + return log_oom(); + } + + if (arg_root_options && !arg_usr_options) { + arg_usr_options = strdup(arg_root_options); + if (!arg_usr_options) + return log_oom(); + } + + if (!arg_usr_what) + return 0; + + what = fstab_node_to_udev_node(arg_usr_what); + if (!what) + return log_oom(); + + if (!arg_usr_options) + opts = arg_root_rw > 0 ? "rw" : "ro"; + else if (!fstab_test_option(arg_usr_options, "ro\0" "rw\0")) + opts = strjoina(arg_usr_options, ",", arg_root_rw > 0 ? "rw" : "ro"); + else + opts = arg_usr_options; + + log_debug("Found entry what=%s where=/sysroot/usr type=%s", what, strna(arg_usr_fstype)); + return add_mount(what, + "/sysroot/usr", + arg_usr_fstype, + opts, + is_device_path(what) ? 1 : 0, /* passno */ + false, /* noauto off */ + false, /* nofail off */ + false, /* automount off */ + SPECIAL_INITRD_FS_TARGET, + "/proc/cmdline"); +} + +static int parse_proc_cmdline_item(const char *key, const char *value) { + int r; + + /* root=, usr=, usrfstype= and roofstype= may occur more than once, the last + * instance should take precedence. In the case of multiple rootflags= + * or usrflags= the arguments should be concatenated */ + + if (STR_IN_SET(key, "fstab", "rd.fstab") && value) { + + r = parse_boolean(value); + if (r < 0) + log_warning("Failed to parse fstab switch %s. Ignoring.", value); + else + arg_fstab_enabled = r; + + } else if (streq(key, "root") && value) { + + if (free_and_strdup(&arg_root_what, value) < 0) + return log_oom(); + + } else if (streq(key, "rootfstype") && value) { + + if (free_and_strdup(&arg_root_fstype, value) < 0) + return log_oom(); + + } else if (streq(key, "rootflags") && value) { + char *o; + + o = arg_root_options ? + strjoin(arg_root_options, ",", value, NULL) : + strdup(value); + if (!o) + return log_oom(); + + free(arg_root_options); + arg_root_options = o; + + } else if (streq(key, "mount.usr") && value) { + + if (free_and_strdup(&arg_usr_what, value) < 0) + return log_oom(); + + } else if (streq(key, "mount.usrfstype") && value) { + + if (free_and_strdup(&arg_usr_fstype, value) < 0) + return log_oom(); + + } else if (streq(key, "mount.usrflags") && value) { + char *o; + + o = arg_usr_options ? + strjoin(arg_usr_options, ",", value, NULL) : + strdup(value); + if (!o) + return log_oom(); + + free(arg_usr_options); + arg_usr_options = o; + + } else if (streq(key, "rw") && !value) + arg_root_rw = true; + else if (streq(key, "ro") && !value) + arg_root_rw = false; + + return 0; +} + +int main(int argc, char *argv[]) { + int r = 0; + + if (argc > 1 && argc != 4) { + log_error("This program takes three or no arguments."); + return EXIT_FAILURE; + } + + if (argc > 1) + arg_dest = argv[1]; + + log_set_target(LOG_TARGET_SAFE); + log_parse_environment(); + log_open(); + + umask(0022); + + r = parse_proc_cmdline(parse_proc_cmdline_item); + if (r < 0) + log_warning_errno(r, "Failed to parse kernel command line, ignoring: %m"); + + /* Always honour root= and usr= in the kernel command line if we are in an initrd */ + if (in_initrd()) { + int k; + + r = add_sysroot_mount(); + + k = add_sysroot_usr_mount(); + if (k < 0) + r = k; + } else + r = 0; + + /* Honour /etc/fstab only when that's enabled */ + if (arg_fstab_enabled) { + int k; + + log_debug("Parsing /etc/fstab"); + + /* Parse the local /etc/fstab, possibly from the initrd */ + k = parse_fstab(false); + if (k < 0) + r = k; + + /* If running in the initrd also parse the /etc/fstab from the host */ + if (in_initrd()) { + log_debug("Parsing /sysroot/etc/fstab"); + + k = parse_fstab(true); + if (k < 0) + r = k; + } + } + + free(arg_root_what); + free(arg_root_fstype); + free(arg_root_options); + + free(arg_usr_what); + free(arg_usr_fstype); + free(arg_usr_options); + + return r < 0 ? EXIT_FAILURE : EXIT_SUCCESS; +} diff --git a/src/grp-system/grp-utils/systemd-fstab-generator/systemd-fstab-generator.xml b/src/grp-system/grp-utils/systemd-fstab-generator/systemd-fstab-generator.xml new file mode 100644 index 0000000000..a971cb3675 --- /dev/null +++ b/src/grp-system/grp-utils/systemd-fstab-generator/systemd-fstab-generator.xml @@ -0,0 +1,183 @@ + + + + + + + + systemd-fstab-generator + systemd + + + + Developer + Lennart + Poettering + lennart@poettering.net + + + + + + systemd-fstab-generator + 8 + + + + systemd-fstab-generator + Unit generator for /etc/fstab + + + + /usr/lib/systemd/system-generators/systemd-fstab-generator + + + + Description + + systemd-fstab-generator is a generator + that translates /etc/fstab (see + fstab5 + for details) into native systemd units early at boot and when + configuration of the system manager is reloaded. This will + instantiate mount and swap units as necessary. + + The passno field is treated like a simple + boolean, and the ordering information is discarded. However, if + the root file system is checked, it is checked before all the + other file systems. + + See + systemd.mount5 + and + systemd.swap5 + for more information about special /etc/fstab + mount options this generator understands. + + systemd-fstab-generator implements + systemd.generator7. + + + + Kernel Command Line + + systemd-fstab-generator understands the + following kernel command line parameters: + + + + + fstab= + rd.fstab= + + Takes a boolean argument. Defaults to + yes. If no, causes the + generator to ignore any mounts or swaps configured in + /etc/fstab. rd.fstab= + is honored only by initial RAM disk (initrd) while + fstab= is honored by both the main system + and the initrd. + + + root= + + Takes the root filesystem to mount in the + initrd. root= is honored by the + initrd. + + + rootfstype= + + Takes the root filesystem type that will be + passed to the mount command. rootfstype= is + honored by the initrd. + + + rootflags= + + Takes the root filesystem mount options to + use. rootflags= is honored by the + initrd. + + + mount.usr= + + Takes the /usr filesystem + to be mounted by the initrd. If + mount.usrfstype= or + mount.usrflags= is set, then + mount.usr= will default to the value set in + root=. + + Otherwise, this parameter defaults to the + /usr entry found in + /etc/fstab on the root filesystem. + + mount.usr= is honored by the initrd. + + + + mount.usrfstype= + + Takes the /usr filesystem + type that will be passed to the mount command. If + mount.usr= or + mount.usrflags= is set, then + mount.usrfstype= will default to the value + set in rootfstype=. + + Otherwise, this value will be read from the + /usr entry in + /etc/fstab on the root filesystem. + + mount.usrfstype= is honored by the + initrd. + + + mount.usrflags= + + Takes the /usr filesystem + mount options to use. If mount.usr= or + mount.usrfstype= is set, then + mount.usrflags= will default to the value + set in rootflags=. + + Otherwise, this value will be read from the + /usr entry in + /etc/fstab on the root filesystem. + + mount.usrflags= is honored by the + initrd. + + + + + + See Also + + systemd1, + fstab5, + systemd.mount5, + systemd.swap5, + systemd-cryptsetup-generator8 + + + + diff --git a/src/grp-system/grp-utils/systemd-run/Makefile b/src/grp-system/grp-utils/systemd-run/Makefile new file mode 100644 index 0000000000..9664591eb6 --- /dev/null +++ b/src/grp-system/grp-utils/systemd-run/Makefile @@ -0,0 +1,33 @@ +# -*- Mode: makefile; indent-tabs-mode: t -*- +# +# This file is part of systemd. +# +# Copyright 2010-2012 Lennart Poettering +# Copyright 2010-2012 Kay Sievers +# Copyright 2013 Zbigniew Jędrzejewski-Szmek +# Copyright 2013 David Strauss +# Copyright 2016 Luke Shumaker +# +# systemd is free software; you can redistribute it and/or modify it +# under the terms of the GNU Lesser General Public License as published by +# the Free Software Foundation; either version 2.1 of the License, or +# (at your option) any later version. +# +# systemd is distributed in the hope that it will be useful, but +# WITHOUT ANY WARRANTY; without even the implied warranty of +# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU +# Lesser General Public License for more details. +# +# You should have received a copy of the GNU Lesser General Public License +# along with systemd; If not, see . +include $(dir $(lastword $(MAKEFILE_LIST)))/../../../../config.mk +include $(topsrcdir)/build-aux/Makefile.head.mk + +bin_PROGRAMS += systemd-run +systemd_run_SOURCES = \ + src/run/run.c + +systemd_run_LDADD = \ + libsystemd-shared.la + +include $(topsrcdir)/build-aux/Makefile.tail.mk diff --git a/src/grp-system/grp-utils/systemd-run/run.c b/src/grp-system/grp-utils/systemd-run/run.c new file mode 100644 index 0000000000..a68938fb71 --- /dev/null +++ b/src/grp-system/grp-utils/systemd-run/run.c @@ -0,0 +1,1261 @@ +/*** + This file is part of systemd. + + Copyright 2013 Lennart Poettering + + systemd is free software; you can redistribute it and/or modify it + under the terms of the GNU Lesser General Public License as published by + the Free Software Foundation; either version 2.1 of the License, or + (at your option) any later version. + + systemd is distributed in the hope that it will be useful, but + WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + Lesser General Public License for more details. + + You should have received a copy of the GNU Lesser General Public License + along with systemd; If not, see . +***/ + +#include +#include + +#include +#include + +#include "sd-bus/bus-error.h" +#include "sd-bus/bus-util.h" +#include "systemd-basic/alloc-util.h" +#include "systemd-basic/calendarspec.h" +#include "systemd-basic/env-util.h" +#include "systemd-basic/fd-util.h" +#include "systemd-basic/formats-util.h" +#include "systemd-basic/parse-util.h" +#include "systemd-basic/path-util.h" +#include "systemd-basic/signal-util.h" +#include "systemd-basic/strv.h" +#include "systemd-basic/terminal-util.h" +#include "systemd-basic/unit-name.h" +#include "systemd-basic/user-util.h" +#include "systemd-shared/bus-unit-util.h" +#include "systemd-shared/ptyfwd.h" +#include "systemd-shared/spawn-polkit-agent.h" + +static bool arg_ask_password = true; +static bool arg_scope = false; +static bool arg_remain_after_exit = false; +static bool arg_no_block = false; +static const char *arg_unit = NULL; +static const char *arg_description = NULL; +static const char *arg_slice = NULL; +static bool arg_send_sighup = false; +static BusTransport arg_transport = BUS_TRANSPORT_LOCAL; +static const char *arg_host = NULL; +static bool arg_user = false; +static const char *arg_service_type = NULL; +static const char *arg_exec_user = NULL; +static const char *arg_exec_group = NULL; +static int arg_nice = 0; +static bool arg_nice_set = false; +static char **arg_environment = NULL; +static char **arg_property = NULL; +static bool arg_pty = false; +static usec_t arg_on_active = 0; +static usec_t arg_on_boot = 0; +static usec_t arg_on_startup = 0; +static usec_t arg_on_unit_active = 0; +static usec_t arg_on_unit_inactive = 0; +static const char *arg_on_calendar = NULL; +static char **arg_timer_property = NULL; +static bool arg_quiet = false; + +static void polkit_agent_open_if_enabled(void) { + + /* Open the polkit agent as a child process if necessary */ + if (!arg_ask_password) + return; + + if (arg_transport != BUS_TRANSPORT_LOCAL) + return; + + polkit_agent_open(); +} + +static void help(void) { + printf("%s [OPTIONS...] {COMMAND} [ARGS...]\n\n" + "Run the specified command in a transient scope or service or timer\n" + "unit. If a timer option is specified and the unit specified with\n" + "the --unit option exists, the command can be omitted.\n\n" + " -h --help Show this help\n" + " --version Show package version\n" + " --no-ask-password Do not prompt for password\n" + " --user Run as user unit\n" + " -H --host=[USER@]HOST Operate on remote host\n" + " -M --machine=CONTAINER Operate on local container\n" + " --scope Run this as scope rather than service\n" + " --unit=UNIT Run under the specified unit name\n" + " -p --property=NAME=VALUE Set unit property\n" + " --description=TEXT Description for unit\n" + " --slice=SLICE Run in the specified slice\n" + " --no-block Do not wait until operation finished\n" + " -r --remain-after-exit Leave service around until explicitly stopped\n" + " --send-sighup Send SIGHUP when terminating\n" + " --service-type=TYPE Service type\n" + " --uid=USER Run as system user\n" + " --gid=GROUP Run as system group\n" + " --nice=NICE Nice level\n" + " -E --setenv=NAME=VALUE Set environment\n" + " -t --pty Run service on pseudo tty\n" + " -q --quiet Suppress information messages during runtime\n\n" + "Timer options:\n\n" + " --on-active=SECONDS Run after SECONDS delay\n" + " --on-boot=SECONDS Run SECONDS after machine was booted up\n" + " --on-startup=SECONDS Run SECONDS after systemd activation\n" + " --on-unit-active=SECONDS Run SECONDS after the last activation\n" + " --on-unit-inactive=SECONDS Run SECONDS after the last deactivation\n" + " --on-calendar=SPEC Realtime timer\n" + " --timer-property=NAME=VALUE Set timer unit property\n", + program_invocation_short_name); +} + +static bool with_timer(void) { + return arg_on_active || arg_on_boot || arg_on_startup || arg_on_unit_active || arg_on_unit_inactive || arg_on_calendar; +} + +static int parse_argv(int argc, char *argv[]) { + + enum { + ARG_VERSION = 0x100, + ARG_USER, + ARG_SYSTEM, + ARG_SCOPE, + ARG_UNIT, + ARG_DESCRIPTION, + ARG_SLICE, + ARG_SEND_SIGHUP, + ARG_SERVICE_TYPE, + ARG_EXEC_USER, + ARG_EXEC_GROUP, + ARG_NICE, + ARG_ON_ACTIVE, + ARG_ON_BOOT, + ARG_ON_STARTUP, + ARG_ON_UNIT_ACTIVE, + ARG_ON_UNIT_INACTIVE, + ARG_ON_CALENDAR, + ARG_TIMER_PROPERTY, + ARG_NO_BLOCK, + ARG_NO_ASK_PASSWORD, + }; + + static const struct option options[] = { + { "help", no_argument, NULL, 'h' }, + { "version", no_argument, NULL, ARG_VERSION }, + { "user", no_argument, NULL, ARG_USER }, + { "system", no_argument, NULL, ARG_SYSTEM }, + { "scope", no_argument, NULL, ARG_SCOPE }, + { "unit", required_argument, NULL, ARG_UNIT }, + { "description", required_argument, NULL, ARG_DESCRIPTION }, + { "slice", required_argument, NULL, ARG_SLICE }, + { "remain-after-exit", no_argument, NULL, 'r' }, + { "send-sighup", no_argument, NULL, ARG_SEND_SIGHUP }, + { "host", required_argument, NULL, 'H' }, + { "machine", required_argument, NULL, 'M' }, + { "service-type", required_argument, NULL, ARG_SERVICE_TYPE }, + { "uid", required_argument, NULL, ARG_EXEC_USER }, + { "gid", required_argument, NULL, ARG_EXEC_GROUP }, + { "nice", required_argument, NULL, ARG_NICE }, + { "setenv", required_argument, NULL, 'E' }, + { "property", required_argument, NULL, 'p' }, + { "tty", no_argument, NULL, 't' }, /* deprecated */ + { "pty", no_argument, NULL, 't' }, + { "quiet", no_argument, NULL, 'q' }, + { "on-active", required_argument, NULL, ARG_ON_ACTIVE }, + { "on-boot", required_argument, NULL, ARG_ON_BOOT }, + { "on-startup", required_argument, NULL, ARG_ON_STARTUP }, + { "on-unit-active", required_argument, NULL, ARG_ON_UNIT_ACTIVE }, + { "on-unit-inactive", required_argument, NULL, ARG_ON_UNIT_INACTIVE }, + { "on-calendar", required_argument, NULL, ARG_ON_CALENDAR }, + { "timer-property", required_argument, NULL, ARG_TIMER_PROPERTY }, + { "no-block", no_argument, NULL, ARG_NO_BLOCK }, + { "no-ask-password", no_argument, NULL, ARG_NO_ASK_PASSWORD }, + {}, + }; + + int r, c; + + assert(argc >= 0); + assert(argv); + + while ((c = getopt_long(argc, argv, "+hrH:M:E:p:tq", options, NULL)) >= 0) + + switch (c) { + + case 'h': + help(); + return 0; + + case ARG_NO_ASK_PASSWORD: + arg_ask_password = false; + break; + + case ARG_VERSION: + return version(); + + case ARG_USER: + arg_user = true; + break; + + case ARG_SYSTEM: + arg_user = false; + break; + + case ARG_SCOPE: + arg_scope = true; + break; + + case ARG_UNIT: + arg_unit = optarg; + break; + + case ARG_DESCRIPTION: + arg_description = optarg; + break; + + case ARG_SLICE: + arg_slice = optarg; + break; + + case ARG_SEND_SIGHUP: + arg_send_sighup = true; + break; + + case 'r': + arg_remain_after_exit = true; + break; + + case 'H': + arg_transport = BUS_TRANSPORT_REMOTE; + arg_host = optarg; + break; + + case 'M': + arg_transport = BUS_TRANSPORT_MACHINE; + arg_host = optarg; + break; + + case ARG_SERVICE_TYPE: + arg_service_type = optarg; + break; + + case ARG_EXEC_USER: + arg_exec_user = optarg; + break; + + case ARG_EXEC_GROUP: + arg_exec_group = optarg; + break; + + case ARG_NICE: + r = safe_atoi(optarg, &arg_nice); + if (r < 0 || arg_nice < PRIO_MIN || arg_nice >= PRIO_MAX) { + log_error("Failed to parse nice value"); + return -EINVAL; + } + + arg_nice_set = true; + break; + + case 'E': + if (strv_extend(&arg_environment, optarg) < 0) + return log_oom(); + + break; + + case 'p': + if (strv_extend(&arg_property, optarg) < 0) + return log_oom(); + + break; + + case 't': + arg_pty = true; + break; + + case 'q': + arg_quiet = true; + break; + + case ARG_ON_ACTIVE: + + r = parse_sec(optarg, &arg_on_active); + if (r < 0) { + log_error("Failed to parse timer value: %s", optarg); + return r; + } + + break; + + case ARG_ON_BOOT: + + r = parse_sec(optarg, &arg_on_boot); + if (r < 0) { + log_error("Failed to parse timer value: %s", optarg); + return r; + } + + break; + + case ARG_ON_STARTUP: + + r = parse_sec(optarg, &arg_on_startup); + if (r < 0) { + log_error("Failed to parse timer value: %s", optarg); + return r; + } + + break; + + case ARG_ON_UNIT_ACTIVE: + + r = parse_sec(optarg, &arg_on_unit_active); + if (r < 0) { + log_error("Failed to parse timer value: %s", optarg); + return r; + } + + break; + + case ARG_ON_UNIT_INACTIVE: + + r = parse_sec(optarg, &arg_on_unit_inactive); + if (r < 0) { + log_error("Failed to parse timer value: %s", optarg); + return r; + } + + break; + + case ARG_ON_CALENDAR: { + CalendarSpec *spec = NULL; + + r = calendar_spec_from_string(optarg, &spec); + if (r < 0) { + log_error("Invalid calendar spec: %s", optarg); + return r; + } + + calendar_spec_free(spec); + arg_on_calendar = optarg; + break; + } + + case ARG_TIMER_PROPERTY: + + if (strv_extend(&arg_timer_property, optarg) < 0) + return log_oom(); + + break; + + case ARG_NO_BLOCK: + arg_no_block = true; + break; + + case '?': + return -EINVAL; + + default: + assert_not_reached("Unhandled option"); + } + + if ((optind >= argc) && (!arg_unit || !with_timer())) { + log_error("Command line to execute required."); + return -EINVAL; + } + + if (arg_user && arg_transport != BUS_TRANSPORT_LOCAL) { + log_error("Execution in user context is not supported on non-local systems."); + return -EINVAL; + } + + if (arg_scope && arg_transport != BUS_TRANSPORT_LOCAL) { + log_error("Scope execution is not supported on non-local systems."); + return -EINVAL; + } + + if (arg_scope && (arg_remain_after_exit || arg_service_type)) { + log_error("--remain-after-exit and --service-type= are not supported in --scope mode."); + return -EINVAL; + } + + if (arg_pty && (with_timer() || arg_scope)) { + log_error("--pty is not compatible in timer or --scope mode."); + return -EINVAL; + } + + if (arg_pty && arg_transport == BUS_TRANSPORT_REMOTE) { + log_error("--pty is only supported when connecting to the local system or containers."); + return -EINVAL; + } + + if (arg_scope && with_timer()) { + log_error("Timer options are not supported in --scope mode."); + return -EINVAL; + } + + if (arg_timer_property && !with_timer()) { + log_error("--timer-property= has no effect without any other timer options."); + return -EINVAL; + } + + return 1; +} + +static int transient_unit_set_properties(sd_bus_message *m, char **properties) { + char **i; + int r; + + r = sd_bus_message_append(m, "(sv)", "Description", "s", arg_description); + if (r < 0) + return r; + + STRV_FOREACH(i, properties) { + r = bus_append_unit_property_assignment(m, *i); + if (r < 0) + return r; + } + + return 0; +} + +static int transient_cgroup_set_properties(sd_bus_message *m) { + int r; + assert(m); + + if (!isempty(arg_slice)) { + _cleanup_free_ char *slice; + + r = unit_name_mangle_with_suffix(arg_slice, UNIT_NAME_NOGLOB, ".slice", &slice); + if (r < 0) + return r; + + r = sd_bus_message_append(m, "(sv)", "Slice", "s", slice); + if (r < 0) + return r; + } + + return 0; +} + +static int transient_kill_set_properties(sd_bus_message *m) { + assert(m); + + if (arg_send_sighup) + return sd_bus_message_append(m, "(sv)", "SendSIGHUP", "b", arg_send_sighup); + else + return 0; +} + +static int transient_service_set_properties(sd_bus_message *m, char **argv, const char *pty_path) { + int r; + + assert(m); + + r = transient_unit_set_properties(m, arg_property); + if (r < 0) + return r; + + r = transient_kill_set_properties(m); + if (r < 0) + return r; + + r = transient_cgroup_set_properties(m); + if (r < 0) + return r; + + if (arg_remain_after_exit) { + r = sd_bus_message_append(m, "(sv)", "RemainAfterExit", "b", arg_remain_after_exit); + if (r < 0) + return r; + } + + if (arg_service_type) { + r = sd_bus_message_append(m, "(sv)", "Type", "s", arg_service_type); + if (r < 0) + return r; + } + + if (arg_exec_user) { + r = sd_bus_message_append(m, "(sv)", "User", "s", arg_exec_user); + if (r < 0) + return r; + } + + if (arg_exec_group) { + r = sd_bus_message_append(m, "(sv)", "Group", "s", arg_exec_group); + if (r < 0) + return r; + } + + if (arg_nice_set) { + r = sd_bus_message_append(m, "(sv)", "Nice", "i", arg_nice); + if (r < 0) + return r; + } + + if (pty_path) { + const char *e; + + r = sd_bus_message_append(m, + "(sv)(sv)(sv)(sv)", + "StandardInput", "s", "tty", + "StandardOutput", "s", "tty", + "StandardError", "s", "tty", + "TTYPath", "s", pty_path); + if (r < 0) + return r; + + e = getenv("TERM"); + if (e) { + char *n; + + n = strjoina("TERM=", e); + r = sd_bus_message_append(m, + "(sv)", + "Environment", "as", 1, n); + if (r < 0) + return r; + } + } + + if (!strv_isempty(arg_environment)) { + r = sd_bus_message_open_container(m, 'r', "sv"); + if (r < 0) + return r; + + r = sd_bus_message_append(m, "s", "Environment"); + if (r < 0) + return r; + + r = sd_bus_message_open_container(m, 'v', "as"); + if (r < 0) + return r; + + r = sd_bus_message_append_strv(m, arg_environment); + if (r < 0) + return r; + + r = sd_bus_message_close_container(m); + if (r < 0) + return r; + + r = sd_bus_message_close_container(m); + if (r < 0) + return r; + } + + /* Exec container */ + { + r = sd_bus_message_open_container(m, 'r', "sv"); + if (r < 0) + return r; + + r = sd_bus_message_append(m, "s", "ExecStart"); + if (r < 0) + return r; + + r = sd_bus_message_open_container(m, 'v', "a(sasb)"); + if (r < 0) + return r; + + r = sd_bus_message_open_container(m, 'a', "(sasb)"); + if (r < 0) + return r; + + r = sd_bus_message_open_container(m, 'r', "sasb"); + if (r < 0) + return r; + + r = sd_bus_message_append(m, "s", argv[0]); + if (r < 0) + return r; + + r = sd_bus_message_append_strv(m, argv); + if (r < 0) + return r; + + r = sd_bus_message_append(m, "b", false); + if (r < 0) + return r; + + r = sd_bus_message_close_container(m); + if (r < 0) + return r; + + r = sd_bus_message_close_container(m); + if (r < 0) + return r; + + r = sd_bus_message_close_container(m); + if (r < 0) + return r; + + r = sd_bus_message_close_container(m); + if (r < 0) + return r; + } + + return 0; +} + +static int transient_scope_set_properties(sd_bus_message *m) { + int r; + + assert(m); + + r = transient_unit_set_properties(m, arg_property); + if (r < 0) + return r; + + r = transient_kill_set_properties(m); + if (r < 0) + return r; + + r = transient_cgroup_set_properties(m); + if (r < 0) + return r; + + r = sd_bus_message_append(m, "(sv)", "PIDs", "au", 1, (uint32_t) getpid()); + if (r < 0) + return r; + + return 0; +} + +static int transient_timer_set_properties(sd_bus_message *m) { + int r; + + assert(m); + + r = transient_unit_set_properties(m, arg_timer_property); + if (r < 0) + return r; + + /* Automatically clean up our transient timers */ + r = sd_bus_message_append(m, "(sv)", "RemainAfterElapse", "b", false); + if (r < 0) + return r; + + if (arg_on_active) { + r = sd_bus_message_append(m, "(sv)", "OnActiveSec", "t", arg_on_active); + if (r < 0) + return r; + } + + if (arg_on_boot) { + r = sd_bus_message_append(m, "(sv)", "OnBootSec", "t", arg_on_boot); + if (r < 0) + return r; + } + + if (arg_on_startup) { + r = sd_bus_message_append(m, "(sv)", "OnStartupSec", "t", arg_on_startup); + if (r < 0) + return r; + } + + if (arg_on_unit_active) { + r = sd_bus_message_append(m, "(sv)", "OnUnitActiveSec", "t", arg_on_unit_active); + if (r < 0) + return r; + } + + if (arg_on_unit_inactive) { + r = sd_bus_message_append(m, "(sv)", "OnUnitInactiveSec", "t", arg_on_unit_inactive); + if (r < 0) + return r; + } + + if (arg_on_calendar) { + r = sd_bus_message_append(m, "(sv)", "OnCalendar", "s", arg_on_calendar); + if (r < 0) + return r; + } + + return 0; +} + +static int make_unit_name(sd_bus *bus, UnitType t, char **ret) { + const char *unique, *id; + char *p; + int r; + + assert(bus); + assert(t >= 0); + assert(t < _UNIT_TYPE_MAX); + + r = sd_bus_get_unique_name(bus, &unique); + if (r < 0) { + sd_id128_t rnd; + + /* We couldn't get the unique name, which is a pretty + * common case if we are connected to systemd + * directly. In that case, just pick a random uuid as + * name */ + + r = sd_id128_randomize(&rnd); + if (r < 0) + return log_error_errno(r, "Failed to generate random run unit name: %m"); + + if (asprintf(ret, "run-r" SD_ID128_FORMAT_STR ".%s", SD_ID128_FORMAT_VAL(rnd), unit_type_to_string(t)) < 0) + return log_oom(); + + return 0; + } + + /* We managed to get the unique name, then let's use that to + * name our transient units. */ + + id = startswith(unique, ":1."); + if (!id) { + log_error("Unique name %s has unexpected format.", unique); + return -EINVAL; + } + + p = strjoin("run-u", id, ".", unit_type_to_string(t), NULL); + if (!p) + return log_oom(); + + *ret = p; + return 0; +} + +static int start_transient_service( + sd_bus *bus, + char **argv) { + + _cleanup_(sd_bus_message_unrefp) sd_bus_message *m = NULL, *reply = NULL; + _cleanup_(sd_bus_error_free) sd_bus_error error = SD_BUS_ERROR_NULL; + _cleanup_(bus_wait_for_jobs_freep) BusWaitForJobs *w = NULL; + _cleanup_free_ char *service = NULL, *pty_path = NULL; + _cleanup_close_ int master = -1; + int r; + + assert(bus); + assert(argv); + + if (arg_pty) { + + if (arg_transport == BUS_TRANSPORT_LOCAL) { + master = posix_openpt(O_RDWR|O_NOCTTY|O_CLOEXEC|O_NDELAY); + if (master < 0) + return log_error_errno(errno, "Failed to acquire pseudo tty: %m"); + + r = ptsname_malloc(master, &pty_path); + if (r < 0) + return log_error_errno(r, "Failed to determine tty name: %m"); + + if (unlockpt(master) < 0) + return log_error_errno(errno, "Failed to unlock tty: %m"); + + } else if (arg_transport == BUS_TRANSPORT_MACHINE) { + _cleanup_(sd_bus_unrefp) sd_bus *system_bus = NULL; + _cleanup_(sd_bus_message_unrefp) sd_bus_message *pty_reply = NULL; + const char *s; + + r = sd_bus_default_system(&system_bus); + if (r < 0) + return log_error_errno(r, "Failed to connect to system bus: %m"); + + r = sd_bus_call_method(system_bus, + "org.freedesktop.machine1", + "/org/freedesktop/machine1", + "org.freedesktop.machine1.Manager", + "OpenMachinePTY", + &error, + &pty_reply, + "s", arg_host); + if (r < 0) { + log_error("Failed to get machine PTY: %s", bus_error_message(&error, -r)); + return r; + } + + r = sd_bus_message_read(pty_reply, "hs", &master, &s); + if (r < 0) + return bus_log_parse_error(r); + + master = fcntl(master, F_DUPFD_CLOEXEC, 3); + if (master < 0) + return log_error_errno(errno, "Failed to duplicate master fd: %m"); + + pty_path = strdup(s); + if (!pty_path) + return log_oom(); + } else + assert_not_reached("Can't allocate tty via ssh"); + } + + if (!arg_no_block) { + r = bus_wait_for_jobs_new(bus, &w); + if (r < 0) + return log_error_errno(r, "Could not watch jobs: %m"); + } + + if (arg_unit) { + r = unit_name_mangle_with_suffix(arg_unit, UNIT_NAME_NOGLOB, ".service", &service); + if (r < 0) + return log_error_errno(r, "Failed to mangle unit name: %m"); + } else { + r = make_unit_name(bus, UNIT_SERVICE, &service); + if (r < 0) + return r; + } + + r = sd_bus_message_new_method_call( + bus, + &m, + "org.freedesktop.systemd1", + "/org/freedesktop/systemd1", + "org.freedesktop.systemd1.Manager", + "StartTransientUnit"); + if (r < 0) + return bus_log_create_error(r); + + r = sd_bus_message_set_allow_interactive_authorization(m, arg_ask_password); + if (r < 0) + return bus_log_create_error(r); + + /* Name and mode */ + r = sd_bus_message_append(m, "ss", service, "fail"); + if (r < 0) + return bus_log_create_error(r); + + /* Properties */ + r = sd_bus_message_open_container(m, 'a', "(sv)"); + if (r < 0) + return bus_log_create_error(r); + + r = transient_service_set_properties(m, argv, pty_path); + if (r < 0) + return bus_log_create_error(r); + + r = sd_bus_message_close_container(m); + if (r < 0) + return bus_log_create_error(r); + + /* Auxiliary units */ + r = sd_bus_message_append(m, "a(sa(sv))", 0); + if (r < 0) + return bus_log_create_error(r); + + polkit_agent_open_if_enabled(); + + r = sd_bus_call(bus, m, 0, &error, &reply); + if (r < 0) + return log_error_errno(r, "Failed to start transient service unit: %s", bus_error_message(&error, r)); + + if (w) { + const char *object; + + r = sd_bus_message_read(reply, "o", &object); + if (r < 0) + return bus_log_parse_error(r); + + r = bus_wait_for_jobs_one(w, object, arg_quiet); + if (r < 0) + return r; + } + + if (master >= 0) { + _cleanup_(pty_forward_freep) PTYForward *forward = NULL; + _cleanup_(sd_event_unrefp) sd_event *event = NULL; + char last_char = 0; + + r = sd_event_default(&event); + if (r < 0) + return log_error_errno(r, "Failed to get event loop: %m"); + + assert_se(sigprocmask_many(SIG_BLOCK, NULL, SIGWINCH, SIGTERM, SIGINT, -1) >= 0); + + (void) sd_event_add_signal(event, NULL, SIGINT, NULL, NULL); + (void) sd_event_add_signal(event, NULL, SIGTERM, NULL, NULL); + + if (!arg_quiet) + log_info("Running as unit: %s\nPress ^] three times within 1s to disconnect TTY.", service); + + r = pty_forward_new(event, master, PTY_FORWARD_IGNORE_INITIAL_VHANGUP, &forward); + if (r < 0) + return log_error_errno(r, "Failed to create PTY forwarder: %m"); + + r = sd_event_loop(event); + if (r < 0) + return log_error_errno(r, "Failed to run event loop: %m"); + + pty_forward_get_last_char(forward, &last_char); + + forward = pty_forward_free(forward); + + if (!arg_quiet && last_char != '\n') + fputc('\n', stdout); + + } else if (!arg_quiet) + log_info("Running as unit: %s", service); + + return 0; +} + +static int start_transient_scope( + sd_bus *bus, + char **argv) { + + _cleanup_(sd_bus_error_free) sd_bus_error error = SD_BUS_ERROR_NULL; + _cleanup_(sd_bus_message_unrefp) sd_bus_message *m = NULL, *reply = NULL; + _cleanup_(bus_wait_for_jobs_freep) BusWaitForJobs *w = NULL; + _cleanup_strv_free_ char **env = NULL, **user_env = NULL; + _cleanup_free_ char *scope = NULL; + const char *object = NULL; + int r; + + assert(bus); + assert(argv); + + r = bus_wait_for_jobs_new(bus, &w); + if (r < 0) + return log_oom(); + + if (arg_unit) { + r = unit_name_mangle_with_suffix(arg_unit, UNIT_NAME_NOGLOB, ".scope", &scope); + if (r < 0) + return log_error_errno(r, "Failed to mangle scope name: %m"); + } else { + r = make_unit_name(bus, UNIT_SCOPE, &scope); + if (r < 0) + return r; + } + + r = sd_bus_message_new_method_call( + bus, + &m, + "org.freedesktop.systemd1", + "/org/freedesktop/systemd1", + "org.freedesktop.systemd1.Manager", + "StartTransientUnit"); + if (r < 0) + return bus_log_create_error(r); + + r = sd_bus_message_set_allow_interactive_authorization(m, arg_ask_password); + if (r < 0) + return bus_log_create_error(r); + + /* Name and Mode */ + r = sd_bus_message_append(m, "ss", scope, "fail"); + if (r < 0) + return bus_log_create_error(r); + + /* Properties */ + r = sd_bus_message_open_container(m, 'a', "(sv)"); + if (r < 0) + return bus_log_create_error(r); + + r = transient_scope_set_properties(m); + if (r < 0) + return bus_log_create_error(r); + + r = sd_bus_message_close_container(m); + if (r < 0) + return bus_log_create_error(r); + + /* Auxiliary units */ + r = sd_bus_message_append(m, "a(sa(sv))", 0); + if (r < 0) + return bus_log_create_error(r); + + polkit_agent_open_if_enabled(); + + r = sd_bus_call(bus, m, 0, &error, &reply); + if (r < 0) { + log_error("Failed to start transient scope unit: %s", bus_error_message(&error, -r)); + return r; + } + + if (arg_nice_set) { + if (setpriority(PRIO_PROCESS, 0, arg_nice) < 0) + return log_error_errno(errno, "Failed to set nice level: %m"); + } + + if (arg_exec_group) { + gid_t gid; + + r = get_group_creds(&arg_exec_group, &gid); + if (r < 0) + return log_error_errno(r, "Failed to resolve group %s: %m", arg_exec_group); + + if (setresgid(gid, gid, gid) < 0) + return log_error_errno(errno, "Failed to change GID to " GID_FMT ": %m", gid); + } + + if (arg_exec_user) { + const char *home, *shell; + uid_t uid; + gid_t gid; + + r = get_user_creds(&arg_exec_user, &uid, &gid, &home, &shell); + if (r < 0) + return log_error_errno(r, "Failed to resolve user %s: %m", arg_exec_user); + + r = strv_extendf(&user_env, "HOME=%s", home); + if (r < 0) + return log_oom(); + + r = strv_extendf(&user_env, "SHELL=%s", shell); + if (r < 0) + return log_oom(); + + r = strv_extendf(&user_env, "USER=%s", arg_exec_user); + if (r < 0) + return log_oom(); + + r = strv_extendf(&user_env, "LOGNAME=%s", arg_exec_user); + if (r < 0) + return log_oom(); + + if (!arg_exec_group) { + if (setresgid(gid, gid, gid) < 0) + return log_error_errno(errno, "Failed to change GID to " GID_FMT ": %m", gid); + } + + if (setresuid(uid, uid, uid) < 0) + return log_error_errno(errno, "Failed to change UID to " UID_FMT ": %m", uid); + } + + env = strv_env_merge(3, environ, user_env, arg_environment); + if (!env) + return log_oom(); + + r = sd_bus_message_read(reply, "o", &object); + if (r < 0) + return bus_log_parse_error(r); + + r = bus_wait_for_jobs_one(w, object, arg_quiet); + if (r < 0) + return r; + + if (!arg_quiet) + log_info("Running scope as unit: %s", scope); + + execvpe(argv[0], argv, env); + + return log_error_errno(errno, "Failed to execute: %m"); +} + +static int start_transient_timer( + sd_bus *bus, + char **argv) { + + _cleanup_(sd_bus_error_free) sd_bus_error error = SD_BUS_ERROR_NULL; + _cleanup_(sd_bus_message_unrefp) sd_bus_message *m = NULL, *reply = NULL; + _cleanup_(bus_wait_for_jobs_freep) BusWaitForJobs *w = NULL; + _cleanup_free_ char *timer = NULL, *service = NULL; + const char *object = NULL; + int r; + + assert(bus); + assert(argv); + + r = bus_wait_for_jobs_new(bus, &w); + if (r < 0) + return log_oom(); + + if (arg_unit) { + switch (unit_name_to_type(arg_unit)) { + + case UNIT_SERVICE: + service = strdup(arg_unit); + if (!service) + return log_oom(); + + r = unit_name_change_suffix(service, ".timer", &timer); + if (r < 0) + return log_error_errno(r, "Failed to change unit suffix: %m"); + break; + + case UNIT_TIMER: + timer = strdup(arg_unit); + if (!timer) + return log_oom(); + + r = unit_name_change_suffix(timer, ".service", &service); + if (r < 0) + return log_error_errno(r, "Failed to change unit suffix: %m"); + break; + + default: + r = unit_name_mangle_with_suffix(arg_unit, UNIT_NAME_NOGLOB, ".service", &service); + if (r < 0) + return log_error_errno(r, "Failed to mangle unit name: %m"); + + r = unit_name_mangle_with_suffix(arg_unit, UNIT_NAME_NOGLOB, ".timer", &timer); + if (r < 0) + return log_error_errno(r, "Failed to mangle unit name: %m"); + + break; + } + } else { + r = make_unit_name(bus, UNIT_SERVICE, &service); + if (r < 0) + return r; + + r = unit_name_change_suffix(service, ".timer", &timer); + if (r < 0) + return log_error_errno(r, "Failed to change unit suffix: %m"); + } + + r = sd_bus_message_new_method_call( + bus, + &m, + "org.freedesktop.systemd1", + "/org/freedesktop/systemd1", + "org.freedesktop.systemd1.Manager", + "StartTransientUnit"); + if (r < 0) + return bus_log_create_error(r); + + r = sd_bus_message_set_allow_interactive_authorization(m, arg_ask_password); + if (r < 0) + return bus_log_create_error(r); + + /* Name and Mode */ + r = sd_bus_message_append(m, "ss", timer, "fail"); + if (r < 0) + return bus_log_create_error(r); + + /* Properties */ + r = sd_bus_message_open_container(m, 'a', "(sv)"); + if (r < 0) + return bus_log_create_error(r); + + r = transient_timer_set_properties(m); + if (r < 0) + return bus_log_create_error(r); + + r = sd_bus_message_close_container(m); + if (r < 0) + return bus_log_create_error(r); + + r = sd_bus_message_open_container(m, 'a', "(sa(sv))"); + if (r < 0) + return bus_log_create_error(r); + + if (argv[0]) { + r = sd_bus_message_open_container(m, 'r', "sa(sv)"); + if (r < 0) + return bus_log_create_error(r); + + r = sd_bus_message_append(m, "s", service); + if (r < 0) + return bus_log_create_error(r); + + r = sd_bus_message_open_container(m, 'a', "(sv)"); + if (r < 0) + return bus_log_create_error(r); + + r = transient_service_set_properties(m, argv, NULL); + if (r < 0) + return bus_log_create_error(r); + + r = sd_bus_message_close_container(m); + if (r < 0) + return bus_log_create_error(r); + + r = sd_bus_message_close_container(m); + if (r < 0) + return bus_log_create_error(r); + } + + r = sd_bus_message_close_container(m); + if (r < 0) + return bus_log_create_error(r); + + polkit_agent_open_if_enabled(); + + r = sd_bus_call(bus, m, 0, &error, &reply); + if (r < 0) { + log_error("Failed to start transient timer unit: %s", bus_error_message(&error, -r)); + return r; + } + + r = sd_bus_message_read(reply, "o", &object); + if (r < 0) + return bus_log_parse_error(r); + + r = bus_wait_for_jobs_one(w, object, arg_quiet); + if (r < 0) + return r; + + log_info("Running timer as unit: %s", timer); + if (argv[0]) + log_info("Will run service as unit: %s", service); + + return 0; +} + +int main(int argc, char* argv[]) { + _cleanup_(sd_bus_flush_close_unrefp) sd_bus *bus = NULL; + _cleanup_free_ char *description = NULL, *command = NULL; + int r; + + log_parse_environment(); + log_open(); + + r = parse_argv(argc, argv); + if (r <= 0) + goto finish; + + if (argc > optind && arg_transport == BUS_TRANSPORT_LOCAL) { + /* Patch in an absolute path */ + + r = find_binary(argv[optind], &command); + if (r < 0) { + log_error_errno(r, "Failed to find executable %s: %m", argv[optind]); + goto finish; + } + + argv[optind] = command; + } + + if (!arg_description) { + description = strv_join(argv + optind, " "); + if (!description) { + r = log_oom(); + goto finish; + } + + if (arg_unit && isempty(description)) { + r = free_and_strdup(&description, arg_unit); + if (r < 0) + goto finish; + } + + arg_description = description; + } + + r = bus_connect_transport_systemd(arg_transport, arg_host, arg_user, &bus); + if (r < 0) { + log_error_errno(r, "Failed to create bus connection: %m"); + goto finish; + } + + if (arg_scope) + r = start_transient_scope(bus, argv + optind); + else if (with_timer()) + r = start_transient_timer(bus, argv + optind); + else + r = start_transient_service(bus, argv + optind); + +finish: + strv_free(arg_environment); + strv_free(arg_property); + strv_free(arg_timer_property); + + return r < 0 ? EXIT_FAILURE : EXIT_SUCCESS; +} diff --git a/src/grp-system/grp-utils/systemd-run/systemd-run.completion.bash b/src/grp-system/grp-utils/systemd-run/systemd-run.completion.bash new file mode 100644 index 0000000000..022331e6a9 --- /dev/null +++ b/src/grp-system/grp-utils/systemd-run/systemd-run.completion.bash @@ -0,0 +1,117 @@ +# systemd-run(1) completion -*- shell-script -*- +# +# This file is part of systemd. +# +# Copyright 2013 Zbigniew Jędrzejewski-Szmek +# +# systemd is free software; you can redistribute it and/or modify it +# under the terms of the GNU Lesser General Public License as published by +# the Free Software Foundation; either version 2.1 of the License, or +# (at your option) any later version. +# +# systemd is distributed in the hope that it will be useful, but +# WITHOUT ANY WARRANTY; without even the implied warranty of +# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU +# General Public License for more details. +# +# You should have received a copy of the GNU Lesser General Public License +# along with systemd; If not, see . + +__systemctl() { + local mode=$1; shift 1 + systemctl $mode --full --no-legend "$@" +} + +__get_slice_units () { __systemctl $1 list-units --all -t slice \ + | { while read -r a b c d; do echo " $a"; done; }; } + +__get_machines() { + local a b + machinectl list --no-legend --no-pager | { while read a b; do echo " $a"; done; }; +} + +_systemd_run() { + local cur=${COMP_WORDS[COMP_CWORD]} prev=${COMP_WORDS[COMP_CWORD-1]} + local OPTS='-h --help --version --user --system --scope --unit --description --slice + -r --remain-after-exit --send-sighup -H --host -M --machine --service-type + --on-active --on-boot --on-startup --on-unit-active --on-unit-inactive + --on-calendar --timer-property -t --pty -q --quiet --no-block + --uid --gid --nice --setenv -p --property --no-ask-password' + + local mode=--system + local i + local opts_with_values=( + --unit --description --slice --service-type -H --host -M --machine -p --property --on-active + --on-boot --on-startup --on-unit-active --on-unit-inactive --on-calendar --timer-property + ) + for (( i=1; i <= COMP_CWORD; i++ )); do + if [[ ${COMP_WORDS[i]} != -* ]]; then + local root_command=${COMP_WORDS[i]} + _command_offset $i + return + fi + + [[ ${COMP_WORDS[i]} == "--user" ]] && mode=--user + + [[ $i -lt $COMP_CWORD && " ${opts_with_values[@]} " =~ " ${COMP_WORDS[i]} " ]] && ((i++)) + done + + case "$prev" in + --unit|--description|--on-active|--on-boot|--on-startup|--on-unit-active|--on-unit-inactive|--on-calendar) + # argument required but no completions available + return + ;; + --slice) + local comps=$(__get_slice_units $mode) + + COMPREPLY=( $(compgen -W '$comps' -- "$cur") ) + return 0 + ;; + --service-type) + local comps='simple forking oneshot dbus notify idle' + + COMPREPLY=( $(compgen -W '$comps' -- "$cur") ) + return 0 + ;; + -p|--property) + local comps='CPUAccounting= MemoryAccounting= BlockIOAccounting= SendSIGHUP= + SendSIGKILL= MemoryLimit= CPUShares= BlockIOWeight= User= Group= + DevicePolicy= KillMode= DeviceAllow= BlockIOReadBandwidth= + BlockIOWriteBandwidth= BlockIODeviceWeight= Nice= Environment= + KillSignal= LimitCPU= LimitFSIZE= LimitDATA= LimitSTACK= + LimitCORE= LimitRSS= LimitNOFILE= LimitAS= LimitNPROC= + LimitMEMLOCK= LimitLOCKS= LimitSIGPENDING= LimitMSGQUEUE= + LimitNICE= LimitRTPRIO= LimitRTTIME= PrivateTmp= PrivateDevices= + PrivateNetwork= NoNewPrivileges= WorkingDirectory= RootDirectory= + TTYPath= SyslogIdentifier= SyslogLevelPrefix= SyslogLevel= + SyslogFacility= TimerSlackNSec= OOMScoreAdjust= ReadWritePaths= + ReadOnlyPaths= InaccessiblePaths= EnvironmentFile= + ProtectSystem= ProtectHome= RuntimeDirectory= PassEnvironment=' + + COMPREPLY=( $(compgen -W '$comps' -- "$cur") ) + return 0 + ;; + -H|--host) + local comps=$(compgen -A hostname) + + COMPREPLY=( $(compgen -W '$comps' -- "$cur") ) + return 0 + ;; + -M|--machine) + local comps=$( __get_machines ) + + COMPREPLY=( $(compgen -W '$comps' -- "$cur") ) + return 0 + ;; + --timer-property) + local comps='AccuracySec= WakeSystem=' + COMPREPLY=( $(compgen -W '$comps' -- "$cur") ) + return 0 + ;; + esac + + COMPREPLY=( $(compgen -W '${OPTS[*]}' -- "$cur") ) + return 0 +} + +complete -F _systemd_run systemd-run diff --git a/src/grp-system/grp-utils/systemd-run/systemd-run.completion.zsh b/src/grp-system/grp-utils/systemd-run/systemd-run.completion.zsh new file mode 100644 index 0000000000..6362b97766 --- /dev/null +++ b/src/grp-system/grp-utils/systemd-run/systemd-run.completion.zsh @@ -0,0 +1,60 @@ +#compdef systemd-run + +__systemctl() { + local -a _modes + _modes=("--user" "--system") + systemctl ${words:*_modes} --full --no-legend --no-pager "$@" 2>/dev/null +} + +__get_slices () { + __systemctl list-units --all -t slice \ + | { while read -r a b; do echo $a; done; }; +} + +__slices () { + local -a _slices + _slices=(${(fo)"$(__get_slices)"}) + typeset -U _slices + _describe 'slices' _slices +} + +_arguments \ + {-h,--help}'[Show help message]' \ + '--version[Show package version]' \ + '--user[Run as user unit]' \ + {-H+,--host=}'[Operate on remote host]:[user@]host:_sd_hosts_or_user_at_host' \ + {-M+,--machine=}'[Operate on local container]:machines:_sd_machines' \ + '--scope[Run this as scope rather than service]' \ + '--unit=[Run under the specified unit name]:unit name' \ + {-p+,--property=}'[Set unit property]:NAME=VALUE:(( \ + CPUAccounting= MemoryAccounting= BlockIOAccounting= SendSIGHUP= \ + SendSIGKILL= MemoryLimit= CPUShares= BlockIOWeight= User= Group= \ + DevicePolicy= KillMode= DeviceAllow= BlockIOReadBandwidth= \ + BlockIOWriteBandwidth= BlockIODeviceWeight= Nice= Environment= \ + KillSignal= LimitCPU= LimitFSIZE= LimitDATA= LimitSTACK= \ + LimitCORE= LimitRSS= LimitNOFILE= LimitAS= LimitNPROC= \ + LimitMEMLOCK= LimitLOCKS= LimitSIGPENDING= LimitMSGQUEUE= \ + LimitNICE= LimitRTPRIO= LimitRTTIME= PrivateTmp= PrivateDevices= \ + PrivateNetwork= NoNewPrivileges= WorkingDirectory= RootDirectory= \ + TTYPath= SyslogIdentifier= SyslogLevelPrefix= SyslogLevel= \ + SyslogFacility= TimerSlackNSec= OOMScoreAdjust= ReadWritePaths= \ + ReadOnlyPaths= InaccessiblePaths= EnvironmentFile= \ + ProtectSystem= ProtectHome= RuntimeDirectory= PassEnvironment= \ + ))' \ + '--description=[Description for unit]:description' \ + '--slice=[Run in the specified slice]:slices:__slices' \ + {-r,--remain-after-exit}'[Leave service around until explicitly stopped]' \ + '--send-sighup[Send SIGHUP when terminating]' \ + '--service-type=[Service type]:type:(simple forking oneshot dbus notify idle)' \ + '--uid=[Run as system user]:user:_users' \ + '--gid=[Run as system group]:group:_groups' \ + '--nice=[Nice level]:nice level' \ + '--setenv=[Set environment]:NAME=VALUE' \ + '--on-active=[Run after SEC seconds]:SEC' \ + '--on-boot=[Run after SEC seconds from machine was booted up]:SEC' \ + '--on-statup=[Run after SEC seconds from systemd was first started]:SEC' \ + '--on-unit-active=[Run after SEC seconds from the last activation]:SEC' \ + '--on-unit-inactive=[Run after SEC seconds from the last deactivation]:SEC' \ + '--on-calendar=[Realtime timer]:SPEC' \ + '--timer-property=[Set timer unit property]:NAME=VALUE' \ + '*::command:_command' diff --git a/src/grp-system/grp-utils/systemd-run/systemd-run.xml b/src/grp-system/grp-utils/systemd-run/systemd-run.xml new file mode 100644 index 0000000000..9c1a29218e --- /dev/null +++ b/src/grp-system/grp-utils/systemd-run/systemd-run.xml @@ -0,0 +1,459 @@ + + + + + + + + + systemd-run + systemd + + + + Developer + Lennart + Poettering + lennart@poettering.net + + + + + + systemd-run + 1 + + + + systemd-run + Run programs in transient scope or service or timer units + + + + + systemd-run + OPTIONS + COMMAND + ARGS + + + + systemd-run + OPTIONS + TIMER OPTIONS + COMMAND + ARGS + + + + + Description + + systemd-run may be used to create and + start a transient .service or + .scope unit and run the specified + COMMAND in it. It may also be used to + create and start transient .timer + units. + + If a command is run as transient service unit, it will be + started and managed by the service manager like any other service, + and thus shows up in the output of systemctl + list-units like any other unit. It will run in a clean + and detached execution environment, with the service manager as + its parent process. In this mode, systemd-run + will start the service asynchronously in the background and return + after the command has begun execution. + + If a command is run as transient scope unit, it will be + started by systemd-run itself as parent process + and will thus inherit the execution environment of the + caller. However, the processes of the command are managed by the + service manager similar to normal services, and will show up in + the output of systemctl list-units. Execution + in this case is synchronous, and will return only when the command + finishes. This mode is enabled via the + switch (see below). + + If a command is run with timer options such as + (see below), a transient timer + unit is created alongside the service unit for the specified + command. Only the transient timer unit is started immediately, the + transient service unit will be started when the transient timer + elapses. If the is specified, the + COMMAND may be omitted. In this case, + systemd-run only creates a + .timer unit that invokes the specified unit + when elapsing. + + + + Options + + The following options are understood: + + + + + + Do not query the user for authentication for + privileged operations. + + + + + + + Create a transient .scope unit instead of + the default transient .service unit. + + + + + + + + Use this unit name instead of an automatically + generated one. + + + + + + + Sets a unit property for the scope or service + unit that is created. This takes an assignment in the same + format as + systemctl1's + set-property command. + + + + + + + Provide a description for the service or scope + unit. If not specified, the command itself will be used as a + description. See Description= in + systemd.unit5. + + + + + + + Make the new .service or + .scope unit part of the specified slice, + instead of the system.slice. + + + + + + + After the service or scope process has + terminated, keep the service around until it is explicitly + stopped. This is useful to collect runtime information about + the service after it finished running. Also see + RemainAfterExit= in + systemd.service5. + + + + + + + + When terminating the scope or service unit, + send a SIGHUP immediately after SIGTERM. This is useful to + indicate to shells and shell-like processes that the + connection has been severed. Also see + SendSIGHUP= in + systemd.kill5. + + + + + + + + Sets the service type. Also see + Type= in + systemd.service5. This + option has no effect in conjunction with + . Defaults to + simple. + + + + + + + + Runs the service process under the UNIX user + and group. Also see User= and + Group= in + systemd.exec5. + + + + + + + Runs the service process with the specified + nice level. Also see Nice= in + systemd.exec5. + + + + + + + + Runs the service process with the specified environment variable set. + Also see Environment= in + systemd.exec5. + + + + + + + + When invoking a command, the service connects + its standard input and output to the invoking tty via a + pseudo TTY device. This allows invoking binaries as services + that expect interactive user input, such as interactive + command shells. + + + + + + + Suppresses additional informational output + while running. This is particularly useful in combination with + when it will suppress the initial + message explaining how to terminate the TTY connection. + + + + + + + + + + Defines monotonic timers relative to different + starting points. Also see OnActiveSec=, + OnBootSec=, + OnStartupSec=, + OnUnitActiveSec= and + OnUnitInactiveSec= in + systemd.timer5. This + options have no effect in conjunction with + . + + + + + + + Defines realtime (i.e. wallclock) timers with + calendar event expressions. Also see + OnCalendar= in + systemd.timer5. This + option has no effect in conjunction with + . + + + + + + + Sets a timer unit property for the timer unit + that is created. It is similar with + but only for created timer + unit. This option only has effect in conjunction with + , , + , + , + , + . This takes an assignment in + the same format as + systemctl1's + set-property command. + + + + + + + Do not synchronously wait for the requested operation + to finish. If this is not specified, the job will be + verified, enqueued and systemd-run will + wait until the unit's start-up is completed. By passing this + argument, it is only verified and enqueued. + + + + + + + + + + + + + All command line arguments after the first non-option + argument become part of the command line of the launched + process. If a command is run as service unit, its first argument + needs to be an absolute binary path. + + + + Exit status + + On success, 0 is returned, a non-zero failure + code otherwise. + + + + Examples + + + Logging environment variables provided by systemd to services + + # systemd-run env +Running as unit: run-19945.service +# journalctl -u run-19945.service +Sep 08 07:37:21 bupkis systemd[1]: Starting /usr/bin/env... +Sep 08 07:37:21 bupkis systemd[1]: Started /usr/bin/env. +Sep 08 07:37:21 bupkis env[19948]: PATH=/usr/local/sbin:/usr/local/bin:/usr/sbin:/usr/bin +Sep 08 07:37:21 bupkis env[19948]: LANG=en_US.UTF-8 +Sep 08 07:37:21 bupkis env[19948]: BOOT_IMAGE=/vmlinuz-3.11.0-0.rc5.git6.2.fc20.x86_64 + + + + Limiting resources available to a command + + # systemd-run -p BlockIOWeight=10 updatedb + + This command invokes the + updatedb8 + tool, but lowers the block I/O weight for it to 10. See + systemd.resource-control5 + for more information on the BlockIOWeight= + property. + + + + Running commands at a specified time + + The following command will touch a file after 30 seconds. + + # date; systemd-run --on-active=30 --timer-property=AccuracySec=100ms /bin/touch /tmp/foo +Mon Dec 8 20:44:24 KST 2014 +Running as unit: run-71.timer +Will run service as unit: run-71.service +# journalctl -b -u run-71.timer +-- Logs begin at Fri 2014-12-05 19:09:21 KST, end at Mon 2014-12-08 20:44:54 KST. -- +Dec 08 20:44:38 container systemd[1]: Starting /bin/touch /tmp/foo. +Dec 08 20:44:38 container systemd[1]: Started /bin/touch /tmp/foo. +# journalctl -b -u run-71.service +-- Logs begin at Fri 2014-12-05 19:09:21 KST, end at Mon 2014-12-08 20:44:54 KST. -- +Dec 08 20:44:48 container systemd[1]: Starting /bin/touch /tmp/foo... +Dec 08 20:44:48 container systemd[1]: Started /bin/touch /tmp/foo. + + + + Allowing access to the tty + + The following command invokes /bin/bash as a service + passing its standard input, output and error to the calling TTY. + + # systemd-run -t --send-sighup /bin/bash + + + + Start <command>screen</command> as a user service + + $ systemd-run --scope --user screen +Running scope as unit run-r14b0047ab6df45bfb45e7786cc839e76.scope. + +$ screen -ls +There is a screen on: + 492..laptop (Detached) +1 Socket in /var/run/screen/S-fatima. + + + This starts the screen process as a child of the + systemd --user process that was started by + user@.service, in a scope unit. A + systemd.scope5 + unit is used instead of a + systemd.service5 + unit, because screen will exit when detaching from the terminal, + and a service unit would be terminated. Running screen + as a user unit has the advantage that it is not part of the session scope. + If KillUserProcesses=yes is configured in + logind.conf5, + the default, the session scope will be terminated when the user logs + out of that session. + + The user@.service is started automatically + when the user first logs in, and stays around as long as at least one + login session is open. After the user logs out of the last session, + user@.service and all services underneath it + are terminated. This behaviour is the default, when "lingering" is + not enabled for that user. Enabling lingering means that + user@.service is started automatically during + boot, even if the user is not logged in, and that the service is + not terminated when the user logs out. + + Enabling lingering allows the user to run processes without being logged in, + for example to allow screen to persist after the user logs out, + even if the session scope is terminated. In the default configuration, users can + enable lingering for themselves: + + $ loginctl enable-linger + + + + + See Also + + systemd1, + systemctl1, + systemd.unit5, + systemd.service5, + systemd.scope5, + systemd.slice5, + systemd.exec5, + systemd.resource-control5, + systemd.timer5, + machinectl1 + + + + diff --git a/src/grp-system/grp-utils/systemd-sysv-generator/.gitignore b/src/grp-system/grp-utils/systemd-sysv-generator/.gitignore new file mode 100644 index 0000000000..c3fea7424f --- /dev/null +++ b/src/grp-system/grp-utils/systemd-sysv-generator/.gitignore @@ -0,0 +1 @@ +/README diff --git a/src/grp-system/grp-utils/systemd-sysv-generator/Makefile b/src/grp-system/grp-utils/systemd-sysv-generator/Makefile new file mode 100644 index 0000000000..9dec62efdc --- /dev/null +++ b/src/grp-system/grp-utils/systemd-sysv-generator/Makefile @@ -0,0 +1,46 @@ +# -*- Mode: makefile; indent-tabs-mode: t -*- +# +# This file is part of systemd. +# +# Copyright 2010-2012 Lennart Poettering +# Copyright 2010-2012 Kay Sievers +# Copyright 2013 Zbigniew Jędrzejewski-Szmek +# Copyright 2013 David Strauss +# Copyright 2016 Luke Shumaker +# +# systemd is free software; you can redistribute it and/or modify it +# under the terms of the GNU Lesser General Public License as published by +# the Free Software Foundation; either version 2.1 of the License, or +# (at your option) any later version. +# +# systemd is distributed in the hope that it will be useful, but +# WITHOUT ANY WARRANTY; without even the implied warranty of +# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU +# Lesser General Public License for more details. +# +# You should have received a copy of the GNU Lesser General Public License +# along with systemd; If not, see . +include $(dir $(lastword $(MAKEFILE_LIST)))/../../../../config.mk +include $(topsrcdir)/build-aux/Makefile.head.mk + +systemd_sysv_generator_SOURCES = \ + src/sysv-generator/sysv-generator.c + +systemd_sysv_generator_LDADD = \ + libcore.la + +ifneq ($(HAVE_SYSV_COMPAT),) +sysvinit_DATA = \ + docs/sysvinit/README + +$(outdir)/README: docs/sysvinit/README.in + $(SED_PROCESS) + +CLEANFILES += \ + docs/sysvinit/README +endif # HAVE_SYSV_COMPAT + +EXTRA_DIST += \ + docs/sysvinit/README.in + +include $(topsrcdir)/build-aux/Makefile.tail.mk diff --git a/src/grp-system/grp-utils/systemd-sysv-generator/README.in b/src/grp-system/grp-utils/systemd-sysv-generator/README.in new file mode 100644 index 0000000000..996402d06b --- /dev/null +++ b/src/grp-system/grp-utils/systemd-sysv-generator/README.in @@ -0,0 +1,27 @@ +You are looking for the traditional init scripts in @SYSTEM_SYSVINIT_PATH@, +and they are gone? + +Here's an explanation on what's going on: + +You are running a systemd-based OS where traditional init scripts have +been replaced by native systemd services files. Service files provide +very similar functionality to init scripts. To make use of service +files simply invoke "systemctl", which will output a list of all +currently running services (and other units). Use "systemctl +list-unit-files" to get a listing of all known unit files, including +stopped, disabled and masked ones. Use "systemctl start +foobar.service" and "systemctl stop foobar.service" to start or stop a +service, respectively. For further details, please refer to +systemctl(1). + +Note that traditional init scripts continue to function on a systemd +system. An init script @SYSTEM_SYSVINIT_PATH@/foobar is implicitly mapped +into a service unit foobar.service during system initialization. + +Thank you! + +Further reading: + man:systemctl(1) + man:systemd(1) + http://0pointer.de/blog/projects/systemd-for-admins-3.html + http://www.freedesktop.org/wiki/Software/systemd/Incompatibilities diff --git a/src/grp-system/grp-utils/systemd-sysv-generator/systemd-sysv-generator.xml b/src/grp-system/grp-utils/systemd-sysv-generator/systemd-sysv-generator.xml new file mode 100644 index 0000000000..2353eb3efe --- /dev/null +++ b/src/grp-system/grp-utils/systemd-sysv-generator/systemd-sysv-generator.xml @@ -0,0 +1,97 @@ + + + + + + + + systemd-sysv-generator + systemd + + + + Documentation + Zbigniew + Jędrzejewski-Szmek + zbyszek@in.waw.pl + + + + + + systemd-sysv-generator + 8 + + + + systemd-sysv-generator + Unit generator for SysV init scripts + + + + /usr/lib/systemd/system-generators/systemd-sysv-generator + + + + Description + + systemd-sysv-generator is a generator + that creates wrapper .service units for + SysV init + scripts in /etc/init.d/* at boot and when + configuration of the system manager is reloaded. This will allow + systemd1 + to support them similarly to native units. + + LSB headers + in SysV init scripts are interpreted, and the ordering specified + in the header is turned into dependencies between the generated + unit and other units. The LSB facilities + $remote_fs, $network, + $named, $portmap, + $time are supported and will be turned into + dependencies on specific native systemd targets. See + systemd.special5 + for more details. + + SysV runlevels have corresponding systemd targets + (runlevelX.target). + The wrapper unit that is generated will be wanted by those targets + which correspond to runlevels for which the script is + enabled. + + systemd does not support SysV scripts as + part of early boot, so all wrapper units are ordered after + basic.target. + + systemd-sysv-generator implements + systemd.generator7. + + + + See Also + + systemd1, + systemd.service5, + systemd.target5 + + + + diff --git a/src/grp-system/grp-utils/systemd-sysv-generator/sysv-generator.c b/src/grp-system/grp-utils/systemd-sysv-generator/sysv-generator.c new file mode 100644 index 0000000000..873cab1c5c --- /dev/null +++ b/src/grp-system/grp-utils/systemd-sysv-generator/sysv-generator.c @@ -0,0 +1,982 @@ +/*** + This file is part of systemd. + + Copyright 2014 Thomas H.P. Andersen + Copyright 2010 Lennart Poettering + Copyright 2011 Michal Schmidt + + systemd is free software; you can redistribute it and/or modify it + under the terms of the GNU Lesser General Public License as published by + the Free Software Foundation; either version 2.1 of the License, or + (at your option) any later version. + + systemd is distributed in the hope that it will be useful, but + WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + Lesser General Public License for more details. + + You should have received a copy of the GNU Lesser General Public License + along with systemd; If not, see . +***/ + +#include +#include +#include + +#include "systemd-basic/alloc-util.h" +#include "systemd-basic/dirent-util.h" +#include "systemd-basic/fd-util.h" +#include "systemd-basic/fileio.h" +#include "systemd-basic/hashmap.h" +#include "systemd-basic/hexdecoct.h" +#include "systemd-basic/log.h" +#include "systemd-basic/mkdir.h" +#include "systemd-basic/path-util.h" +#include "systemd-basic/set.h" +#include "systemd-basic/special.h" +#include "systemd-basic/stat-util.h" +#include "systemd-basic/string-util.h" +#include "systemd-basic/strv.h" +#include "systemd-basic/unit-name.h" +#include "systemd-basic/util.h" +#include "systemd-shared/install.h" +#include "systemd-shared/path-lookup.h" + +static const struct { + const char *path; + const char *target; +} rcnd_table[] = { + /* Standard SysV runlevels for start-up */ + { "rc1.d", SPECIAL_RESCUE_TARGET }, + { "rc2.d", SPECIAL_MULTI_USER_TARGET }, + { "rc3.d", SPECIAL_MULTI_USER_TARGET }, + { "rc4.d", SPECIAL_MULTI_USER_TARGET }, + { "rc5.d", SPECIAL_GRAPHICAL_TARGET }, + + /* We ignore the SysV runlevels for shutdown here, as SysV services get default dependencies anyway, and that + * means they are shut down anyway at system power off if running. */ +}; + +static const char *arg_dest = "/tmp"; + +typedef struct SysvStub { + char *name; + char *path; + char *description; + int sysv_start_priority; + char *pid_file; + char **before; + char **after; + char **wants; + char **wanted_by; + bool has_lsb; + bool reload; + bool loaded; +} SysvStub; + +static void free_sysvstub(SysvStub *s) { + if (!s) + return; + + free(s->name); + free(s->path); + free(s->description); + free(s->pid_file); + strv_free(s->before); + strv_free(s->after); + strv_free(s->wants); + strv_free(s->wanted_by); + free(s); +} + +DEFINE_TRIVIAL_CLEANUP_FUNC(SysvStub*, free_sysvstub); + +static void free_sysvstub_hashmapp(Hashmap **h) { + SysvStub *stub; + + while ((stub = hashmap_steal_first(*h))) + free_sysvstub(stub); + + hashmap_free(*h); +} + +static int add_symlink(const char *service, const char *where) { + const char *from, *to; + int r; + + assert(service); + assert(where); + + from = strjoina(arg_dest, "/", service); + to = strjoina(arg_dest, "/", where, ".wants/", service); + + mkdir_parents_label(to, 0755); + + r = symlink(from, to); + if (r < 0) { + if (errno == EEXIST) + return 0; + + return -errno; + } + + return 1; +} + +static int add_alias(const char *service, const char *alias) { + const char *link; + int r; + + assert(service); + assert(alias); + + link = strjoina(arg_dest, "/", alias); + + r = symlink(service, link); + if (r < 0) { + if (errno == EEXIST) + return 0; + + return -errno; + } + + return 1; +} + +static int generate_unit_file(SysvStub *s) { + _cleanup_fclose_ FILE *f = NULL; + const char *unit; + char **p; + int r; + + assert(s); + + if (!s->loaded) + return 0; + + unit = strjoina(arg_dest, "/", s->name); + + /* We might already have a symlink with the same name from a Provides:, + * or from backup files like /etc/init.d/foo.bak. Real scripts always win, + * so remove an existing link */ + if (is_symlink(unit) > 0) { + log_warning("Overwriting existing symlink %s with real service.", unit); + (void) unlink(unit); + } + + f = fopen(unit, "wxe"); + if (!f) + return log_error_errno(errno, "Failed to create unit file %s: %m", unit); + + fprintf(f, + "# Automatically generated by systemd-sysv-generator\n\n" + "[Unit]\n" + "Documentation=man:systemd-sysv-generator(8)\n" + "SourcePath=%s\n", + s->path); + + if (s->description) + fprintf(f, "Description=%s\n", s->description); + + STRV_FOREACH(p, s->before) + fprintf(f, "Before=%s\n", *p); + STRV_FOREACH(p, s->after) + fprintf(f, "After=%s\n", *p); + STRV_FOREACH(p, s->wants) + fprintf(f, "Wants=%s\n", *p); + + fprintf(f, + "\n[Service]\n" + "Type=forking\n" + "Restart=no\n" + "TimeoutSec=5min\n" + "IgnoreSIGPIPE=no\n" + "KillMode=process\n" + "GuessMainPID=no\n" + "RemainAfterExit=%s\n", + yes_no(!s->pid_file)); + + if (s->pid_file) + fprintf(f, "PIDFile=%s\n", s->pid_file); + + fprintf(f, + "ExecStart=%s start\n" + "ExecStop=%s stop\n", + s->path, s->path); + + if (s->reload) + fprintf(f, "ExecReload=%s reload\n", s->path); + + r = fflush_and_check(f); + if (r < 0) + return log_error_errno(r, "Failed to write unit %s: %m", unit); + + STRV_FOREACH(p, s->wanted_by) { + r = add_symlink(s->name, *p); + if (r < 0) + log_warning_errno(r, "Failed to create 'Wants' symlink to %s, ignoring: %m", *p); + } + + return 1; +} + +static bool usage_contains_reload(const char *line) { + return (strcasestr(line, "{reload|") || + strcasestr(line, "{reload}") || + strcasestr(line, "{reload\"") || + strcasestr(line, "|reload|") || + strcasestr(line, "|reload}") || + strcasestr(line, "|reload\"")); +} + +static char *sysv_translate_name(const char *name) { + _cleanup_free_ char *c = NULL; + char *res; + + c = strdup(name); + if (!c) + return NULL; + + res = endswith(c, ".sh"); + if (res) + *res = 0; + + if (unit_name_mangle(c, UNIT_NAME_NOGLOB, &res) < 0) + return NULL; + + return res; +} + +static int sysv_translate_facility(const char *name, const char *filename, char **ret) { + + /* We silently ignore the $ prefix here. According to the LSB + * spec it simply indicates whether something is a + * standardized name or a distribution-specific one. Since we + * just follow what already exists and do not introduce new + * uses or names we don't care who introduced a new name. */ + + static const char * const table[] = { + /* LSB defined facilities */ + "local_fs", NULL, + "network", SPECIAL_NETWORK_ONLINE_TARGET, + "named", SPECIAL_NSS_LOOKUP_TARGET, + "portmap", SPECIAL_RPCBIND_TARGET, + "remote_fs", SPECIAL_REMOTE_FS_TARGET, + "syslog", NULL, + "time", SPECIAL_TIME_SYNC_TARGET, + }; + + char *filename_no_sh, *e, *m; + const char *n; + unsigned i; + int r; + + assert(name); + assert(filename); + assert(ret); + + n = *name == '$' ? name + 1 : name; + + for (i = 0; i < ELEMENTSOF(table); i += 2) { + if (!streq(table[i], n)) + continue; + + if (!table[i+1]) + return 0; + + m = strdup(table[i+1]); + if (!m) + return log_oom(); + + *ret = m; + return 1; + } + + /* If we don't know this name, fallback heuristics to figure + * out whether something is a target or a service alias. */ + + /* Facilities starting with $ are most likely targets */ + if (*name == '$') { + r = unit_name_build(n, NULL, ".target", ret); + if (r < 0) + return log_error_errno(r, "Failed to build name: %m"); + + return r; + } + + /* Strip ".sh" suffix from file name for comparison */ + filename_no_sh = strdupa(filename); + e = endswith(filename_no_sh, ".sh"); + if (e) { + *e = '\0'; + filename = filename_no_sh; + } + + /* Names equaling the file name of the services are redundant */ + if (streq_ptr(n, filename)) + return 0; + + /* Everything else we assume to be normal service names */ + m = sysv_translate_name(n); + if (!m) + return log_oom(); + + *ret = m; + return 1; +} + +static int handle_provides(SysvStub *s, unsigned line, const char *full_text, const char *text) { + int r; + + assert(s); + assert(full_text); + assert(text); + + for (;;) { + _cleanup_free_ char *word = NULL, *m = NULL; + + r = extract_first_word(&text, &word, NULL, EXTRACT_QUOTES|EXTRACT_RELAX); + if (r < 0) + return log_error_errno(r, "Failed to parse word from provides string: %m"); + if (r == 0) + break; + + r = sysv_translate_facility(word, basename(s->path), &m); + if (r <= 0) /* continue on error */ + continue; + + switch (unit_name_to_type(m)) { + + case UNIT_SERVICE: + log_debug("Adding Provides: alias '%s' for '%s'", m, s->name); + r = add_alias(s->name, m); + if (r < 0) + log_warning_errno(r, "[%s:%u] Failed to add LSB Provides name %s, ignoring: %m", s->path, line, m); + break; + + case UNIT_TARGET: + + /* NB: SysV targets which are provided by a + * service are pulled in by the services, as + * an indication that the generic service is + * now available. This is strictly one-way. + * The targets do NOT pull in SysV services! */ + + r = strv_extend(&s->before, m); + if (r < 0) + return log_oom(); + + r = strv_extend(&s->wants, m); + if (r < 0) + return log_oom(); + + if (streq(m, SPECIAL_NETWORK_ONLINE_TARGET)) { + r = strv_extend(&s->before, SPECIAL_NETWORK_TARGET); + if (r < 0) + return log_oom(); + } + + break; + + case _UNIT_TYPE_INVALID: + log_warning("Unit name '%s' is invalid", m); + break; + + default: + log_warning("Unknown unit type for unit '%s'", m); + } + } + + return 0; +} + +static int handle_dependencies(SysvStub *s, unsigned line, const char *full_text, const char *text) { + int r; + + assert(s); + assert(full_text); + assert(text); + + for (;;) { + _cleanup_free_ char *word = NULL, *m = NULL; + bool is_before; + + r = extract_first_word(&text, &word, NULL, EXTRACT_QUOTES|EXTRACT_RELAX); + if (r < 0) + return log_error_errno(r, "Failed to parse word from provides string: %m"); + if (r == 0) + break; + + r = sysv_translate_facility(word, basename(s->path), &m); + if (r <= 0) /* continue on error */ + continue; + + is_before = startswith_no_case(full_text, "X-Start-Before:"); + + if (streq(m, SPECIAL_NETWORK_ONLINE_TARGET) && !is_before) { + /* the network-online target is special, as it needs to be actively pulled in */ + r = strv_extend(&s->after, m); + if (r < 0) + return log_oom(); + + r = strv_extend(&s->wants, m); + } else + r = strv_extend(is_before ? &s->before : &s->after, m); + if (r < 0) + return log_oom(); + } + + return 0; +} + +static int load_sysv(SysvStub *s) { + _cleanup_fclose_ FILE *f; + unsigned line = 0; + int r; + enum { + NORMAL, + DESCRIPTION, + LSB, + LSB_DESCRIPTION, + USAGE_CONTINUATION + } state = NORMAL; + _cleanup_free_ char *short_description = NULL, *long_description = NULL, *chkconfig_description = NULL; + char *description; + bool supports_reload = false; + char l[LINE_MAX]; + + assert(s); + + f = fopen(s->path, "re"); + if (!f) { + if (errno == ENOENT) + return 0; + + return log_error_errno(errno, "Failed to open %s: %m", s->path); + } + + log_debug("Loading SysV script %s", s->path); + + FOREACH_LINE(l, f, goto fail) { + char *t; + + line++; + + t = strstrip(l); + if (*t != '#') { + /* Try to figure out whether this init script supports + * the reload operation. This heuristic looks for + * "Usage" lines which include the reload option. */ + if ( state == USAGE_CONTINUATION || + (state == NORMAL && strcasestr(t, "usage"))) { + if (usage_contains_reload(t)) { + supports_reload = true; + state = NORMAL; + } else if (t[strlen(t)-1] == '\\') + state = USAGE_CONTINUATION; + else + state = NORMAL; + } + + continue; + } + + if (state == NORMAL && streq(t, "### BEGIN INIT INFO")) { + state = LSB; + s->has_lsb = true; + continue; + } + + if ((state == LSB_DESCRIPTION || state == LSB) && streq(t, "### END INIT INFO")) { + state = NORMAL; + continue; + } + + t++; + t += strspn(t, WHITESPACE); + + if (state == NORMAL) { + + /* Try to parse Red Hat style description */ + + if (startswith_no_case(t, "description:")) { + + size_t k; + const char *j; + + k = strlen(t); + if (k > 0 && t[k-1] == '\\') { + state = DESCRIPTION; + t[k-1] = 0; + } + + j = empty_to_null(strstrip(t+12)); + + r = free_and_strdup(&chkconfig_description, j); + if (r < 0) + return log_oom(); + + } else if (startswith_no_case(t, "pidfile:")) { + const char *fn; + + state = NORMAL; + + fn = strstrip(t+8); + if (!path_is_absolute(fn)) { + log_error("[%s:%u] PID file not absolute. Ignoring.", s->path, line); + continue; + } + + r = free_and_strdup(&s->pid_file, fn); + if (r < 0) + return log_oom(); + } + + } else if (state == DESCRIPTION) { + + /* Try to parse Red Hat style description + * continuation */ + + size_t k; + char *j; + + k = strlen(t); + if (k > 0 && t[k-1] == '\\') + t[k-1] = 0; + else + state = NORMAL; + + j = strstrip(t); + if (!isempty(j)) { + char *d = NULL; + + if (chkconfig_description) + d = strjoin(chkconfig_description, " ", j, NULL); + else + d = strdup(j); + if (!d) + return log_oom(); + + free(chkconfig_description); + chkconfig_description = d; + } + + } else if (state == LSB || state == LSB_DESCRIPTION) { + + if (startswith_no_case(t, "Provides:")) { + state = LSB; + + r = handle_provides(s, line, t, t + 9); + if (r < 0) + return r; + + } else if (startswith_no_case(t, "Required-Start:") || + startswith_no_case(t, "Should-Start:") || + startswith_no_case(t, "X-Start-Before:") || + startswith_no_case(t, "X-Start-After:")) { + + state = LSB; + + r = handle_dependencies(s, line, t, strchr(t, ':') + 1); + if (r < 0) + return r; + + } else if (startswith_no_case(t, "Description:")) { + const char *j; + + state = LSB_DESCRIPTION; + + j = empty_to_null(strstrip(t+12)); + + r = free_and_strdup(&long_description, j); + if (r < 0) + return log_oom(); + + } else if (startswith_no_case(t, "Short-Description:")) { + const char *j; + + state = LSB; + + j = empty_to_null(strstrip(t+18)); + + r = free_and_strdup(&short_description, j); + if (r < 0) + return log_oom(); + + } else if (state == LSB_DESCRIPTION) { + + if (startswith(l, "#\t") || startswith(l, "# ")) { + const char *j; + + j = strstrip(t); + if (!isempty(j)) { + char *d = NULL; + + if (long_description) + d = strjoin(long_description, " ", t, NULL); + else + d = strdup(j); + if (!d) + return log_oom(); + + free(long_description); + long_description = d; + } + + } else + state = LSB; + } + } + } + + s->reload = supports_reload; + + /* We use the long description only if + * no short description is set. */ + + if (short_description) + description = short_description; + else if (chkconfig_description) + description = chkconfig_description; + else if (long_description) + description = long_description; + else + description = NULL; + + if (description) { + char *d; + + d = strappend(s->has_lsb ? "LSB: " : "SYSV: ", description); + if (!d) + return log_oom(); + + s->description = d; + } + + s->loaded = true; + return 0; + +fail: + return log_error_errno(errno, "Failed to read configuration file '%s': %m", s->path); +} + +static int fix_order(SysvStub *s, Hashmap *all_services) { + SysvStub *other; + Iterator j; + int r; + + assert(s); + + if (!s->loaded) + return 0; + + if (s->sysv_start_priority < 0) + return 0; + + HASHMAP_FOREACH(other, all_services, j) { + if (s == other) + continue; + + if (!other->loaded) + continue; + + if (other->sysv_start_priority < 0) + continue; + + /* If both units have modern headers we don't care + * about the priorities */ + if (s->has_lsb && other->has_lsb) + continue; + + if (other->sysv_start_priority < s->sysv_start_priority) { + r = strv_extend(&s->after, other->name); + if (r < 0) + return log_oom(); + + } else if (other->sysv_start_priority > s->sysv_start_priority) { + r = strv_extend(&s->before, other->name); + if (r < 0) + return log_oom(); + } else + continue; + + /* FIXME: Maybe we should compare the name here lexicographically? */ + } + + return 0; +} + +static int acquire_search_path(const char *def, const char *envvar, char ***ret) { + _cleanup_strv_free_ char **l = NULL; + const char *e; + int r; + + assert(def); + assert(envvar); + + e = getenv(envvar); + if (e) { + r = path_split_and_make_absolute(e, &l); + if (r < 0) + return log_error_errno(r, "Failed to make $%s search path absolute: %m", envvar); + } + + if (strv_isempty(l)) { + strv_free(l); + + l = strv_new(def, NULL); + if (!l) + return log_oom(); + } + + if (!path_strv_resolve_uniq(l, NULL)) + return log_oom(); + + *ret = l; + l = NULL; + + return 0; +} + +static int enumerate_sysv(const LookupPaths *lp, Hashmap *all_services) { + _cleanup_strv_free_ char **sysvinit_path = NULL; + char **path; + int r; + + assert(lp); + + r = acquire_search_path(SYSTEM_SYSVINIT_PATH, "SYSTEMD_SYSVINIT_PATH", &sysvinit_path); + if (r < 0) + return r; + + STRV_FOREACH(path, sysvinit_path) { + _cleanup_closedir_ DIR *d = NULL; + struct dirent *de; + + d = opendir(*path); + if (!d) { + if (errno != ENOENT) + log_warning_errno(errno, "Opening %s failed, ignoring: %m", *path); + continue; + } + + FOREACH_DIRENT(de, d, log_error_errno(errno, "Failed to enumerate directory %s, ignoring: %m", *path)) { + _cleanup_free_ char *fpath = NULL, *name = NULL; + _cleanup_(free_sysvstubp) SysvStub *service = NULL; + struct stat st; + + if (fstatat(dirfd(d), de->d_name, &st, 0) < 0) { + log_warning_errno(errno, "stat() failed on %s/%s, ignoring: %m", *path, de->d_name); + continue; + } + + if (!(st.st_mode & S_IXUSR)) + continue; + + if (!S_ISREG(st.st_mode)) + continue; + + name = sysv_translate_name(de->d_name); + if (!name) + return log_oom(); + + if (hashmap_contains(all_services, name)) + continue; + + r = unit_file_exists(UNIT_FILE_SYSTEM, lp, name); + if (r < 0 && !IN_SET(r, -ELOOP, -ERFKILL, -EADDRNOTAVAIL)) { + log_debug_errno(r, "Failed to detect whether %s exists, skipping: %m", name); + continue; + } else if (r != 0) { + log_debug("Native unit for %s already exists, skipping.", name); + continue; + } + + fpath = strjoin(*path, "/", de->d_name, NULL); + if (!fpath) + return log_oom(); + + service = new0(SysvStub, 1); + if (!service) + return log_oom(); + + service->sysv_start_priority = -1; + service->name = name; + service->path = fpath; + name = fpath = NULL; + + r = hashmap_put(all_services, service->name, service); + if (r < 0) + return log_oom(); + + service = NULL; + } + } + + return 0; +} + +static int set_dependencies_from_rcnd(const LookupPaths *lp, Hashmap *all_services) { + Set *runlevel_services[ELEMENTSOF(rcnd_table)] = {}; + _cleanup_strv_free_ char **sysvrcnd_path = NULL; + SysvStub *service; + unsigned i; + Iterator j; + char **p; + int r; + + assert(lp); + + r = acquire_search_path(SYSTEM_SYSVRCND_PATH, "SYSTEMD_SYSVRCND_PATH", &sysvrcnd_path); + if (r < 0) + return r; + + STRV_FOREACH(p, sysvrcnd_path) { + for (i = 0; i < ELEMENTSOF(rcnd_table); i ++) { + + _cleanup_closedir_ DIR *d = NULL; + _cleanup_free_ char *path = NULL; + struct dirent *de; + + path = strjoin(*p, "/", rcnd_table[i].path, NULL); + if (!path) { + r = log_oom(); + goto finish; + } + + d = opendir(path); + if (!d) { + if (errno != ENOENT) + log_warning_errno(errno, "Opening %s failed, ignoring: %m", path); + + continue; + } + + FOREACH_DIRENT(de, d, log_error_errno(errno, "Failed to enumerate directory %s, ignoring: %m", path)) { + _cleanup_free_ char *name = NULL, *fpath = NULL; + int a, b; + + if (de->d_name[0] != 'S') + continue; + + if (strlen(de->d_name) < 4) + continue; + + a = undecchar(de->d_name[1]); + b = undecchar(de->d_name[2]); + + if (a < 0 || b < 0) + continue; + + fpath = strjoin(*p, "/", de->d_name, NULL); + if (!fpath) { + r = log_oom(); + goto finish; + } + + name = sysv_translate_name(de->d_name + 3); + if (!name) { + r = log_oom(); + goto finish; + } + + service = hashmap_get(all_services, name); + if (!service) { + log_debug("Ignoring %s symlink in %s, not generating %s.", de->d_name, rcnd_table[i].path, name); + continue; + } + + service->sysv_start_priority = MAX(a*10 + b, service->sysv_start_priority); + + r = set_ensure_allocated(&runlevel_services[i], NULL); + if (r < 0) { + log_oom(); + goto finish; + } + + r = set_put(runlevel_services[i], service); + if (r < 0) { + log_oom(); + goto finish; + } + } + } + } + + for (i = 0; i < ELEMENTSOF(rcnd_table); i ++) + SET_FOREACH(service, runlevel_services[i], j) { + r = strv_extend(&service->before, rcnd_table[i].target); + if (r < 0) { + log_oom(); + goto finish; + } + r = strv_extend(&service->wanted_by, rcnd_table[i].target); + if (r < 0) { + log_oom(); + goto finish; + } + } + + r = 0; + +finish: + for (i = 0; i < ELEMENTSOF(rcnd_table); i++) + set_free(runlevel_services[i]); + + return r; +} + +int main(int argc, char *argv[]) { + _cleanup_(free_sysvstub_hashmapp) Hashmap *all_services = NULL; + _cleanup_lookup_paths_free_ LookupPaths lp = {}; + SysvStub *service; + Iterator j; + int r; + + if (argc > 1 && argc != 4) { + log_error("This program takes three or no arguments."); + return EXIT_FAILURE; + } + + if (argc > 1) + arg_dest = argv[3]; + + log_set_target(LOG_TARGET_SAFE); + log_parse_environment(); + log_open(); + + umask(0022); + + r = lookup_paths_init(&lp, UNIT_FILE_SYSTEM, LOOKUP_PATHS_EXCLUDE_GENERATED, NULL); + if (r < 0) { + log_error_errno(r, "Failed to find lookup paths: %m"); + goto finish; + } + + all_services = hashmap_new(&string_hash_ops); + if (!all_services) { + r = log_oom(); + goto finish; + } + + r = enumerate_sysv(&lp, all_services); + if (r < 0) + goto finish; + + r = set_dependencies_from_rcnd(&lp, all_services); + if (r < 0) + goto finish; + + HASHMAP_FOREACH(service, all_services, j) + (void) load_sysv(service); + + HASHMAP_FOREACH(service, all_services, j) { + (void) fix_order(service, all_services); + (void) generate_unit_file(service); + } + + r = 0; + +finish: + return r < 0 ? EXIT_FAILURE : EXIT_SUCCESS; +} diff --git a/src/grp-system/kernel-command-line.xml b/src/grp-system/kernel-command-line.xml new file mode 100644 index 0000000000..3ecc969c10 --- /dev/null +++ b/src/grp-system/kernel-command-line.xml @@ -0,0 +1,384 @@ + + + + + + + + + kernel-command-line + systemd + + + + Developer + Lennart + Poettering + lennart@poettering.net + + + + + + kernel-command-line + 7 + + + + kernel-command-line + Kernel command line parameters + + + + /proc/cmdline + + + + Description + + The kernel, the initial RAM disk (initrd) and + basic userspace functionality may be configured at boot via + kernel command line arguments. + + For command line parameters understood by the kernel, please + see kernel-parameters.txt + and + bootparam7. + + For command line parameters understood by the initial RAM + disk, please see + dracut.cmdline7, + or the documentation of the specific initrd implementation of your + installation. + + + + Core OS Command Line Arguments + + + + systemd.unit= + rd.systemd.unit= + systemd.dump_core= + systemd.crash_chvt= + systemd.crash_shell= + systemd.crash_reboot= + systemd.confirm_spawn= + systemd.show_status= + systemd.log_target= + systemd.log_level= + systemd.log_color= + systemd.log_location= + systemd.default_standard_output= + systemd.default_standard_error= + systemd.setenv= + systemd.machine_id= + + Parameters understood by the system and service + manager to control system behavior. For details, see + systemd1. + + + + + systemd.mask= + systemd.wants= + systemd.debug-shell + + Additional parameters understood by + systemd-debug-generator8, + to mask or start specific units at boot, or invoke a debug + shell on tty9. + + + + + systemd.restore_state= + + This parameter is understood by several system tools + to control whether or not they should restore system state + from the previous boot. For details, see + systemd-backlight@.service8 + and + systemd-rfkill.service8. + + + + + + quiet + + Parameter understood by both the kernel and the system + and service manager to control console log verbosity. For + details, see + systemd1. + + + + + debug + + Parameter understood by both the kernel and the system + and service manager to control console log verbosity. For + details, see + systemd1. + + + + + -b + rd.emergency + emergency + rd.rescue + rescue + single + s + S + 1 + 2 + 3 + 4 + 5 + + Parameters understood by the system and service + manager, as compatibility and convenience options. For details, see + systemd1. + + + + + locale.LANG= + locale.LANGUAGE= + locale.LC_CTYPE= + locale.LC_NUMERIC= + locale.LC_TIME= + locale.LC_COLLATE= + locale.LC_MONETARY= + locale.LC_MESSAGES= + locale.LC_PAPER= + locale.LC_NAME= + locale.LC_ADDRESS= + locale.LC_TELEPHONE= + locale.LC_MEASUREMENT= + locale.LC_IDENTIFICATION= + + Parameters understood by the system and service + manager to control locale and language settings. For + details, see + systemd1. + + + + + fsck.mode= + fsck.repair= + + + Parameters understood by the file system checker + services. For details, see + systemd-fsck@.service8. + + + + + quotacheck.mode= + + + Parameter understood by the file quota checker + service. For details, see + systemd-quotacheck.service8. + + + + + systemd.journald.forward_to_syslog= + systemd.journald.forward_to_kmsg= + systemd.journald.forward_to_console= + systemd.journald.forward_to_wall= + + + Parameters understood by the journal service. For + details, see + systemd-journald.service8. + + + + + vconsole.keymap= + vconsole.keymap.toggle= + vconsole.font= + vconsole.font.map= + vconsole.font.unimap= + + + Parameters understood by the virtual console setup + logic. For details, see + systemd-vconsole-setup.service8. + + + + + udev.log-priority= + rd.udev.log-priority= + udev.children-max= + rd.udev.children-max= + udev.exec-delay= + rd.udev.exec-delay= + udev.event-timeout= + rd.udev.event-timeout= + net.ifnames= + + + Parameters understood by the device event managing + daemon. For details, see + systemd-udevd.service8. + + + + + plymouth.enable= + + + May be used to disable the Plymouth boot splash. For + details, see + plymouth8. + + + + + luks= + rd.luks= + luks.crypttab= + rd.luks.crypttab= + luks.name= + rd.luks.name= + luks.uuid= + rd.luks.uuid= + luks.options= + rd.luks.options= + luks.key= + rd.luks.key= + + + Configures the LUKS full-disk encryption logic at + boot. For details, see + systemd-cryptsetup-generator8. + + + + + fstab= + rd.fstab= + + + Configures the /etc/fstab logic + at boot. For details, see + systemd-fstab-generator8. + + + + + root= + rootfstype= + rootflags= + ro + rw + + + Configures the root file system and its file system + type and mount options, as well as whether it shall be + mounted read-only or read-writable initially. For details, + see + systemd-fstab-generator8. + + + + + systemd.gpt_auto= + rd.systemd.gpt_auto= + + + Configures whether GPT based partition auto-discovery + shall be attempted. For details, see + systemd-gpt-auto-generator8. + + + + + systemd.default_timeout_start_sec= + + + Overwrites the default start job timeout DefaultTimeoutStartSec= at boot. For details, + see systemd-system.conf5. + + + + + modules-load= + rd.modules-load= + + + Load a specific kernel module early at boot. For + details, see + systemd-modules-load.service8. + + + + + resume= + + + Enables resume from hibernation using the specified + device. All + fstab5-like + paths are supported. For details, see + systemd-hibernate-resume-generator8. + + + + + + + + See Also + + systemd1, + bootparam7, + dracut.cmdline7, + systemd-debug-generator8, + systemd-fsck@.service8, + systemd-quotacheck.service8, + systemd-journald.service8, + systemd-vconsole-setup.service8, + systemd-udevd.service8, + plymouth8, + systemd-cryptsetup-generator8, + systemd-fstab-generator8, + systemd-gpt-auto-generator8, + systemd-modules-load.service8, + systemd-backlight@.service8, + systemd-rfkill.service8, + systemd-hibernate-resume-generator8 + + + + diff --git a/src/grp-system/libcore/.gitignore b/src/grp-system/libcore/.gitignore new file mode 100644 index 0000000000..465b4fcc20 --- /dev/null +++ b/src/grp-system/libcore/.gitignore @@ -0,0 +1,3 @@ +/macros.systemd +/triggers.systemd +/systemd.pc diff --git a/src/grp-system/libcore/Makefile b/src/grp-system/libcore/Makefile new file mode 100644 index 0000000000..9468b89e74 --- /dev/null +++ b/src/grp-system/libcore/Makefile @@ -0,0 +1,170 @@ +# -*- Mode: makefile; indent-tabs-mode: t -*- +# +# This file is part of systemd. +# +# Copyright 2010-2012 Lennart Poettering +# Copyright 2010-2012 Kay Sievers +# Copyright 2013 Zbigniew Jędrzejewski-Szmek +# Copyright 2013 David Strauss +# Copyright 2016 Luke Shumaker +# +# systemd is free software; you can redistribute it and/or modify it +# under the terms of the GNU Lesser General Public License as published by +# the Free Software Foundation; either version 2.1 of the License, or +# (at your option) any later version. +# +# systemd is distributed in the hope that it will be useful, but +# WITHOUT ANY WARRANTY; without even the implied warranty of +# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU +# Lesser General Public License for more details. +# +# You should have received a copy of the GNU Lesser General Public License +# along with systemd; If not, see . +include $(dir $(lastword $(MAKEFILE_LIST)))/../../../config.mk +include $(topsrcdir)/build-aux/Makefile.head.mk + +noinst_LTLIBRARIES += \ + libcore.la + +libcore_la_SOURCES = \ + src/core/unit.c \ + src/core/unit.h \ + src/core/unit-printf.c \ + src/core/unit-printf.h \ + src/core/job.c \ + src/core/job.h \ + src/core/manager.c \ + src/core/manager.h \ + src/core/transaction.c \ + src/core/transaction.h \ + src/core/load-fragment.c \ + src/core/load-fragment.h \ + src/core/service.c \ + src/core/service.h \ + src/core/socket.c \ + src/core/socket.h \ + src/core/busname.c \ + src/core/busname.h \ + src/core/bus-policy.c \ + src/core/bus-policy.h \ + src/core/target.c \ + src/core/target.h \ + src/core/device.c \ + src/core/device.h \ + src/core/mount.c \ + src/core/mount.h \ + src/core/automount.c \ + src/core/automount.h \ + src/core/swap.c \ + src/core/swap.h \ + src/core/timer.c \ + src/core/timer.h \ + src/core/path.c \ + src/core/path.h \ + src/core/slice.c \ + src/core/slice.h \ + src/core/scope.c \ + src/core/scope.h \ + src/core/load-dropin.c \ + src/core/load-dropin.h \ + src/core/execute.c \ + src/core/execute.h \ + src/core/kill.c \ + src/core/kill.h \ + src/core/dbus.c \ + src/core/dbus.h \ + src/core/dbus-manager.c \ + src/core/dbus-manager.h \ + src/core/dbus-unit.c \ + src/core/dbus-unit.h \ + src/core/dbus-job.c \ + src/core/dbus-job.h \ + src/core/dbus-service.c \ + src/core/dbus-service.h \ + src/core/dbus-socket.c \ + src/core/dbus-socket.h \ + src/core/dbus-busname.c \ + src/core/dbus-busname.h \ + src/core/dbus-target.c \ + src/core/dbus-target.h \ + src/core/dbus-device.c \ + src/core/dbus-device.h \ + src/core/dbus-mount.c \ + src/core/dbus-mount.h \ + src/core/dbus-automount.c \ + src/core/dbus-automount.h \ + src/core/dbus-swap.c \ + src/core/dbus-swap.h \ + src/core/dbus-timer.c \ + src/core/dbus-timer.h \ + src/core/dbus-path.c \ + src/core/dbus-path.h \ + src/core/dbus-slice.c \ + src/core/dbus-slice.h \ + src/core/dbus-scope.c \ + src/core/dbus-scope.h \ + src/core/dbus-execute.c \ + src/core/dbus-execute.h \ + src/core/dbus-kill.c \ + src/core/dbus-kill.h \ + src/core/dbus-cgroup.c \ + src/core/dbus-cgroup.h \ + src/core/cgroup.c \ + src/core/cgroup.h \ + src/core/selinux-access.c \ + src/core/selinux-access.h \ + src/core/selinux-setup.c \ + src/core/selinux-setup.h \ + src/core/smack-setup.c \ + src/core/smack-setup.h \ + src/core/ima-setup.c \ + src/core/ima-setup.h \ + src/core/locale-setup.h \ + src/core/locale-setup.c \ + src/core/hostname-setup.c \ + src/core/hostname-setup.h \ + src/core/machine-id-setup.c \ + src/core/machine-id-setup.h \ + src/core/mount-setup.c \ + src/core/mount-setup.h \ + src/core/kmod-setup.c \ + src/core/kmod-setup.h \ + src/core/loopback-setup.h \ + src/core/loopback-setup.c \ + src/core/namespace.c \ + src/core/namespace.h \ + src/core/killall.h \ + src/core/killall.c \ + src/core/audit-fd.c \ + src/core/audit-fd.h \ + src/core/show-status.c \ + src/core/show-status.h \ + src/core/failure-action.c \ + src/core/failure-action.h + +nodist_libcore_la_SOURCES = \ + src/core/load-fragment-gperf.c \ + src/core/load-fragment-gperf-nulstr.c + +libcore_la_CFLAGS = \ + $(PAM_CFLAGS) \ + $(AUDIT_CFLAGS) \ + $(KMOD_CFLAGS) \ + $(APPARMOR_CFLAGS) \ + $(MOUNT_CFLAGS) \ + $(SECCOMP_CFLAGS) + +libcore_la_LIBADD = \ + libsystemd-shared.la \ + $(PAM_LIBS) \ + $(AUDIT_LIBS) \ + $(KMOD_LIBS) \ + $(APPARMOR_LIBS) \ + $(MOUNT_LIBS) + +$(outdir)/load-fragment-gperf-nulstr.c: src/core/load-fragment-gperf.gperf + $(AM_V_GEN)$(AWK) 'BEGIN{ keywords=0 ; FS="," ; print "extern const char load_fragment_gperf_nulstr[];" ; print "const char load_fragment_gperf_nulstr[] ="} ; keyword==1 { print "\"" $$1 "\\0\"" } ; /%%/ { keyword=1} ; END { print ";" }' < $< > $@ + +nested.subdirs += src + +include $(topsrcdir)/build-aux/Makefile.tail.mk diff --git a/src/grp-system/libcore/include/core/automount.h b/src/grp-system/libcore/include/core/automount.h new file mode 100644 index 0000000000..76a201178e --- /dev/null +++ b/src/grp-system/libcore/include/core/automount.h @@ -0,0 +1,59 @@ +#pragma once + +/*** + This file is part of systemd. + + Copyright 2010 Lennart Poettering + + systemd is free software; you can redistribute it and/or modify it + under the terms of the GNU Lesser General Public License as published by + the Free Software Foundation; either version 2.1 of the License, or + (at your option) any later version. + + systemd is distributed in the hope that it will be useful, but + WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + Lesser General Public License for more details. + + You should have received a copy of the GNU Lesser General Public License + along with systemd; If not, see . +***/ + +typedef struct Automount Automount; + +#include "unit.h" + +typedef enum AutomountResult { + AUTOMOUNT_SUCCESS, + AUTOMOUNT_FAILURE_RESOURCES, + AUTOMOUNT_FAILURE_START_LIMIT_HIT, + AUTOMOUNT_FAILURE_MOUNT_START_LIMIT_HIT, + _AUTOMOUNT_RESULT_MAX, + _AUTOMOUNT_RESULT_INVALID = -1 +} AutomountResult; + +struct Automount { + Unit meta; + + AutomountState state, deserialized_state; + + char *where; + usec_t timeout_idle_usec; + + int pipe_fd; + sd_event_source *pipe_event_source; + mode_t directory_mode; + dev_t dev_id; + + Set *tokens; + Set *expire_tokens; + + sd_event_source *expire_event_source; + + AutomountResult result; +}; + +extern const UnitVTable automount_vtable; + +const char* automount_result_to_string(AutomountResult i) _const_; +AutomountResult automount_result_from_string(const char *s) _pure_; diff --git a/src/grp-system/libcore/include/core/bus-policy.h b/src/grp-system/libcore/include/core/bus-policy.h new file mode 100644 index 0000000000..a338f29af6 --- /dev/null +++ b/src/grp-system/libcore/include/core/bus-policy.h @@ -0,0 +1,64 @@ +#pragma once + +/*** + This file is part of systemd. + + Copyright 2014 Daniel Mack + + systemd is free software; you can redistribute it and/or modify it + under the terms of the GNU Lesser General Public License as published by + the Free Software Foundation; either version 2.1 of the License, or + (at your option) any later version. + + systemd is distributed in the hope that it will be useful, but + WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + Lesser General Public License for more details. + + You should have received a copy of the GNU Lesser General Public License + along with systemd; If not, see . +***/ + +#include "sd-bus/kdbus.h" +#include "systemd-basic/list.h" +#include "systemd-basic/macro.h" + +typedef struct BusNamePolicy BusNamePolicy; + +typedef enum BusPolicyAccess { + BUS_POLICY_ACCESS_SEE, + BUS_POLICY_ACCESS_TALK, + BUS_POLICY_ACCESS_OWN, + _BUS_POLICY_ACCESS_MAX, + _BUS_POLICY_ACCESS_INVALID = -1 +} BusPolicyAccess; + +typedef enum BusNamePolicyType { + BUSNAME_POLICY_TYPE_USER, + BUSNAME_POLICY_TYPE_GROUP, + _BUSNAME_POLICY_TYPE_MAX, + _BUSNAME_POLICY_TYPE_INVALID = -1 +} BusNamePolicyType; + +struct BusNamePolicy { + BusNamePolicyType type; + BusPolicyAccess access; + + char *name; + + LIST_FIELDS(BusNamePolicy, policy); +}; + +int bus_kernel_translate_access(BusPolicyAccess access); +int bus_kernel_translate_policy(const BusNamePolicy *policy, struct kdbus_item *item); + +const char* bus_policy_access_to_string(BusPolicyAccess i) _const_; +BusPolicyAccess bus_policy_access_from_string(const char *s) _pure_; + +int bus_kernel_make_starter( + int fd, + const char *name, + bool activating, + bool accept_fd, + BusNamePolicy *policy, + BusPolicyAccess world_policy); diff --git a/src/grp-system/libcore/include/core/busname.h b/src/grp-system/libcore/include/core/busname.h new file mode 100644 index 0000000000..aa7f0ecb1b --- /dev/null +++ b/src/grp-system/libcore/include/core/busname.h @@ -0,0 +1,69 @@ +#pragma once + +/*** + This file is part of systemd. + + Copyright 2013 Lennart Poettering + + systemd is free software; you can redistribute it and/or modify it + under the terms of the GNU Lesser General Public License as published by + the Free Software Foundation; either version 2.1 of the License, or + (at your option) any later version. + + systemd is distributed in the hope that it will be useful, but + WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + Lesser General Public License for more details. + + You should have received a copy of the GNU Lesser General Public License + along with systemd; If not, see . +***/ + +typedef struct BusName BusName; +typedef struct BusNamePolicy BusNamePolicy; + +#include "bus-policy.h" +#include "unit.h" + +typedef enum BusNameResult { + BUSNAME_SUCCESS, + BUSNAME_FAILURE_RESOURCES, + BUSNAME_FAILURE_TIMEOUT, + BUSNAME_FAILURE_EXIT_CODE, + BUSNAME_FAILURE_SIGNAL, + BUSNAME_FAILURE_CORE_DUMP, + BUSNAME_FAILURE_START_LIMIT_HIT, + BUSNAME_FAILURE_SERVICE_START_LIMIT_HIT, + _BUSNAME_RESULT_MAX, + _BUSNAME_RESULT_INVALID = -1 +} BusNameResult; + +struct BusName { + Unit meta; + + char *name; + int starter_fd; + + bool activating; + bool accept_fd; + + UnitRef service; + + BusNameState state, deserialized_state; + BusNameResult result; + + usec_t timeout_usec; + + sd_event_source *starter_event_source; + sd_event_source *timer_event_source; + + pid_t control_pid; + + LIST_HEAD(BusNamePolicy, policy); + BusPolicyAccess policy_world; +}; + +extern const UnitVTable busname_vtable; + +const char* busname_result_to_string(BusNameResult i) _const_; +BusNameResult busname_result_from_string(const char *s) _pure_; diff --git a/src/grp-system/libcore/include/core/cgroup.h b/src/grp-system/libcore/include/core/cgroup.h new file mode 100644 index 0000000000..0026609570 --- /dev/null +++ b/src/grp-system/libcore/include/core/cgroup.h @@ -0,0 +1,183 @@ +#pragma once + +/*** + This file is part of systemd. + + Copyright 2013 Lennart Poettering + + systemd is free software; you can redistribute it and/or modify it + under the terms of the GNU Lesser General Public License as published by + the Free Software Foundation; either version 2.1 of the License, or + (at your option) any later version. + + systemd is distributed in the hope that it will be useful, but + WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + Lesser General Public License for more details. + + You should have received a copy of the GNU Lesser General Public License + along with systemd; If not, see . +***/ + +#include + +#include "systemd-basic/cgroup-util.h" +#include "systemd-basic/list.h" +#include "systemd-basic/time-util.h" + +typedef struct CGroupContext CGroupContext; +typedef struct CGroupDeviceAllow CGroupDeviceAllow; +typedef struct CGroupIODeviceWeight CGroupIODeviceWeight; +typedef struct CGroupIODeviceLimit CGroupIODeviceLimit; +typedef struct CGroupBlockIODeviceWeight CGroupBlockIODeviceWeight; +typedef struct CGroupBlockIODeviceBandwidth CGroupBlockIODeviceBandwidth; + +typedef enum CGroupDevicePolicy { + + /* When devices listed, will allow those, plus built-in ones, + if none are listed will allow everything. */ + CGROUP_AUTO, + + /* Everything forbidden, except built-in ones and listed ones. */ + CGROUP_CLOSED, + + /* Everythings forbidden, except for the listed devices */ + CGROUP_STRICT, + + _CGROUP_DEVICE_POLICY_MAX, + _CGROUP_DEVICE_POLICY_INVALID = -1 +} CGroupDevicePolicy; + +struct CGroupDeviceAllow { + LIST_FIELDS(CGroupDeviceAllow, device_allow); + char *path; + bool r:1; + bool w:1; + bool m:1; +}; + +struct CGroupIODeviceWeight { + LIST_FIELDS(CGroupIODeviceWeight, device_weights); + char *path; + uint64_t weight; +}; + +struct CGroupIODeviceLimit { + LIST_FIELDS(CGroupIODeviceLimit, device_limits); + char *path; + uint64_t limits[_CGROUP_IO_LIMIT_TYPE_MAX]; +}; + +struct CGroupBlockIODeviceWeight { + LIST_FIELDS(CGroupBlockIODeviceWeight, device_weights); + char *path; + uint64_t weight; +}; + +struct CGroupBlockIODeviceBandwidth { + LIST_FIELDS(CGroupBlockIODeviceBandwidth, device_bandwidths); + char *path; + uint64_t rbps; + uint64_t wbps; +}; + +struct CGroupContext { + bool cpu_accounting; + bool io_accounting; + bool blockio_accounting; + bool memory_accounting; + bool tasks_accounting; + + /* For unified hierarchy */ + uint64_t io_weight; + uint64_t startup_io_weight; + LIST_HEAD(CGroupIODeviceWeight, io_device_weights); + LIST_HEAD(CGroupIODeviceLimit, io_device_limits); + + uint64_t memory_low; + uint64_t memory_high; + uint64_t memory_max; + + /* For legacy hierarchies */ + uint64_t cpu_shares; + uint64_t startup_cpu_shares; + usec_t cpu_quota_per_sec_usec; + + uint64_t blockio_weight; + uint64_t startup_blockio_weight; + LIST_HEAD(CGroupBlockIODeviceWeight, blockio_device_weights); + LIST_HEAD(CGroupBlockIODeviceBandwidth, blockio_device_bandwidths); + + uint64_t memory_limit; + + CGroupDevicePolicy device_policy; + LIST_HEAD(CGroupDeviceAllow, device_allow); + + /* Common */ + uint64_t tasks_max; + + bool delegate; +}; + +#include "unit.h" + +void cgroup_context_init(CGroupContext *c); +void cgroup_context_done(CGroupContext *c); +void cgroup_context_dump(CGroupContext *c, FILE* f, const char *prefix); + +CGroupMask cgroup_context_get_mask(CGroupContext *c); + +void cgroup_context_free_device_allow(CGroupContext *c, CGroupDeviceAllow *a); +void cgroup_context_free_io_device_weight(CGroupContext *c, CGroupIODeviceWeight *w); +void cgroup_context_free_io_device_limit(CGroupContext *c, CGroupIODeviceLimit *l); +void cgroup_context_free_blockio_device_weight(CGroupContext *c, CGroupBlockIODeviceWeight *w); +void cgroup_context_free_blockio_device_bandwidth(CGroupContext *c, CGroupBlockIODeviceBandwidth *b); + +CGroupMask unit_get_own_mask(Unit *u); +CGroupMask unit_get_siblings_mask(Unit *u); +CGroupMask unit_get_members_mask(Unit *u); +CGroupMask unit_get_subtree_mask(Unit *u); + +CGroupMask unit_get_target_mask(Unit *u); +CGroupMask unit_get_enable_mask(Unit *u); + +void unit_update_cgroup_members_masks(Unit *u); + +char *unit_default_cgroup_path(Unit *u); +int unit_set_cgroup_path(Unit *u, const char *path); + +int unit_realize_cgroup(Unit *u); +void unit_release_cgroup(Unit *u); +void unit_prune_cgroup(Unit *u); +int unit_watch_cgroup(Unit *u); + +int unit_attach_pids_to_cgroup(Unit *u); + +int manager_setup_cgroup(Manager *m); +void manager_shutdown_cgroup(Manager *m, bool delete); + +unsigned manager_dispatch_cgroup_queue(Manager *m); + +Unit *manager_get_unit_by_cgroup(Manager *m, const char *cgroup); +Unit *manager_get_unit_by_pid_cgroup(Manager *m, pid_t pid); +Unit* manager_get_unit_by_pid(Manager *m, pid_t pid); + +int unit_search_main_pid(Unit *u, pid_t *ret); +int unit_watch_all_pids(Unit *u); + +int unit_get_memory_current(Unit *u, uint64_t *ret); +int unit_get_tasks_current(Unit *u, uint64_t *ret); +int unit_get_cpu_usage(Unit *u, nsec_t *ret); +int unit_reset_cpu_usage(Unit *u); + +bool unit_cgroup_delegate(Unit *u); + +int unit_notify_cgroup_empty(Unit *u); +int manager_notify_cgroup_empty(Manager *m, const char *group); + +void unit_invalidate_cgroup(Unit *u, CGroupMask m); + +void manager_invalidate_startup_units(Manager *m); + +const char* cgroup_device_policy_to_string(CGroupDevicePolicy i) _const_; +CGroupDevicePolicy cgroup_device_policy_from_string(const char *s) _pure_; diff --git a/src/grp-system/libcore/include/core/dbus-manager.h b/src/grp-system/libcore/include/core/dbus-manager.h new file mode 100644 index 0000000000..36a2e9481b --- /dev/null +++ b/src/grp-system/libcore/include/core/dbus-manager.h @@ -0,0 +1,28 @@ +#pragma once + +/*** + This file is part of systemd. + + Copyright 2010 Lennart Poettering + + systemd is free software; you can redistribute it and/or modify it + under the terms of the GNU Lesser General Public License as published by + the Free Software Foundation; either version 2.1 of the License, or + (at your option) any later version. + + systemd is distributed in the hope that it will be useful, but + WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + Lesser General Public License for more details. + + You should have received a copy of the GNU Lesser General Public License + along with systemd; If not, see . +***/ + +#include "manager.h" + +extern const sd_bus_vtable bus_manager_vtable[]; + +void bus_manager_send_finished(Manager *m, usec_t firmware_usec, usec_t loader_usec, usec_t kernel_usec, usec_t initrd_usec, usec_t userspace_usec, usec_t total_usec); +void bus_manager_send_reloading(Manager *m, bool active); +void bus_manager_send_change_signal(Manager *m); diff --git a/src/grp-system/libcore/include/core/device.h b/src/grp-system/libcore/include/core/device.h new file mode 100644 index 0000000000..184a1a349b --- /dev/null +++ b/src/grp-system/libcore/include/core/device.h @@ -0,0 +1,47 @@ +#pragma once + +/*** + This file is part of systemd. + + Copyright 2010 Lennart Poettering + + systemd is free software; you can redistribute it and/or modify it + under the terms of the GNU Lesser General Public License as published by + the Free Software Foundation; either version 2.1 of the License, or + (at your option) any later version. + + systemd is distributed in the hope that it will be useful, but + WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + Lesser General Public License for more details. + + You should have received a copy of the GNU Lesser General Public License + along with systemd; If not, see . +***/ + +typedef struct Device Device; + +typedef enum DeviceFound { + DEVICE_NOT_FOUND = 0, + DEVICE_FOUND_UDEV = 1, + DEVICE_FOUND_MOUNT = 2, + DEVICE_FOUND_SWAP = 4, +} DeviceFound; + +struct Device { + Unit meta; + + char *sysfs; + DeviceFound found; + + /* In order to be able to distinguish dependencies on + different device nodes we might end up creating multiple + devices for the same sysfs path. We chain them up here. */ + LIST_FIELDS(struct Device, same_sysfs); + + DeviceState state, deserialized_state; +}; + +extern const UnitVTable device_vtable; + +int device_found_node(Manager *m, const char *node, bool add, DeviceFound found, bool now); diff --git a/src/grp-system/libcore/include/core/execute.h b/src/grp-system/libcore/include/core/execute.h new file mode 100644 index 0000000000..178adc00ae --- /dev/null +++ b/src/grp-system/libcore/include/core/execute.h @@ -0,0 +1,292 @@ +#pragma once + +/*** + This file is part of systemd. + + Copyright 2010 Lennart Poettering + + systemd is free software; you can redistribute it and/or modify it + under the terms of the GNU Lesser General Public License as published by + the Free Software Foundation; either version 2.1 of the License, or + (at your option) any later version. + + systemd is distributed in the hope that it will be useful, but + WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + Lesser General Public License for more details. + + You should have received a copy of the GNU Lesser General Public License + along with systemd; If not, see . +***/ + +#include +#include +#include +#include + +#include "systemd-basic/cgroup-util.h" +#include "systemd-basic/list.h" +#include "systemd-basic/missing.h" +#include "systemd-shared/fdset.h" + +typedef struct ExecCommand ExecCommand; +typedef struct ExecContext ExecContext; +typedef struct ExecParameters ExecParameters; +typedef struct ExecRuntime ExecRuntime; +typedef struct ExecStatus ExecStatus; + +#include "namespace.h" + +typedef enum ExecUtmpMode { + EXEC_UTMP_INIT, + EXEC_UTMP_LOGIN, + EXEC_UTMP_USER, + _EXEC_UTMP_MODE_MAX, + _EXEC_UTMP_MODE_INVALID = -1 +} ExecUtmpMode; + +typedef enum ExecInput { + EXEC_INPUT_NULL, + EXEC_INPUT_TTY, + EXEC_INPUT_TTY_FORCE, + EXEC_INPUT_TTY_FAIL, + EXEC_INPUT_SOCKET, + _EXEC_INPUT_MAX, + _EXEC_INPUT_INVALID = -1 +} ExecInput; + +typedef enum ExecOutput { + EXEC_OUTPUT_INHERIT, + EXEC_OUTPUT_NULL, + EXEC_OUTPUT_TTY, + EXEC_OUTPUT_SYSLOG, + EXEC_OUTPUT_SYSLOG_AND_CONSOLE, + EXEC_OUTPUT_KMSG, + EXEC_OUTPUT_KMSG_AND_CONSOLE, + EXEC_OUTPUT_JOURNAL, + EXEC_OUTPUT_JOURNAL_AND_CONSOLE, + EXEC_OUTPUT_SOCKET, + _EXEC_OUTPUT_MAX, + _EXEC_OUTPUT_INVALID = -1 +} ExecOutput; + +struct ExecStatus { + dual_timestamp start_timestamp; + dual_timestamp exit_timestamp; + pid_t pid; + int code; /* as in siginfo_t::si_code */ + int status; /* as in sigingo_t::si_status */ +}; + +struct ExecCommand { + char *path; + char **argv; + ExecStatus exec_status; + LIST_FIELDS(ExecCommand, command); /* useful for chaining commands */ + bool ignore:1; + bool privileged:1; +}; + +struct ExecRuntime { + int n_ref; + + char *tmp_dir; + char *var_tmp_dir; + + int netns_storage_socket[2]; +}; + +struct ExecContext { + char **environment; + char **environment_files; + char **pass_environment; + + struct rlimit *rlimit[_RLIMIT_MAX]; + char *working_directory, *root_directory; + bool working_directory_missing_ok; + bool working_directory_home; + + mode_t umask; + int oom_score_adjust; + int nice; + int ioprio; + int cpu_sched_policy; + int cpu_sched_priority; + + cpu_set_t *cpuset; + unsigned cpuset_ncpus; + + ExecInput std_input; + ExecOutput std_output; + ExecOutput std_error; + + nsec_t timer_slack_nsec; + + bool stdio_as_fds; + + char *tty_path; + + bool tty_reset; + bool tty_vhangup; + bool tty_vt_disallocate; + + bool ignore_sigpipe; + + /* Since resolving these names might involve socket + * connections and we don't want to deadlock ourselves these + * names are resolved on execution only and in the child + * process. */ + char *user; + char *group; + char **supplementary_groups; + + char *pam_name; + + char *utmp_id; + ExecUtmpMode utmp_mode; + + bool selinux_context_ignore; + char *selinux_context; + + bool apparmor_profile_ignore; + char *apparmor_profile; + + bool smack_process_label_ignore; + char *smack_process_label; + + char **read_write_paths, **read_only_paths, **inaccessible_paths; + unsigned long mount_flags; + + uint64_t capability_bounding_set; + uint64_t capability_ambient_set; + int secure_bits; + + int syslog_priority; + char *syslog_identifier; + bool syslog_level_prefix; + + bool cpu_sched_reset_on_fork; + bool non_blocking; + bool private_tmp; + bool private_network; + bool private_devices; + ProtectSystem protect_system; + ProtectHome protect_home; + + bool no_new_privileges; + + /* This is not exposed to the user but available + * internally. We need it to make sure that whenever we spawn + * /usr/bin/mount it is run in the same process group as us so + * that the autofs logic detects that it belongs to us and we + * don't enter a trigger loop. */ + bool same_pgrp; + + unsigned long personality; + + Set *syscall_filter; + Set *syscall_archs; + int syscall_errno; + bool syscall_whitelist:1; + + Set *address_families; + bool address_families_whitelist:1; + + char **runtime_directory; + mode_t runtime_directory_mode; + + bool memory_deny_write_execute; + bool restrict_realtime; + + bool oom_score_adjust_set:1; + bool nice_set:1; + bool ioprio_set:1; + bool cpu_sched_set:1; + bool no_new_privileges_set:1; +}; + +struct ExecParameters { + char **argv; + char **environment; + + int *fds; + char **fd_names; + unsigned n_fds; + + bool apply_permissions:1; + bool apply_chroot:1; + bool apply_tty_stdin:1; + + bool confirm_spawn:1; + bool selinux_context_net:1; + + bool cgroup_delegate:1; + CGroupMask cgroup_supported; + const char *cgroup_path; + + const char *runtime_prefix; + + usec_t watchdog_usec; + + int *idle_pipe; + + int stdin_fd; + int stdout_fd; + int stderr_fd; +}; + +#include "unit.h" + +int exec_spawn(Unit *unit, + ExecCommand *command, + const ExecContext *context, + const ExecParameters *exec_params, + ExecRuntime *runtime, + pid_t *ret); + +void exec_command_done(ExecCommand *c); +void exec_command_done_array(ExecCommand *c, unsigned n); + +ExecCommand* exec_command_free_list(ExecCommand *c); +void exec_command_free_array(ExecCommand **c, unsigned n); + +char *exec_command_line(char **argv); + +void exec_command_dump(ExecCommand *c, FILE *f, const char *prefix); +void exec_command_dump_list(ExecCommand *c, FILE *f, const char *prefix); +void exec_command_append_list(ExecCommand **l, ExecCommand *e); +int exec_command_set(ExecCommand *c, const char *path, ...); +int exec_command_append(ExecCommand *c, const char *path, ...); + +void exec_context_init(ExecContext *c); +void exec_context_done(ExecContext *c); +void exec_context_dump(ExecContext *c, FILE* f, const char *prefix); + +int exec_context_destroy_runtime_directory(ExecContext *c, const char *runtime_root); + +int exec_context_load_environment(Unit *unit, const ExecContext *c, char ***l); + +bool exec_context_may_touch_console(ExecContext *c); +bool exec_context_maintains_privileges(ExecContext *c); + +void exec_status_start(ExecStatus *s, pid_t pid); +void exec_status_exit(ExecStatus *s, ExecContext *context, pid_t pid, int code, int status); +void exec_status_dump(ExecStatus *s, FILE *f, const char *prefix); + +int exec_runtime_make(ExecRuntime **rt, ExecContext *c, const char *id); +ExecRuntime *exec_runtime_ref(ExecRuntime *r); +ExecRuntime *exec_runtime_unref(ExecRuntime *r); + +int exec_runtime_serialize(Unit *unit, ExecRuntime *rt, FILE *f, FDSet *fds); +int exec_runtime_deserialize_item(Unit *unit, ExecRuntime **rt, const char *key, const char *value, FDSet *fds); + +void exec_runtime_destroy(ExecRuntime *rt); + +const char* exec_output_to_string(ExecOutput i) _const_; +ExecOutput exec_output_from_string(const char *s) _pure_; + +const char* exec_input_to_string(ExecInput i) _const_; +ExecInput exec_input_from_string(const char *s) _pure_; + +const char* exec_utmp_mode_to_string(ExecUtmpMode i) _const_; +ExecUtmpMode exec_utmp_mode_from_string(const char *s) _pure_; diff --git a/src/grp-system/libcore/include/core/failure-action.h b/src/grp-system/libcore/include/core/failure-action.h new file mode 100644 index 0000000000..47c08f4006 --- /dev/null +++ b/src/grp-system/libcore/include/core/failure-action.h @@ -0,0 +1,42 @@ +#pragma once + +/*** + This file is part of systemd. + + Copyright 2014 Lennart Poettering + Copyright 2012 Michael Olbrich + + systemd is free software; you can redistribute it and/or modify it + under the terms of the GNU Lesser General Public License as published by + the Free Software Foundation; either version 2.1 of the License, or + (at your option) any later version. + + systemd is distributed in the hope that it will be useful, but + WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + Lesser General Public License for more details. + + You should have received a copy of the GNU Lesser General Public License + along with systemd; If not, see . +***/ + +typedef enum FailureAction { + FAILURE_ACTION_NONE, + FAILURE_ACTION_REBOOT, + FAILURE_ACTION_REBOOT_FORCE, + FAILURE_ACTION_REBOOT_IMMEDIATE, + FAILURE_ACTION_POWEROFF, + FAILURE_ACTION_POWEROFF_FORCE, + FAILURE_ACTION_POWEROFF_IMMEDIATE, + _FAILURE_ACTION_MAX, + _FAILURE_ACTION_INVALID = -1 +} FailureAction; + +#include "systemd-basic/macro.h" + +#include "manager.h" + +int failure_action(Manager *m, FailureAction action, const char *reboot_arg); + +const char* failure_action_to_string(FailureAction i) _const_; +FailureAction failure_action_from_string(const char *s) _pure_; diff --git a/src/grp-system/libcore/include/core/hostname-setup.h b/src/grp-system/libcore/include/core/hostname-setup.h new file mode 100644 index 0000000000..73e8c75c71 --- /dev/null +++ b/src/grp-system/libcore/include/core/hostname-setup.h @@ -0,0 +1,22 @@ +#pragma once + +/*** + This file is part of systemd. + + Copyright 2010 Lennart Poettering + + systemd is free software; you can redistribute it and/or modify it + under the terms of the GNU Lesser General Public License as published by + the Free Software Foundation; either version 2.1 of the License, or + (at your option) any later version. + + systemd is distributed in the hope that it will be useful, but + WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + Lesser General Public License for more details. + + You should have received a copy of the GNU Lesser General Public License + along with systemd; If not, see . +***/ + +int hostname_setup(void); diff --git a/src/grp-system/libcore/include/core/ima-setup.h b/src/grp-system/libcore/include/core/ima-setup.h new file mode 100644 index 0000000000..472b58cb00 --- /dev/null +++ b/src/grp-system/libcore/include/core/ima-setup.h @@ -0,0 +1,24 @@ +#pragma once + +/*** + This file is part of systemd. + + Copyright 2010 Lennart Poettering + Copyright (C) 2012 Roberto Sassu - Politecnico di Torino, Italy + TORSEC group — http://security.polito.it + + systemd is free software; you can redistribute it and/or modify it + under the terms of the GNU Lesser General Public License as published by + the Free Software Foundation; either version 2.1 of the License, or + (at your option) any later version. + + systemd is distributed in the hope that it will be useful, but + WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + Lesser General Public License for more details. + + You should have received a copy of the GNU Lesser General Public License + along with systemd; If not, see . +***/ + +int ima_setup(void); diff --git a/src/grp-system/libcore/include/core/job.h b/src/grp-system/libcore/include/core/job.h new file mode 100644 index 0000000000..7338601bb7 --- /dev/null +++ b/src/grp-system/libcore/include/core/job.h @@ -0,0 +1,242 @@ +#pragma once + +/*** + This file is part of systemd. + + Copyright 2010 Lennart Poettering + + systemd is free software; you can redistribute it and/or modify it + under the terms of the GNU Lesser General Public License as published by + the Free Software Foundation; either version 2.1 of the License, or + (at your option) any later version. + + systemd is distributed in the hope that it will be useful, but + WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + Lesser General Public License for more details. + + You should have received a copy of the GNU Lesser General Public License + along with systemd; If not, see . +***/ + +#include + +#include + +#include "systemd-basic/list.h" +#include "systemd-basic/unit-name.h" + +typedef struct Job Job; +typedef struct JobDependency JobDependency; +typedef enum JobType JobType; +typedef enum JobState JobState; +typedef enum JobMode JobMode; +typedef enum JobResult JobResult; + +/* Be careful when changing the job types! Adjust job_merging_table[] accordingly! */ +enum JobType { + JOB_START, /* if a unit does not support being started, we'll just wait until it becomes active */ + JOB_VERIFY_ACTIVE, + + JOB_STOP, + + JOB_RELOAD, /* if running, reload */ + + /* Note that restarts are first treated like JOB_STOP, but + * then instead of finishing are patched to become + * JOB_START. */ + JOB_RESTART, /* If running, stop. Then start unconditionally. */ + + _JOB_TYPE_MAX_MERGING, + + /* JOB_NOP can enter into a transaction, but as it won't pull in + * any dependencies and it uses the special 'nop_job' slot in Unit, + * it won't have to merge with anything (except possibly into another + * JOB_NOP, previously installed). JOB_NOP is special-cased in + * job_type_is_*() functions so that the transaction can be + * activated. */ + JOB_NOP = _JOB_TYPE_MAX_MERGING, /* do nothing */ + + _JOB_TYPE_MAX_IN_TRANSACTION, + + /* JOB_TRY_RESTART can never appear in a transaction, because + * it always collapses into JOB_RESTART or JOB_NOP before entering. + * Thus we never need to merge it with anything. */ + JOB_TRY_RESTART = _JOB_TYPE_MAX_IN_TRANSACTION, /* if running, stop and then start */ + + /* Similar to JOB_TRY_RESTART but collapses to JOB_RELOAD or JOB_NOP */ + JOB_TRY_RELOAD, + + /* JOB_RELOAD_OR_START won't enter into a transaction and cannot result + * from transaction merging (there's no way for JOB_RELOAD and + * JOB_START to meet in one transaction). It can result from a merge + * during job installation, but then it will immediately collapse into + * one of the two simpler types. */ + JOB_RELOAD_OR_START, /* if running, reload, otherwise start */ + + _JOB_TYPE_MAX, + _JOB_TYPE_INVALID = -1 +}; + +enum JobState { + JOB_WAITING, + JOB_RUNNING, + _JOB_STATE_MAX, + _JOB_STATE_INVALID = -1 +}; + +enum JobMode { + JOB_FAIL, /* Fail if a conflicting job is already queued */ + JOB_REPLACE, /* Replace an existing conflicting job */ + JOB_REPLACE_IRREVERSIBLY,/* Like JOB_REPLACE + produce irreversible jobs */ + JOB_ISOLATE, /* Start a unit, and stop all others */ + JOB_FLUSH, /* Flush out all other queued jobs when queing this one */ + JOB_IGNORE_DEPENDENCIES, /* Ignore both requirement and ordering dependencies */ + JOB_IGNORE_REQUIREMENTS, /* Ignore requirement dependencies */ + _JOB_MODE_MAX, + _JOB_MODE_INVALID = -1 +}; + +enum JobResult { + JOB_DONE, /* Job completed successfully */ + JOB_CANCELED, /* Job canceled by a conflicting job installation or by explicit cancel request */ + JOB_TIMEOUT, /* Job timeout elapsed */ + JOB_FAILED, /* Job failed */ + JOB_DEPENDENCY, /* A required dependency job did not result in JOB_DONE */ + JOB_SKIPPED, /* Negative result of JOB_VERIFY_ACTIVE */ + JOB_INVALID, /* JOB_RELOAD of inactive unit */ + JOB_ASSERT, /* Couldn't start a unit, because an assert didn't hold */ + JOB_UNSUPPORTED, /* Couldn't start a unit, because the unit type is not supported on the system */ + _JOB_RESULT_MAX, + _JOB_RESULT_INVALID = -1 +}; + +#include "unit.h" + +struct JobDependency { + /* Encodes that the 'subject' job needs the 'object' job in + * some way. This structure is used only while building a transaction. */ + Job *subject; + Job *object; + + LIST_FIELDS(JobDependency, subject); + LIST_FIELDS(JobDependency, object); + + bool matters; + bool conflicts; +}; + +struct Job { + Manager *manager; + Unit *unit; + + LIST_FIELDS(Job, transaction); + LIST_FIELDS(Job, run_queue); + LIST_FIELDS(Job, dbus_queue); + + LIST_HEAD(JobDependency, subject_list); + LIST_HEAD(JobDependency, object_list); + + /* Used for graph algs as a "I have been here" marker */ + Job* marker; + unsigned generation; + + uint32_t id; + + JobType type; + JobState state; + + sd_event_source *timer_event_source; + usec_t begin_usec; + + /* + * This tracks where to send signals, and also which clients + * are allowed to call DBus methods on the job (other than + * root). + * + * There can be more than one client, because of job merging. + */ + sd_bus_track *clients; + char **deserialized_clients; + + JobResult result; + + bool installed:1; + bool in_run_queue:1; + bool matters_to_anchor:1; + bool in_dbus_queue:1; + bool sent_dbus_new_signal:1; + bool ignore_order:1; + bool irreversible:1; +}; + +Job* job_new(Unit *unit, JobType type); +Job* job_new_raw(Unit *unit); +void job_free(Job *job); +Job* job_install(Job *j); +int job_install_deserialized(Job *j); +void job_uninstall(Job *j); +void job_dump(Job *j, FILE*f, const char *prefix); +int job_serialize(Job *j, FILE *f, FDSet *fds); +int job_deserialize(Job *j, FILE *f, FDSet *fds); +int job_coldplug(Job *j); + +JobDependency* job_dependency_new(Job *subject, Job *object, bool matters, bool conflicts); +void job_dependency_free(JobDependency *l); + +int job_merge(Job *j, Job *other); + +JobType job_type_lookup_merge(JobType a, JobType b) _pure_; + +_pure_ static inline bool job_type_is_mergeable(JobType a, JobType b) { + return job_type_lookup_merge(a, b) >= 0; +} + +_pure_ static inline bool job_type_is_conflicting(JobType a, JobType b) { + return a != JOB_NOP && b != JOB_NOP && !job_type_is_mergeable(a, b); +} + +_pure_ static inline bool job_type_is_superset(JobType a, JobType b) { + /* Checks whether operation a is a "superset" of b in its actions */ + if (b == JOB_NOP) + return true; + if (a == JOB_NOP) + return false; + return a == job_type_lookup_merge(a, b); +} + +bool job_type_is_redundant(JobType a, UnitActiveState b) _pure_; + +/* Collapses a state-dependent job type into a simpler type by observing + * the state of the unit which it is going to be applied to. */ +JobType job_type_collapse(JobType t, Unit *u); + +int job_type_merge_and_collapse(JobType *a, JobType b, Unit *u); + +void job_add_to_run_queue(Job *j); +void job_add_to_dbus_queue(Job *j); + +int job_start_timer(Job *j); + +int job_run_and_invalidate(Job *j); +int job_finish_and_invalidate(Job *j, JobResult result, bool recursive, bool already); + +char *job_dbus_path(Job *j); + +void job_shutdown_magic(Job *j); + +int job_get_timeout(Job *j, usec_t *timeout) _pure_; + +const char* job_type_to_string(JobType t) _const_; +JobType job_type_from_string(const char *s) _pure_; + +const char* job_state_to_string(JobState t) _const_; +JobState job_state_from_string(const char *s) _pure_; + +const char* job_mode_to_string(JobMode t) _const_; +JobMode job_mode_from_string(const char *s) _pure_; + +const char* job_result_to_string(JobResult t) _const_; +JobResult job_result_from_string(const char *s) _pure_; + +const char* job_type_to_access_method(JobType t); diff --git a/src/grp-system/libcore/include/core/kill.h b/src/grp-system/libcore/include/core/kill.h new file mode 100644 index 0000000000..ad8583b9b0 --- /dev/null +++ b/src/grp-system/libcore/include/core/kill.h @@ -0,0 +1,65 @@ +#pragma once + +/*** + This file is part of systemd. + + Copyright 2012 Lennart Poettering + + systemd is free software; you can redistribute it and/or modify it + under the terms of the GNU Lesser General Public License as published by + the Free Software Foundation; either version 2.1 of the License, or + (at your option) any later version. + + systemd is distributed in the hope that it will be useful, but + WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + Lesser General Public License for more details. + + You should have received a copy of the GNU Lesser General Public License + along with systemd; If not, see . +***/ + +#include +#include + +#include "systemd-basic/macro.h" + +typedef struct KillContext KillContext; + +typedef enum KillMode { + /* The kill mode is a property of a unit. */ + KILL_CONTROL_GROUP = 0, + KILL_PROCESS, + KILL_MIXED, + KILL_NONE, + _KILL_MODE_MAX, + _KILL_MODE_INVALID = -1 +} KillMode; + +struct KillContext { + KillMode kill_mode; + int kill_signal; + bool send_sigkill; + bool send_sighup; +}; + +typedef enum KillWho { + /* Kill who is a property of an operation */ + KILL_MAIN, + KILL_CONTROL, + KILL_ALL, + KILL_MAIN_FAIL, + KILL_CONTROL_FAIL, + KILL_ALL_FAIL, + _KILL_WHO_MAX, + _KILL_WHO_INVALID = -1 +} KillWho; + +void kill_context_init(KillContext *c); +void kill_context_dump(KillContext *c, FILE *f, const char *prefix); + +const char *kill_mode_to_string(KillMode k) _const_; +KillMode kill_mode_from_string(const char *s) _pure_; + +const char *kill_who_to_string(KillWho k) _const_; +KillWho kill_who_from_string(const char *s) _pure_; diff --git a/src/grp-system/libcore/include/core/killall.h b/src/grp-system/libcore/include/core/killall.h new file mode 100644 index 0000000000..acc2439f00 --- /dev/null +++ b/src/grp-system/libcore/include/core/killall.h @@ -0,0 +1,22 @@ +#pragma once + +/*** + This file is part of systemd. + + Copyright 2012 Lennart Poettering + + systemd is free software; you can redistribute it and/or modify it + under the terms of the GNU Lesser General Public License as published by + the Free Software Foundation; either version 2.1 of the License, or + (at your option) any later version. + + systemd is distributed in the hope that it will be useful, but + WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + Lesser General Public License for more details. + + You should have received a copy of the GNU Lesser General Public License + along with systemd; If not, see . +***/ + +void broadcast_signal(int sig, bool wait_for_exit, bool send_sighup); diff --git a/src/grp-system/libcore/include/core/kmod-setup.h b/src/grp-system/libcore/include/core/kmod-setup.h new file mode 100644 index 0000000000..685f4df301 --- /dev/null +++ b/src/grp-system/libcore/include/core/kmod-setup.h @@ -0,0 +1,22 @@ +#pragma once + +/*** + This file is part of systemd. + + Copyright 2010 Lennart Poettering + + systemd is free software; you can redistribute it and/or modify it + under the terms of the GNU Lesser General Public License as published by + the Free Software Foundation; either version 2.1 of the License, or + (at your option) any later version. + + systemd is distributed in the hope that it will be useful, but + WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + Lesser General Public License for more details. + + You should have received a copy of the GNU Lesser General Public License + along with systemd; If not, see . +***/ + +int kmod_setup(void); diff --git a/src/grp-system/libcore/include/core/load-fragment.h b/src/grp-system/libcore/include/core/load-fragment.h new file mode 100644 index 0000000000..b36a2e3a02 --- /dev/null +++ b/src/grp-system/libcore/include/core/load-fragment.h @@ -0,0 +1,123 @@ +#pragma once + +/*** + This file is part of systemd. + + Copyright 2010 Lennart Poettering + + systemd is free software; you can redistribute it and/or modify it + under the terms of the GNU Lesser General Public License as published by + the Free Software Foundation; either version 2.1 of the License, or + (at your option) any later version. + + systemd is distributed in the hope that it will be useful, but + WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + Lesser General Public License for more details. + + You should have received a copy of the GNU Lesser General Public License + along with systemd; If not, see . +***/ + +#include "unit.h" + +/* Read service data from .desktop file style configuration fragments */ + +int unit_load_fragment(Unit *u); + +void unit_dump_config_items(FILE *f); + +int config_parse_warn_compat(const char *unit, const char *filename, unsigned line, const char *section, unsigned section_line, const char *lvalue, int ltype, const char *rvalue, void *data, void *userdata); +int config_parse_unit_deps(const char *unit, const char *filename, unsigned line, const char *section, unsigned section_line, const char *lvalue, int ltype, const char *rvalue, void *data, void *userdata); +int config_parse_obsolete_unit_deps(const char *unit, const char *filename, unsigned line, const char *section, unsigned section_line, const char *lvalue, int ltype, const char *rvalue, void *data, void *userdata); +int config_parse_unit_string_printf(const char *unit, const char *filename, unsigned line, const char *section, unsigned section_line, const char *lvalue, int ltype, const char *rvalue, void *data, void *userdata); +int config_parse_unit_strv_printf(const char *unit, const char *filename, unsigned line, const char *section, unsigned section_line, const char *lvalue, int ltype, const char *rvalue, void *data, void *userdata); +int config_parse_unit_path_printf(const char *unit, const char *filename, unsigned line, const char *section, unsigned section_line, const char *lvalue, int ltype, const char *rvalue, void *data, void *userdata); +int config_parse_unit_path_strv_printf(const char *unit, const char *filename, unsigned line, const char *section, unsigned section_line, const char *lvalue, int ltype, const char *rvalue, void *data, void *userdata); +int config_parse_documentation(const char *unit, const char *filename, unsigned line, const char *section, unsigned section_line, const char *lvalue, int ltype, const char *rvalue, void *data, void *userdata); +int config_parse_socket_listen(const char *unit, const char *filename, unsigned line, const char *section, unsigned section_line, const char *lvalue, int ltype, const char *rvalue, void *data, void *userdata); +int config_parse_socket_protocol(const char *unit, const char *filename, unsigned line, const char *section, unsigned section_line, const char *lvalue, int ltype, const char *rvalue, void *data, void *userdata); +int config_parse_socket_bind(const char *unit, const char *filename, unsigned line, const char *section, unsigned section_line, const char *lvalue, int ltype, const char *rvalue, void *data, void *userdata); +int config_parse_exec_nice(const char *unit, const char *filename, unsigned line, const char *section, unsigned section_line, const char *lvalue, int ltype, const char *rvalue, void *data, void *userdata); +int config_parse_exec_oom_score_adjust(const char *unit, const char *filename, unsigned line, const char *section, unsigned section_line, const char *lvalue, int ltype, const char *rvalue, void *data, void *userdata); +int config_parse_exec(const char *unit, const char *filename, unsigned line, const char *section, unsigned section_line, const char *lvalue, int ltype, const char *rvalue, void *data, void *userdata); +int config_parse_service_timeout(const char *unit, const char *filename, unsigned line, const char *section, unsigned section_line, const char *lvalue, int ltype, const char *rvalue, void *data, void *userdata); +int config_parse_service_type(const char *unit, const char *filename, unsigned line, const char *section, unsigned section_line, const char *lvalue, int ltype, const char *rvalue, void *data, void *userdata); +int config_parse_service_restart(const char *unit, const char *filename, unsigned line, const char *section, unsigned section_line, const char *lvalue, int ltype, const char *rvalue, void *data, void *userdata); +int config_parse_socket_bindtodevice(const char *unit, const char *filename, unsigned line, const char *section, unsigned section_line, const char *lvalue, int ltype, const char *rvalue, void *data, void *userdata); +int config_parse_output(const char *unit, const char *filename, unsigned line, const char *section, unsigned section_line, const char *lvalue, int ltype, const char *rvalue, void *data, void *userdata); +int config_parse_input(const char *unit, const char *filename, unsigned line, const char *section, unsigned section_line, const char *lvalue, int ltype, const char *rvalue, void *data, void *userdata); +int config_parse_exec_io_class(const char *unit, const char *filename, unsigned line, const char *section, unsigned section_line, const char *lvalue, int ltype, const char *rvalue, void *data, void *userdata); +int config_parse_exec_io_priority(const char *unit, const char *filename, unsigned line, const char *section, unsigned section_line, const char *lvalue, int ltype, const char *rvalue, void *data, void *userdata); +int config_parse_exec_cpu_sched_policy(const char *unit, const char *filename, unsigned line, const char *section, unsigned section_line, const char *lvalue, int ltype, const char *rvalue, void *data, void *userdata); +int config_parse_exec_cpu_sched_prio(const char *unit, const char *filename, unsigned line, const char *section, unsigned section_line, const char *lvalue, int ltype, const char *rvalue, void *data, void *userdata); +int config_parse_exec_cpu_affinity(const char *unit, const char *filename, unsigned line, const char *section, unsigned section_line, const char *lvalue, int ltype, const char *rvalue, void *data, void *userdata); +int config_parse_exec_secure_bits(const char *unit, const char *filename, unsigned line, const char *section, unsigned section_line, const char *lvalue, int ltype, const char *rvalue, void *data, void *userdata); +int config_parse_capability_set(const char *unit, const char *filename, unsigned line, const char *section, unsigned section_line, const char *lvalue, int ltype, const char *rvalue, void *data, void *userdata); +int config_parse_limit(const char *unit, const char *filename, unsigned line, const char *section, unsigned section_line, const char *lvalue, int ltype, const char *rvalue, void *data, void *userdata); +int config_parse_sysv_priority(const char *unit, const char *filename, unsigned line, const char *section, unsigned section_line, const char *lvalue, int ltype, const char *rvalue, void *data, void *userdata); +int config_parse_kill_signal(const char *unit, const char *filename, unsigned line, const char *section, unsigned section_line, const char *lvalue, int ltype, const char *rvalue, void *data, void *userdata); +int config_parse_exec_mount_flags(const char *unit, const char *filename, unsigned line, const char *section, unsigned section_line, const char *lvalue, int ltype, const char *rvalue, void *data, void *userdata); +int config_parse_timer(const char *unit, const char *filename, unsigned line, const char *section, unsigned section_line, const char *lvalue, int ltype, const char *rvalue, void *data, void *userdata); +int config_parse_trigger_unit(const char *unit, const char *filename, unsigned line, const char *section, unsigned section_line, const char *lvalue, int ltype, const char *rvalue, void *data, void *userdata); +int config_parse_path_spec(const char *unit, const char *filename, unsigned line, const char *section, unsigned section_line, const char *lvalue, int ltype, const char *rvalue, void *data, void *userdata); +int config_parse_socket_service(const char *unit, const char *filename, unsigned line, const char *section, unsigned section_line, const char *lvalue, int ltype, const char *rvalue, void *data, void *userdata); +int config_parse_service_sockets(const char *unit, const char *filename, unsigned line, const char *section, unsigned section_line, const char *lvalue, int ltype, const char *rvalue, void *data, void *userdata); +int config_parse_busname_service(const char *unit, const char *filename, unsigned line, const char *section, unsigned section_line, const char *lvalue, int ltype, const char *rvalue, void *data, void *userdata); +int config_parse_bus_policy(const char *unit, const char *filename, unsigned line, const char *section, unsigned section_line, const char *lvalue, int ltype, const char *rvalue, void *data, void *userdata); +int config_parse_bus_policy_world(const char *unit, const char *filename, unsigned line, const char *section, unsigned section_line, const char *lvalue, int ltype, const char *rvalue, void *data, void *userdata); +int config_parse_unit_env_file(const char *unit, const char *filename, unsigned line, const char *section, unsigned section_line, const char *lvalue, int ltype, const char *rvalue, void *data, void *userdata); +int config_parse_ip_tos(const char *unit, const char *filename, unsigned line, const char *section, unsigned section_line, const char *lvalue, int ltype, const char *rvalue, void *data, void *userdata); +int config_parse_unit_condition_path(const char *unit, const char *filename, unsigned line, const char *section, unsigned section_line, const char *lvalue, int ltype, const char *rvalue, void *data, void *userdata); +int config_parse_unit_condition_string(const char *unit, const char *filename, unsigned line, const char *section, unsigned section_line, const char *lvalue, int ltype, const char *rvalue, void *data, void *userdata); +int config_parse_unit_condition_null(const char *unit, const char *filename, unsigned line, const char *section, unsigned section_line, const char *lvalue, int ltype, const char *rvalue, void *data, void *userdata); +int config_parse_kill_mode(const char *unit, const char *filename, unsigned line, const char *section, unsigned section_line, const char *lvalue, int ltype, const char *rvalue, void *data, void *userdata); +int config_parse_notify_access(const char *unit, const char *filename, unsigned line, const char *section, unsigned section_line, const char *lvalue, int ltype, const char *rvalue, void *data, void *userdata); +int config_parse_failure_action(const char *unit, const char *filename, unsigned line, const char *section, unsigned section_line, const char *lvalue, int ltype, const char *rvalue, void *data, void *userdata); +int config_parse_unit_requires_mounts_for(const char *unit, const char *filename, unsigned line, const char *section, unsigned section_line, const char *lvalue, int ltype, const char *rvalue, void *data, void *userdata); +int config_parse_syscall_filter(const char *unit, const char *filename, unsigned line, const char *section, unsigned section_line, const char *lvalue, int ltype, const char *rvalue, void *data, void *userdata); +int config_parse_syscall_archs(const char *unit, const char *filename, unsigned line, const char *section, unsigned section_line, const char *lvalue, int ltype, const char *rvalue, void *data, void *userdata); +int config_parse_syscall_errno(const char *unit, const char *filename, unsigned line, const char *section, unsigned section_line, const char *lvalue, int ltype, const char *rvalue, void *data, void *userdata); +int config_parse_environ(const char *unit, const char *filename, unsigned line, const char *section, unsigned section_line, const char *lvalue, int ltype, const char *rvalue, void *data, void *userdata); +int config_parse_pass_environ(const char *unit, const char *filename, unsigned line, const char *section, unsigned section_line, const char *lvalue, int ltype, const char *rvalue, void *data, void *userdata); +int config_parse_unit_slice(const char *unit, const char *filename, unsigned line, const char *section, unsigned section_line, const char *lvalue, int ltype, const char *rvalue, void *data, void *userdata); +int config_parse_cpu_shares(const char *unit, const char *filename, unsigned line, const char *section, unsigned section_line, const char *lvalue, int ltype, const char *rvalue, void *data, void *userdata); +int config_parse_memory_limit(const char *unit, const char *filename, unsigned line, const char *section, unsigned section_line, const char *lvalue, int ltype, const char *rvalue, void *data, void *userdata); +int config_parse_tasks_max(const char *unit, const char *filename, unsigned line, const char *section, unsigned section_line, const char *lvalue, int ltype, const char *rvalue, void *data, void *userdata); +int config_parse_device_policy(const char *unit, const char *filename, unsigned line, const char *section, unsigned section_line, const char *lvalue, int ltype, const char *rvalue, void *data, void *userdata); +int config_parse_device_allow(const char *unit, const char *filename, unsigned line, const char *section, unsigned section_line, const char *lvalue, int ltype, const char *rvalue, void *data, void *userdata); +int config_parse_io_weight(const char *unit, const char *filename, unsigned line, const char *section, unsigned section_line, const char *lvalue, int ltype, const char *rvalue, void *data, void *userdata); +int config_parse_io_device_weight(const char *unit, const char *filename, unsigned line, const char *section, unsigned section_line, const char *lvalue, int ltype, const char *rvalue, void *data, void *userdata); +int config_parse_io_limit(const char *unit, const char *filename, unsigned line, const char *section, unsigned section_line, const char *lvalue, int ltype, const char *rvalue, void *data, void *userdata); +int config_parse_blockio_weight(const char *unit, const char *filename, unsigned line, const char *section, unsigned section_line, const char *lvalue, int ltype, const char *rvalue, void *data, void *userdata); +int config_parse_blockio_device_weight(const char *unit, const char *filename, unsigned line, const char *section, unsigned section_line, const char *lvalue, int ltype, const char *rvalue, void *data, void *userdata); +int config_parse_blockio_bandwidth(const char *unit, const char *filename, unsigned line, const char *section, unsigned section_line, const char *lvalue, int ltype, const char *rvalue, void *data, void *userdata); +int config_parse_netclass(const char *unit, const char *filename, unsigned line, const char *section, unsigned section_line, const char *lvalue, int ltype, const char *rvalue, void *data, void *userdata); +int config_parse_job_mode(const char *unit, const char *filename, unsigned line, const char *section, unsigned section_line, const char *lvalue, int ltype, const char *rvalue, void *data, void *userdata); +int config_parse_job_mode_isolate(const char *unit, const char *filename, unsigned line, const char *section, unsigned section_line, const char *lvalue, int ltype, const char *rvalue, void *data, void *userdata); +int config_parse_exec_selinux_context(const char *unit, const char *filename, unsigned line, const char *section, unsigned section_line, const char *lvalue, int ltype, const char *rvalue, void *data, void *userdata); +int config_parse_exec_apparmor_profile(const char *unit, const char *filename, unsigned line, const char *section, unsigned section_line, const char *lvalue, int ltype, const char *rvalue, void *data, void *userdata); +int config_parse_exec_smack_process_label(const char *unit, const char *filename, unsigned line, const char *section, unsigned section_line, const char *lvalue, int ltype, const char *rvalue, void *data, void *userdata); +int config_parse_address_families(const char *unit, const char *filename, unsigned line, const char *section, unsigned section_line, const char *lvalue, int ltype, const char *rvalue, void *data, void *userdata); +int config_parse_runtime_directory(const char *unit, const char *filename, unsigned line, const char *section, unsigned section_line, const char *lvalue, int ltype, const char *rvalue, void *data, void *userdata); +int config_parse_set_status(const char *unit, const char *filename, unsigned line, const char *section, unsigned section_line, const char *lvalue, int ltype, const char *rvalue, void *data, void *userdata); +int config_parse_namespace_path_strv(const char *unit, const char *filename, unsigned line, const char *section, unsigned section_line, const char *lvalue, int ltype, const char *rvalue, void *data, void *userdata); +int config_parse_no_new_privileges(const char *unit, const char *filename, unsigned line, const char *section, unsigned section_line, const char *lvalue, int ltype, const char *rvalue, void *data, void *userdata); +int config_parse_cpu_quota(const char *unit, const char *filename, unsigned line, const char *section, unsigned section_line, const char *lvalue, int ltype, const char *rvalue, void *data, void *userdata); +int config_parse_protect_home(const char* unit, const char *filename, unsigned line, const char *section, unsigned section_line, const char *lvalue, int ltype, const char *rvalue, void *data, void *userdata); +int config_parse_protect_system(const char* unit, const char *filename, unsigned line, const char *section, unsigned section_line, const char *lvalue, int ltype, const char *rvalue, void *data, void *userdata); +int config_parse_bus_name(const char* unit, const char *filename, unsigned line, const char *section, unsigned section_line, const char *lvalue, int ltype, const char *rvalue, void *data, void *userdata); +int config_parse_exec_utmp_mode(const char *unit, const char *filename, unsigned line, const char *section, unsigned section_line, const char *lvalue, int ltype, const char *rvalue, void *data, void *userdata); +int config_parse_working_directory(const char *unit, const char *filename, unsigned line, const char *section, unsigned section_line, const char *lvalue, int ltype, const char *rvalue, void *data, void *userdata); +int config_parse_fdname(const char *unit, const char *filename, unsigned line, const char *section, unsigned section_line, const char *lvalue, int ltype, const char *rvalue, void *data, void *userdata); +int config_parse_sec_fix_0(const char *unit, const char *filename, unsigned line, const char *section, unsigned section_line, const char *lvalue, int ltype, const char *rvalue, void *data, void *userdata); + +/* gperf prototypes */ +const struct ConfigPerfItem* load_fragment_gperf_lookup(const char *key, unsigned length); +extern const char load_fragment_gperf_nulstr[]; + +typedef enum Disabled { + DISABLED_CONFIGURATION, + DISABLED_LEGACY, + DISABLED_EXPERIMENTAL, +} Disabled; diff --git a/src/grp-system/libcore/include/core/loopback-setup.h b/src/grp-system/libcore/include/core/loopback-setup.h new file mode 100644 index 0000000000..e7547b8a26 --- /dev/null +++ b/src/grp-system/libcore/include/core/loopback-setup.h @@ -0,0 +1,22 @@ +#pragma once + +/*** + This file is part of systemd. + + Copyright 2010 Lennart Poettering + + systemd is free software; you can redistribute it and/or modify it + under the terms of the GNU Lesser General Public License as published by + the Free Software Foundation; either version 2.1 of the License, or + (at your option) any later version. + + systemd is distributed in the hope that it will be useful, but + WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + Lesser General Public License for more details. + + You should have received a copy of the GNU Lesser General Public License + along with systemd; If not, see . +***/ + +int loopback_setup(void); diff --git a/src/grp-system/libcore/include/core/machine-id-setup.h b/src/grp-system/libcore/include/core/machine-id-setup.h new file mode 100644 index 0000000000..29f4620646 --- /dev/null +++ b/src/grp-system/libcore/include/core/machine-id-setup.h @@ -0,0 +1,23 @@ +#pragma once + +/*** + This file is part of systemd. + + Copyright 2010 Lennart Poettering + + systemd is free software; you can redistribute it and/or modify it + under the terms of the GNU Lesser General Public License as published by + the Free Software Foundation; either version 2.1 of the License, or + (at your option) any later version. + + systemd is distributed in the hope that it will be useful, but + WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + Lesser General Public License for more details. + + You should have received a copy of the GNU Lesser General Public License + along with systemd; If not, see . +***/ + +int machine_id_commit(const char *root); +int machine_id_setup(const char *root, sd_id128_t requested, sd_id128_t *ret); diff --git a/src/grp-system/libcore/include/core/manager.h b/src/grp-system/libcore/include/core/manager.h new file mode 100644 index 0000000000..ae56cf1a9a --- /dev/null +++ b/src/grp-system/libcore/include/core/manager.h @@ -0,0 +1,380 @@ +#pragma once + +/*** + This file is part of systemd. + + Copyright 2010 Lennart Poettering + + systemd is free software; you can redistribute it and/or modify it + under the terms of the GNU Lesser General Public License as published by + the Free Software Foundation; either version 2.1 of the License, or + (at your option) any later version. + + systemd is distributed in the hope that it will be useful, but + WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + Lesser General Public License for more details. + + You should have received a copy of the GNU Lesser General Public License + along with systemd; If not, see . +***/ + +#include +#include +#include + +#include +#include + +#include "systemd-basic/cgroup-util.h" +#include "systemd-basic/hashmap.h" +#include "systemd-basic/list.h" +#include "systemd-basic/ratelimit.h" +#include "systemd-shared/fdset.h" + +/* Enforce upper limit how many names we allow */ +#define MANAGER_MAX_NAMES 131072 /* 128K */ + +typedef struct Manager Manager; + +typedef enum ManagerState { + MANAGER_INITIALIZING, + MANAGER_STARTING, + MANAGER_RUNNING, + MANAGER_DEGRADED, + MANAGER_MAINTENANCE, + MANAGER_STOPPING, + _MANAGER_STATE_MAX, + _MANAGER_STATE_INVALID = -1 +} ManagerState; + +typedef enum ManagerExitCode { + MANAGER_OK, + MANAGER_EXIT, + MANAGER_RELOAD, + MANAGER_REEXECUTE, + MANAGER_REBOOT, + MANAGER_POWEROFF, + MANAGER_HALT, + MANAGER_KEXEC, + MANAGER_SWITCH_ROOT, + _MANAGER_EXIT_CODE_MAX, + _MANAGER_EXIT_CODE_INVALID = -1 +} ManagerExitCode; + +typedef enum StatusType { + STATUS_TYPE_EPHEMERAL, + STATUS_TYPE_NORMAL, + STATUS_TYPE_EMERGENCY, +} StatusType; + +#include "systemd-basic/unit-name.h" +#include "systemd-shared/path-lookup.h" + +#include "execute.h" +#include "job.h" +#include "show-status.h" + +struct Manager { + /* Note that the set of units we know of is allowed to be + * inconsistent. However the subset of it that is loaded may + * not, and the list of jobs may neither. */ + + /* Active jobs and units */ + Hashmap *units; /* name string => Unit object n:1 */ + Hashmap *jobs; /* job id => Job object 1:1 */ + + /* To make it easy to iterate through the units of a specific + * type we maintain a per type linked list */ + LIST_HEAD(Unit, units_by_type[_UNIT_TYPE_MAX]); + + /* Units that need to be loaded */ + LIST_HEAD(Unit, load_queue); /* this is actually more a stack than a queue, but uh. */ + + /* Jobs that need to be run */ + LIST_HEAD(Job, run_queue); /* more a stack than a queue, too */ + + /* Units and jobs that have not yet been announced via + * D-Bus. When something about a job changes it is added here + * if it is not in there yet. This allows easy coalescing of + * D-Bus change signals. */ + LIST_HEAD(Unit, dbus_unit_queue); + LIST_HEAD(Job, dbus_job_queue); + + /* Units to remove */ + LIST_HEAD(Unit, cleanup_queue); + + /* Units to check when doing GC */ + LIST_HEAD(Unit, gc_queue); + + /* Units that should be realized */ + LIST_HEAD(Unit, cgroup_queue); + + sd_event *event; + + /* We use two hash tables here, since the same PID might be + * watched by two different units: once the unit that forked + * it off, and possibly a different unit to which it was + * joined as cgroup member. Since we know that it is either + * one or two units for each PID we just use to hashmaps + * here. */ + Hashmap *watch_pids1; /* pid => Unit object n:1 */ + Hashmap *watch_pids2; /* pid => Unit object n:1 */ + + /* A set contains all units which cgroup should be refreshed after startup */ + Set *startup_units; + + /* A set which contains all currently failed units */ + Set *failed_units; + + sd_event_source *run_queue_event_source; + + char *notify_socket; + int notify_fd; + sd_event_source *notify_event_source; + + int cgroups_agent_fd; + sd_event_source *cgroups_agent_event_source; + + int signal_fd; + sd_event_source *signal_event_source; + + int time_change_fd; + sd_event_source *time_change_event_source; + + sd_event_source *jobs_in_progress_event_source; + + UnitFileScope unit_file_scope; + LookupPaths lookup_paths; + Set *unit_path_cache; + + char **environment; + + usec_t runtime_watchdog; + usec_t shutdown_watchdog; + + dual_timestamp firmware_timestamp; + dual_timestamp loader_timestamp; + dual_timestamp kernel_timestamp; + dual_timestamp initrd_timestamp; + dual_timestamp userspace_timestamp; + dual_timestamp finish_timestamp; + + dual_timestamp security_start_timestamp; + dual_timestamp security_finish_timestamp; + dual_timestamp generators_start_timestamp; + dual_timestamp generators_finish_timestamp; + dual_timestamp units_load_start_timestamp; + dual_timestamp units_load_finish_timestamp; + + struct udev* udev; + + /* Data specific to the device subsystem */ + struct udev_monitor* udev_monitor; + sd_event_source *udev_event_source; + Hashmap *devices_by_sysfs; + + /* Data specific to the mount subsystem */ + struct libmnt_monitor *mount_monitor; + sd_event_source *mount_event_source; + + /* Data specific to the swap filesystem */ + FILE *proc_swaps; + sd_event_source *swap_event_source; + Hashmap *swaps_by_devnode; + + /* Data specific to the D-Bus subsystem */ + sd_bus *api_bus, *system_bus; + Set *private_buses; + int private_listen_fd; + sd_event_source *private_listen_event_source; + + /* Contains all the clients that are subscribed to signals via + the API bus. Note that private bus connections are always + considered subscribes, since they last for very short only, + and it is much simpler that way. */ + sd_bus_track *subscribed; + char **deserialized_subscribed; + + /* This is used during reloading: before the reload we queue + * the reply message here, and afterwards we send it */ + sd_bus_message *queued_message; + + Hashmap *watch_bus; /* D-Bus names => Unit object n:1 */ + + bool send_reloading_done; + + uint32_t current_job_id; + uint32_t default_unit_job_id; + + /* Data specific to the Automount subsystem */ + int dev_autofs_fd; + + /* Data specific to the cgroup subsystem */ + Hashmap *cgroup_unit; + CGroupMask cgroup_supported; + char *cgroup_root; + + /* Notifications from cgroups, when the unified hierarchy is + * used is done via inotify. */ + int cgroup_inotify_fd; + sd_event_source *cgroup_inotify_event_source; + Hashmap *cgroup_inotify_wd_unit; + + /* Make sure the user cannot accidentally unmount our cgroup + * file system */ + int pin_cgroupfs_fd; + + int gc_marker; + unsigned n_in_gc_queue; + + /* Flags */ + ManagerExitCode exit_code:5; + + bool dispatching_load_queue:1; + bool dispatching_dbus_queue:1; + + bool taint_usr:1; + + bool test_run:1; + + /* If non-zero, exit with the following value when the systemd + * process terminate. Useful for containers: systemd-nspawn could get + * the return value. */ + uint8_t return_value; + + ShowStatus show_status; + bool confirm_spawn; + bool no_console_output; + + ExecOutput default_std_output, default_std_error; + + usec_t default_restart_usec, default_timeout_start_usec, default_timeout_stop_usec; + + usec_t default_start_limit_interval; + unsigned default_start_limit_burst; + + bool default_cpu_accounting; + bool default_memory_accounting; + bool default_io_accounting; + bool default_blockio_accounting; + bool default_tasks_accounting; + + uint64_t default_tasks_max; + usec_t default_timer_accuracy_usec; + + struct rlimit *rlimit[_RLIMIT_MAX]; + + /* non-zero if we are reloading or reexecuting, */ + int n_reloading; + + unsigned n_installed_jobs; + unsigned n_failed_jobs; + + /* Jobs in progress watching */ + unsigned n_running_jobs; + unsigned n_on_console; + unsigned jobs_in_progress_iteration; + + /* Do we have any outstanding password prompts? */ + int have_ask_password; + int ask_password_inotify_fd; + sd_event_source *ask_password_event_source; + + /* Type=idle pipes */ + int idle_pipe[4]; + sd_event_source *idle_pipe_event_source; + + char *switch_root; + char *switch_root_init; + + /* This maps all possible path prefixes to the units needing + * them. It's a hashmap with a path string as key and a Set as + * value where Unit objects are contained. */ + Hashmap *units_requiring_mounts_for; + + /* Reference to the kdbus bus control fd */ + int kdbus_fd; + + /* Used for processing polkit authorization responses */ + Hashmap *polkit_registry; + + /* When the user hits C-A-D more than 7 times per 2s, reboot immediately... */ + RateLimit ctrl_alt_del_ratelimit; + + const char *unit_log_field; + const char *unit_log_format_string; + + int first_boot; /* tri-state */ +}; + +#define MANAGER_IS_SYSTEM(m) ((m)->unit_file_scope == UNIT_FILE_SYSTEM) +#define MANAGER_IS_USER(m) ((m)->unit_file_scope != UNIT_FILE_SYSTEM) + +#define MANAGER_IS_RELOADING(m) ((m)->n_reloading > 0) + +int manager_new(UnitFileScope scope, bool test_run, Manager **m); +Manager* manager_free(Manager *m); + +void manager_enumerate(Manager *m); +int manager_startup(Manager *m, FILE *serialization, FDSet *fds); + +Job *manager_get_job(Manager *m, uint32_t id); +Unit *manager_get_unit(Manager *m, const char *name); + +int manager_get_job_from_dbus_path(Manager *m, const char *s, Job **_j); + +int manager_load_unit_prepare(Manager *m, const char *name, const char *path, sd_bus_error *e, Unit **_ret); +int manager_load_unit(Manager *m, const char *name, const char *path, sd_bus_error *e, Unit **_ret); +int manager_load_unit_from_dbus_path(Manager *m, const char *s, sd_bus_error *e, Unit **_u); + +int manager_add_job(Manager *m, JobType type, Unit *unit, JobMode mode, sd_bus_error *e, Job **_ret); +int manager_add_job_by_name(Manager *m, JobType type, const char *name, JobMode mode, sd_bus_error *e, Job **_ret); +int manager_add_job_by_name_and_warn(Manager *m, JobType type, const char *name, JobMode mode, Job **ret); + +void manager_dump_units(Manager *s, FILE *f, const char *prefix); +void manager_dump_jobs(Manager *s, FILE *f, const char *prefix); + +void manager_clear_jobs(Manager *m); + +unsigned manager_dispatch_load_queue(Manager *m); + +int manager_environment_add(Manager *m, char **minus, char **plus); +int manager_set_default_rlimits(Manager *m, struct rlimit **default_rlimit); + +int manager_loop(Manager *m); + +int manager_open_serialization(Manager *m, FILE **_f); + +int manager_serialize(Manager *m, FILE *f, FDSet *fds, bool switching_root); +int manager_deserialize(Manager *m, FILE *f, FDSet *fds); + +int manager_reload(Manager *m); + +void manager_reset_failed(Manager *m); + +void manager_send_unit_audit(Manager *m, Unit *u, int type, bool success); +void manager_send_unit_plymouth(Manager *m, Unit *u); + +bool manager_unit_inactive_or_pending(Manager *m, const char *name); + +void manager_check_finished(Manager *m); + +void manager_recheck_journal(Manager *m); + +void manager_set_show_status(Manager *m, ShowStatus mode); +void manager_set_first_boot(Manager *m, bool b); + +void manager_status_printf(Manager *m, StatusType type, const char *status, const char *format, ...) _printf_(4,5); +void manager_flip_auto_status(Manager *m, bool enable); + +Set *manager_get_units_requiring_mounts_for(Manager *m, const char *path); + +const char *manager_get_runtime_prefix(Manager *m); + +ManagerState manager_state(Manager *m); + +int manager_update_failed_units(Manager *m, Unit *u, bool failed); + +const char *manager_state_to_string(ManagerState m) _const_; +ManagerState manager_state_from_string(const char *s) _pure_; diff --git a/src/grp-system/libcore/include/core/mount-setup.h b/src/grp-system/libcore/include/core/mount-setup.h new file mode 100644 index 0000000000..647bd770ae --- /dev/null +++ b/src/grp-system/libcore/include/core/mount-setup.h @@ -0,0 +1,30 @@ +#pragma once + +/*** + This file is part of systemd. + + Copyright 2010 Lennart Poettering + + systemd is free software; you can redistribute it and/or modify it + under the terms of the GNU Lesser General Public License as published by + the Free Software Foundation; either version 2.1 of the License, or + (at your option) any later version. + + systemd is distributed in the hope that it will be useful, but + WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + Lesser General Public License for more details. + + You should have received a copy of the GNU Lesser General Public License + along with systemd; If not, see . +***/ + +#include + +int mount_setup_early(void); +int mount_setup(bool loaded_policy); + +int mount_cgroup_controllers(char ***join_controllers); + +bool mount_point_is_api(const char *path); +bool mount_point_ignore(const char *path); diff --git a/src/grp-system/libcore/include/core/mount.h b/src/grp-system/libcore/include/core/mount.h new file mode 100644 index 0000000000..da529c44f4 --- /dev/null +++ b/src/grp-system/libcore/include/core/mount.h @@ -0,0 +1,108 @@ +#pragma once + +/*** + This file is part of systemd. + + Copyright 2010 Lennart Poettering + + systemd is free software; you can redistribute it and/or modify it + under the terms of the GNU Lesser General Public License as published by + the Free Software Foundation; either version 2.1 of the License, or + (at your option) any later version. + + systemd is distributed in the hope that it will be useful, but + WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + Lesser General Public License for more details. + + You should have received a copy of the GNU Lesser General Public License + along with systemd; If not, see . +***/ + +typedef struct Mount Mount; + +#include "execute.h" +#include "kill.h" + +typedef enum MountExecCommand { + MOUNT_EXEC_MOUNT, + MOUNT_EXEC_UNMOUNT, + MOUNT_EXEC_REMOUNT, + _MOUNT_EXEC_COMMAND_MAX, + _MOUNT_EXEC_COMMAND_INVALID = -1 +} MountExecCommand; + +typedef enum MountResult { + MOUNT_SUCCESS, + MOUNT_FAILURE_RESOURCES, + MOUNT_FAILURE_TIMEOUT, + MOUNT_FAILURE_EXIT_CODE, + MOUNT_FAILURE_SIGNAL, + MOUNT_FAILURE_CORE_DUMP, + MOUNT_FAILURE_START_LIMIT_HIT, + _MOUNT_RESULT_MAX, + _MOUNT_RESULT_INVALID = -1 +} MountResult; + +typedef struct MountParameters { + char *what; + char *options; + char *fstype; +} MountParameters; + +struct Mount { + Unit meta; + + char *where; + + MountParameters parameters_proc_self_mountinfo; + MountParameters parameters_fragment; + + bool from_proc_self_mountinfo:1; + bool from_fragment:1; + + /* Used while looking for mount points that vanished or got + * added from/to /proc/self/mountinfo */ + bool is_mounted:1; + bool just_mounted:1; + bool just_changed:1; + + bool reset_cpu_usage:1; + + bool sloppy_options; + + MountResult result; + MountResult reload_result; + + mode_t directory_mode; + + usec_t timeout_usec; + + ExecCommand exec_command[_MOUNT_EXEC_COMMAND_MAX]; + + ExecContext exec_context; + KillContext kill_context; + CGroupContext cgroup_context; + + ExecRuntime *exec_runtime; + + MountState state, deserialized_state; + + ExecCommand* control_command; + MountExecCommand control_command_id; + pid_t control_pid; + + sd_event_source *timer_event_source; + + unsigned n_retry_umount; +}; + +extern const UnitVTable mount_vtable; + +void mount_fd_event(Manager *m, int events); + +const char* mount_exec_command_to_string(MountExecCommand i) _const_; +MountExecCommand mount_exec_command_from_string(const char *s) _pure_; + +const char* mount_result_to_string(MountResult i) _const_; +MountResult mount_result_from_string(const char *s) _pure_; diff --git a/src/grp-system/libcore/include/core/namespace.h b/src/grp-system/libcore/include/core/namespace.h new file mode 100644 index 0000000000..89c6225c39 --- /dev/null +++ b/src/grp-system/libcore/include/core/namespace.h @@ -0,0 +1,63 @@ +#pragma once + +/*** + This file is part of systemd. + + Copyright 2010 Lennart Poettering + + systemd is free software; you can redistribute it and/or modify it + under the terms of the GNU Lesser General Public License as published by + the Free Software Foundation; either version 2.1 of the License, or + (at your option) any later version. + + systemd is distributed in the hope that it will be useful, but + WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + Lesser General Public License for more details. + + You should have received a copy of the GNU Lesser General Public License + along with systemd; If not, see . +***/ + +#include + +#include "systemd-basic/macro.h" + +typedef enum ProtectHome { + PROTECT_HOME_NO, + PROTECT_HOME_YES, + PROTECT_HOME_READ_ONLY, + _PROTECT_HOME_MAX, + _PROTECT_HOME_INVALID = -1 +} ProtectHome; + +typedef enum ProtectSystem { + PROTECT_SYSTEM_NO, + PROTECT_SYSTEM_YES, + PROTECT_SYSTEM_FULL, + _PROTECT_SYSTEM_MAX, + _PROTECT_SYSTEM_INVALID = -1 +} ProtectSystem; + +int setup_namespace(const char *chroot, + char **read_write_paths, + char **read_only_paths, + char **inaccessible_paths, + const char *tmp_dir, + const char *var_tmp_dir, + bool private_dev, + ProtectHome protect_home, + ProtectSystem protect_system, + unsigned long mount_flags); + +int setup_tmp_dirs(const char *id, + char **tmp_dir, + char **var_tmp_dir); + +int setup_netns(int netns_storage_socket[2]); + +const char* protect_home_to_string(ProtectHome p) _const_; +ProtectHome protect_home_from_string(const char *s) _pure_; + +const char* protect_system_to_string(ProtectSystem p) _const_; +ProtectSystem protect_system_from_string(const char *s) _pure_; diff --git a/src/grp-system/libcore/include/core/path.h b/src/grp-system/libcore/include/core/path.h new file mode 100644 index 0000000000..4230c8fb99 --- /dev/null +++ b/src/grp-system/libcore/include/core/path.h @@ -0,0 +1,93 @@ +#pragma once + +/*** + This file is part of systemd. + + Copyright 2010 Lennart Poettering + + systemd is free software; you can redistribute it and/or modify it + under the terms of the GNU Lesser General Public License as published by + the Free Software Foundation; either version 2.1 of the License, or + (at your option) any later version. + + systemd is distributed in the hope that it will be useful, but + WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + Lesser General Public License for more details. + + You should have received a copy of the GNU Lesser General Public License + along with systemd; If not, see . +***/ + +typedef struct Path Path; +typedef struct PathSpec PathSpec; + +#include "unit.h" + +typedef enum PathType { + PATH_EXISTS, + PATH_EXISTS_GLOB, + PATH_DIRECTORY_NOT_EMPTY, + PATH_CHANGED, + PATH_MODIFIED, + _PATH_TYPE_MAX, + _PATH_TYPE_INVALID = -1 +} PathType; + +typedef struct PathSpec { + Unit *unit; + + char *path; + + sd_event_source *event_source; + + LIST_FIELDS(struct PathSpec, spec); + + PathType type; + int inotify_fd; + int primary_wd; + + bool previous_exists; +} PathSpec; + +int path_spec_watch(PathSpec *s, sd_event_io_handler_t handler); +void path_spec_unwatch(PathSpec *s); +int path_spec_fd_event(PathSpec *s, uint32_t events); +void path_spec_done(PathSpec *s); + +static inline bool path_spec_owns_inotify_fd(PathSpec *s, int fd) { + return s->inotify_fd == fd; +} + +typedef enum PathResult { + PATH_SUCCESS, + PATH_FAILURE_RESOURCES, + PATH_FAILURE_START_LIMIT_HIT, + _PATH_RESULT_MAX, + _PATH_RESULT_INVALID = -1 +} PathResult; + +struct Path { + Unit meta; + + LIST_HEAD(PathSpec, specs); + + PathState state, deserialized_state; + + bool inotify_triggered; + + bool make_directory; + mode_t directory_mode; + + PathResult result; +}; + +void path_free_specs(Path *p); + +extern const UnitVTable path_vtable; + +const char* path_type_to_string(PathType i) _const_; +PathType path_type_from_string(const char *s) _pure_; + +const char* path_result_to_string(PathResult i) _const_; +PathResult path_result_from_string(const char *s) _pure_; diff --git a/src/grp-system/libcore/include/core/scope.h b/src/grp-system/libcore/include/core/scope.h new file mode 100644 index 0000000000..eaf8e8b447 --- /dev/null +++ b/src/grp-system/libcore/include/core/scope.h @@ -0,0 +1,58 @@ +#pragma once + +/*** + This file is part of systemd. + + Copyright 2013 Lennart Poettering + + systemd is free software; you can redistribute it and/or modify it + under the terms of the GNU Lesser General Public License as published by + the Free Software Foundation; either version 2.1 of the License, or + (at your option) any later version. + + systemd is distributed in the hope that it will be useful, but + WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + Lesser General Public License for more details. + + You should have received a copy of the GNU Lesser General Public License + along with systemd; If not, see . +***/ + +typedef struct Scope Scope; + +#include "cgroup.h" +#include "kill.h" +#include "unit.h" + +typedef enum ScopeResult { + SCOPE_SUCCESS, + SCOPE_FAILURE_RESOURCES, + SCOPE_FAILURE_TIMEOUT, + _SCOPE_RESULT_MAX, + _SCOPE_RESULT_INVALID = -1 +} ScopeResult; + +struct Scope { + Unit meta; + + CGroupContext cgroup_context; + KillContext kill_context; + + ScopeState state, deserialized_state; + ScopeResult result; + + usec_t timeout_stop_usec; + + char *controller; + bool was_abandoned; + + sd_event_source *timer_event_source; +}; + +extern const UnitVTable scope_vtable; + +int scope_abandon(Scope *s); + +const char* scope_result_to_string(ScopeResult i) _const_; +ScopeResult scope_result_from_string(const char *s) _pure_; diff --git a/src/grp-system/libcore/include/core/selinux-setup.h b/src/grp-system/libcore/include/core/selinux-setup.h new file mode 100644 index 0000000000..7b613249b0 --- /dev/null +++ b/src/grp-system/libcore/include/core/selinux-setup.h @@ -0,0 +1,24 @@ +#pragma once + +/*** + This file is part of systemd. + + Copyright 2010 Lennart Poettering + + systemd is free software; you can redistribute it and/or modify it + under the terms of the GNU Lesser General Public License as published by + the Free Software Foundation; either version 2.1 of the License, or + (at your option) any later version. + + systemd is distributed in the hope that it will be useful, but + WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + Lesser General Public License for more details. + + You should have received a copy of the GNU Lesser General Public License + along with systemd; If not, see . +***/ + +#include + +int mac_selinux_setup(bool *loaded_policy); diff --git a/src/grp-system/libcore/include/core/service.h b/src/grp-system/libcore/include/core/service.h new file mode 100644 index 0000000000..528df23faf --- /dev/null +++ b/src/grp-system/libcore/include/core/service.h @@ -0,0 +1,223 @@ +#pragma once + +/*** + This file is part of systemd. + + Copyright 2010 Lennart Poettering + + systemd is free software; you can redistribute it and/or modify it + under the terms of the GNU Lesser General Public License as published by + the Free Software Foundation; either version 2.1 of the License, or + (at your option) any later version. + + systemd is distributed in the hope that it will be useful, but + WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + Lesser General Public License for more details. + + You should have received a copy of the GNU Lesser General Public License + along with systemd; If not, see . +***/ + +#include "systemd-basic/exit-status.h" +#include "systemd-basic/ratelimit.h" + +typedef struct Service Service; +typedef struct ServiceFDStore ServiceFDStore; + +#include "kill.h" +#include "path.h" + +typedef enum ServiceRestart { + SERVICE_RESTART_NO, + SERVICE_RESTART_ON_SUCCESS, + SERVICE_RESTART_ON_FAILURE, + SERVICE_RESTART_ON_ABNORMAL, + SERVICE_RESTART_ON_WATCHDOG, + SERVICE_RESTART_ON_ABORT, + SERVICE_RESTART_ALWAYS, + _SERVICE_RESTART_MAX, + _SERVICE_RESTART_INVALID = -1 +} ServiceRestart; + +typedef enum ServiceType { + SERVICE_SIMPLE, /* we fork and go on right-away (i.e. modern socket activated daemons) */ + SERVICE_FORKING, /* forks by itself (i.e. traditional daemons) */ + SERVICE_ONESHOT, /* we fork and wait until the program finishes (i.e. programs like fsck which run and need to finish before we continue) */ + SERVICE_DBUS, /* we fork and wait until a specific D-Bus name appears on the bus */ + SERVICE_NOTIFY, /* we fork and wait until a daemon sends us a ready message with sd_notify() */ + SERVICE_IDLE, /* much like simple, but delay exec() until all jobs are dispatched. */ + _SERVICE_TYPE_MAX, + _SERVICE_TYPE_INVALID = -1 +} ServiceType; + +typedef enum ServiceExecCommand { + SERVICE_EXEC_START_PRE, + SERVICE_EXEC_START, + SERVICE_EXEC_START_POST, + SERVICE_EXEC_RELOAD, + SERVICE_EXEC_STOP, + SERVICE_EXEC_STOP_POST, + _SERVICE_EXEC_COMMAND_MAX, + _SERVICE_EXEC_COMMAND_INVALID = -1 +} ServiceExecCommand; + +typedef enum NotifyAccess { + NOTIFY_NONE, + NOTIFY_ALL, + NOTIFY_MAIN, + _NOTIFY_ACCESS_MAX, + _NOTIFY_ACCESS_INVALID = -1 +} NotifyAccess; + +typedef enum NotifyState { + NOTIFY_UNKNOWN, + NOTIFY_READY, + NOTIFY_RELOADING, + NOTIFY_STOPPING, + _NOTIFY_STATE_MAX, + _NOTIFY_STATE_INVALID = -1 +} NotifyState; + +typedef enum ServiceResult { + SERVICE_SUCCESS, + SERVICE_FAILURE_RESOURCES, /* a bit of a misnomer, just our catch-all error for errnos we didn't expect */ + SERVICE_FAILURE_TIMEOUT, + SERVICE_FAILURE_EXIT_CODE, + SERVICE_FAILURE_SIGNAL, + SERVICE_FAILURE_CORE_DUMP, + SERVICE_FAILURE_WATCHDOG, + SERVICE_FAILURE_START_LIMIT_HIT, + _SERVICE_RESULT_MAX, + _SERVICE_RESULT_INVALID = -1 +} ServiceResult; + +struct ServiceFDStore { + Service *service; + + int fd; + char *fdname; + sd_event_source *event_source; + + LIST_FIELDS(ServiceFDStore, fd_store); +}; + +struct Service { + Unit meta; + + ServiceType type; + ServiceRestart restart; + ExitStatusSet restart_prevent_status; + ExitStatusSet restart_force_status; + ExitStatusSet success_status; + + /* If set we'll read the main daemon PID from this file */ + char *pid_file; + + usec_t restart_usec; + usec_t timeout_start_usec; + usec_t timeout_stop_usec; + usec_t runtime_max_usec; + + dual_timestamp watchdog_timestamp; + usec_t watchdog_usec; + usec_t watchdog_override_usec; + bool watchdog_override_enable; + sd_event_source *watchdog_event_source; + + ExecCommand* exec_command[_SERVICE_EXEC_COMMAND_MAX]; + + ExecContext exec_context; + KillContext kill_context; + CGroupContext cgroup_context; + + ServiceState state, deserialized_state; + + /* The exit status of the real main process */ + ExecStatus main_exec_status; + + /* The currently executed control process */ + ExecCommand *control_command; + + /* The currently executed main process, which may be NULL if + * the main process got started via forking mode and not by + * us */ + ExecCommand *main_command; + + /* The ID of the control command currently being executed */ + ServiceExecCommand control_command_id; + + /* Runtime data of the execution context */ + ExecRuntime *exec_runtime; + + pid_t main_pid, control_pid; + int socket_fd; + bool socket_fd_selinux_context_net; + + bool permissions_start_only; + bool root_directory_start_only; + bool remain_after_exit; + bool guess_main_pid; + + /* If we shut down, remember why */ + ServiceResult result; + ServiceResult reload_result; + + bool main_pid_known:1; + bool main_pid_alien:1; + bool bus_name_good:1; + bool forbid_restart:1; + bool start_timeout_defined:1; + + bool reset_cpu_usage:1; + + char *bus_name; + char *bus_name_owner; /* unique name of the current owner */ + + char *status_text; + int status_errno; + + FailureAction failure_action; + + UnitRef accept_socket; + + sd_event_source *timer_event_source; + PathSpec *pid_file_pathspec; + + NotifyAccess notify_access; + NotifyState notify_state; + + ServiceFDStore *fd_store; + unsigned n_fd_store; + unsigned n_fd_store_max; + + char *usb_function_descriptors; + char *usb_function_strings; + + int stdin_fd; + int stdout_fd; + int stderr_fd; +}; + +extern const UnitVTable service_vtable; + +int service_set_socket_fd(Service *s, int fd, struct Socket *socket, bool selinux_context_net); +void service_close_socket_fd(Service *s); + +const char* service_restart_to_string(ServiceRestart i) _const_; +ServiceRestart service_restart_from_string(const char *s) _pure_; + +const char* service_type_to_string(ServiceType i) _const_; +ServiceType service_type_from_string(const char *s) _pure_; + +const char* service_exec_command_to_string(ServiceExecCommand i) _const_; +ServiceExecCommand service_exec_command_from_string(const char *s) _pure_; + +const char* notify_access_to_string(NotifyAccess i) _const_; +NotifyAccess notify_access_from_string(const char *s) _pure_; + +const char* notify_state_to_string(NotifyState i) _const_; +NotifyState notify_state_from_string(const char *s) _pure_; + +const char* service_result_to_string(ServiceResult i) _const_; +ServiceResult service_result_from_string(const char *s) _pure_; diff --git a/src/grp-system/libcore/include/core/show-status.h b/src/grp-system/libcore/include/core/show-status.h new file mode 100644 index 0000000000..08d6b7f6e1 --- /dev/null +++ b/src/grp-system/libcore/include/core/show-status.h @@ -0,0 +1,39 @@ +#pragma once + +/*** + This file is part of systemd. + + Copyright 2014 Lennart Poettering + + systemd is free software; you can redistribute it and/or modify it + under the terms of the GNU Lesser General Public License as published by + the Free Software Foundation; either version 2.1 of the License, or + (at your option) any later version. + + systemd is distributed in the hope that it will be useful, but + WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + Lesser General Public License for more details. + + You should have received a copy of the GNU Lesser General Public License + along with systemd; If not, see . +***/ + +#include + +#include "systemd-basic/macro.h" + +/* Manager status */ + +typedef enum ShowStatus { + _SHOW_STATUS_UNSET = -2, + SHOW_STATUS_AUTO = -1, + SHOW_STATUS_NO = 0, + SHOW_STATUS_YES = 1, + SHOW_STATUS_TEMPORARY = 2, +} ShowStatus; + +int parse_show_status(const char *v, ShowStatus *ret); + +int status_vprintf(const char *status, bool ellipse, bool ephemeral, const char *format, va_list ap) _printf_(4,0); +int status_printf(const char *status, bool ellipse, bool ephemeral, const char *format, ...) _printf_(4,5); diff --git a/src/grp-system/libcore/include/core/slice.h b/src/grp-system/libcore/include/core/slice.h new file mode 100644 index 0000000000..c9f3f61067 --- /dev/null +++ b/src/grp-system/libcore/include/core/slice.h @@ -0,0 +1,32 @@ +#pragma once + +/*** + This file is part of systemd. + + Copyright 2013 Lennart Poettering + + systemd is free software; you can redistribute it and/or modify it + under the terms of the GNU Lesser General Public License as published by + the Free Software Foundation; either version 2.1 of the License, or + (at your option) any later version. + + systemd is distributed in the hope that it will be useful, but + WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + Lesser General Public License for more details. + + You should have received a copy of the GNU Lesser General Public License + along with systemd; If not, see . +***/ + +typedef struct Slice Slice; + +struct Slice { + Unit meta; + + SliceState state, deserialized_state; + + CGroupContext cgroup_context; +}; + +extern const UnitVTable slice_vtable; diff --git a/src/grp-system/libcore/include/core/smack-setup.h b/src/grp-system/libcore/include/core/smack-setup.h new file mode 100644 index 0000000000..78164c85e6 --- /dev/null +++ b/src/grp-system/libcore/include/core/smack-setup.h @@ -0,0 +1,24 @@ +#pragma once + +/*** + This file is part of systemd. + + Copyright (C) 2013 Intel Corporation + Authors: + Nathaniel Chen + + systemd is free software; you can redistribute it and/or modify it + under the terms of the GNU Lesser General Public License as published + by the Free Software Foundation; either version 2.1 of the License, + or (at your option) any later version. + + systemd is distributed in the hope that it will be useful, but + WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + Lesser General Public License for more details. + + You should have received a copy of the GNU Lesser General Public License + along with systemd; If not, see . +***/ + +int mac_smack_setup(bool *loaded_policy); diff --git a/src/grp-system/libcore/include/core/socket.h b/src/grp-system/libcore/include/core/socket.h new file mode 100644 index 0000000000..c22500c69c --- /dev/null +++ b/src/grp-system/libcore/include/core/socket.h @@ -0,0 +1,186 @@ +#pragma once + +/*** + This file is part of systemd. + + Copyright 2010 Lennart Poettering + + systemd is free software; you can redistribute it and/or modify it + under the terms of the GNU Lesser General Public License as published by + the Free Software Foundation; either version 2.1 of the License, or + (at your option) any later version. + + systemd is distributed in the hope that it will be useful, but + WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + Lesser General Public License for more details. + + You should have received a copy of the GNU Lesser General Public License + along with systemd; If not, see . +***/ + +#include "systemd-basic/socket-util.h" + +typedef struct Socket Socket; + +#include "mount.h" +#include "service.h" + +typedef enum SocketExecCommand { + SOCKET_EXEC_START_PRE, + SOCKET_EXEC_START_CHOWN, + SOCKET_EXEC_START_POST, + SOCKET_EXEC_STOP_PRE, + SOCKET_EXEC_STOP_POST, + _SOCKET_EXEC_COMMAND_MAX, + _SOCKET_EXEC_COMMAND_INVALID = -1 +} SocketExecCommand; + +typedef enum SocketType { + SOCKET_SOCKET, + SOCKET_FIFO, + SOCKET_SPECIAL, + SOCKET_MQUEUE, + SOCKET_USB_FUNCTION, + _SOCKET_FIFO_MAX, + _SOCKET_FIFO_INVALID = -1 +} SocketType; + +typedef enum SocketResult { + SOCKET_SUCCESS, + SOCKET_FAILURE_RESOURCES, + SOCKET_FAILURE_TIMEOUT, + SOCKET_FAILURE_EXIT_CODE, + SOCKET_FAILURE_SIGNAL, + SOCKET_FAILURE_CORE_DUMP, + SOCKET_FAILURE_START_LIMIT_HIT, + SOCKET_FAILURE_TRIGGER_LIMIT_HIT, + SOCKET_FAILURE_SERVICE_START_LIMIT_HIT, + _SOCKET_RESULT_MAX, + _SOCKET_RESULT_INVALID = -1 +} SocketResult; + +typedef struct SocketPort { + Socket *socket; + + SocketType type; + int fd; + int *auxiliary_fds; + int n_auxiliary_fds; + + SocketAddress address; + char *path; + sd_event_source *event_source; + + LIST_FIELDS(struct SocketPort, port); +} SocketPort; + +struct Socket { + Unit meta; + + LIST_HEAD(SocketPort, ports); + + unsigned n_accepted; + unsigned n_connections; + unsigned max_connections; + + unsigned backlog; + unsigned keep_alive_cnt; + usec_t timeout_usec; + usec_t keep_alive_time; + usec_t keep_alive_interval; + usec_t defer_accept; + + ExecCommand* exec_command[_SOCKET_EXEC_COMMAND_MAX]; + ExecContext exec_context; + KillContext kill_context; + CGroupContext cgroup_context; + ExecRuntime *exec_runtime; + + /* For Accept=no sockets refers to the one service we'll + activate. For Accept=yes sockets is either NULL, or filled + when the next service we spawn. */ + UnitRef service; + + SocketState state, deserialized_state; + + sd_event_source *timer_event_source; + + ExecCommand* control_command; + SocketExecCommand control_command_id; + pid_t control_pid; + + mode_t directory_mode; + mode_t socket_mode; + + SocketResult result; + + char **symlinks; + + bool accept; + bool remove_on_stop; + bool writable; + + int socket_protocol; + + /* Socket options */ + bool keep_alive; + bool no_delay; + bool free_bind; + bool transparent; + bool broadcast; + bool pass_cred; + bool pass_sec; + + /* Only for INET6 sockets: issue IPV6_V6ONLY sockopt */ + SocketAddressBindIPv6Only bind_ipv6_only; + + int priority; + int mark; + size_t receive_buffer; + size_t send_buffer; + int ip_tos; + int ip_ttl; + size_t pipe_size; + char *bind_to_device; + char *tcp_congestion; + bool reuse_port; + long mq_maxmsg; + long mq_msgsize; + + char *smack; + char *smack_ip_in; + char *smack_ip_out; + + bool selinux_context_from_net; + + char *user, *group; + + bool reset_cpu_usage:1; + + char *fdname; + + RateLimit trigger_limit; +}; + +/* Called from the service code when collecting fds */ +int socket_collect_fds(Socket *s, int **fds); + +/* Called from the service code when a per-connection service ended */ +void socket_connection_unref(Socket *s); + +void socket_free_ports(Socket *s); + +int socket_instantiate_service(Socket *s); + +char *socket_fdname(Socket *s); + +extern const UnitVTable socket_vtable; + +const char* socket_exec_command_to_string(SocketExecCommand i) _const_; +SocketExecCommand socket_exec_command_from_string(const char *s) _pure_; + +const char* socket_result_to_string(SocketResult i) _const_; +SocketResult socket_result_from_string(const char *s) _pure_; + +const char* socket_port_type_to_string(SocketPort *p) _pure_; diff --git a/src/grp-system/libcore/include/core/swap.h b/src/grp-system/libcore/include/core/swap.h new file mode 100644 index 0000000000..c2cbfc5d2a --- /dev/null +++ b/src/grp-system/libcore/include/core/swap.h @@ -0,0 +1,110 @@ +#pragma once + +/*** + This file is part of systemd. + + Copyright 2010 Lennart Poettering + Copyright 2010 Maarten Lankhorst + + systemd is free software; you can redistribute it and/or modify it + under the terms of the GNU Lesser General Public License as published by + the Free Software Foundation; either version 2.1 of the License, or + (at your option) any later version. + + systemd is distributed in the hope that it will be useful, but + WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + Lesser General Public License for more details. + + You should have received a copy of the GNU Lesser General Public License + along with systemd; If not, see . +***/ + +#include + +typedef struct Swap Swap; + +typedef enum SwapExecCommand { + SWAP_EXEC_ACTIVATE, + SWAP_EXEC_DEACTIVATE, + _SWAP_EXEC_COMMAND_MAX, + _SWAP_EXEC_COMMAND_INVALID = -1 +} SwapExecCommand; + +typedef enum SwapResult { + SWAP_SUCCESS, + SWAP_FAILURE_RESOURCES, + SWAP_FAILURE_TIMEOUT, + SWAP_FAILURE_EXIT_CODE, + SWAP_FAILURE_SIGNAL, + SWAP_FAILURE_CORE_DUMP, + SWAP_FAILURE_START_LIMIT_HIT, + _SWAP_RESULT_MAX, + _SWAP_RESULT_INVALID = -1 +} SwapResult; + +typedef struct SwapParameters { + char *what; + char *options; + int priority; +} SwapParameters; + +struct Swap { + Unit meta; + + char *what; + + /* If the device has already shown up, this is the device + * node, which might be different from what, due to + * symlinks */ + char *devnode; + + SwapParameters parameters_proc_swaps; + SwapParameters parameters_fragment; + + bool from_proc_swaps:1; + bool from_fragment:1; + + /* Used while looking for swaps that vanished or got added + * from/to /proc/swaps */ + bool is_active:1; + bool just_activated:1; + + bool reset_cpu_usage:1; + + SwapResult result; + + usec_t timeout_usec; + + ExecCommand exec_command[_SWAP_EXEC_COMMAND_MAX]; + ExecContext exec_context; + KillContext kill_context; + CGroupContext cgroup_context; + + ExecRuntime *exec_runtime; + + SwapState state, deserialized_state; + + ExecCommand* control_command; + SwapExecCommand control_command_id; + pid_t control_pid; + + sd_event_source *timer_event_source; + + /* In order to be able to distinguish dependencies on + different device nodes we might end up creating multiple + devices for the same swap. We chain them up here. */ + + LIST_FIELDS(struct Swap, same_devnode); +}; + +extern const UnitVTable swap_vtable; + +int swap_process_device_new(Manager *m, struct udev_device *dev); +int swap_process_device_remove(Manager *m, struct udev_device *dev); + +const char* swap_exec_command_to_string(SwapExecCommand i) _const_; +SwapExecCommand swap_exec_command_from_string(const char *s) _pure_; + +const char* swap_result_to_string(SwapResult i) _const_; +SwapResult swap_result_from_string(const char *s) _pure_; diff --git a/src/grp-system/libcore/include/core/target.h b/src/grp-system/libcore/include/core/target.h new file mode 100644 index 0000000000..339aea154e --- /dev/null +++ b/src/grp-system/libcore/include/core/target.h @@ -0,0 +1,30 @@ +#pragma once + +/*** + This file is part of systemd. + + Copyright 2010 Lennart Poettering + + systemd is free software; you can redistribute it and/or modify it + under the terms of the GNU Lesser General Public License as published by + the Free Software Foundation; either version 2.1 of the License, or + (at your option) any later version. + + systemd is distributed in the hope that it will be useful, but + WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + Lesser General Public License for more details. + + You should have received a copy of the GNU Lesser General Public License + along with systemd; If not, see . +***/ + +typedef struct Target Target; + +struct Target { + Unit meta; + + TargetState state, deserialized_state; +}; + +extern const UnitVTable target_vtable; diff --git a/src/grp-system/libcore/include/core/timer.h b/src/grp-system/libcore/include/core/timer.h new file mode 100644 index 0000000000..9cb30249e3 --- /dev/null +++ b/src/grp-system/libcore/include/core/timer.h @@ -0,0 +1,89 @@ +#pragma once + +/*** + This file is part of systemd. + + Copyright 2010 Lennart Poettering + + systemd is free software; you can redistribute it and/or modify it + under the terms of the GNU Lesser General Public License as published by + the Free Software Foundation; either version 2.1 of the License, or + (at your option) any later version. + + systemd is distributed in the hope that it will be useful, but + WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + Lesser General Public License for more details. + + You should have received a copy of the GNU Lesser General Public License + along with systemd; If not, see . +***/ + +#include "systemd-basic/calendarspec.h" + +typedef struct Timer Timer; + +typedef enum TimerBase { + TIMER_ACTIVE, + TIMER_BOOT, + TIMER_STARTUP, + TIMER_UNIT_ACTIVE, + TIMER_UNIT_INACTIVE, + TIMER_CALENDAR, + _TIMER_BASE_MAX, + _TIMER_BASE_INVALID = -1 +} TimerBase; + +typedef struct TimerValue { + TimerBase base; + bool disabled; + + usec_t value; /* only for monotonic events */ + CalendarSpec *calendar_spec; /* only for calendar events */ + usec_t next_elapse; + + LIST_FIELDS(struct TimerValue, value); +} TimerValue; + +typedef enum TimerResult { + TIMER_SUCCESS, + TIMER_FAILURE_RESOURCES, + TIMER_FAILURE_START_LIMIT_HIT, + _TIMER_RESULT_MAX, + _TIMER_RESULT_INVALID = -1 +} TimerResult; + +struct Timer { + Unit meta; + + usec_t accuracy_usec; + usec_t random_usec; + + LIST_HEAD(TimerValue, values); + usec_t next_elapse_realtime; + usec_t next_elapse_monotonic_or_boottime; + dual_timestamp last_trigger; + + TimerState state, deserialized_state; + + sd_event_source *monotonic_event_source; + sd_event_source *realtime_event_source; + + TimerResult result; + + bool persistent; + bool wake_system; + bool remain_after_elapse; + + char *stamp_path; +}; + +void timer_free_values(Timer *t); + +extern const UnitVTable timer_vtable; + +const char *timer_base_to_string(TimerBase i) _const_; +TimerBase timer_base_from_string(const char *s) _pure_; + +const char* timer_result_to_string(TimerResult i) _const_; +TimerResult timer_result_from_string(const char *s) _pure_; diff --git a/src/grp-system/libcore/include/core/unit.h b/src/grp-system/libcore/include/core/unit.h new file mode 100644 index 0000000000..6c76e352d4 --- /dev/null +++ b/src/grp-system/libcore/include/core/unit.h @@ -0,0 +1,644 @@ +#pragma once + +/*** + This file is part of systemd. + + Copyright 2010 Lennart Poettering + + systemd is free software; you can redistribute it and/or modify it + under the terms of the GNU Lesser General Public License as published by + the Free Software Foundation; either version 2.1 of the License, or + (at your option) any later version. + + systemd is distributed in the hope that it will be useful, but + WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + Lesser General Public License for more details. + + You should have received a copy of the GNU Lesser General Public License + along with systemd; If not, see . +***/ + +#include +#include +#include + +#include "systemd-basic/list.h" +#include "systemd-basic/unit-name.h" +#include "systemd-shared/condition.h" +#include "systemd-shared/install.h" + +typedef struct Unit Unit; +typedef struct UnitRef UnitRef; +typedef struct UnitStatusMessageFormats UnitStatusMessageFormats; +typedef struct UnitVTable UnitVTable; + +#include "failure-action.h" + +typedef enum KillOperation { + KILL_TERMINATE, + KILL_TERMINATE_AND_LOG, + KILL_KILL, + KILL_ABORT, + _KILL_OPERATION_MAX, + _KILL_OPERATION_INVALID = -1 +} KillOperation; + +static inline bool UNIT_IS_ACTIVE_OR_RELOADING(UnitActiveState t) { + return t == UNIT_ACTIVE || t == UNIT_RELOADING; +} + +static inline bool UNIT_IS_ACTIVE_OR_ACTIVATING(UnitActiveState t) { + return t == UNIT_ACTIVE || t == UNIT_ACTIVATING || t == UNIT_RELOADING; +} + +static inline bool UNIT_IS_INACTIVE_OR_DEACTIVATING(UnitActiveState t) { + return t == UNIT_INACTIVE || t == UNIT_FAILED || t == UNIT_DEACTIVATING; +} + +static inline bool UNIT_IS_INACTIVE_OR_FAILED(UnitActiveState t) { + return t == UNIT_INACTIVE || t == UNIT_FAILED; +} + +#include "job.h" + +struct UnitRef { + /* Keeps tracks of references to a unit. This is useful so + * that we can merge two units if necessary and correct all + * references to them */ + + Unit* unit; + LIST_FIELDS(UnitRef, refs); +}; + +struct Unit { + Manager *manager; + + UnitType type; + UnitLoadState load_state; + Unit *merged_into; + + char *id; /* One name is special because we use it for identification. Points to an entry in the names set */ + char *instance; + + Set *names; + Set *dependencies[_UNIT_DEPENDENCY_MAX]; + + char **requires_mounts_for; + + char *description; + char **documentation; + + char *fragment_path; /* if loaded from a config file this is the primary path to it */ + char *source_path; /* if converted, the source file */ + char **dropin_paths; + + usec_t fragment_mtime; + usec_t source_mtime; + usec_t dropin_mtime; + + /* If this is a transient unit we are currently writing, this is where we are writing it to */ + FILE *transient_file; + + /* If there is something to do with this unit, then this is the installed job for it */ + Job *job; + + /* JOB_NOP jobs are special and can be installed without disturbing the real job. */ + Job *nop_job; + + /* The slot used for watching NameOwnerChanged signals */ + sd_bus_slot *match_bus_slot; + + /* Job timeout and action to take */ + usec_t job_timeout; + FailureAction job_timeout_action; + char *job_timeout_reboot_arg; + + /* References to this */ + LIST_HEAD(UnitRef, refs); + + /* Conditions to check */ + LIST_HEAD(Condition, conditions); + LIST_HEAD(Condition, asserts); + + dual_timestamp condition_timestamp; + dual_timestamp assert_timestamp; + + /* Updated whenever the low-level state changes */ + dual_timestamp state_change_timestamp; + + /* Updated whenever the (high-level) active state enters or leaves the active or inactive states */ + dual_timestamp inactive_exit_timestamp; + dual_timestamp active_enter_timestamp; + dual_timestamp active_exit_timestamp; + dual_timestamp inactive_enter_timestamp; + + UnitRef slice; + + /* Per type list */ + LIST_FIELDS(Unit, units_by_type); + + /* All units which have requires_mounts_for set */ + LIST_FIELDS(Unit, has_requires_mounts_for); + + /* Load queue */ + LIST_FIELDS(Unit, load_queue); + + /* D-Bus queue */ + LIST_FIELDS(Unit, dbus_queue); + + /* Cleanup queue */ + LIST_FIELDS(Unit, cleanup_queue); + + /* GC queue */ + LIST_FIELDS(Unit, gc_queue); + + /* CGroup realize members queue */ + LIST_FIELDS(Unit, cgroup_queue); + + /* Units with the same CGroup netclass */ + LIST_FIELDS(Unit, cgroup_netclass); + + /* PIDs we keep an eye on. Note that a unit might have many + * more, but these are the ones we care enough about to + * process SIGCHLD for */ + Set *pids; + + /* Used in sigchld event invocation to avoid repeat events being invoked */ + uint64_t sigchldgen; + + /* Used during GC sweeps */ + unsigned gc_marker; + + /* Error code when we didn't manage to load the unit (negative) */ + int load_error; + + /* Put a ratelimit on unit starting */ + RateLimit start_limit; + FailureAction start_limit_action; + char *reboot_arg; + + /* Make sure we never enter endless loops with the check unneeded logic, or the BindsTo= logic */ + RateLimit auto_stop_ratelimit; + + /* Cached unit file state and preset */ + UnitFileState unit_file_state; + int unit_file_preset; + + /* Where the cpuacct.usage cgroup counter was at the time the unit was started */ + nsec_t cpuacct_usage_base; + + /* Counterparts in the cgroup filesystem */ + char *cgroup_path; + CGroupMask cgroup_realized_mask; + CGroupMask cgroup_enabled_mask; + CGroupMask cgroup_subtree_mask; + CGroupMask cgroup_members_mask; + int cgroup_inotify_wd; + + uint32_t cgroup_netclass_id; + + /* How to start OnFailure units */ + JobMode on_failure_job_mode; + + /* Garbage collect us we nobody wants or requires us anymore */ + bool stop_when_unneeded; + + /* Create default dependencies */ + bool default_dependencies; + + /* Refuse manual starting, allow starting only indirectly via dependency. */ + bool refuse_manual_start; + + /* Don't allow the user to stop this unit manually, allow stopping only indirectly via dependency. */ + bool refuse_manual_stop; + + /* Allow isolation requests */ + bool allow_isolate; + + /* Ignore this unit when isolating */ + bool ignore_on_isolate; + + /* Did the last condition check succeed? */ + bool condition_result; + bool assert_result; + + /* Is this a transient unit? */ + bool transient; + + bool in_load_queue:1; + bool in_dbus_queue:1; + bool in_cleanup_queue:1; + bool in_gc_queue:1; + bool in_cgroup_queue:1; + + bool sent_dbus_new_signal:1; + + bool no_gc:1; + + bool in_audit:1; + + bool cgroup_realized:1; + bool cgroup_members_mask_valid:1; + bool cgroup_subtree_mask_valid:1; + + bool start_limit_hit:1; + + /* Did we already invoke unit_coldplug() for this unit? */ + bool coldplugged:1; +}; + +struct UnitStatusMessageFormats { + const char *starting_stopping[2]; + const char *finished_start_job[_JOB_RESULT_MAX]; + const char *finished_stop_job[_JOB_RESULT_MAX]; +}; + +typedef enum UnitSetPropertiesMode { + UNIT_CHECK = 0, + UNIT_RUNTIME = 1, + UNIT_PERSISTENT = 2, +} UnitSetPropertiesMode; + +#include "automount.h" +#include "busname.h" +#include "device.h" +#include "path.h" +#include "scope.h" +#include "slice.h" +#include "socket.h" +#include "swap.h" +#include "target.h" +#include "timer.h" + +struct UnitVTable { + /* How much memory does an object of this unit type need */ + size_t object_size; + + /* If greater than 0, the offset into the object where + * ExecContext is found, if the unit type has that */ + size_t exec_context_offset; + + /* If greater than 0, the offset into the object where + * CGroupContext is found, if the unit type has that */ + size_t cgroup_context_offset; + + /* If greater than 0, the offset into the object where + * KillContext is found, if the unit type has that */ + size_t kill_context_offset; + + /* If greater than 0, the offset into the object where the + * pointer to ExecRuntime is found, if the unit type has + * that */ + size_t exec_runtime_offset; + + /* The name of the configuration file section with the private settings of this unit */ + const char *private_section; + + /* Config file sections this unit type understands, separated + * by NUL chars */ + const char *sections; + + /* This should reset all type-specific variables. This should + * not allocate memory, and is called with zero-initialized + * data. It should hence only initialize variables that need + * to be set != 0. */ + void (*init)(Unit *u); + + /* This should free all type-specific variables. It should be + * idempotent. */ + void (*done)(Unit *u); + + /* Actually load data from disk. This may fail, and should set + * load_state to UNIT_LOADED, UNIT_MERGED or leave it at + * UNIT_STUB if no configuration could be found. */ + int (*load)(Unit *u); + + /* If a lot of units got created via enumerate(), this is + * where to actually set the state and call unit_notify(). */ + int (*coldplug)(Unit *u); + + void (*dump)(Unit *u, FILE *f, const char *prefix); + + int (*start)(Unit *u); + int (*stop)(Unit *u); + int (*reload)(Unit *u); + + int (*kill)(Unit *u, KillWho w, int signo, sd_bus_error *error); + + bool (*can_reload)(Unit *u); + + /* Write all data that cannot be restored from other sources + * away using unit_serialize_item() */ + int (*serialize)(Unit *u, FILE *f, FDSet *fds); + + /* Restore one item from the serialization */ + int (*deserialize_item)(Unit *u, const char *key, const char *data, FDSet *fds); + + /* Try to match up fds with what we need for this unit */ + void (*distribute_fds)(Unit *u, FDSet *fds); + + /* Boils down the more complex internal state of this unit to + * a simpler one that the engine can understand */ + UnitActiveState (*active_state)(Unit *u); + + /* Returns the substate specific to this unit type as + * string. This is purely information so that we can give the + * user a more fine grained explanation in which actual state a + * unit is in. */ + const char* (*sub_state_to_string)(Unit *u); + + /* Return true when there is reason to keep this entry around + * even nothing references it and it isn't active in any + * way */ + bool (*check_gc)(Unit *u); + + /* When the unit is not running and no job for it queued we + * shall release its runtime resources */ + void (*release_resources)(Unit *u); + + /* Invoked on every child that died */ + void (*sigchld_event)(Unit *u, pid_t pid, int code, int status); + + /* Reset failed state if we are in failed state */ + void (*reset_failed)(Unit *u); + + /* Called whenever any of the cgroups this unit watches for + * ran empty */ + void (*notify_cgroup_empty)(Unit *u); + + /* Called whenever a process of this unit sends us a message */ + void (*notify_message)(Unit *u, pid_t pid, char **tags, FDSet *fds); + + /* Called whenever a name this Unit registered for comes or + * goes away. */ + void (*bus_name_owner_change)(Unit *u, const char *name, const char *old_owner, const char *new_owner); + + /* Called for each property that is being set */ + int (*bus_set_property)(Unit *u, const char *name, sd_bus_message *message, UnitSetPropertiesMode mode, sd_bus_error *error); + + /* Called after at least one property got changed to apply the necessary change */ + int (*bus_commit_properties)(Unit *u); + + /* Return the unit this unit is following */ + Unit *(*following)(Unit *u); + + /* Return the set of units that are following each other */ + int (*following_set)(Unit *u, Set **s); + + /* Invoked each time a unit this unit is triggering changes + * state or gains/loses a job */ + void (*trigger_notify)(Unit *u, Unit *trigger); + + /* Called whenever CLOCK_REALTIME made a jump */ + void (*time_change)(Unit *u); + + /* Returns the next timeout of a unit */ + int (*get_timeout)(Unit *u, usec_t *timeout); + + /* Returns the main PID if there is any defined, or 0. */ + pid_t (*main_pid)(Unit *u); + + /* Returns the main PID if there is any defined, or 0. */ + pid_t (*control_pid)(Unit *u); + + /* This is called for each unit type and should be used to + * enumerate existing devices and load them. However, + * everything that is loaded here should still stay in + * inactive state. It is the job of the coldplug() call above + * to put the units into the initial state. */ + void (*enumerate)(Manager *m); + + /* Type specific cleanups. */ + void (*shutdown)(Manager *m); + + /* If this function is set and return false all jobs for units + * of this type will immediately fail. */ + bool (*supported)(void); + + /* The bus vtable */ + const sd_bus_vtable *bus_vtable; + + /* The strings to print in status messages */ + UnitStatusMessageFormats status_message_formats; + + /* True if transient units of this type are OK */ + bool can_transient:1; +}; + +extern const UnitVTable * const unit_vtable[_UNIT_TYPE_MAX]; + +#define UNIT_VTABLE(u) unit_vtable[(u)->type] + +/* For casting a unit into the various unit types */ +#define DEFINE_CAST(UPPERCASE, MixedCase) \ + static inline MixedCase* UPPERCASE(Unit *u) { \ + if (_unlikely_(!u || u->type != UNIT_##UPPERCASE)) \ + return NULL; \ + \ + return (MixedCase*) u; \ + } + +/* For casting the various unit types into a unit */ +#define UNIT(u) (&(u)->meta) + +#define UNIT_HAS_EXEC_CONTEXT(u) (UNIT_VTABLE(u)->exec_context_offset > 0) +#define UNIT_HAS_CGROUP_CONTEXT(u) (UNIT_VTABLE(u)->cgroup_context_offset > 0) +#define UNIT_HAS_KILL_CONTEXT(u) (UNIT_VTABLE(u)->kill_context_offset > 0) + +#define UNIT_TRIGGER(u) ((Unit*) set_first((u)->dependencies[UNIT_TRIGGERS])) + +DEFINE_CAST(SERVICE, Service); +DEFINE_CAST(SOCKET, Socket); +DEFINE_CAST(BUSNAME, BusName); +DEFINE_CAST(TARGET, Target); +DEFINE_CAST(DEVICE, Device); +DEFINE_CAST(MOUNT, Mount); +DEFINE_CAST(AUTOMOUNT, Automount); +DEFINE_CAST(SWAP, Swap); +DEFINE_CAST(TIMER, Timer); +DEFINE_CAST(PATH, Path); +DEFINE_CAST(SLICE, Slice); +DEFINE_CAST(SCOPE, Scope); + +Unit *unit_new(Manager *m, size_t size); +void unit_free(Unit *u); + +int unit_add_name(Unit *u, const char *name); + +int unit_add_dependency(Unit *u, UnitDependency d, Unit *other, bool add_reference); +int unit_add_two_dependencies(Unit *u, UnitDependency d, UnitDependency e, Unit *other, bool add_reference); + +int unit_add_dependency_by_name(Unit *u, UnitDependency d, const char *name, const char *filename, bool add_reference); +int unit_add_two_dependencies_by_name(Unit *u, UnitDependency d, UnitDependency e, const char *name, const char *path, bool add_reference); + +int unit_add_exec_dependencies(Unit *u, ExecContext *c); + +int unit_choose_id(Unit *u, const char *name); +int unit_set_description(Unit *u, const char *description); + +bool unit_check_gc(Unit *u); + +void unit_add_to_load_queue(Unit *u); +void unit_add_to_dbus_queue(Unit *u); +void unit_add_to_cleanup_queue(Unit *u); +void unit_add_to_gc_queue(Unit *u); + +int unit_merge(Unit *u, Unit *other); +int unit_merge_by_name(Unit *u, const char *other); + +Unit *unit_follow_merge(Unit *u) _pure_; + +int unit_load_fragment_and_dropin(Unit *u); +int unit_load_fragment_and_dropin_optional(Unit *u); +int unit_load(Unit *unit); + +int unit_set_slice(Unit *u, Unit *slice); +int unit_set_default_slice(Unit *u); + +const char *unit_description(Unit *u) _pure_; + +bool unit_has_name(Unit *u, const char *name); + +UnitActiveState unit_active_state(Unit *u); + +const char* unit_sub_state_to_string(Unit *u); + +void unit_dump(Unit *u, FILE *f, const char *prefix); + +bool unit_can_reload(Unit *u) _pure_; +bool unit_can_start(Unit *u) _pure_; +bool unit_can_isolate(Unit *u) _pure_; + +int unit_start(Unit *u); +int unit_stop(Unit *u); +int unit_reload(Unit *u); + +int unit_kill(Unit *u, KillWho w, int signo, sd_bus_error *error); +int unit_kill_common(Unit *u, KillWho who, int signo, pid_t main_pid, pid_t control_pid, sd_bus_error *error); + +void unit_notify(Unit *u, UnitActiveState os, UnitActiveState ns, bool reload_success); + +int unit_watch_pid(Unit *u, pid_t pid); +void unit_unwatch_pid(Unit *u, pid_t pid); +void unit_unwatch_all_pids(Unit *u); + +void unit_tidy_watch_pids(Unit *u, pid_t except1, pid_t except2); + +int unit_install_bus_match(Unit *u, sd_bus *bus, const char *name); +int unit_watch_bus_name(Unit *u, const char *name); +void unit_unwatch_bus_name(Unit *u, const char *name); + +bool unit_job_is_applicable(Unit *u, JobType j); + +int set_unit_path(const char *p); + +char *unit_dbus_path(Unit *u); + +int unit_load_related_unit(Unit *u, const char *type, Unit **_found); + +bool unit_can_serialize(Unit *u) _pure_; + +int unit_serialize(Unit *u, FILE *f, FDSet *fds, bool serialize_jobs); +int unit_deserialize(Unit *u, FILE *f, FDSet *fds); + +int unit_serialize_item(Unit *u, FILE *f, const char *key, const char *value); +int unit_serialize_item_escaped(Unit *u, FILE *f, const char *key, const char *value); +int unit_serialize_item_fd(Unit *u, FILE *f, FDSet *fds, const char *key, int fd); +void unit_serialize_item_format(Unit *u, FILE *f, const char *key, const char *value, ...) _printf_(4,5); + +int unit_add_node_link(Unit *u, const char *what, bool wants, UnitDependency d); + +int unit_coldplug(Unit *u); + +void unit_status_printf(Unit *u, const char *status, const char *unit_status_msg_format) _printf_(3, 0); +void unit_status_emit_starting_stopping_reloading(Unit *u, JobType t); + +bool unit_need_daemon_reload(Unit *u); + +void unit_reset_failed(Unit *u); + +Unit *unit_following(Unit *u); +int unit_following_set(Unit *u, Set **s); + +const char *unit_slice_name(Unit *u); + +bool unit_stop_pending(Unit *u) _pure_; +bool unit_inactive_or_pending(Unit *u) _pure_; +bool unit_active_or_pending(Unit *u); + +int unit_add_default_target_dependency(Unit *u, Unit *target); + +void unit_start_on_failure(Unit *u); +void unit_trigger_notify(Unit *u); + +UnitFileState unit_get_unit_file_state(Unit *u); +int unit_get_unit_file_preset(Unit *u); + +Unit* unit_ref_set(UnitRef *ref, Unit *u); +void unit_ref_unset(UnitRef *ref); + +#define UNIT_DEREF(ref) ((ref).unit) +#define UNIT_ISSET(ref) (!!(ref).unit) + +int unit_patch_contexts(Unit *u); + +ExecContext *unit_get_exec_context(Unit *u) _pure_; +KillContext *unit_get_kill_context(Unit *u) _pure_; +CGroupContext *unit_get_cgroup_context(Unit *u) _pure_; + +ExecRuntime *unit_get_exec_runtime(Unit *u) _pure_; + +int unit_setup_exec_runtime(Unit *u); + +int unit_write_drop_in(Unit *u, UnitSetPropertiesMode mode, const char *name, const char *data); +int unit_write_drop_in_format(Unit *u, UnitSetPropertiesMode mode, const char *name, const char *format, ...) _printf_(4,5); + +int unit_write_drop_in_private(Unit *u, UnitSetPropertiesMode mode, const char *name, const char *data); +int unit_write_drop_in_private_format(Unit *u, UnitSetPropertiesMode mode, const char *name, const char *format, ...) _printf_(4,5); + +int unit_kill_context(Unit *u, KillContext *c, KillOperation k, pid_t main_pid, pid_t control_pid, bool main_pid_alien); + +int unit_make_transient(Unit *u); + +int unit_require_mounts_for(Unit *u, const char *path); + +bool unit_type_supported(UnitType t); + +bool unit_is_pristine(Unit *u); + +pid_t unit_control_pid(Unit *u); +pid_t unit_main_pid(Unit *u); + +static inline bool unit_supported(Unit *u) { + return unit_type_supported(u->type); +} + +void unit_warn_if_dir_nonempty(Unit *u, const char* where); +int unit_fail_if_symlink(Unit *u, const char* where); + +int unit_start_limit_test(Unit *u); + +/* Macros which append UNIT= or USER_UNIT= to the message */ + +#define log_unit_full(unit, level, error, ...) \ + ({ \ + Unit *_u = (unit); \ + _u ? log_object_internal(level, error, __FILE__, __LINE__, __func__, _u->manager->unit_log_field, _u->id, ##__VA_ARGS__) : \ + log_internal(level, error, __FILE__, __LINE__, __func__, ##__VA_ARGS__); \ + }) + +#define log_unit_debug(unit, ...) log_unit_full(unit, LOG_DEBUG, 0, ##__VA_ARGS__) +#define log_unit_info(unit, ...) log_unit_full(unit, LOG_INFO, 0, ##__VA_ARGS__) +#define log_unit_notice(unit, ...) log_unit_full(unit, LOG_NOTICE, 0, ##__VA_ARGS__) +#define log_unit_warning(unit, ...) log_unit_full(unit, LOG_WARNING, 0, ##__VA_ARGS__) +#define log_unit_error(unit, ...) log_unit_full(unit, LOG_ERR, 0, ##__VA_ARGS__) + +#define log_unit_debug_errno(unit, error, ...) log_unit_full(unit, LOG_DEBUG, error, ##__VA_ARGS__) +#define log_unit_info_errno(unit, error, ...) log_unit_full(unit, LOG_INFO, error, ##__VA_ARGS__) +#define log_unit_notice_errno(unit, error, ...) log_unit_full(unit, LOG_NOTICE, error, ##__VA_ARGS__) +#define log_unit_warning_errno(unit, error, ...) log_unit_full(unit, LOG_WARNING, error, ##__VA_ARGS__) +#define log_unit_error_errno(unit, error, ...) log_unit_full(unit, LOG_ERR, error, ##__VA_ARGS__) + +#define LOG_UNIT_MESSAGE(unit, fmt, ...) "MESSAGE=%s: " fmt, (unit)->id, ##__VA_ARGS__ +#define LOG_UNIT_ID(unit) (unit)->manager->unit_log_format_string, (unit)->id diff --git a/src/grp-system/libcore/src/Makefile b/src/grp-system/libcore/src/Makefile new file mode 100644 index 0000000000..60150788d8 --- /dev/null +++ b/src/grp-system/libcore/src/Makefile @@ -0,0 +1,26 @@ +# -*- Mode: makefile; indent-tabs-mode: t -*- +# +# This file is part of systemd. +# +# Copyright 2010-2012 Lennart Poettering +# Copyright 2010-2012 Kay Sievers +# Copyright 2013 Zbigniew Jędrzejewski-Szmek +# Copyright 2013 David Strauss +# Copyright 2016 Luke Shumaker +# +# systemd is free software; you can redistribute it and/or modify it +# under the terms of the GNU Lesser General Public License as published by +# the Free Software Foundation; either version 2.1 of the License, or +# (at your option) any later version. +# +# systemd is distributed in the hope that it will be useful, but +# WITHOUT ANY WARRANTY; without even the implied warranty of +# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU +# Lesser General Public License for more details. +# +# You should have received a copy of the GNU Lesser General Public License +# along with systemd; If not, see . +include $(dir $(lastword $(MAKEFILE_LIST)))/../../../../config.mk +include $(topsrcdir)/build-aux/Makefile.head.mk + +include $(topsrcdir)/build-aux/Makefile.tail.mk diff --git a/src/grp-system/libcore/src/audit-fd.c b/src/grp-system/libcore/src/audit-fd.c new file mode 100644 index 0000000000..12509951b7 --- /dev/null +++ b/src/grp-system/libcore/src/audit-fd.c @@ -0,0 +1,73 @@ +/*** + This file is part of systemd. + + Copyright 2012 Lennart Poettering + + systemd is free software; you can redistribute it and/or modify it + under the terms of the GNU Lesser General Public License as published by + the Free Software Foundation; either version 2.1 of the License, or + (at your option) any later version. + + systemd is distributed in the hope that it will be useful, but + WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + Lesser General Public License for more details. + + You should have received a copy of the GNU Lesser General Public License + along with systemd; If not, see . +***/ + + +#include + +#include "audit-fd.h" + +#ifdef HAVE_AUDIT + +#include +#include + +#include "systemd-basic/fd-util.h" +#include "systemd-basic/log.h" +#include "systemd-basic/util.h" + +static bool initialized = false; +static int audit_fd; + +int get_audit_fd(void) { + + if (!initialized) { + audit_fd = audit_open(); + + if (audit_fd < 0) { + if (errno != EAFNOSUPPORT && errno != EPROTONOSUPPORT) + log_error_errno(errno, "Failed to connect to audit log: %m"); + + audit_fd = errno ? -errno : -EINVAL; + } + + initialized = true; + } + + return audit_fd; +} + +void close_audit_fd(void) { + + if (initialized && audit_fd >= 0) + safe_close(audit_fd); + + initialized = true; + audit_fd = -ECONNRESET; +} + +#else + +int get_audit_fd(void) { + return -EAFNOSUPPORT; +} + +void close_audit_fd(void) { +} + +#endif diff --git a/src/grp-system/libcore/src/audit-fd.h b/src/grp-system/libcore/src/audit-fd.h new file mode 100644 index 0000000000..0eccb59210 --- /dev/null +++ b/src/grp-system/libcore/src/audit-fd.h @@ -0,0 +1,23 @@ +#pragma once + +/*** + This file is part of systemd. + + Copyright 2012 Lennart Poettering + + systemd is free software; you can redistribute it and/or modify it + under the terms of the GNU Lesser General Public License as published by + the Free Software Foundation; either version 2.1 of the License, or + (at your option) any later version. + + systemd is distributed in the hope that it will be useful, but + WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + Lesser General Public License for more details. + + You should have received a copy of the GNU Lesser General Public License + along with systemd; If not, see . +***/ + +int get_audit_fd(void); +void close_audit_fd(void); diff --git a/src/grp-system/libcore/src/automount.c b/src/grp-system/libcore/src/automount.c new file mode 100644 index 0000000000..2f116977de --- /dev/null +++ b/src/grp-system/libcore/src/automount.c @@ -0,0 +1,1124 @@ +/*** + This file is part of systemd. + + Copyright 2010 Lennart Poettering + + systemd is free software; you can redistribute it and/or modify it + under the terms of the GNU Lesser General Public License as published by + the Free Software Foundation; either version 2.1 of the License, or + (at your option) any later version. + + systemd is distributed in the hope that it will be useful, but + WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + Lesser General Public License for more details. + + You should have received a copy of the GNU Lesser General Public License + along with systemd; If not, see . +***/ + +#include +#include +#include +#include +#include +#include +#include + +#include +#include + +#include "core/automount.h" +#include "core/mount.h" +#include "core/unit.h" +#include "sd-bus/bus-error.h" +#include "sd-bus/bus-util.h" +#include "systemd-basic/alloc-util.h" +#include "systemd-basic/async.h" +#include "systemd-basic/fd-util.h" +#include "systemd-basic/formats-util.h" +#include "systemd-basic/io-util.h" +#include "systemd-basic/label.h" +#include "systemd-basic/mkdir.h" +#include "systemd-basic/mount-util.h" +#include "systemd-basic/parse-util.h" +#include "systemd-basic/path-util.h" +#include "systemd-basic/process-util.h" +#include "systemd-basic/special.h" +#include "systemd-basic/stdio-util.h" +#include "systemd-basic/string-table.h" +#include "systemd-basic/string-util.h" +#include "systemd-basic/unit-name.h" + +#include "dbus-automount.h" + +static const UnitActiveState state_translation_table[_AUTOMOUNT_STATE_MAX] = { + [AUTOMOUNT_DEAD] = UNIT_INACTIVE, + [AUTOMOUNT_WAITING] = UNIT_ACTIVE, + [AUTOMOUNT_RUNNING] = UNIT_ACTIVE, + [AUTOMOUNT_FAILED] = UNIT_FAILED +}; + +struct expire_data { + int dev_autofs_fd; + int ioctl_fd; +}; + +static inline void expire_data_free(struct expire_data *data) { + if (!data) + return; + + safe_close(data->dev_autofs_fd); + safe_close(data->ioctl_fd); + free(data); +} + +DEFINE_TRIVIAL_CLEANUP_FUNC(struct expire_data*, expire_data_free); + +static int open_dev_autofs(Manager *m); +static int automount_dispatch_io(sd_event_source *s, int fd, uint32_t events, void *userdata); +static int automount_start_expire(Automount *a); +static void automount_stop_expire(Automount *a); +static int automount_send_ready(Automount *a, Set *tokens, int status); + +static void automount_init(Unit *u) { + Automount *a = AUTOMOUNT(u); + + assert(u); + assert(u->load_state == UNIT_STUB); + + a->pipe_fd = -1; + a->directory_mode = 0755; + UNIT(a)->ignore_on_isolate = true; +} + +static void unmount_autofs(Automount *a) { + int r; + + assert(a); + + if (a->pipe_fd < 0) + return; + + a->pipe_event_source = sd_event_source_unref(a->pipe_event_source); + a->pipe_fd = safe_close(a->pipe_fd); + + /* If we reload/reexecute things we keep the mount point + * around */ + if (a->where && + (UNIT(a)->manager->exit_code != MANAGER_RELOAD && + UNIT(a)->manager->exit_code != MANAGER_REEXECUTE)) { + automount_send_ready(a, a->tokens, -EHOSTDOWN); + automount_send_ready(a, a->expire_tokens, -EHOSTDOWN); + + r = repeat_unmount(a->where, MNT_DETACH); + if (r < 0) + log_error_errno(r, "Failed to unmount: %m"); + } +} + +static void automount_done(Unit *u) { + Automount *a = AUTOMOUNT(u); + + assert(a); + + unmount_autofs(a); + + a->where = mfree(a->where); + + a->tokens = set_free(a->tokens); + a->expire_tokens = set_free(a->expire_tokens); + + a->expire_event_source = sd_event_source_unref(a->expire_event_source); +} + +static int automount_add_mount_links(Automount *a) { + _cleanup_free_ char *parent = NULL; + + assert(a); + + parent = dirname_malloc(a->where); + if (!parent) + return -ENOMEM; + + return unit_require_mounts_for(UNIT(a), parent); +} + +static int automount_add_default_dependencies(Automount *a) { + int r; + + assert(a); + + if (!UNIT(a)->default_dependencies) + return 0; + + if (!MANAGER_IS_SYSTEM(UNIT(a)->manager)) + return 0; + + r = unit_add_two_dependencies_by_name(UNIT(a), UNIT_BEFORE, UNIT_CONFLICTS, SPECIAL_UMOUNT_TARGET, NULL, true); + if (r < 0) + return r; + + return 0; +} + +static int automount_verify(Automount *a) { + _cleanup_free_ char *e = NULL; + int r; + + assert(a); + + if (UNIT(a)->load_state != UNIT_LOADED) + return 0; + + if (path_equal(a->where, "/")) { + log_unit_error(UNIT(a), "Cannot have an automount unit for the root directory. Refusing."); + return -EINVAL; + } + + r = unit_name_from_path(a->where, ".automount", &e); + if (r < 0) + return log_unit_error(UNIT(a), "Failed to generate unit name from path: %m"); + + if (!unit_has_name(UNIT(a), e)) { + log_unit_error(UNIT(a), "Where= setting doesn't match unit name. Refusing."); + return -EINVAL; + } + + return 0; +} + +static int automount_load(Unit *u) { + Automount *a = AUTOMOUNT(u); + int r; + + assert(u); + assert(u->load_state == UNIT_STUB); + + /* Load a .automount file */ + r = unit_load_fragment_and_dropin_optional(u); + if (r < 0) + return r; + + if (u->load_state == UNIT_LOADED) { + Unit *x; + + if (!a->where) { + r = unit_name_to_path(u->id, &a->where); + if (r < 0) + return r; + } + + path_kill_slashes(a->where); + + r = unit_load_related_unit(u, ".mount", &x); + if (r < 0) + return r; + + r = unit_add_two_dependencies(u, UNIT_BEFORE, UNIT_TRIGGERS, x, true); + if (r < 0) + return r; + + r = automount_add_mount_links(a); + if (r < 0) + return r; + + r = automount_add_default_dependencies(a); + if (r < 0) + return r; + } + + return automount_verify(a); +} + +static void automount_set_state(Automount *a, AutomountState state) { + AutomountState old_state; + assert(a); + + old_state = a->state; + a->state = state; + + if (state != AUTOMOUNT_RUNNING) + automount_stop_expire(a); + + if (state != AUTOMOUNT_WAITING && + state != AUTOMOUNT_RUNNING) + unmount_autofs(a); + + if (state != old_state) + log_unit_debug(UNIT(a), "Changed %s -> %s", automount_state_to_string(old_state), automount_state_to_string(state)); + + unit_notify(UNIT(a), state_translation_table[old_state], state_translation_table[state], true); +} + +static int automount_coldplug(Unit *u) { + Automount *a = AUTOMOUNT(u); + int r; + + assert(a); + assert(a->state == AUTOMOUNT_DEAD); + + if (a->deserialized_state != a->state) { + + r = open_dev_autofs(u->manager); + if (r < 0) + return r; + + if (a->deserialized_state == AUTOMOUNT_WAITING || + a->deserialized_state == AUTOMOUNT_RUNNING) { + assert(a->pipe_fd >= 0); + + r = sd_event_add_io(u->manager->event, &a->pipe_event_source, a->pipe_fd, EPOLLIN, automount_dispatch_io, u); + if (r < 0) + return r; + + (void) sd_event_source_set_description(a->pipe_event_source, "automount-io"); + } + + automount_set_state(a, a->deserialized_state); + } + + return 0; +} + +static void automount_dump(Unit *u, FILE *f, const char *prefix) { + char time_string[FORMAT_TIMESPAN_MAX]; + Automount *a = AUTOMOUNT(u); + + assert(a); + + fprintf(f, + "%sAutomount State: %s\n" + "%sResult: %s\n" + "%sWhere: %s\n" + "%sDirectoryMode: %04o\n" + "%sTimeoutIdleUSec: %s\n", + prefix, automount_state_to_string(a->state), + prefix, automount_result_to_string(a->result), + prefix, a->where, + prefix, a->directory_mode, + prefix, format_timespan(time_string, FORMAT_TIMESPAN_MAX, a->timeout_idle_usec, USEC_PER_SEC)); +} + +static void automount_enter_dead(Automount *a, AutomountResult f) { + assert(a); + + if (f != AUTOMOUNT_SUCCESS) + a->result = f; + + automount_set_state(a, a->result != AUTOMOUNT_SUCCESS ? AUTOMOUNT_FAILED : AUTOMOUNT_DEAD); +} + +static int open_dev_autofs(Manager *m) { + struct autofs_dev_ioctl param; + + assert(m); + + if (m->dev_autofs_fd >= 0) + return m->dev_autofs_fd; + + label_fix("/dev/autofs", false, false); + + m->dev_autofs_fd = open("/dev/autofs", O_CLOEXEC|O_RDONLY); + if (m->dev_autofs_fd < 0) + return log_error_errno(errno, "Failed to open /dev/autofs: %m"); + + init_autofs_dev_ioctl(¶m); + if (ioctl(m->dev_autofs_fd, AUTOFS_DEV_IOCTL_VERSION, ¶m) < 0) { + m->dev_autofs_fd = safe_close(m->dev_autofs_fd); + return -errno; + } + + log_debug("Autofs kernel version %i.%i", param.ver_major, param.ver_minor); + + return m->dev_autofs_fd; +} + +static int open_ioctl_fd(int dev_autofs_fd, const char *where, dev_t devid) { + struct autofs_dev_ioctl *param; + size_t l; + + assert(dev_autofs_fd >= 0); + assert(where); + + l = sizeof(struct autofs_dev_ioctl) + strlen(where) + 1; + param = alloca(l); + + init_autofs_dev_ioctl(param); + param->size = l; + param->ioctlfd = -1; + param->openmount.devid = devid; + strcpy(param->path, where); + + if (ioctl(dev_autofs_fd, AUTOFS_DEV_IOCTL_OPENMOUNT, param) < 0) + return -errno; + + if (param->ioctlfd < 0) + return -EIO; + + (void) fd_cloexec(param->ioctlfd, true); + return param->ioctlfd; +} + +static int autofs_protocol(int dev_autofs_fd, int ioctl_fd) { + uint32_t major, minor; + struct autofs_dev_ioctl param; + + assert(dev_autofs_fd >= 0); + assert(ioctl_fd >= 0); + + init_autofs_dev_ioctl(¶m); + param.ioctlfd = ioctl_fd; + + if (ioctl(dev_autofs_fd, AUTOFS_DEV_IOCTL_PROTOVER, ¶m) < 0) + return -errno; + + major = param.protover.version; + + init_autofs_dev_ioctl(¶m); + param.ioctlfd = ioctl_fd; + + if (ioctl(dev_autofs_fd, AUTOFS_DEV_IOCTL_PROTOSUBVER, ¶m) < 0) + return -errno; + + minor = param.protosubver.sub_version; + + log_debug("Autofs protocol version %i.%i", major, minor); + return 0; +} + +static int autofs_set_timeout(int dev_autofs_fd, int ioctl_fd, usec_t usec) { + struct autofs_dev_ioctl param; + + assert(dev_autofs_fd >= 0); + assert(ioctl_fd >= 0); + + init_autofs_dev_ioctl(¶m); + param.ioctlfd = ioctl_fd; + + /* Convert to seconds, rounding up. */ + param.timeout.timeout = (usec + USEC_PER_SEC - 1) / USEC_PER_SEC; + + if (ioctl(dev_autofs_fd, AUTOFS_DEV_IOCTL_TIMEOUT, ¶m) < 0) + return -errno; + + return 0; +} + +static int autofs_send_ready(int dev_autofs_fd, int ioctl_fd, uint32_t token, int status) { + struct autofs_dev_ioctl param; + + assert(dev_autofs_fd >= 0); + assert(ioctl_fd >= 0); + + init_autofs_dev_ioctl(¶m); + param.ioctlfd = ioctl_fd; + + if (status != 0) { + param.fail.token = token; + param.fail.status = status; + } else + param.ready.token = token; + + if (ioctl(dev_autofs_fd, status ? AUTOFS_DEV_IOCTL_FAIL : AUTOFS_DEV_IOCTL_READY, ¶m) < 0) + return -errno; + + return 0; +} + +static int automount_send_ready(Automount *a, Set *tokens, int status) { + _cleanup_close_ int ioctl_fd = -1; + unsigned token; + int r; + + assert(a); + assert(status <= 0); + + if (set_isempty(tokens)) + return 0; + + ioctl_fd = open_ioctl_fd(UNIT(a)->manager->dev_autofs_fd, a->where, a->dev_id); + if (ioctl_fd < 0) + return ioctl_fd; + + if (status != 0) + log_unit_debug_errno(UNIT(a), status, "Sending failure: %m"); + else + log_unit_debug(UNIT(a), "Sending success."); + + r = 0; + + /* Autofs thankfully does not hand out 0 as a token */ + while ((token = PTR_TO_UINT(set_steal_first(tokens)))) { + int k; + + /* Autofs fun fact II: + * + * if you pass a positive status code here, the kernel will + * freeze! Yay! */ + + k = autofs_send_ready(UNIT(a)->manager->dev_autofs_fd, + ioctl_fd, + token, + status); + if (k < 0) + r = k; + } + + return r; +} + +static void automount_trigger_notify(Unit *u, Unit *other) { + Automount *a = AUTOMOUNT(u); + int r; + + assert(a); + assert(other); + + /* Filter out invocations with bogus state */ + if (other->load_state != UNIT_LOADED || other->type != UNIT_MOUNT) + return; + + /* Don't propagate state changes from the mount if we are already down */ + if (!IN_SET(a->state, AUTOMOUNT_WAITING, AUTOMOUNT_RUNNING)) + return; + + /* Propagate start limit hit state */ + if (other->start_limit_hit) { + automount_enter_dead(a, AUTOMOUNT_FAILURE_MOUNT_START_LIMIT_HIT); + return; + } + + /* Don't propagate anything if there's still a job queued */ + if (other->job) + return; + + /* The mount is successfully established */ + if (IN_SET(MOUNT(other)->state, MOUNT_MOUNTED, MOUNT_REMOUNTING)) { + (void) automount_send_ready(a, a->tokens, 0); + + r = automount_start_expire(a); + if (r < 0) + log_unit_warning_errno(UNIT(a), r, "Failed to start expiration timer, ignoring: %m"); + + automount_set_state(a, AUTOMOUNT_RUNNING); + } + + if (IN_SET(MOUNT(other)->state, + MOUNT_MOUNTING, MOUNT_MOUNTING_DONE, + MOUNT_MOUNTED, MOUNT_REMOUNTING, + MOUNT_MOUNTING_SIGTERM, MOUNT_MOUNTING_SIGKILL, + MOUNT_REMOUNTING_SIGTERM, MOUNT_REMOUNTING_SIGKILL, + MOUNT_UNMOUNTING_SIGTERM, MOUNT_UNMOUNTING_SIGKILL, + MOUNT_FAILED)) { + + (void) automount_send_ready(a, a->expire_tokens, -ENODEV); + } + + if (MOUNT(other)->state == MOUNT_DEAD) + (void) automount_send_ready(a, a->expire_tokens, 0); + + /* The mount is in some unhappy state now, let's unfreeze any waiting clients */ + if (IN_SET(MOUNT(other)->state, + MOUNT_DEAD, MOUNT_UNMOUNTING, + MOUNT_MOUNTING_SIGTERM, MOUNT_MOUNTING_SIGKILL, + MOUNT_REMOUNTING_SIGTERM, MOUNT_REMOUNTING_SIGKILL, + MOUNT_UNMOUNTING_SIGTERM, MOUNT_UNMOUNTING_SIGKILL, + MOUNT_FAILED)) { + + (void) automount_send_ready(a, a->tokens, -ENODEV); + + automount_set_state(a, AUTOMOUNT_WAITING); + } +} + +static void automount_enter_waiting(Automount *a) { + _cleanup_close_ int ioctl_fd = -1; + int p[2] = { -1, -1 }; + char name[sizeof("systemd-")-1 + DECIMAL_STR_MAX(pid_t) + 1]; + char options[sizeof("fd=,pgrp=,minproto=5,maxproto=5,direct")-1 + + DECIMAL_STR_MAX(int) + DECIMAL_STR_MAX(gid_t) + 1]; + bool mounted = false; + int r, dev_autofs_fd; + struct stat st; + + assert(a); + assert(a->pipe_fd < 0); + assert(a->where); + + set_clear(a->tokens); + + r = unit_fail_if_symlink(UNIT(a), a->where); + if (r < 0) + goto fail; + + (void) mkdir_p_label(a->where, 0555); + + unit_warn_if_dir_nonempty(UNIT(a), a->where); + + dev_autofs_fd = open_dev_autofs(UNIT(a)->manager); + if (dev_autofs_fd < 0) { + r = dev_autofs_fd; + goto fail; + } + + if (pipe2(p, O_NONBLOCK|O_CLOEXEC) < 0) { + r = -errno; + goto fail; + } + + xsprintf(options, "fd=%i,pgrp="PID_FMT",minproto=5,maxproto=5,direct", p[1], getpgrp()); + xsprintf(name, "systemd-"PID_FMT, getpid()); + if (mount(name, a->where, "autofs", 0, options) < 0) { + r = -errno; + goto fail; + } + + mounted = true; + + p[1] = safe_close(p[1]); + + if (stat(a->where, &st) < 0) { + r = -errno; + goto fail; + } + + ioctl_fd = open_ioctl_fd(dev_autofs_fd, a->where, st.st_dev); + if (ioctl_fd < 0) { + r = ioctl_fd; + goto fail; + } + + r = autofs_protocol(dev_autofs_fd, ioctl_fd); + if (r < 0) + goto fail; + + r = autofs_set_timeout(dev_autofs_fd, ioctl_fd, a->timeout_idle_usec); + if (r < 0) + goto fail; + + /* Autofs fun fact: + * + * Unless we close the ioctl fd here, for some weird reason + * the direct mount will not receive events from the + * kernel. */ + + r = sd_event_add_io(UNIT(a)->manager->event, &a->pipe_event_source, p[0], EPOLLIN, automount_dispatch_io, a); + if (r < 0) + goto fail; + + (void) sd_event_source_set_description(a->pipe_event_source, "automount-io"); + + a->pipe_fd = p[0]; + a->dev_id = st.st_dev; + + automount_set_state(a, AUTOMOUNT_WAITING); + + return; + +fail: + log_unit_error_errno(UNIT(a), r, "Failed to initialize automounter: %m"); + + safe_close_pair(p); + + if (mounted) { + r = repeat_unmount(a->where, MNT_DETACH); + if (r < 0) + log_error_errno(r, "Failed to unmount, ignoring: %m"); + } + + automount_enter_dead(a, AUTOMOUNT_FAILURE_RESOURCES); +} + +static void *expire_thread(void *p) { + struct autofs_dev_ioctl param; + _cleanup_(expire_data_freep) struct expire_data *data = (struct expire_data*)p; + int r; + + assert(data->dev_autofs_fd >= 0); + assert(data->ioctl_fd >= 0); + + init_autofs_dev_ioctl(¶m); + param.ioctlfd = data->ioctl_fd; + + do { + r = ioctl(data->dev_autofs_fd, AUTOFS_DEV_IOCTL_EXPIRE, ¶m); + } while (r >= 0); + + if (errno != EAGAIN) + log_warning_errno(errno, "Failed to expire automount, ignoring: %m"); + + return NULL; +} + +static int automount_dispatch_expire(sd_event_source *source, usec_t usec, void *userdata) { + Automount *a = AUTOMOUNT(userdata); + _cleanup_(expire_data_freep) struct expire_data *data = NULL; + int r; + + assert(a); + assert(source == a->expire_event_source); + + data = new0(struct expire_data, 1); + if (!data) + return log_oom(); + + data->ioctl_fd = -1; + + data->dev_autofs_fd = fcntl(UNIT(a)->manager->dev_autofs_fd, F_DUPFD_CLOEXEC, 3); + if (data->dev_autofs_fd < 0) + return log_unit_error_errno(UNIT(a), errno, "Failed to duplicate autofs fd: %m"); + + data->ioctl_fd = open_ioctl_fd(UNIT(a)->manager->dev_autofs_fd, a->where, a->dev_id); + if (data->ioctl_fd < 0) + return log_unit_error_errno(UNIT(a), data->ioctl_fd, "Couldn't open autofs ioctl fd: %m"); + + r = asynchronous_job(expire_thread, data); + if (r < 0) + return log_unit_error_errno(UNIT(a), r, "Failed to start expire job: %m"); + + data = NULL; + + return automount_start_expire(a); +} + +static int automount_start_expire(Automount *a) { + int r; + usec_t timeout; + + assert(a); + + if (a->timeout_idle_usec == 0) + return 0; + + timeout = now(CLOCK_MONOTONIC) + MAX(a->timeout_idle_usec/3, USEC_PER_SEC); + + if (a->expire_event_source) { + r = sd_event_source_set_time(a->expire_event_source, timeout); + if (r < 0) + return r; + + return sd_event_source_set_enabled(a->expire_event_source, SD_EVENT_ONESHOT); + } + + r = sd_event_add_time( + UNIT(a)->manager->event, + &a->expire_event_source, + CLOCK_MONOTONIC, timeout, 0, + automount_dispatch_expire, a); + if (r < 0) + return r; + + (void) sd_event_source_set_description(a->expire_event_source, "automount-expire"); + + return 0; +} + +static void automount_stop_expire(Automount *a) { + assert(a); + + if (!a->expire_event_source) + return; + + (void) sd_event_source_set_enabled(a->expire_event_source, SD_EVENT_OFF); +} + +static void automount_enter_runnning(Automount *a) { + _cleanup_(sd_bus_error_free) sd_bus_error error = SD_BUS_ERROR_NULL; + struct stat st; + int r; + + assert(a); + + /* We don't take mount requests anymore if we are supposed to + * shut down anyway */ + if (unit_stop_pending(UNIT(a))) { + log_unit_debug(UNIT(a), "Suppressing automount request since unit stop is scheduled."); + automount_send_ready(a, a->tokens, -EHOSTDOWN); + automount_send_ready(a, a->expire_tokens, -EHOSTDOWN); + return; + } + + mkdir_p_label(a->where, a->directory_mode); + + /* Before we do anything, let's see if somebody is playing games with us? */ + if (lstat(a->where, &st) < 0) { + log_unit_warning_errno(UNIT(a), errno, "Failed to stat automount point: %m"); + goto fail; + } + + if (!S_ISDIR(st.st_mode) || st.st_dev != a->dev_id) + log_unit_info(UNIT(a), "Automount point already active?"); + else { + Unit *trigger; + + trigger = UNIT_TRIGGER(UNIT(a)); + if (!trigger) { + log_unit_error(UNIT(a), "Unit to trigger vanished."); + goto fail; + } + + r = manager_add_job(UNIT(a)->manager, JOB_START, trigger, JOB_REPLACE, &error, NULL); + if (r < 0) { + log_unit_warning(UNIT(a), "Failed to queue mount startup job: %s", bus_error_message(&error, r)); + goto fail; + } + } + + automount_set_state(a, AUTOMOUNT_RUNNING); + return; + +fail: + automount_enter_dead(a, AUTOMOUNT_FAILURE_RESOURCES); +} + +static int automount_start(Unit *u) { + Automount *a = AUTOMOUNT(u); + Unit *trigger; + int r; + + assert(a); + assert(a->state == AUTOMOUNT_DEAD || a->state == AUTOMOUNT_FAILED); + + if (path_is_mount_point(a->where, 0) > 0) { + log_unit_error(u, "Path %s is already a mount point, refusing start.", a->where); + return -EEXIST; + } + + trigger = UNIT_TRIGGER(u); + if (!trigger || trigger->load_state != UNIT_LOADED) { + log_unit_error(u, "Refusing to start, unit to trigger not loaded."); + return -ENOENT; + } + + r = unit_start_limit_test(u); + if (r < 0) { + automount_enter_dead(a, AUTOMOUNT_FAILURE_START_LIMIT_HIT); + return r; + } + + a->result = AUTOMOUNT_SUCCESS; + automount_enter_waiting(a); + return 1; +} + +static int automount_stop(Unit *u) { + Automount *a = AUTOMOUNT(u); + + assert(a); + assert(a->state == AUTOMOUNT_WAITING || a->state == AUTOMOUNT_RUNNING); + + automount_enter_dead(a, AUTOMOUNT_SUCCESS); + return 1; +} + +static int automount_serialize(Unit *u, FILE *f, FDSet *fds) { + Automount *a = AUTOMOUNT(u); + Iterator i; + void *p; + int r; + + assert(a); + assert(f); + assert(fds); + + unit_serialize_item(u, f, "state", automount_state_to_string(a->state)); + unit_serialize_item(u, f, "result", automount_result_to_string(a->result)); + unit_serialize_item_format(u, f, "dev-id", "%u", (unsigned) a->dev_id); + + SET_FOREACH(p, a->tokens, i) + unit_serialize_item_format(u, f, "token", "%u", PTR_TO_UINT(p)); + SET_FOREACH(p, a->expire_tokens, i) + unit_serialize_item_format(u, f, "expire-token", "%u", PTR_TO_UINT(p)); + + r = unit_serialize_item_fd(u, f, fds, "pipe-fd", a->pipe_fd); + if (r < 0) + return r; + + return 0; +} + +static int automount_deserialize_item(Unit *u, const char *key, const char *value, FDSet *fds) { + Automount *a = AUTOMOUNT(u); + int r; + + assert(a); + assert(fds); + + if (streq(key, "state")) { + AutomountState state; + + state = automount_state_from_string(value); + if (state < 0) + log_unit_debug(u, "Failed to parse state value: %s", value); + else + a->deserialized_state = state; + } else if (streq(key, "result")) { + AutomountResult f; + + f = automount_result_from_string(value); + if (f < 0) + log_unit_debug(u, "Failed to parse result value: %s", value); + else if (f != AUTOMOUNT_SUCCESS) + a->result = f; + + } else if (streq(key, "dev-id")) { + unsigned d; + + if (safe_atou(value, &d) < 0) + log_unit_debug(u, "Failed to parse dev-id value: %s", value); + else + a->dev_id = (unsigned) d; + } else if (streq(key, "token")) { + unsigned token; + + if (safe_atou(value, &token) < 0) + log_unit_debug(u, "Failed to parse token value: %s", value); + else { + r = set_ensure_allocated(&a->tokens, NULL); + if (r < 0) { + log_oom(); + return 0; + } + + r = set_put(a->tokens, UINT_TO_PTR(token)); + if (r < 0) + log_unit_error_errno(u, r, "Failed to add token to set: %m"); + } + } else if (streq(key, "expire-token")) { + unsigned token; + + if (safe_atou(value, &token) < 0) + log_unit_debug(u, "Failed to parse token value: %s", value); + else { + r = set_ensure_allocated(&a->expire_tokens, NULL); + if (r < 0) { + log_oom(); + return 0; + } + + r = set_put(a->expire_tokens, UINT_TO_PTR(token)); + if (r < 0) + log_unit_error_errno(u, r, "Failed to add expire token to set: %m"); + } + } else if (streq(key, "pipe-fd")) { + int fd; + + if (safe_atoi(value, &fd) < 0 || fd < 0 || !fdset_contains(fds, fd)) + log_unit_debug(u, "Failed to parse pipe-fd value: %s", value); + else { + safe_close(a->pipe_fd); + a->pipe_fd = fdset_remove(fds, fd); + } + } else + log_unit_debug(u, "Unknown serialization key: %s", key); + + return 0; +} + +static UnitActiveState automount_active_state(Unit *u) { + assert(u); + + return state_translation_table[AUTOMOUNT(u)->state]; +} + +static const char *automount_sub_state_to_string(Unit *u) { + assert(u); + + return automount_state_to_string(AUTOMOUNT(u)->state); +} + +static bool automount_check_gc(Unit *u) { + assert(u); + + if (!UNIT_TRIGGER(u)) + return false; + + return UNIT_VTABLE(UNIT_TRIGGER(u))->check_gc(UNIT_TRIGGER(u)); +} + +static int automount_dispatch_io(sd_event_source *s, int fd, uint32_t events, void *userdata) { + _cleanup_(sd_bus_error_free) sd_bus_error error = SD_BUS_ERROR_NULL; + union autofs_v5_packet_union packet; + Automount *a = AUTOMOUNT(userdata); + struct stat st; + Unit *trigger; + int r; + + assert(a); + assert(fd == a->pipe_fd); + + if (events != EPOLLIN) { + log_unit_error(UNIT(a), "Got invalid poll event %"PRIu32" on pipe (fd=%d)", events, fd); + goto fail; + } + + r = loop_read_exact(a->pipe_fd, &packet, sizeof(packet), true); + if (r < 0) { + log_unit_error_errno(UNIT(a), r, "Invalid read from pipe: %m"); + goto fail; + } + + switch (packet.hdr.type) { + + case autofs_ptype_missing_direct: + + if (packet.v5_packet.pid > 0) { + _cleanup_free_ char *p = NULL; + + get_process_comm(packet.v5_packet.pid, &p); + log_unit_info(UNIT(a), "Got automount request for %s, triggered by %"PRIu32" (%s)", a->where, packet.v5_packet.pid, strna(p)); + } else + log_unit_debug(UNIT(a), "Got direct mount request on %s", a->where); + + r = set_ensure_allocated(&a->tokens, NULL); + if (r < 0) { + log_unit_error(UNIT(a), "Failed to allocate token set."); + goto fail; + } + + r = set_put(a->tokens, UINT_TO_PTR(packet.v5_packet.wait_queue_token)); + if (r < 0) { + log_unit_error_errno(UNIT(a), r, "Failed to remember token: %m"); + goto fail; + } + + automount_enter_runnning(a); + break; + + case autofs_ptype_expire_direct: + log_unit_debug(UNIT(a), "Got direct umount request on %s", a->where); + + automount_stop_expire(a); + + r = set_ensure_allocated(&a->expire_tokens, NULL); + if (r < 0) { + log_unit_error(UNIT(a), "Failed to allocate token set."); + goto fail; + } + + r = set_put(a->expire_tokens, UINT_TO_PTR(packet.v5_packet.wait_queue_token)); + if (r < 0) { + log_unit_error_errno(UNIT(a), r, "Failed to remember token: %m"); + goto fail; + } + + /* Before we do anything, let's see if somebody is playing games with us? */ + if (lstat(a->where, &st) < 0) { + log_unit_warning_errno(UNIT(a), errno, "Failed to stat automount point: %m"); + goto fail; + } + + if (!S_ISDIR(st.st_mode) || st.st_dev == a->dev_id) { + log_unit_info(UNIT(a), "Automount point already unmounted?"); + automount_send_ready(a, a->expire_tokens, 0); + break; + } + + trigger = UNIT_TRIGGER(UNIT(a)); + if (!trigger) { + log_unit_error(UNIT(a), "Unit to trigger vanished."); + goto fail; + } + + r = manager_add_job(UNIT(a)->manager, JOB_STOP, trigger, JOB_REPLACE, &error, NULL); + if (r < 0) { + log_unit_warning(UNIT(a), "Failed to queue umount startup job: %s", bus_error_message(&error, r)); + goto fail; + } + break; + + default: + log_unit_error(UNIT(a), "Received unknown automount request %i", packet.hdr.type); + break; + } + + return 0; + +fail: + automount_enter_dead(a, AUTOMOUNT_FAILURE_RESOURCES); + return 0; +} + +static void automount_shutdown(Manager *m) { + assert(m); + + m->dev_autofs_fd = safe_close(m->dev_autofs_fd); +} + +static void automount_reset_failed(Unit *u) { + Automount *a = AUTOMOUNT(u); + + assert(a); + + if (a->state == AUTOMOUNT_FAILED) + automount_set_state(a, AUTOMOUNT_DEAD); + + a->result = AUTOMOUNT_SUCCESS; +} + +static bool automount_supported(void) { + static int supported = -1; + + if (supported < 0) + supported = access("/dev/autofs", F_OK) >= 0; + + return supported; +} + +static const char* const automount_result_table[_AUTOMOUNT_RESULT_MAX] = { + [AUTOMOUNT_SUCCESS] = "success", + [AUTOMOUNT_FAILURE_RESOURCES] = "resources", + [AUTOMOUNT_FAILURE_START_LIMIT_HIT] = "start-limit-hit", + [AUTOMOUNT_FAILURE_MOUNT_START_LIMIT_HIT] = "mount-start-limit-hit", +}; + +DEFINE_STRING_TABLE_LOOKUP(automount_result, AutomountResult); + +const UnitVTable automount_vtable = { + .object_size = sizeof(Automount), + + .sections = + "Unit\0" + "Automount\0" + "Install\0", + + .init = automount_init, + .load = automount_load, + .done = automount_done, + + .coldplug = automount_coldplug, + + .dump = automount_dump, + + .start = automount_start, + .stop = automount_stop, + + .serialize = automount_serialize, + .deserialize_item = automount_deserialize_item, + + .active_state = automount_active_state, + .sub_state_to_string = automount_sub_state_to_string, + + .check_gc = automount_check_gc, + + .trigger_notify = automount_trigger_notify, + + .reset_failed = automount_reset_failed, + + .bus_vtable = bus_automount_vtable, + + .shutdown = automount_shutdown, + .supported = automount_supported, + + .status_message_formats = { + .finished_start_job = { + [JOB_DONE] = "Set up automount %s.", + [JOB_FAILED] = "Failed to set up automount %s.", + }, + .finished_stop_job = { + [JOB_DONE] = "Unset automount %s.", + [JOB_FAILED] = "Failed to unset automount %s.", + }, + }, +}; diff --git a/src/grp-system/libcore/src/bus-policy.c b/src/grp-system/libcore/src/bus-policy.c new file mode 100644 index 0000000000..d5bba61a93 --- /dev/null +++ b/src/grp-system/libcore/src/bus-policy.c @@ -0,0 +1,180 @@ +/*** + This file is part of systemd. + + Copyright 2014 Daniel Mack + + systemd is free software; you can redistribute it and/or modify it + under the terms of the GNU Lesser General Public License as published by + the Free Software Foundation; either version 2.1 of the License, or + (at your option) any later version. + + systemd is distributed in the hope that it will be useful, but + WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + Lesser General Public License for more details. + + You should have received a copy of the GNU Lesser General Public License + along with systemd; If not, see . +***/ + +#include + +#include "core/bus-policy.h" +#include "sd-bus/bus-kernel.h" +#include "sd-bus/kdbus.h" +#include "systemd-basic/alloc-util.h" +#include "systemd-basic/string-table.h" +#include "systemd-basic/user-util.h" +#include "systemd-basic/util.h" + +int bus_kernel_translate_access(BusPolicyAccess access) { + assert(access >= 0); + assert(access < _BUS_POLICY_ACCESS_MAX); + + switch (access) { + + case BUS_POLICY_ACCESS_SEE: + return KDBUS_POLICY_SEE; + + case BUS_POLICY_ACCESS_TALK: + return KDBUS_POLICY_TALK; + + case BUS_POLICY_ACCESS_OWN: + return KDBUS_POLICY_OWN; + + default: + assert_not_reached("Unknown policy access"); + } +} + +int bus_kernel_translate_policy(const BusNamePolicy *policy, struct kdbus_item *item) { + int r; + + assert(policy); + assert(item); + + switch (policy->type) { + + case BUSNAME_POLICY_TYPE_USER: { + const char *user = policy->name; + uid_t uid; + + r = get_user_creds(&user, &uid, NULL, NULL, NULL); + if (r < 0) + return r; + + item->policy_access.type = KDBUS_POLICY_ACCESS_USER; + item->policy_access.id = uid; + break; + } + + case BUSNAME_POLICY_TYPE_GROUP: { + const char *group = policy->name; + gid_t gid; + + r = get_group_creds(&group, &gid); + if (r < 0) + return r; + + item->policy_access.type = KDBUS_POLICY_ACCESS_GROUP; + item->policy_access.id = gid; + break; + } + + default: + assert_not_reached("Unknown policy type"); + } + + item->policy_access.access = bus_kernel_translate_access(policy->access); + + return 0; +} + +int bus_kernel_make_starter( + int fd, + const char *name, + bool activating, + bool accept_fd, + BusNamePolicy *policy, + BusPolicyAccess world_policy) { + + struct kdbus_cmd_free cmd_free = { .size = sizeof(cmd_free) }; + struct kdbus_cmd_hello *hello; + struct kdbus_item *n; + size_t policy_cnt = 0; + BusNamePolicy *po; + size_t size; + int r; + + assert(fd >= 0); + assert(name); + + LIST_FOREACH(policy, po, policy) + policy_cnt++; + + if (world_policy >= 0) + policy_cnt++; + + size = offsetof(struct kdbus_cmd_hello, items) + + ALIGN8(offsetof(struct kdbus_item, str) + strlen(name) + 1) + + policy_cnt * ALIGN8(offsetof(struct kdbus_item, policy_access) + sizeof(struct kdbus_policy_access)); + + hello = alloca0_align(size, 8); + + n = hello->items; + strcpy(n->str, name); + n->size = offsetof(struct kdbus_item, str) + strlen(n->str) + 1; + n->type = KDBUS_ITEM_NAME; + n = KDBUS_ITEM_NEXT(n); + + LIST_FOREACH(policy, po, policy) { + n->type = KDBUS_ITEM_POLICY_ACCESS; + n->size = offsetof(struct kdbus_item, policy_access) + sizeof(struct kdbus_policy_access); + + r = bus_kernel_translate_policy(po, n); + if (r < 0) + return r; + + n = KDBUS_ITEM_NEXT(n); + } + + if (world_policy >= 0) { + n->type = KDBUS_ITEM_POLICY_ACCESS; + n->size = offsetof(struct kdbus_item, policy_access) + sizeof(struct kdbus_policy_access); + n->policy_access.type = KDBUS_POLICY_ACCESS_WORLD; + n->policy_access.access = bus_kernel_translate_access(world_policy); + } + + hello->size = size; + hello->flags = + (activating ? KDBUS_HELLO_ACTIVATOR : KDBUS_HELLO_POLICY_HOLDER) | + (accept_fd ? KDBUS_HELLO_ACCEPT_FD : 0); + hello->pool_size = KDBUS_POOL_SIZE; + hello->attach_flags_send = _KDBUS_ATTACH_ANY; + hello->attach_flags_recv = _KDBUS_ATTACH_ANY; + + if (ioctl(fd, KDBUS_CMD_HELLO, hello) < 0) { + if (errno == ENOTTY) /* Major API change */ + return -ESOCKTNOSUPPORT; + return -errno; + } + + /* not interested in any output values */ + cmd_free.offset = hello->offset; + (void) ioctl(fd, KDBUS_CMD_FREE, &cmd_free); + + /* The higher 32bit of the bus_flags fields are considered + * 'incompatible flags'. Refuse them all for now. */ + if (hello->bus_flags > 0xFFFFFFFFULL) + return -ESOCKTNOSUPPORT; + + return fd; +} + +static const char* const bus_policy_access_table[_BUS_POLICY_ACCESS_MAX] = { + [BUS_POLICY_ACCESS_SEE] = "see", + [BUS_POLICY_ACCESS_TALK] = "talk", + [BUS_POLICY_ACCESS_OWN] = "own", +}; + +DEFINE_STRING_TABLE_LOOKUP(bus_policy_access, BusPolicyAccess); diff --git a/src/grp-system/libcore/src/busname.c b/src/grp-system/libcore/src/busname.c new file mode 100644 index 0000000000..178866fc15 --- /dev/null +++ b/src/grp-system/libcore/src/busname.c @@ -0,0 +1,1078 @@ +/*** + This file is part of systemd. + + Copyright 2013 Lennart Poettering + + systemd is free software; you can redistribute it and/or modify it + under the terms of the GNU Lesser General Public License as published by + the Free Software Foundation; either version 2.1 of the License, or + (at your option) any later version. + + systemd is distributed in the hope that it will be useful, but + WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + Lesser General Public License for more details. + + You should have received a copy of the GNU Lesser General Public License + along with systemd; If not, see . +***/ + +#include + +#include "core/bus-policy.h" +#include "core/busname.h" +#include "core/service.h" +#include "sd-bus/bus-internal.h" +#include "sd-bus/bus-kernel.h" +#include "sd-bus/bus-util.h" +#include "sd-bus/kdbus.h" +#include "systemd-basic/alloc-util.h" +#include "systemd-basic/fd-util.h" +#include "systemd-basic/formats-util.h" +#include "systemd-basic/parse-util.h" +#include "systemd-basic/process-util.h" +#include "systemd-basic/signal-util.h" +#include "systemd-basic/special.h" +#include "systemd-basic/string-table.h" +#include "systemd-basic/string-util.h" + +#include "dbus-busname.h" + +static const UnitActiveState state_translation_table[_BUSNAME_STATE_MAX] = { + [BUSNAME_DEAD] = UNIT_INACTIVE, + [BUSNAME_MAKING] = UNIT_ACTIVATING, + [BUSNAME_REGISTERED] = UNIT_ACTIVE, + [BUSNAME_LISTENING] = UNIT_ACTIVE, + [BUSNAME_RUNNING] = UNIT_ACTIVE, + [BUSNAME_SIGTERM] = UNIT_DEACTIVATING, + [BUSNAME_SIGKILL] = UNIT_DEACTIVATING, + [BUSNAME_FAILED] = UNIT_FAILED +}; + +static int busname_dispatch_io(sd_event_source *source, int fd, uint32_t revents, void *userdata); +static int busname_dispatch_timer(sd_event_source *source, usec_t usec, void *userdata); + +static void busname_init(Unit *u) { + BusName *n = BUSNAME(u); + + assert(u); + assert(u->load_state == UNIT_STUB); + + n->starter_fd = -1; + n->accept_fd = true; + n->activating = true; + + n->timeout_usec = u->manager->default_timeout_start_usec; +} + +static void busname_unwatch_control_pid(BusName *n) { + assert(n); + + if (n->control_pid <= 0) + return; + + unit_unwatch_pid(UNIT(n), n->control_pid); + n->control_pid = 0; +} + +static void busname_free_policy(BusName *n) { + BusNamePolicy *p; + + assert(n); + + while ((p = n->policy)) { + LIST_REMOVE(policy, n->policy, p); + + free(p->name); + free(p); + } +} + +static void busname_close_fd(BusName *n) { + assert(n); + + n->starter_event_source = sd_event_source_unref(n->starter_event_source); + n->starter_fd = safe_close(n->starter_fd); +} + +static void busname_done(Unit *u) { + BusName *n = BUSNAME(u); + + assert(n); + + n->name = mfree(n->name); + + busname_free_policy(n); + busname_unwatch_control_pid(n); + busname_close_fd(n); + + unit_ref_unset(&n->service); + + n->timer_event_source = sd_event_source_unref(n->timer_event_source); +} + +static int busname_arm_timer(BusName *n, usec_t usec) { + int r; + + assert(n); + + if (n->timer_event_source) { + r = sd_event_source_set_time(n->timer_event_source, usec); + if (r < 0) + return r; + + return sd_event_source_set_enabled(n->timer_event_source, SD_EVENT_ONESHOT); + } + + if (usec == USEC_INFINITY) + return 0; + + r = sd_event_add_time( + UNIT(n)->manager->event, + &n->timer_event_source, + CLOCK_MONOTONIC, + usec, 0, + busname_dispatch_timer, n); + if (r < 0) + return r; + + (void) sd_event_source_set_description(n->timer_event_source, "busname-timer"); + + return 0; +} + +static int busname_add_default_default_dependencies(BusName *n) { + int r; + + assert(n); + + r = unit_add_dependency_by_name(UNIT(n), UNIT_BEFORE, SPECIAL_BUSNAMES_TARGET, NULL, true); + if (r < 0) + return r; + + if (MANAGER_IS_SYSTEM(UNIT(n)->manager)) { + r = unit_add_two_dependencies_by_name(UNIT(n), UNIT_AFTER, UNIT_REQUIRES, SPECIAL_SYSINIT_TARGET, NULL, true); + if (r < 0) + return r; + } + + return unit_add_two_dependencies_by_name(UNIT(n), UNIT_BEFORE, UNIT_CONFLICTS, SPECIAL_SHUTDOWN_TARGET, NULL, true); +} + +static int busname_add_extras(BusName *n) { + Unit *u = UNIT(n); + int r; + + assert(n); + + if (!n->name) { + r = unit_name_to_prefix(u->id, &n->name); + if (r < 0) + return r; + } + + if (!u->description) { + r = unit_set_description(u, n->name); + if (r < 0) + return r; + } + + if (n->activating) { + if (!UNIT_DEREF(n->service)) { + Unit *x; + + r = unit_load_related_unit(u, ".service", &x); + if (r < 0) + return r; + + unit_ref_set(&n->service, x); + } + + r = unit_add_two_dependencies(u, UNIT_BEFORE, UNIT_TRIGGERS, UNIT_DEREF(n->service), true); + if (r < 0) + return r; + } + + if (u->default_dependencies) { + r = busname_add_default_default_dependencies(n); + if (r < 0) + return r; + } + + return 0; +} + +static int busname_verify(BusName *n) { + char *e; + + assert(n); + + if (UNIT(n)->load_state != UNIT_LOADED) + return 0; + + if (!service_name_is_valid(n->name)) { + log_unit_error(UNIT(n), "Name= setting is not a valid service name Refusing."); + return -EINVAL; + } + + e = strjoina(n->name, ".busname"); + if (!unit_has_name(UNIT(n), e)) { + log_unit_error(UNIT(n), "Name= setting doesn't match unit name. Refusing."); + return -EINVAL; + } + + return 0; +} + +static int busname_load(Unit *u) { + BusName *n = BUSNAME(u); + int r; + + assert(u); + assert(u->load_state == UNIT_STUB); + + r = unit_load_fragment_and_dropin(u); + if (r < 0) + return r; + + if (u->load_state == UNIT_LOADED) { + /* This is a new unit? Then let's add in some extras */ + r = busname_add_extras(n); + if (r < 0) + return r; + } + + return busname_verify(n); +} + +static void busname_dump(Unit *u, FILE *f, const char *prefix) { + BusName *n = BUSNAME(u); + + assert(n); + assert(f); + + fprintf(f, + "%sBus Name State: %s\n" + "%sResult: %s\n" + "%sName: %s\n" + "%sActivating: %s\n" + "%sAccept FD: %s\n", + prefix, busname_state_to_string(n->state), + prefix, busname_result_to_string(n->result), + prefix, n->name, + prefix, yes_no(n->activating), + prefix, yes_no(n->accept_fd)); + + if (n->control_pid > 0) + fprintf(f, + "%sControl PID: "PID_FMT"\n", + prefix, n->control_pid); +} + +static void busname_unwatch_fd(BusName *n) { + int r; + + assert(n); + + if (!n->starter_event_source) + return; + + r = sd_event_source_set_enabled(n->starter_event_source, SD_EVENT_OFF); + if (r < 0) + log_unit_debug_errno(UNIT(n), r, "Failed to disable event source: %m"); +} + +static int busname_watch_fd(BusName *n) { + int r; + + assert(n); + + if (n->starter_fd < 0) + return 0; + + if (n->starter_event_source) { + r = sd_event_source_set_enabled(n->starter_event_source, SD_EVENT_ON); + if (r < 0) + goto fail; + } else { + r = sd_event_add_io(UNIT(n)->manager->event, &n->starter_event_source, n->starter_fd, EPOLLIN, busname_dispatch_io, n); + if (r < 0) + goto fail; + + (void) sd_event_source_set_description(n->starter_event_source, "busname-starter"); + } + + return 0; + +fail: + log_unit_warning_errno(UNIT(n), r, "Failed to watch starter fd: %m"); + busname_unwatch_fd(n); + return r; +} + +static int busname_open_fd(BusName *n) { + _cleanup_free_ char *path = NULL; + const char *mode; + + assert(n); + + if (n->starter_fd >= 0) + return 0; + + mode = MANAGER_IS_SYSTEM(UNIT(n)->manager) ? "system" : "user"; + n->starter_fd = bus_kernel_open_bus_fd(mode, &path); + if (n->starter_fd < 0) + return log_unit_warning_errno(UNIT(n), n->starter_fd, "Failed to open %s: %m", path ?: "kdbus"); + + return 0; +} + +static void busname_set_state(BusName *n, BusNameState state) { + BusNameState old_state; + assert(n); + + old_state = n->state; + n->state = state; + + if (!IN_SET(state, BUSNAME_MAKING, BUSNAME_SIGTERM, BUSNAME_SIGKILL)) { + n->timer_event_source = sd_event_source_unref(n->timer_event_source); + busname_unwatch_control_pid(n); + } + + if (state != BUSNAME_LISTENING) + busname_unwatch_fd(n); + + if (!IN_SET(state, BUSNAME_LISTENING, BUSNAME_MAKING, BUSNAME_REGISTERED, BUSNAME_RUNNING)) + busname_close_fd(n); + + if (state != old_state) + log_unit_debug(UNIT(n), "Changed %s -> %s", busname_state_to_string(old_state), busname_state_to_string(state)); + + unit_notify(UNIT(n), state_translation_table[old_state], state_translation_table[state], true); +} + +static int busname_coldplug(Unit *u) { + BusName *n = BUSNAME(u); + int r; + + assert(n); + assert(n->state == BUSNAME_DEAD); + + if (n->deserialized_state == n->state) + return 0; + + if (n->control_pid > 0 && + pid_is_unwaited(n->control_pid) && + IN_SET(n->deserialized_state, BUSNAME_MAKING, BUSNAME_SIGTERM, BUSNAME_SIGKILL)) { + + r = unit_watch_pid(UNIT(n), n->control_pid); + if (r < 0) + return r; + + r = busname_arm_timer(n, usec_add(u->state_change_timestamp.monotonic, n->timeout_usec)); + if (r < 0) + return r; + } + + if (IN_SET(n->deserialized_state, BUSNAME_MAKING, BUSNAME_LISTENING, BUSNAME_REGISTERED, BUSNAME_RUNNING)) { + r = busname_open_fd(n); + if (r < 0) + return r; + } + + if (n->deserialized_state == BUSNAME_LISTENING) { + r = busname_watch_fd(n); + if (r < 0) + return r; + } + + busname_set_state(n, n->deserialized_state); + return 0; +} + +static int busname_make_starter(BusName *n, pid_t *_pid) { + pid_t pid; + int r; + + r = busname_arm_timer(n, usec_add(now(CLOCK_MONOTONIC), n->timeout_usec)); + if (r < 0) + goto fail; + + /* We have to resolve the user/group names out-of-process, + * hence let's fork here. It's messy, but well, what can we + * do? */ + + pid = fork(); + if (pid < 0) + return -errno; + + if (pid == 0) { + int ret; + + (void) default_signals(SIGNALS_CRASH_HANDLER, SIGNALS_IGNORE, -1); + (void) ignore_signals(SIGPIPE, -1); + log_forget_fds(); + + r = bus_kernel_make_starter(n->starter_fd, n->name, n->activating, n->accept_fd, n->policy, n->policy_world); + if (r < 0) { + ret = EXIT_MAKE_STARTER; + goto fail_child; + } + + _exit(0); + + fail_child: + log_open(); + log_error_errno(r, "Failed to create starter connection at step %s: %m", exit_status_to_string(ret, EXIT_STATUS_SYSTEMD)); + + _exit(ret); + } + + r = unit_watch_pid(UNIT(n), pid); + if (r < 0) + goto fail; + + *_pid = pid; + return 0; + +fail: + n->timer_event_source = sd_event_source_unref(n->timer_event_source); + return r; +} + +static void busname_enter_dead(BusName *n, BusNameResult f) { + assert(n); + + if (f != BUSNAME_SUCCESS) + n->result = f; + + busname_set_state(n, n->result != BUSNAME_SUCCESS ? BUSNAME_FAILED : BUSNAME_DEAD); +} + +static void busname_enter_signal(BusName *n, BusNameState state, BusNameResult f) { + KillContext kill_context = {}; + int r; + + assert(n); + + if (f != BUSNAME_SUCCESS) + n->result = f; + + kill_context_init(&kill_context); + + r = unit_kill_context(UNIT(n), + &kill_context, + state != BUSNAME_SIGTERM ? KILL_KILL : KILL_TERMINATE, + -1, + n->control_pid, + false); + if (r < 0) { + log_unit_warning_errno(UNIT(n), r, "Failed to kill control process: %m"); + goto fail; + } + + if (r > 0) { + r = busname_arm_timer(n, usec_add(now(CLOCK_MONOTONIC), n->timeout_usec)); + if (r < 0) { + log_unit_warning_errno(UNIT(n), r, "Failed to arm timer: %m"); + goto fail; + } + + busname_set_state(n, state); + } else if (state == BUSNAME_SIGTERM) + busname_enter_signal(n, BUSNAME_SIGKILL, BUSNAME_SUCCESS); + else + busname_enter_dead(n, BUSNAME_SUCCESS); + + return; + +fail: + busname_enter_dead(n, BUSNAME_FAILURE_RESOURCES); +} + +static void busname_enter_listening(BusName *n) { + int r; + + assert(n); + + if (n->activating) { + r = busname_watch_fd(n); + if (r < 0) { + log_unit_warning_errno(UNIT(n), r, "Failed to watch names: %m"); + goto fail; + } + + busname_set_state(n, BUSNAME_LISTENING); + } else + busname_set_state(n, BUSNAME_REGISTERED); + + return; + +fail: + busname_enter_signal(n, BUSNAME_SIGTERM, BUSNAME_FAILURE_RESOURCES); +} + +static void busname_enter_making(BusName *n) { + int r; + + assert(n); + + r = busname_open_fd(n); + if (r < 0) + goto fail; + + if (n->policy) { + /* If there is a policy, we need to resolve user/group + * names, which we can't do from PID1, hence let's + * fork. */ + busname_unwatch_control_pid(n); + + r = busname_make_starter(n, &n->control_pid); + if (r < 0) { + log_unit_warning_errno(UNIT(n), r, "Failed to fork 'making' task: %m"); + goto fail; + } + + busname_set_state(n, BUSNAME_MAKING); + } else { + /* If there is no policy, we can do everything + * directly from PID 1, hence do so. */ + + r = bus_kernel_make_starter(n->starter_fd, n->name, n->activating, n->accept_fd, NULL, n->policy_world); + if (r < 0) { + log_unit_warning_errno(UNIT(n), r, "Failed to make starter: %m"); + goto fail; + } + + busname_enter_listening(n); + } + + return; + +fail: + busname_enter_dead(n, BUSNAME_FAILURE_RESOURCES); +} + +static void busname_enter_running(BusName *n) { + _cleanup_(sd_bus_error_free) sd_bus_error error = SD_BUS_ERROR_NULL; + bool pending = false; + Unit *other; + Iterator i; + int r; + + assert(n); + + if (!n->activating) + return; + + /* We don't take connections anymore if we are supposed to + * shut down anyway */ + + if (unit_stop_pending(UNIT(n))) { + log_unit_debug(UNIT(n), "Suppressing activation request since unit stop is scheduled."); + + /* Flush all queued activation reqeuest by closing and reopening the connection */ + bus_kernel_drop_one(n->starter_fd); + + busname_enter_listening(n); + return; + } + + /* If there's already a start pending don't bother to do + * anything */ + SET_FOREACH(other, UNIT(n)->dependencies[UNIT_TRIGGERS], i) + if (unit_active_or_pending(other)) { + pending = true; + break; + } + + if (!pending) { + if (!UNIT_ISSET(n->service)) { + log_unit_error(UNIT(n), "Service to activate vanished, refusing activation."); + r = -ENOENT; + goto fail; + } + + r = manager_add_job(UNIT(n)->manager, JOB_START, UNIT_DEREF(n->service), JOB_REPLACE, &error, NULL); + if (r < 0) + goto fail; + } + + busname_set_state(n, BUSNAME_RUNNING); + return; + +fail: + log_unit_warning(UNIT(n), "Failed to queue service startup job: %s", bus_error_message(&error, r)); + busname_enter_dead(n, BUSNAME_FAILURE_RESOURCES); +} + +static int busname_start(Unit *u) { + BusName *n = BUSNAME(u); + int r; + + assert(n); + + /* We cannot fulfill this request right now, try again later + * please! */ + if (IN_SET(n->state, BUSNAME_SIGTERM, BUSNAME_SIGKILL)) + return -EAGAIN; + + /* Already on it! */ + if (n->state == BUSNAME_MAKING) + return 0; + + if (n->activating && UNIT_ISSET(n->service)) { + Service *service; + + service = SERVICE(UNIT_DEREF(n->service)); + + if (UNIT(service)->load_state != UNIT_LOADED) { + log_unit_error(u, "Bus service %s not loaded, refusing.", UNIT(service)->id); + return -ENOENT; + } + } + + assert(IN_SET(n->state, BUSNAME_DEAD, BUSNAME_FAILED)); + + r = unit_start_limit_test(u); + if (r < 0) { + busname_enter_dead(n, BUSNAME_FAILURE_START_LIMIT_HIT); + return r; + } + + n->result = BUSNAME_SUCCESS; + busname_enter_making(n); + + return 1; +} + +static int busname_stop(Unit *u) { + BusName *n = BUSNAME(u); + + assert(n); + + /* Already on it */ + if (IN_SET(n->state, BUSNAME_SIGTERM, BUSNAME_SIGKILL)) + return 0; + + /* If there's already something running, we go directly into + * kill mode. */ + + if (n->state == BUSNAME_MAKING) { + busname_enter_signal(n, BUSNAME_SIGTERM, BUSNAME_SUCCESS); + return -EAGAIN; + } + + assert(IN_SET(n->state, BUSNAME_REGISTERED, BUSNAME_LISTENING, BUSNAME_RUNNING)); + + busname_enter_dead(n, BUSNAME_SUCCESS); + return 1; +} + +static int busname_serialize(Unit *u, FILE *f, FDSet *fds) { + BusName *n = BUSNAME(u); + int r; + + assert(n); + assert(f); + assert(fds); + + unit_serialize_item(u, f, "state", busname_state_to_string(n->state)); + unit_serialize_item(u, f, "result", busname_result_to_string(n->result)); + + if (n->control_pid > 0) + unit_serialize_item_format(u, f, "control-pid", PID_FMT, n->control_pid); + + r = unit_serialize_item_fd(u, f, fds, "starter-fd", n->starter_fd); + if (r < 0) + return r; + + return 0; +} + +static int busname_deserialize_item(Unit *u, const char *key, const char *value, FDSet *fds) { + BusName *n = BUSNAME(u); + + assert(n); + assert(key); + assert(value); + + if (streq(key, "state")) { + BusNameState state; + + state = busname_state_from_string(value); + if (state < 0) + log_unit_debug(u, "Failed to parse state value: %s", value); + else + n->deserialized_state = state; + + } else if (streq(key, "result")) { + BusNameResult f; + + f = busname_result_from_string(value); + if (f < 0) + log_unit_debug(u, "Failed to parse result value: %s", value); + else if (f != BUSNAME_SUCCESS) + n->result = f; + + } else if (streq(key, "control-pid")) { + pid_t pid; + + if (parse_pid(value, &pid) < 0) + log_unit_debug(u, "Failed to parse control-pid value: %s", value); + else + n->control_pid = pid; + } else if (streq(key, "starter-fd")) { + int fd; + + if (safe_atoi(value, &fd) < 0 || fd < 0 || !fdset_contains(fds, fd)) + log_unit_debug(u, "Failed to parse starter fd value: %s", value); + else { + safe_close(n->starter_fd); + n->starter_fd = fdset_remove(fds, fd); + } + } else + log_unit_debug(u, "Unknown serialization key: %s", key); + + return 0; +} + +_pure_ static UnitActiveState busname_active_state(Unit *u) { + assert(u); + + return state_translation_table[BUSNAME(u)->state]; +} + +_pure_ static const char *busname_sub_state_to_string(Unit *u) { + assert(u); + + return busname_state_to_string(BUSNAME(u)->state); +} + +static int busname_peek_message(BusName *n) { + struct kdbus_cmd_recv cmd_recv = { + .size = sizeof(cmd_recv), + .flags = KDBUS_RECV_PEEK, + }; + struct kdbus_cmd_free cmd_free = { + .size = sizeof(cmd_free), + }; + const char *comm = NULL; + struct kdbus_item *d; + struct kdbus_msg *k; + size_t start, ps, sz, delta; + void *p = NULL; + pid_t pid = 0; + int r; + + /* Generate a friendly debug log message about which process + * caused triggering of this bus name. This simply peeks the + * metadata of the first queued message and logs it. */ + + assert(n); + + /* Let's shortcut things a bit, if debug logging is turned off + * anyway. */ + + if (log_get_max_level() < LOG_DEBUG) + return 0; + + r = ioctl(n->starter_fd, KDBUS_CMD_RECV, &cmd_recv); + if (r < 0) { + if (errno == EINTR || errno == EAGAIN) + return 0; + + return log_unit_error_errno(UNIT(n), errno, "Failed to query activation message: %m"); + } + + /* We map as late as possible, and unmap imemdiately after + * use. On 32bit address space is scarce and we want to be + * able to handle a lot of activator connections at the same + * time, and hence shouldn't keep the mmap()s around for + * longer than necessary. */ + + ps = page_size(); + start = (cmd_recv.msg.offset / ps) * ps; + delta = cmd_recv.msg.offset - start; + sz = PAGE_ALIGN(delta + cmd_recv.msg.msg_size); + + p = mmap(NULL, sz, PROT_READ, MAP_SHARED, n->starter_fd, start); + if (p == MAP_FAILED) { + r = log_unit_error_errno(UNIT(n), errno, "Failed to map activation message: %m"); + goto finish; + } + + k = (struct kdbus_msg *) ((uint8_t *) p + delta); + KDBUS_ITEM_FOREACH(d, k, items) { + switch (d->type) { + + case KDBUS_ITEM_PIDS: + pid = d->pids.pid; + break; + + case KDBUS_ITEM_PID_COMM: + comm = d->str; + break; + } + } + + if (pid > 0) + log_unit_debug(UNIT(n), "Activation triggered by process " PID_FMT " (%s)", pid, strna(comm)); + + r = 0; + +finish: + if (p) + (void) munmap(p, sz); + + cmd_free.offset = cmd_recv.msg.offset; + if (ioctl(n->starter_fd, KDBUS_CMD_FREE, &cmd_free) < 0) + log_unit_warning(UNIT(n), "Failed to free peeked message, ignoring: %m"); + + return r; +} + +static int busname_dispatch_io(sd_event_source *source, int fd, uint32_t revents, void *userdata) { + BusName *n = userdata; + + assert(n); + assert(fd >= 0); + + if (n->state != BUSNAME_LISTENING) + return 0; + + log_unit_debug(UNIT(n), "Activation request"); + + if (revents != EPOLLIN) { + log_unit_error(UNIT(n), "Got unexpected poll event (0x%x) on starter fd.", revents); + goto fail; + } + + busname_peek_message(n); + busname_enter_running(n); + return 0; +fail: + + busname_enter_dead(n, BUSNAME_FAILURE_RESOURCES); + return 0; +} + +static void busname_sigchld_event(Unit *u, pid_t pid, int code, int status) { + BusName *n = BUSNAME(u); + BusNameResult f; + + assert(n); + assert(pid >= 0); + + if (pid != n->control_pid) + return; + + n->control_pid = 0; + + if (is_clean_exit(code, status, NULL)) + f = BUSNAME_SUCCESS; + else if (code == CLD_EXITED) + f = BUSNAME_FAILURE_EXIT_CODE; + else if (code == CLD_KILLED) + f = BUSNAME_FAILURE_SIGNAL; + else if (code == CLD_DUMPED) + f = BUSNAME_FAILURE_CORE_DUMP; + else + assert_not_reached("Unknown sigchld code"); + + log_unit_full(u, f == BUSNAME_SUCCESS ? LOG_DEBUG : LOG_NOTICE, 0, + "Control process exited, code=%s status=%i", sigchld_code_to_string(code), status); + + if (f != BUSNAME_SUCCESS) + n->result = f; + + switch (n->state) { + + case BUSNAME_MAKING: + if (f == BUSNAME_SUCCESS) + busname_enter_listening(n); + else + busname_enter_signal(n, BUSNAME_SIGTERM, f); + break; + + case BUSNAME_SIGTERM: + case BUSNAME_SIGKILL: + busname_enter_dead(n, f); + break; + + default: + assert_not_reached("Uh, control process died at wrong time."); + } + + /* Notify clients about changed exit status */ + unit_add_to_dbus_queue(u); +} + +static int busname_dispatch_timer(sd_event_source *source, usec_t usec, void *userdata) { + BusName *n = BUSNAME(userdata); + + assert(n); + assert(n->timer_event_source == source); + + switch (n->state) { + + case BUSNAME_MAKING: + log_unit_warning(UNIT(n), "Making timed out. Terminating."); + busname_enter_signal(n, BUSNAME_SIGTERM, BUSNAME_FAILURE_TIMEOUT); + break; + + case BUSNAME_SIGTERM: + log_unit_warning(UNIT(n), "Stopping timed out. Killing."); + busname_enter_signal(n, BUSNAME_SIGKILL, BUSNAME_FAILURE_TIMEOUT); + break; + + case BUSNAME_SIGKILL: + log_unit_warning(UNIT(n), "Processes still around after SIGKILL. Ignoring."); + busname_enter_dead(n, BUSNAME_FAILURE_TIMEOUT); + break; + + default: + assert_not_reached("Timeout at wrong time."); + } + + return 0; +} + +static void busname_reset_failed(Unit *u) { + BusName *n = BUSNAME(u); + + assert(n); + + if (n->state == BUSNAME_FAILED) + busname_set_state(n, BUSNAME_DEAD); + + n->result = BUSNAME_SUCCESS; +} + +static void busname_trigger_notify(Unit *u, Unit *other) { + BusName *n = BUSNAME(u); + + assert(n); + assert(other); + + if (!IN_SET(n->state, BUSNAME_RUNNING, BUSNAME_LISTENING)) + return; + + if (other->start_limit_hit) { + busname_enter_dead(n, BUSNAME_FAILURE_SERVICE_START_LIMIT_HIT); + return; + } + + if (other->load_state != UNIT_LOADED || other->type != UNIT_SERVICE) + return; + + if (IN_SET(SERVICE(other)->state, + SERVICE_DEAD, SERVICE_FAILED, + SERVICE_FINAL_SIGTERM, SERVICE_FINAL_SIGKILL, + SERVICE_AUTO_RESTART)) + busname_enter_listening(n); + + if (SERVICE(other)->state == SERVICE_RUNNING) + busname_set_state(n, BUSNAME_RUNNING); +} + +static int busname_kill(Unit *u, KillWho who, int signo, sd_bus_error *error) { + return unit_kill_common(u, who, signo, -1, BUSNAME(u)->control_pid, error); +} + +static int busname_get_timeout(Unit *u, usec_t *timeout) { + BusName *n = BUSNAME(u); + usec_t t; + int r; + + if (!n->timer_event_source) + return 0; + + r = sd_event_source_get_time(n->timer_event_source, &t); + if (r < 0) + return r; + if (t == USEC_INFINITY) + return 0; + + *timeout = t; + return 1; +} + +static bool busname_supported(void) { + return false; +} + +static int busname_control_pid(Unit *u) { + BusName *n = BUSNAME(u); + + assert(n); + + return n->control_pid; +} + +static const char* const busname_result_table[_BUSNAME_RESULT_MAX] = { + [BUSNAME_SUCCESS] = "success", + [BUSNAME_FAILURE_RESOURCES] = "resources", + [BUSNAME_FAILURE_TIMEOUT] = "timeout", + [BUSNAME_FAILURE_EXIT_CODE] = "exit-code", + [BUSNAME_FAILURE_SIGNAL] = "signal", + [BUSNAME_FAILURE_CORE_DUMP] = "core-dump", + [BUSNAME_FAILURE_START_LIMIT_HIT] = "start-limit-hit", + [BUSNAME_FAILURE_SERVICE_START_LIMIT_HIT] = "service-start-limit-hit", +}; + +DEFINE_STRING_TABLE_LOOKUP(busname_result, BusNameResult); + +const UnitVTable busname_vtable = { + .object_size = sizeof(BusName), + + .sections = + "Unit\0" + "BusName\0" + "Install\0", + .private_section = "BusName", + + .init = busname_init, + .done = busname_done, + .load = busname_load, + + .coldplug = busname_coldplug, + + .dump = busname_dump, + + .start = busname_start, + .stop = busname_stop, + + .kill = busname_kill, + + .get_timeout = busname_get_timeout, + + .serialize = busname_serialize, + .deserialize_item = busname_deserialize_item, + + .active_state = busname_active_state, + .sub_state_to_string = busname_sub_state_to_string, + + .sigchld_event = busname_sigchld_event, + + .trigger_notify = busname_trigger_notify, + + .reset_failed = busname_reset_failed, + + .supported = busname_supported, + + .control_pid = busname_control_pid, + + .bus_vtable = bus_busname_vtable, + + .status_message_formats = { + .finished_start_job = { + [JOB_DONE] = "Listening on %s.", + [JOB_FAILED] = "Failed to listen on %s.", + }, + .finished_stop_job = { + [JOB_DONE] = "Closed %s.", + [JOB_FAILED] = "Failed stopping %s.", + }, + }, +}; diff --git a/src/grp-system/libcore/src/cgroup.c b/src/grp-system/libcore/src/cgroup.c new file mode 100644 index 0000000000..fca369c4ff --- /dev/null +++ b/src/grp-system/libcore/src/cgroup.c @@ -0,0 +1,1991 @@ +/*** + This file is part of systemd. + + Copyright 2013 Lennart Poettering + + systemd is free software; you can redistribute it and/or modify it + under the terms of the GNU Lesser General Public License as published by + the Free Software Foundation; either version 2.1 of the License, or + (at your option) any later version. + + systemd is distributed in the hope that it will be useful, but + WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + Lesser General Public License for more details. + + You should have received a copy of the GNU Lesser General Public License + along with systemd; If not, see . +***/ + +#include +#include + +#include "core/cgroup.h" +#include "systemd-basic/alloc-util.h" +#include "systemd-basic/cgroup-util.h" +#include "systemd-basic/fd-util.h" +#include "systemd-basic/fileio.h" +#include "systemd-basic/fs-util.h" +#include "systemd-basic/parse-util.h" +#include "systemd-basic/path-util.h" +#include "systemd-basic/process-util.h" +#include "systemd-basic/special.h" +#include "systemd-basic/stdio-util.h" +#include "systemd-basic/string-table.h" +#include "systemd-basic/string-util.h" + +#define CGROUP_CPU_QUOTA_PERIOD_USEC ((usec_t) 100 * USEC_PER_MSEC) + +static void cgroup_compat_warn(void) { + static bool cgroup_compat_warned = false; + + if (cgroup_compat_warned) + return; + + log_warning("cgroup compatibility translation between legacy and unified hierarchy settings activated. See cgroup-compat debug messages for details."); + cgroup_compat_warned = true; +} + +#define log_cgroup_compat(unit, fmt, ...) do { \ + cgroup_compat_warn(); \ + log_unit_debug(unit, "cgroup-compat: " fmt, ##__VA_ARGS__); \ + } while (false) + +void cgroup_context_init(CGroupContext *c) { + assert(c); + + /* Initialize everything to the kernel defaults, assuming the + * structure is preinitialized to 0 */ + + c->cpu_shares = CGROUP_CPU_SHARES_INVALID; + c->startup_cpu_shares = CGROUP_CPU_SHARES_INVALID; + c->cpu_quota_per_sec_usec = USEC_INFINITY; + + c->memory_high = CGROUP_LIMIT_MAX; + c->memory_max = CGROUP_LIMIT_MAX; + + c->memory_limit = CGROUP_LIMIT_MAX; + + c->io_weight = CGROUP_WEIGHT_INVALID; + c->startup_io_weight = CGROUP_WEIGHT_INVALID; + + c->blockio_weight = CGROUP_BLKIO_WEIGHT_INVALID; + c->startup_blockio_weight = CGROUP_BLKIO_WEIGHT_INVALID; + + c->tasks_max = (uint64_t) -1; +} + +void cgroup_context_free_device_allow(CGroupContext *c, CGroupDeviceAllow *a) { + assert(c); + assert(a); + + LIST_REMOVE(device_allow, c->device_allow, a); + free(a->path); + free(a); +} + +void cgroup_context_free_io_device_weight(CGroupContext *c, CGroupIODeviceWeight *w) { + assert(c); + assert(w); + + LIST_REMOVE(device_weights, c->io_device_weights, w); + free(w->path); + free(w); +} + +void cgroup_context_free_io_device_limit(CGroupContext *c, CGroupIODeviceLimit *l) { + assert(c); + assert(l); + + LIST_REMOVE(device_limits, c->io_device_limits, l); + free(l->path); + free(l); +} + +void cgroup_context_free_blockio_device_weight(CGroupContext *c, CGroupBlockIODeviceWeight *w) { + assert(c); + assert(w); + + LIST_REMOVE(device_weights, c->blockio_device_weights, w); + free(w->path); + free(w); +} + +void cgroup_context_free_blockio_device_bandwidth(CGroupContext *c, CGroupBlockIODeviceBandwidth *b) { + assert(c); + assert(b); + + LIST_REMOVE(device_bandwidths, c->blockio_device_bandwidths, b); + free(b->path); + free(b); +} + +void cgroup_context_done(CGroupContext *c) { + assert(c); + + while (c->io_device_weights) + cgroup_context_free_io_device_weight(c, c->io_device_weights); + + while (c->io_device_limits) + cgroup_context_free_io_device_limit(c, c->io_device_limits); + + while (c->blockio_device_weights) + cgroup_context_free_blockio_device_weight(c, c->blockio_device_weights); + + while (c->blockio_device_bandwidths) + cgroup_context_free_blockio_device_bandwidth(c, c->blockio_device_bandwidths); + + while (c->device_allow) + cgroup_context_free_device_allow(c, c->device_allow); +} + +void cgroup_context_dump(CGroupContext *c, FILE* f, const char *prefix) { + CGroupIODeviceLimit *il; + CGroupIODeviceWeight *iw; + CGroupBlockIODeviceBandwidth *b; + CGroupBlockIODeviceWeight *w; + CGroupDeviceAllow *a; + char u[FORMAT_TIMESPAN_MAX]; + + assert(c); + assert(f); + + prefix = strempty(prefix); + + fprintf(f, + "%sCPUAccounting=%s\n" + "%sIOAccounting=%s\n" + "%sBlockIOAccounting=%s\n" + "%sMemoryAccounting=%s\n" + "%sTasksAccounting=%s\n" + "%sCPUShares=%" PRIu64 "\n" + "%sStartupCPUShares=%" PRIu64 "\n" + "%sCPUQuotaPerSecSec=%s\n" + "%sIOWeight=%" PRIu64 "\n" + "%sStartupIOWeight=%" PRIu64 "\n" + "%sBlockIOWeight=%" PRIu64 "\n" + "%sStartupBlockIOWeight=%" PRIu64 "\n" + "%sMemoryLow=%" PRIu64 "\n" + "%sMemoryHigh=%" PRIu64 "\n" + "%sMemoryMax=%" PRIu64 "\n" + "%sMemoryLimit=%" PRIu64 "\n" + "%sTasksMax=%" PRIu64 "\n" + "%sDevicePolicy=%s\n" + "%sDelegate=%s\n", + prefix, yes_no(c->cpu_accounting), + prefix, yes_no(c->io_accounting), + prefix, yes_no(c->blockio_accounting), + prefix, yes_no(c->memory_accounting), + prefix, yes_no(c->tasks_accounting), + prefix, c->cpu_shares, + prefix, c->startup_cpu_shares, + prefix, format_timespan(u, sizeof(u), c->cpu_quota_per_sec_usec, 1), + prefix, c->io_weight, + prefix, c->startup_io_weight, + prefix, c->blockio_weight, + prefix, c->startup_blockio_weight, + prefix, c->memory_low, + prefix, c->memory_high, + prefix, c->memory_max, + prefix, c->memory_limit, + prefix, c->tasks_max, + prefix, cgroup_device_policy_to_string(c->device_policy), + prefix, yes_no(c->delegate)); + + LIST_FOREACH(device_allow, a, c->device_allow) + fprintf(f, + "%sDeviceAllow=%s %s%s%s\n", + prefix, + a->path, + a->r ? "r" : "", a->w ? "w" : "", a->m ? "m" : ""); + + LIST_FOREACH(device_weights, iw, c->io_device_weights) + fprintf(f, + "%sIODeviceWeight=%s %" PRIu64, + prefix, + iw->path, + iw->weight); + + LIST_FOREACH(device_limits, il, c->io_device_limits) { + char buf[FORMAT_BYTES_MAX]; + CGroupIOLimitType type; + + for (type = 0; type < _CGROUP_IO_LIMIT_TYPE_MAX; type++) + if (il->limits[type] != cgroup_io_limit_defaults[type]) + fprintf(f, + "%s%s=%s %s\n", + prefix, + cgroup_io_limit_type_to_string(type), + il->path, + format_bytes(buf, sizeof(buf), il->limits[type])); + } + + LIST_FOREACH(device_weights, w, c->blockio_device_weights) + fprintf(f, + "%sBlockIODeviceWeight=%s %" PRIu64, + prefix, + w->path, + w->weight); + + LIST_FOREACH(device_bandwidths, b, c->blockio_device_bandwidths) { + char buf[FORMAT_BYTES_MAX]; + + if (b->rbps != CGROUP_LIMIT_MAX) + fprintf(f, + "%sBlockIOReadBandwidth=%s %s\n", + prefix, + b->path, + format_bytes(buf, sizeof(buf), b->rbps)); + if (b->wbps != CGROUP_LIMIT_MAX) + fprintf(f, + "%sBlockIOWriteBandwidth=%s %s\n", + prefix, + b->path, + format_bytes(buf, sizeof(buf), b->wbps)); + } +} + +static int lookup_block_device(const char *p, dev_t *dev) { + struct stat st; + int r; + + assert(p); + assert(dev); + + r = stat(p, &st); + if (r < 0) + return log_warning_errno(errno, "Couldn't stat device %s: %m", p); + + if (S_ISBLK(st.st_mode)) + *dev = st.st_rdev; + else if (major(st.st_dev) != 0) { + /* If this is not a device node then find the block + * device this file is stored on */ + *dev = st.st_dev; + + /* If this is a partition, try to get the originating + * block device */ + block_get_whole_disk(*dev, dev); + } else { + log_warning("%s is not a block device and file system block device cannot be determined or is not local.", p); + return -ENODEV; + } + + return 0; +} + +static int whitelist_device(const char *path, const char *node, const char *acc) { + char buf[2+DECIMAL_STR_MAX(dev_t)*2+2+4]; + struct stat st; + int r; + + assert(path); + assert(acc); + + if (stat(node, &st) < 0) { + log_warning("Couldn't stat device %s", node); + return -errno; + } + + if (!S_ISCHR(st.st_mode) && !S_ISBLK(st.st_mode)) { + log_warning("%s is not a device.", node); + return -ENODEV; + } + + sprintf(buf, + "%c %u:%u %s", + S_ISCHR(st.st_mode) ? 'c' : 'b', + major(st.st_rdev), minor(st.st_rdev), + acc); + + r = cg_set_attribute("devices", path, "devices.allow", buf); + if (r < 0) + log_full_errno(IN_SET(r, -ENOENT, -EROFS, -EINVAL, -EACCES) ? LOG_DEBUG : LOG_WARNING, r, + "Failed to set devices.allow on %s: %m", path); + + return r; +} + +static int whitelist_major(const char *path, const char *name, char type, const char *acc) { + _cleanup_fclose_ FILE *f = NULL; + char line[LINE_MAX]; + bool good = false; + int r; + + assert(path); + assert(acc); + assert(type == 'b' || type == 'c'); + + f = fopen("/proc/devices", "re"); + if (!f) + return log_warning_errno(errno, "Cannot open /proc/devices to resolve %s (%c): %m", name, type); + + FOREACH_LINE(line, f, goto fail) { + char buf[2+DECIMAL_STR_MAX(unsigned)+3+4], *p, *w; + unsigned maj; + + truncate_nl(line); + + if (type == 'c' && streq(line, "Character devices:")) { + good = true; + continue; + } + + if (type == 'b' && streq(line, "Block devices:")) { + good = true; + continue; + } + + if (isempty(line)) { + good = false; + continue; + } + + if (!good) + continue; + + p = strstrip(line); + + w = strpbrk(p, WHITESPACE); + if (!w) + continue; + *w = 0; + + r = safe_atou(p, &maj); + if (r < 0) + continue; + if (maj <= 0) + continue; + + w++; + w += strspn(w, WHITESPACE); + + if (fnmatch(name, w, 0) != 0) + continue; + + sprintf(buf, + "%c %u:* %s", + type, + maj, + acc); + + r = cg_set_attribute("devices", path, "devices.allow", buf); + if (r < 0) + log_full_errno(IN_SET(r, -ENOENT, -EROFS, -EINVAL, -EACCES) ? LOG_DEBUG : LOG_WARNING, r, + "Failed to set devices.allow on %s: %m", path); + } + + return 0; + +fail: + log_warning_errno(errno, "Failed to read /proc/devices: %m"); + return -errno; +} + +static bool cgroup_context_has_io_config(CGroupContext *c) { + return c->io_accounting || + c->io_weight != CGROUP_WEIGHT_INVALID || + c->startup_io_weight != CGROUP_WEIGHT_INVALID || + c->io_device_weights || + c->io_device_limits; +} + +static bool cgroup_context_has_blockio_config(CGroupContext *c) { + return c->blockio_accounting || + c->blockio_weight != CGROUP_BLKIO_WEIGHT_INVALID || + c->startup_blockio_weight != CGROUP_BLKIO_WEIGHT_INVALID || + c->blockio_device_weights || + c->blockio_device_bandwidths; +} + +static uint64_t cgroup_context_io_weight(CGroupContext *c, ManagerState state) { + if (IN_SET(state, MANAGER_STARTING, MANAGER_INITIALIZING) && + c->startup_io_weight != CGROUP_WEIGHT_INVALID) + return c->startup_io_weight; + else if (c->io_weight != CGROUP_WEIGHT_INVALID) + return c->io_weight; + else + return CGROUP_WEIGHT_DEFAULT; +} + +static uint64_t cgroup_context_blkio_weight(CGroupContext *c, ManagerState state) { + if (IN_SET(state, MANAGER_STARTING, MANAGER_INITIALIZING) && + c->startup_blockio_weight != CGROUP_BLKIO_WEIGHT_INVALID) + return c->startup_blockio_weight; + else if (c->blockio_weight != CGROUP_BLKIO_WEIGHT_INVALID) + return c->blockio_weight; + else + return CGROUP_BLKIO_WEIGHT_DEFAULT; +} + +static uint64_t cgroup_weight_blkio_to_io(uint64_t blkio_weight) { + return CLAMP(blkio_weight * CGROUP_WEIGHT_DEFAULT / CGROUP_BLKIO_WEIGHT_DEFAULT, + CGROUP_WEIGHT_MIN, CGROUP_WEIGHT_MAX); +} + +static uint64_t cgroup_weight_io_to_blkio(uint64_t io_weight) { + return CLAMP(io_weight * CGROUP_BLKIO_WEIGHT_DEFAULT / CGROUP_WEIGHT_DEFAULT, + CGROUP_BLKIO_WEIGHT_MIN, CGROUP_BLKIO_WEIGHT_MAX); +} + +static void cgroup_apply_io_device_weight(Unit *u, const char *dev_path, uint64_t io_weight) { + char buf[DECIMAL_STR_MAX(dev_t)*2+2+DECIMAL_STR_MAX(uint64_t)+1]; + dev_t dev; + int r; + + r = lookup_block_device(dev_path, &dev); + if (r < 0) + return; + + xsprintf(buf, "%u:%u %" PRIu64 "\n", major(dev), minor(dev), io_weight); + r = cg_set_attribute("io", u->cgroup_path, "io.weight", buf); + if (r < 0) + log_unit_full(u, IN_SET(r, -ENOENT, -EROFS, -EACCES) ? LOG_DEBUG : LOG_WARNING, r, + "Failed to set io.weight: %m"); +} + +static void cgroup_apply_blkio_device_weight(Unit *u, const char *dev_path, uint64_t blkio_weight) { + char buf[DECIMAL_STR_MAX(dev_t)*2+2+DECIMAL_STR_MAX(uint64_t)+1]; + dev_t dev; + int r; + + r = lookup_block_device(dev_path, &dev); + if (r < 0) + return; + + xsprintf(buf, "%u:%u %" PRIu64 "\n", major(dev), minor(dev), blkio_weight); + r = cg_set_attribute("blkio", u->cgroup_path, "blkio.weight_device", buf); + if (r < 0) + log_unit_full(u, IN_SET(r, -ENOENT, -EROFS, -EACCES) ? LOG_DEBUG : LOG_WARNING, r, + "Failed to set blkio.weight_device: %m"); +} + +static unsigned cgroup_apply_io_device_limit(Unit *u, const char *dev_path, uint64_t *limits) { + char limit_bufs[_CGROUP_IO_LIMIT_TYPE_MAX][DECIMAL_STR_MAX(uint64_t)]; + char buf[DECIMAL_STR_MAX(dev_t)*2+2+(6+DECIMAL_STR_MAX(uint64_t)+1)*4]; + CGroupIOLimitType type; + dev_t dev; + unsigned n = 0; + int r; + + r = lookup_block_device(dev_path, &dev); + if (r < 0) + return 0; + + for (type = 0; type < _CGROUP_IO_LIMIT_TYPE_MAX; type++) { + if (limits[type] != cgroup_io_limit_defaults[type]) { + xsprintf(limit_bufs[type], "%" PRIu64, limits[type]); + n++; + } else { + xsprintf(limit_bufs[type], "%s", limits[type] == CGROUP_LIMIT_MAX ? "max" : "0"); + } + } + + xsprintf(buf, "%u:%u rbps=%s wbps=%s riops=%s wiops=%s\n", major(dev), minor(dev), + limit_bufs[CGROUP_IO_RBPS_MAX], limit_bufs[CGROUP_IO_WBPS_MAX], + limit_bufs[CGROUP_IO_RIOPS_MAX], limit_bufs[CGROUP_IO_WIOPS_MAX]); + r = cg_set_attribute("io", u->cgroup_path, "io.max", buf); + if (r < 0) + log_unit_full(u, IN_SET(r, -ENOENT, -EROFS, -EACCES) ? LOG_DEBUG : LOG_WARNING, r, + "Failed to set io.max: %m"); + return n; +} + +static unsigned cgroup_apply_blkio_device_limit(Unit *u, const char *dev_path, uint64_t rbps, uint64_t wbps) { + char buf[DECIMAL_STR_MAX(dev_t)*2+2+DECIMAL_STR_MAX(uint64_t)+1]; + dev_t dev; + unsigned n = 0; + int r; + + r = lookup_block_device(dev_path, &dev); + if (r < 0) + return 0; + + if (rbps != CGROUP_LIMIT_MAX) + n++; + sprintf(buf, "%u:%u %" PRIu64 "\n", major(dev), minor(dev), rbps); + r = cg_set_attribute("blkio", u->cgroup_path, "blkio.throttle.read_bps_device", buf); + if (r < 0) + log_unit_full(u, IN_SET(r, -ENOENT, -EROFS, -EACCES) ? LOG_DEBUG : LOG_WARNING, r, + "Failed to set blkio.throttle.read_bps_device: %m"); + + if (wbps != CGROUP_LIMIT_MAX) + n++; + sprintf(buf, "%u:%u %" PRIu64 "\n", major(dev), minor(dev), wbps); + r = cg_set_attribute("blkio", u->cgroup_path, "blkio.throttle.write_bps_device", buf); + if (r < 0) + log_unit_full(u, IN_SET(r, -ENOENT, -EROFS, -EACCES) ? LOG_DEBUG : LOG_WARNING, r, + "Failed to set blkio.throttle.write_bps_device: %m"); + + return n; +} + +static bool cgroup_context_has_unified_memory_config(CGroupContext *c) { + return c->memory_low > 0 || c->memory_high != CGROUP_LIMIT_MAX || c->memory_max != CGROUP_LIMIT_MAX; +} + +static void cgroup_apply_unified_memory_limit(Unit *u, const char *file, uint64_t v) { + char buf[DECIMAL_STR_MAX(uint64_t) + 1] = "max"; + int r; + + if (v != CGROUP_LIMIT_MAX) + xsprintf(buf, "%" PRIu64 "\n", v); + + r = cg_set_attribute("memory", u->cgroup_path, file, buf); + if (r < 0) + log_unit_full(u, IN_SET(r, -ENOENT, -EROFS, -EACCES) ? LOG_DEBUG : LOG_WARNING, r, + "Failed to set %s: %m", file); +} + +static void cgroup_context_apply(Unit *u, CGroupMask mask, ManagerState state) { + const char *path; + CGroupContext *c; + bool is_root; + int r; + + assert(u); + + c = unit_get_cgroup_context(u); + path = u->cgroup_path; + + assert(c); + assert(path); + + if (mask == 0) + return; + + /* Some cgroup attributes are not supported on the root cgroup, + * hence silently ignore */ + is_root = isempty(path) || path_equal(path, "/"); + if (is_root) + /* Make sure we don't try to display messages with an empty path. */ + path = "/"; + + /* We generally ignore errors caused by read-only mounted + * cgroup trees (assuming we are running in a container then), + * and missing cgroups, i.e. EROFS and ENOENT. */ + + if ((mask & CGROUP_MASK_CPU) && !is_root) { + char buf[MAX(DECIMAL_STR_MAX(uint64_t), DECIMAL_STR_MAX(usec_t)) + 1]; + + sprintf(buf, "%" PRIu64 "\n", + IN_SET(state, MANAGER_STARTING, MANAGER_INITIALIZING) && c->startup_cpu_shares != CGROUP_CPU_SHARES_INVALID ? c->startup_cpu_shares : + c->cpu_shares != CGROUP_CPU_SHARES_INVALID ? c->cpu_shares : CGROUP_CPU_SHARES_DEFAULT); + r = cg_set_attribute("cpu", path, "cpu.shares", buf); + if (r < 0) + log_unit_full(u, IN_SET(r, -ENOENT, -EROFS, -EACCES) ? LOG_DEBUG : LOG_WARNING, r, + "Failed to set cpu.shares: %m"); + + sprintf(buf, USEC_FMT "\n", CGROUP_CPU_QUOTA_PERIOD_USEC); + r = cg_set_attribute("cpu", path, "cpu.cfs_period_us", buf); + if (r < 0) + log_unit_full(u, IN_SET(r, -ENOENT, -EROFS, -EACCES) ? LOG_DEBUG : LOG_WARNING, r, + "Failed to set cpu.cfs_period_us: %m"); + + if (c->cpu_quota_per_sec_usec != USEC_INFINITY) { + sprintf(buf, USEC_FMT "\n", c->cpu_quota_per_sec_usec * CGROUP_CPU_QUOTA_PERIOD_USEC / USEC_PER_SEC); + r = cg_set_attribute("cpu", path, "cpu.cfs_quota_us", buf); + } else + r = cg_set_attribute("cpu", path, "cpu.cfs_quota_us", "-1"); + if (r < 0) + log_unit_full(u, IN_SET(r, -ENOENT, -EROFS, -EACCES) ? LOG_DEBUG : LOG_WARNING, r, + "Failed to set cpu.cfs_quota_us: %m"); + } + + if (mask & CGROUP_MASK_IO) { + bool has_io = cgroup_context_has_io_config(c); + bool has_blockio = cgroup_context_has_blockio_config(c); + + if (!is_root) { + char buf[8+DECIMAL_STR_MAX(uint64_t)+1]; + uint64_t weight; + + if (has_io) + weight = cgroup_context_io_weight(c, state); + else if (has_blockio) { + uint64_t blkio_weight = cgroup_context_blkio_weight(c, state); + + weight = cgroup_weight_blkio_to_io(blkio_weight); + + log_cgroup_compat(u, "Applying [Startup]BlockIOWeight %" PRIu64 " as [Startup]IOWeight %" PRIu64, + blkio_weight, weight); + } else + weight = CGROUP_WEIGHT_DEFAULT; + + xsprintf(buf, "default %" PRIu64 "\n", weight); + r = cg_set_attribute("io", path, "io.weight", buf); + if (r < 0) + log_unit_full(u, IN_SET(r, -ENOENT, -EROFS, -EACCES) ? LOG_DEBUG : LOG_WARNING, r, + "Failed to set io.weight: %m"); + + if (has_io) { + CGroupIODeviceWeight *w; + + /* FIXME: no way to reset this list */ + LIST_FOREACH(device_weights, w, c->io_device_weights) + cgroup_apply_io_device_weight(u, w->path, w->weight); + } else if (has_blockio) { + CGroupBlockIODeviceWeight *w; + + /* FIXME: no way to reset this list */ + LIST_FOREACH(device_weights, w, c->blockio_device_weights) { + weight = cgroup_weight_blkio_to_io(w->weight); + + log_cgroup_compat(u, "Applying BlockIODeviceWeight %" PRIu64 " as IODeviceWeight %" PRIu64 " for %s", + w->weight, weight, w->path); + + cgroup_apply_io_device_weight(u, w->path, weight); + } + } + } + + /* Apply limits and free ones without config. */ + if (has_io) { + CGroupIODeviceLimit *l, *next; + + LIST_FOREACH_SAFE(device_limits, l, next, c->io_device_limits) { + if (!cgroup_apply_io_device_limit(u, l->path, l->limits)) + cgroup_context_free_io_device_limit(c, l); + } + } else if (has_blockio) { + CGroupBlockIODeviceBandwidth *b, *next; + + LIST_FOREACH_SAFE(device_bandwidths, b, next, c->blockio_device_bandwidths) { + uint64_t limits[_CGROUP_IO_LIMIT_TYPE_MAX]; + CGroupIOLimitType type; + + for (type = 0; type < _CGROUP_IO_LIMIT_TYPE_MAX; type++) + limits[type] = cgroup_io_limit_defaults[type]; + + limits[CGROUP_IO_RBPS_MAX] = b->rbps; + limits[CGROUP_IO_WBPS_MAX] = b->wbps; + + log_cgroup_compat(u, "Applying BlockIO{Read|Write}Bandwidth %" PRIu64 " %" PRIu64 " as IO{Read|Write}BandwidthMax for %s", + b->rbps, b->wbps, b->path); + + if (!cgroup_apply_io_device_limit(u, b->path, limits)) + cgroup_context_free_blockio_device_bandwidth(c, b); + } + } + } + + if (mask & CGROUP_MASK_BLKIO) { + bool has_io = cgroup_context_has_io_config(c); + bool has_blockio = cgroup_context_has_blockio_config(c); + + if (!is_root) { + char buf[DECIMAL_STR_MAX(uint64_t)+1]; + uint64_t weight; + + if (has_blockio) + weight = cgroup_context_blkio_weight(c, state); + else if (has_io) { + uint64_t io_weight = cgroup_context_io_weight(c, state); + + weight = cgroup_weight_io_to_blkio(cgroup_context_io_weight(c, state)); + + log_cgroup_compat(u, "Applying [Startup]IOWeight %" PRIu64 " as [Startup]BlockIOWeight %" PRIu64, + io_weight, weight); + } else + weight = CGROUP_BLKIO_WEIGHT_DEFAULT; + + xsprintf(buf, "%" PRIu64 "\n", weight); + r = cg_set_attribute("blkio", path, "blkio.weight", buf); + if (r < 0) + log_unit_full(u, IN_SET(r, -ENOENT, -EROFS, -EACCES) ? LOG_DEBUG : LOG_WARNING, r, + "Failed to set blkio.weight: %m"); + + if (has_blockio) { + CGroupBlockIODeviceWeight *w; + + /* FIXME: no way to reset this list */ + LIST_FOREACH(device_weights, w, c->blockio_device_weights) + cgroup_apply_blkio_device_weight(u, w->path, w->weight); + } else if (has_io) { + CGroupIODeviceWeight *w; + + /* FIXME: no way to reset this list */ + LIST_FOREACH(device_weights, w, c->io_device_weights) { + weight = cgroup_weight_io_to_blkio(w->weight); + + log_cgroup_compat(u, "Applying IODeviceWeight %" PRIu64 " as BlockIODeviceWeight %" PRIu64 " for %s", + w->weight, weight, w->path); + + cgroup_apply_blkio_device_weight(u, w->path, weight); + } + } + } + + /* Apply limits and free ones without config. */ + if (has_blockio) { + CGroupBlockIODeviceBandwidth *b, *next; + + LIST_FOREACH_SAFE(device_bandwidths, b, next, c->blockio_device_bandwidths) { + if (!cgroup_apply_blkio_device_limit(u, b->path, b->rbps, b->wbps)) + cgroup_context_free_blockio_device_bandwidth(c, b); + } + } else if (has_io) { + CGroupIODeviceLimit *l, *next; + + LIST_FOREACH_SAFE(device_limits, l, next, c->io_device_limits) { + log_cgroup_compat(u, "Applying IO{Read|Write}Bandwidth %" PRIu64 " %" PRIu64 " as BlockIO{Read|Write}BandwidthMax for %s", + l->limits[CGROUP_IO_RBPS_MAX], l->limits[CGROUP_IO_WBPS_MAX], l->path); + + if (!cgroup_apply_blkio_device_limit(u, l->path, l->limits[CGROUP_IO_RBPS_MAX], l->limits[CGROUP_IO_WBPS_MAX])) + cgroup_context_free_io_device_limit(c, l); + } + } + } + + if ((mask & CGROUP_MASK_MEMORY) && !is_root) { + if (cg_unified() > 0) { + uint64_t max = c->memory_max; + + if (cgroup_context_has_unified_memory_config(c)) + max = c->memory_max; + else { + max = c->memory_limit; + + if (max != CGROUP_LIMIT_MAX) + log_cgroup_compat(u, "Applying MemoryLimit %" PRIu64 " as MemoryMax", max); + } + + cgroup_apply_unified_memory_limit(u, "memory.low", c->memory_low); + cgroup_apply_unified_memory_limit(u, "memory.high", c->memory_high); + cgroup_apply_unified_memory_limit(u, "memory.max", max); + } else { + char buf[DECIMAL_STR_MAX(uint64_t) + 1]; + uint64_t val = c->memory_limit; + + if (val == CGROUP_LIMIT_MAX) { + val = c->memory_max; + + if (val != CGROUP_LIMIT_MAX) + log_cgroup_compat(u, "Applying MemoryMax %" PRIi64 " as MemoryLimit", c->memory_max); + } + + if (val == CGROUP_LIMIT_MAX) + strncpy(buf, "-1\n", sizeof(buf)); + else + xsprintf(buf, "%" PRIu64 "\n", val); + + r = cg_set_attribute("memory", path, "memory.limit_in_bytes", buf); + if (r < 0) + log_unit_full(u, IN_SET(r, -ENOENT, -EROFS, -EACCES) ? LOG_DEBUG : LOG_WARNING, r, + "Failed to set memory.limit_in_bytes: %m"); + } + } + + if ((mask & CGROUP_MASK_DEVICES) && !is_root) { + CGroupDeviceAllow *a; + + /* Changing the devices list of a populated cgroup + * might result in EINVAL, hence ignore EINVAL + * here. */ + + if (c->device_allow || c->device_policy != CGROUP_AUTO) + r = cg_set_attribute("devices", path, "devices.deny", "a"); + else + r = cg_set_attribute("devices", path, "devices.allow", "a"); + if (r < 0) + log_unit_full(u, IN_SET(r, -ENOENT, -EROFS, -EINVAL, -EACCES) ? LOG_DEBUG : LOG_WARNING, r, + "Failed to reset devices.list: %m"); + + if (c->device_policy == CGROUP_CLOSED || + (c->device_policy == CGROUP_AUTO && c->device_allow)) { + static const char auto_devices[] = + "/dev/null\0" "rwm\0" + "/dev/zero\0" "rwm\0" + "/dev/full\0" "rwm\0" + "/dev/random\0" "rwm\0" + "/dev/urandom\0" "rwm\0" + "/dev/tty\0" "rwm\0" + "/dev/pts/ptmx\0" "rw\0" /* /dev/pts/ptmx may not be duplicated, but accessed */ + /* Allow /run/systemd/inaccessible/{chr,blk} devices for mapping InaccessiblePaths */ + "/run/systemd/inaccessible/chr\0" "rwm\0" + "/run/systemd/inaccessible/blk\0" "rwm\0"; + + const char *x, *y; + + NULSTR_FOREACH_PAIR(x, y, auto_devices) + whitelist_device(path, x, y); + + whitelist_major(path, "pts", 'c', "rw"); + whitelist_major(path, "kdbus", 'c', "rw"); + whitelist_major(path, "kdbus/*", 'c', "rw"); + } + + LIST_FOREACH(device_allow, a, c->device_allow) { + char acc[4]; + unsigned k = 0; + + if (a->r) + acc[k++] = 'r'; + if (a->w) + acc[k++] = 'w'; + if (a->m) + acc[k++] = 'm'; + + if (k == 0) + continue; + + acc[k++] = 0; + + if (startswith(a->path, "/dev/")) + whitelist_device(path, a->path, acc); + else if (startswith(a->path, "block-")) + whitelist_major(path, a->path + 6, 'b', acc); + else if (startswith(a->path, "char-")) + whitelist_major(path, a->path + 5, 'c', acc); + else + log_unit_debug(u, "Ignoring device %s while writing cgroup attribute.", a->path); + } + } + + if ((mask & CGROUP_MASK_PIDS) && !is_root) { + + if (c->tasks_max != (uint64_t) -1) { + char buf[DECIMAL_STR_MAX(uint64_t) + 2]; + + sprintf(buf, "%" PRIu64 "\n", c->tasks_max); + r = cg_set_attribute("pids", path, "pids.max", buf); + } else + r = cg_set_attribute("pids", path, "pids.max", "max"); + + if (r < 0) + log_unit_full(u, IN_SET(r, -ENOENT, -EROFS, -EACCES) ? LOG_DEBUG : LOG_WARNING, r, + "Failed to set pids.max: %m"); + } +} + +CGroupMask cgroup_context_get_mask(CGroupContext *c) { + CGroupMask mask = 0; + + /* Figure out which controllers we need */ + + if (c->cpu_accounting || + c->cpu_shares != CGROUP_CPU_SHARES_INVALID || + c->startup_cpu_shares != CGROUP_CPU_SHARES_INVALID || + c->cpu_quota_per_sec_usec != USEC_INFINITY) + mask |= CGROUP_MASK_CPUACCT | CGROUP_MASK_CPU; + + if (cgroup_context_has_io_config(c) || cgroup_context_has_blockio_config(c)) + mask |= CGROUP_MASK_IO | CGROUP_MASK_BLKIO; + + if (c->memory_accounting || + c->memory_limit != CGROUP_LIMIT_MAX || + cgroup_context_has_unified_memory_config(c)) + mask |= CGROUP_MASK_MEMORY; + + if (c->device_allow || + c->device_policy != CGROUP_AUTO) + mask |= CGROUP_MASK_DEVICES; + + if (c->tasks_accounting || + c->tasks_max != (uint64_t) -1) + mask |= CGROUP_MASK_PIDS; + + return mask; +} + +CGroupMask unit_get_own_mask(Unit *u) { + CGroupContext *c; + + /* Returns the mask of controllers the unit needs for itself */ + + c = unit_get_cgroup_context(u); + if (!c) + return 0; + + /* If delegation is turned on, then turn on all cgroups, + * unless we are on the legacy hierarchy and the process we + * fork into it is known to drop privileges, and hence + * shouldn't get access to the controllers. + * + * Note that on the unified hierarchy it is safe to delegate + * controllers to unprivileged services. */ + + if (c->delegate) { + ExecContext *e; + + e = unit_get_exec_context(u); + if (!e || + exec_context_maintains_privileges(e) || + cg_unified() > 0) + return _CGROUP_MASK_ALL; + } + + return cgroup_context_get_mask(c); +} + +CGroupMask unit_get_members_mask(Unit *u) { + assert(u); + + /* Returns the mask of controllers all of the unit's children + * require, merged */ + + if (u->cgroup_members_mask_valid) + return u->cgroup_members_mask; + + u->cgroup_members_mask = 0; + + if (u->type == UNIT_SLICE) { + Unit *member; + Iterator i; + + SET_FOREACH(member, u->dependencies[UNIT_BEFORE], i) { + + if (member == u) + continue; + + if (UNIT_DEREF(member->slice) != u) + continue; + + u->cgroup_members_mask |= + unit_get_own_mask(member) | + unit_get_members_mask(member); + } + } + + u->cgroup_members_mask_valid = true; + return u->cgroup_members_mask; +} + +CGroupMask unit_get_siblings_mask(Unit *u) { + assert(u); + + /* Returns the mask of controllers all of the unit's siblings + * require, i.e. the members mask of the unit's parent slice + * if there is one. */ + + if (UNIT_ISSET(u->slice)) + return unit_get_members_mask(UNIT_DEREF(u->slice)); + + return unit_get_own_mask(u) | unit_get_members_mask(u); +} + +CGroupMask unit_get_subtree_mask(Unit *u) { + + /* Returns the mask of this subtree, meaning of the group + * itself and its children. */ + + return unit_get_own_mask(u) | unit_get_members_mask(u); +} + +CGroupMask unit_get_target_mask(Unit *u) { + CGroupMask mask; + + /* This returns the cgroup mask of all controllers to enable + * for a specific cgroup, i.e. everything it needs itself, + * plus all that its children need, plus all that its siblings + * need. This is primarily useful on the legacy cgroup + * hierarchy, where we need to duplicate each cgroup in each + * hierarchy that shall be enabled for it. */ + + mask = unit_get_own_mask(u) | unit_get_members_mask(u) | unit_get_siblings_mask(u); + mask &= u->manager->cgroup_supported; + + return mask; +} + +CGroupMask unit_get_enable_mask(Unit *u) { + CGroupMask mask; + + /* This returns the cgroup mask of all controllers to enable + * for the children of a specific cgroup. This is primarily + * useful for the unified cgroup hierarchy, where each cgroup + * controls which controllers are enabled for its children. */ + + mask = unit_get_members_mask(u); + mask &= u->manager->cgroup_supported; + + return mask; +} + +/* Recurse from a unit up through its containing slices, propagating + * mask bits upward. A unit is also member of itself. */ +void unit_update_cgroup_members_masks(Unit *u) { + CGroupMask m; + bool more; + + assert(u); + + /* Calculate subtree mask */ + m = unit_get_subtree_mask(u); + + /* See if anything changed from the previous invocation. If + * not, we're done. */ + if (u->cgroup_subtree_mask_valid && m == u->cgroup_subtree_mask) + return; + + more = + u->cgroup_subtree_mask_valid && + ((m & ~u->cgroup_subtree_mask) != 0) && + ((~m & u->cgroup_subtree_mask) == 0); + + u->cgroup_subtree_mask = m; + u->cgroup_subtree_mask_valid = true; + + if (UNIT_ISSET(u->slice)) { + Unit *s = UNIT_DEREF(u->slice); + + if (more) + /* There's more set now than before. We + * propagate the new mask to the parent's mask + * (not caring if it actually was valid or + * not). */ + + s->cgroup_members_mask |= m; + + else + /* There's less set now than before (or we + * don't know), we need to recalculate + * everything, so let's invalidate the + * parent's members mask */ + + s->cgroup_members_mask_valid = false; + + /* And now make sure that this change also hits our + * grandparents */ + unit_update_cgroup_members_masks(s); + } +} + +static const char *migrate_callback(CGroupMask mask, void *userdata) { + Unit *u = userdata; + + assert(mask != 0); + assert(u); + + while (u) { + if (u->cgroup_path && + u->cgroup_realized && + (u->cgroup_realized_mask & mask) == mask) + return u->cgroup_path; + + u = UNIT_DEREF(u->slice); + } + + return NULL; +} + +char *unit_default_cgroup_path(Unit *u) { + _cleanup_free_ char *escaped = NULL, *slice = NULL; + int r; + + assert(u); + + if (unit_has_name(u, SPECIAL_ROOT_SLICE)) + return strdup(u->manager->cgroup_root); + + if (UNIT_ISSET(u->slice) && !unit_has_name(UNIT_DEREF(u->slice), SPECIAL_ROOT_SLICE)) { + r = cg_slice_to_path(UNIT_DEREF(u->slice)->id, &slice); + if (r < 0) + return NULL; + } + + escaped = cg_escape(u->id); + if (!escaped) + return NULL; + + if (slice) + return strjoin(u->manager->cgroup_root, "/", slice, "/", escaped, NULL); + else + return strjoin(u->manager->cgroup_root, "/", escaped, NULL); +} + +int unit_set_cgroup_path(Unit *u, const char *path) { + _cleanup_free_ char *p = NULL; + int r; + + assert(u); + + if (path) { + p = strdup(path); + if (!p) + return -ENOMEM; + } else + p = NULL; + + if (streq_ptr(u->cgroup_path, p)) + return 0; + + if (p) { + r = hashmap_put(u->manager->cgroup_unit, p, u); + if (r < 0) + return r; + } + + unit_release_cgroup(u); + + u->cgroup_path = p; + p = NULL; + + return 1; +} + +int unit_watch_cgroup(Unit *u) { + _cleanup_free_ char *events = NULL; + int r; + + assert(u); + + if (!u->cgroup_path) + return 0; + + if (u->cgroup_inotify_wd >= 0) + return 0; + + /* Only applies to the unified hierarchy */ + r = cg_unified(); + if (r < 0) + return log_unit_error_errno(u, r, "Failed detect whether the unified hierarchy is used: %m"); + if (r == 0) + return 0; + + /* Don't watch the root slice, it's pointless. */ + if (unit_has_name(u, SPECIAL_ROOT_SLICE)) + return 0; + + r = hashmap_ensure_allocated(&u->manager->cgroup_inotify_wd_unit, &trivial_hash_ops); + if (r < 0) + return log_oom(); + + r = cg_get_path(SYSTEMD_CGROUP_CONTROLLER, u->cgroup_path, "cgroup.events", &events); + if (r < 0) + return log_oom(); + + u->cgroup_inotify_wd = inotify_add_watch(u->manager->cgroup_inotify_fd, events, IN_MODIFY); + if (u->cgroup_inotify_wd < 0) { + + if (errno == ENOENT) /* If the directory is already + * gone we don't need to track + * it, so this is not an error */ + return 0; + + return log_unit_error_errno(u, errno, "Failed to add inotify watch descriptor for control group %s: %m", u->cgroup_path); + } + + r = hashmap_put(u->manager->cgroup_inotify_wd_unit, INT_TO_PTR(u->cgroup_inotify_wd), u); + if (r < 0) + return log_unit_error_errno(u, r, "Failed to add inotify watch descriptor to hash map: %m"); + + return 0; +} + +static int unit_create_cgroup( + Unit *u, + CGroupMask target_mask, + CGroupMask enable_mask) { + + CGroupContext *c; + int r; + + assert(u); + + c = unit_get_cgroup_context(u); + if (!c) + return 0; + + if (!u->cgroup_path) { + _cleanup_free_ char *path = NULL; + + path = unit_default_cgroup_path(u); + if (!path) + return log_oom(); + + r = unit_set_cgroup_path(u, path); + if (r == -EEXIST) + return log_unit_error_errno(u, r, "Control group %s exists already.", path); + if (r < 0) + return log_unit_error_errno(u, r, "Failed to set unit's control group path to %s: %m", path); + } + + /* First, create our own group */ + r = cg_create_everywhere(u->manager->cgroup_supported, target_mask, u->cgroup_path); + if (r < 0) + return log_unit_error_errno(u, r, "Failed to create cgroup %s: %m", u->cgroup_path); + + /* Start watching it */ + (void) unit_watch_cgroup(u); + + /* Enable all controllers we need */ + r = cg_enable_everywhere(u->manager->cgroup_supported, enable_mask, u->cgroup_path); + if (r < 0) + log_unit_warning_errno(u, r, "Failed to enable controllers on cgroup %s, ignoring: %m", u->cgroup_path); + + /* Keep track that this is now realized */ + u->cgroup_realized = true; + u->cgroup_realized_mask = target_mask; + u->cgroup_enabled_mask = enable_mask; + + if (u->type != UNIT_SLICE && !c->delegate) { + + /* Then, possibly move things over, but not if + * subgroups may contain processes, which is the case + * for slice and delegation units. */ + r = cg_migrate_everywhere(u->manager->cgroup_supported, u->cgroup_path, u->cgroup_path, migrate_callback, u); + if (r < 0) + log_unit_warning_errno(u, r, "Failed to migrate cgroup from to %s, ignoring: %m", u->cgroup_path); + } + + return 0; +} + +int unit_attach_pids_to_cgroup(Unit *u) { + int r; + assert(u); + + r = unit_realize_cgroup(u); + if (r < 0) + return r; + + r = cg_attach_many_everywhere(u->manager->cgroup_supported, u->cgroup_path, u->pids, migrate_callback, u); + if (r < 0) + return r; + + return 0; +} + +static bool unit_has_mask_realized(Unit *u, CGroupMask target_mask, CGroupMask enable_mask) { + assert(u); + + return u->cgroup_realized && u->cgroup_realized_mask == target_mask && u->cgroup_enabled_mask == enable_mask; +} + +/* Check if necessary controllers and attributes for a unit are in place. + * + * If so, do nothing. + * If not, create paths, move processes over, and set attributes. + * + * Returns 0 on success and < 0 on failure. */ +static int unit_realize_cgroup_now(Unit *u, ManagerState state) { + CGroupMask target_mask, enable_mask; + int r; + + assert(u); + + if (u->in_cgroup_queue) { + LIST_REMOVE(cgroup_queue, u->manager->cgroup_queue, u); + u->in_cgroup_queue = false; + } + + target_mask = unit_get_target_mask(u); + enable_mask = unit_get_enable_mask(u); + + if (unit_has_mask_realized(u, target_mask, enable_mask)) + return 0; + + /* First, realize parents */ + if (UNIT_ISSET(u->slice)) { + r = unit_realize_cgroup_now(UNIT_DEREF(u->slice), state); + if (r < 0) + return r; + } + + /* And then do the real work */ + r = unit_create_cgroup(u, target_mask, enable_mask); + if (r < 0) + return r; + + /* Finally, apply the necessary attributes. */ + cgroup_context_apply(u, target_mask, state); + + return 0; +} + +static void unit_add_to_cgroup_queue(Unit *u) { + + if (u->in_cgroup_queue) + return; + + LIST_PREPEND(cgroup_queue, u->manager->cgroup_queue, u); + u->in_cgroup_queue = true; +} + +unsigned manager_dispatch_cgroup_queue(Manager *m) { + ManagerState state; + unsigned n = 0; + Unit *i; + int r; + + state = manager_state(m); + + while ((i = m->cgroup_queue)) { + assert(i->in_cgroup_queue); + + r = unit_realize_cgroup_now(i, state); + if (r < 0) + log_warning_errno(r, "Failed to realize cgroups for queued unit %s, ignoring: %m", i->id); + + n++; + } + + return n; +} + +static void unit_queue_siblings(Unit *u) { + Unit *slice; + + /* This adds the siblings of the specified unit and the + * siblings of all parent units to the cgroup queue. (But + * neither the specified unit itself nor the parents.) */ + + while ((slice = UNIT_DEREF(u->slice))) { + Iterator i; + Unit *m; + + SET_FOREACH(m, slice->dependencies[UNIT_BEFORE], i) { + if (m == u) + continue; + + /* Skip units that have a dependency on the slice + * but aren't actually in it. */ + if (UNIT_DEREF(m->slice) != slice) + continue; + + /* No point in doing cgroup application for units + * without active processes. */ + if (UNIT_IS_INACTIVE_OR_FAILED(unit_active_state(m))) + continue; + + /* If the unit doesn't need any new controllers + * and has current ones realized, it doesn't need + * any changes. */ + if (unit_has_mask_realized(m, unit_get_target_mask(m), unit_get_enable_mask(m))) + continue; + + unit_add_to_cgroup_queue(m); + } + + u = slice; + } +} + +int unit_realize_cgroup(Unit *u) { + assert(u); + + if (!UNIT_HAS_CGROUP_CONTEXT(u)) + return 0; + + /* So, here's the deal: when realizing the cgroups for this + * unit, we need to first create all parents, but there's more + * actually: for the weight-based controllers we also need to + * make sure that all our siblings (i.e. units that are in the + * same slice as we are) have cgroups, too. Otherwise, things + * would become very uneven as each of their processes would + * get as much resources as all our group together. This call + * will synchronously create the parent cgroups, but will + * defer work on the siblings to the next event loop + * iteration. */ + + /* Add all sibling slices to the cgroup queue. */ + unit_queue_siblings(u); + + /* And realize this one now (and apply the values) */ + return unit_realize_cgroup_now(u, manager_state(u->manager)); +} + +void unit_release_cgroup(Unit *u) { + assert(u); + + /* Forgets all cgroup details for this cgroup */ + + if (u->cgroup_path) { + (void) hashmap_remove(u->manager->cgroup_unit, u->cgroup_path); + u->cgroup_path = mfree(u->cgroup_path); + } + + if (u->cgroup_inotify_wd >= 0) { + if (inotify_rm_watch(u->manager->cgroup_inotify_fd, u->cgroup_inotify_wd) < 0) + log_unit_debug_errno(u, errno, "Failed to remove cgroup inotify watch %i for %s, ignoring", u->cgroup_inotify_wd, u->id); + + (void) hashmap_remove(u->manager->cgroup_inotify_wd_unit, INT_TO_PTR(u->cgroup_inotify_wd)); + u->cgroup_inotify_wd = -1; + } +} + +void unit_prune_cgroup(Unit *u) { + int r; + bool is_root_slice; + + assert(u); + + /* Removes the cgroup, if empty and possible, and stops watching it. */ + + if (!u->cgroup_path) + return; + + is_root_slice = unit_has_name(u, SPECIAL_ROOT_SLICE); + + r = cg_trim_everywhere(u->manager->cgroup_supported, u->cgroup_path, !is_root_slice); + if (r < 0) { + log_unit_debug_errno(u, r, "Failed to destroy cgroup %s, ignoring: %m", u->cgroup_path); + return; + } + + if (is_root_slice) + return; + + unit_release_cgroup(u); + + u->cgroup_realized = false; + u->cgroup_realized_mask = 0; + u->cgroup_enabled_mask = 0; +} + +int unit_search_main_pid(Unit *u, pid_t *ret) { + _cleanup_fclose_ FILE *f = NULL; + pid_t pid = 0, npid, mypid; + int r; + + assert(u); + assert(ret); + + if (!u->cgroup_path) + return -ENXIO; + + r = cg_enumerate_processes(SYSTEMD_CGROUP_CONTROLLER, u->cgroup_path, &f); + if (r < 0) + return r; + + mypid = getpid(); + while (cg_read_pid(f, &npid) > 0) { + pid_t ppid; + + if (npid == pid) + continue; + + /* Ignore processes that aren't our kids */ + if (get_process_ppid(npid, &ppid) >= 0 && ppid != mypid) + continue; + + if (pid != 0) + /* Dang, there's more than one daemonized PID + in this group, so we don't know what process + is the main process. */ + + return -ENODATA; + + pid = npid; + } + + *ret = pid; + return 0; +} + +static int unit_watch_pids_in_path(Unit *u, const char *path) { + _cleanup_closedir_ DIR *d = NULL; + _cleanup_fclose_ FILE *f = NULL; + int ret = 0, r; + + assert(u); + assert(path); + + r = cg_enumerate_processes(SYSTEMD_CGROUP_CONTROLLER, path, &f); + if (r < 0) + ret = r; + else { + pid_t pid; + + while ((r = cg_read_pid(f, &pid)) > 0) { + r = unit_watch_pid(u, pid); + if (r < 0 && ret >= 0) + ret = r; + } + + if (r < 0 && ret >= 0) + ret = r; + } + + r = cg_enumerate_subgroups(SYSTEMD_CGROUP_CONTROLLER, path, &d); + if (r < 0) { + if (ret >= 0) + ret = r; + } else { + char *fn; + + while ((r = cg_read_subgroup(d, &fn)) > 0) { + _cleanup_free_ char *p = NULL; + + p = strjoin(path, "/", fn, NULL); + free(fn); + + if (!p) + return -ENOMEM; + + r = unit_watch_pids_in_path(u, p); + if (r < 0 && ret >= 0) + ret = r; + } + + if (r < 0 && ret >= 0) + ret = r; + } + + return ret; +} + +int unit_watch_all_pids(Unit *u) { + assert(u); + + /* Adds all PIDs from our cgroup to the set of PIDs we + * watch. This is a fallback logic for cases where we do not + * get reliable cgroup empty notifications: we try to use + * SIGCHLD as replacement. */ + + if (!u->cgroup_path) + return -ENOENT; + + if (cg_unified() > 0) /* On unified we can use proper notifications */ + return 0; + + return unit_watch_pids_in_path(u, u->cgroup_path); +} + +int unit_notify_cgroup_empty(Unit *u) { + int r; + + assert(u); + + if (!u->cgroup_path) + return 0; + + r = cg_is_empty_recursive(SYSTEMD_CGROUP_CONTROLLER, u->cgroup_path); + if (r <= 0) + return r; + + unit_add_to_gc_queue(u); + + if (UNIT_VTABLE(u)->notify_cgroup_empty) + UNIT_VTABLE(u)->notify_cgroup_empty(u); + + return 0; +} + +static int on_cgroup_inotify_event(sd_event_source *s, int fd, uint32_t revents, void *userdata) { + Manager *m = userdata; + + assert(s); + assert(fd >= 0); + assert(m); + + for (;;) { + union inotify_event_buffer buffer; + struct inotify_event *e; + ssize_t l; + + l = read(fd, &buffer, sizeof(buffer)); + if (l < 0) { + if (errno == EINTR || errno == EAGAIN) + return 0; + + return log_error_errno(errno, "Failed to read control group inotify events: %m"); + } + + FOREACH_INOTIFY_EVENT(e, buffer, l) { + Unit *u; + + if (e->wd < 0) + /* Queue overflow has no watch descriptor */ + continue; + + if (e->mask & IN_IGNORED) + /* The watch was just removed */ + continue; + + u = hashmap_get(m->cgroup_inotify_wd_unit, INT_TO_PTR(e->wd)); + if (!u) /* Not that inotify might deliver + * events for a watch even after it + * was removed, because it was queued + * before the removal. Let's ignore + * this here safely. */ + continue; + + (void) unit_notify_cgroup_empty(u); + } + } +} + +int manager_setup_cgroup(Manager *m) { + _cleanup_free_ char *path = NULL; + CGroupController c; + int r, unified; + char *e; + + assert(m); + + /* 1. Determine hierarchy */ + m->cgroup_root = mfree(m->cgroup_root); + r = cg_pid_get_path(SYSTEMD_CGROUP_CONTROLLER, 0, &m->cgroup_root); + if (r < 0) + return log_error_errno(r, "Cannot determine cgroup we are running in: %m"); + + /* Chop off the init scope, if we are already located in it */ + e = endswith(m->cgroup_root, "/" SPECIAL_INIT_SCOPE); + + /* LEGACY: Also chop off the system slice if we are in + * it. This is to support live upgrades from older systemd + * versions where PID 1 was moved there. Also see + * cg_get_root_path(). */ + if (!e && MANAGER_IS_SYSTEM(m)) { + e = endswith(m->cgroup_root, "/" SPECIAL_SYSTEM_SLICE); + if (!e) + e = endswith(m->cgroup_root, "/system"); /* even more legacy */ + } + if (e) + *e = 0; + + /* And make sure to store away the root value without trailing + * slash, even for the root dir, so that we can easily prepend + * it everywhere. */ + while ((e = endswith(m->cgroup_root, "/"))) + *e = 0; + + /* 2. Show data */ + r = cg_get_path(SYSTEMD_CGROUP_CONTROLLER, m->cgroup_root, NULL, &path); + if (r < 0) + return log_error_errno(r, "Cannot find cgroup mount point: %m"); + + unified = cg_unified(); + if (unified < 0) + return log_error_errno(r, "Couldn't determine if we are running in the unified hierarchy: %m"); + if (unified > 0) + log_debug("Unified cgroup hierarchy is located at %s.", path); + else + log_debug("Using cgroup controller " SYSTEMD_CGROUP_CONTROLLER ". File system hierarchy is at %s.", path); + + if (!m->test_run) { + const char *scope_path; + + /* 3. Install agent */ + if (unified) { + + /* In the unified hierarchy we can get + * cgroup empty notifications via inotify. */ + + m->cgroup_inotify_event_source = sd_event_source_unref(m->cgroup_inotify_event_source); + safe_close(m->cgroup_inotify_fd); + + m->cgroup_inotify_fd = inotify_init1(IN_NONBLOCK|IN_CLOEXEC); + if (m->cgroup_inotify_fd < 0) + return log_error_errno(errno, "Failed to create control group inotify object: %m"); + + r = sd_event_add_io(m->event, &m->cgroup_inotify_event_source, m->cgroup_inotify_fd, EPOLLIN, on_cgroup_inotify_event, m); + if (r < 0) + return log_error_errno(r, "Failed to watch control group inotify object: %m"); + + /* Process cgroup empty notifications early, but after service notifications and SIGCHLD. Also + * see handling of cgroup agent notifications, for the classic cgroup hierarchy support. */ + r = sd_event_source_set_priority(m->cgroup_inotify_event_source, SD_EVENT_PRIORITY_NORMAL-5); + if (r < 0) + return log_error_errno(r, "Failed to set priority of inotify event source: %m"); + + (void) sd_event_source_set_description(m->cgroup_inotify_event_source, "cgroup-inotify"); + + } else if (MANAGER_IS_SYSTEM(m)) { + + /* On the legacy hierarchy we only get + * notifications via cgroup agents. (Which + * isn't really reliable, since it does not + * generate events when control groups with + * children run empty. */ + + r = cg_install_release_agent(SYSTEMD_CGROUP_CONTROLLER, SYSTEMD_CGROUP_AGENT_PATH); + if (r < 0) + log_warning_errno(r, "Failed to install release agent, ignoring: %m"); + else if (r > 0) + log_debug("Installed release agent."); + else if (r == 0) + log_debug("Release agent already installed."); + } + + /* 4. Make sure we are in the special "init.scope" unit in the root slice. */ + scope_path = strjoina(m->cgroup_root, "/" SPECIAL_INIT_SCOPE); + r = cg_create_and_attach(SYSTEMD_CGROUP_CONTROLLER, scope_path, 0); + if (r < 0) + return log_error_errno(r, "Failed to create %s control group: %m", scope_path); + + /* also, move all other userspace processes remaining + * in the root cgroup into that scope. */ + r = cg_migrate(SYSTEMD_CGROUP_CONTROLLER, m->cgroup_root, SYSTEMD_CGROUP_CONTROLLER, scope_path, 0); + if (r < 0) + log_warning_errno(r, "Couldn't move remaining userspace processes, ignoring: %m"); + + /* 5. And pin it, so that it cannot be unmounted */ + safe_close(m->pin_cgroupfs_fd); + m->pin_cgroupfs_fd = open(path, O_RDONLY|O_CLOEXEC|O_DIRECTORY|O_NOCTTY|O_NONBLOCK); + if (m->pin_cgroupfs_fd < 0) + return log_error_errno(errno, "Failed to open pin file: %m"); + + /* 6. Always enable hierarchical support if it exists... */ + if (!unified) + (void) cg_set_attribute("memory", "/", "memory.use_hierarchy", "1"); + } + + /* 7. Figure out which controllers are supported */ + r = cg_mask_supported(&m->cgroup_supported); + if (r < 0) + return log_error_errno(r, "Failed to determine supported controllers: %m"); + + for (c = 0; c < _CGROUP_CONTROLLER_MAX; c++) + log_debug("Controller '%s' supported: %s", cgroup_controller_to_string(c), yes_no(m->cgroup_supported & CGROUP_CONTROLLER_TO_MASK(c))); + + return 0; +} + +void manager_shutdown_cgroup(Manager *m, bool delete) { + assert(m); + + /* We can't really delete the group, since we are in it. But + * let's trim it. */ + if (delete && m->cgroup_root) + (void) cg_trim(SYSTEMD_CGROUP_CONTROLLER, m->cgroup_root, false); + + m->cgroup_inotify_wd_unit = hashmap_free(m->cgroup_inotify_wd_unit); + + m->cgroup_inotify_event_source = sd_event_source_unref(m->cgroup_inotify_event_source); + m->cgroup_inotify_fd = safe_close(m->cgroup_inotify_fd); + + m->pin_cgroupfs_fd = safe_close(m->pin_cgroupfs_fd); + + m->cgroup_root = mfree(m->cgroup_root); +} + +Unit* manager_get_unit_by_cgroup(Manager *m, const char *cgroup) { + char *p; + Unit *u; + + assert(m); + assert(cgroup); + + u = hashmap_get(m->cgroup_unit, cgroup); + if (u) + return u; + + p = strdupa(cgroup); + for (;;) { + char *e; + + e = strrchr(p, '/'); + if (!e || e == p) + return hashmap_get(m->cgroup_unit, SPECIAL_ROOT_SLICE); + + *e = 0; + + u = hashmap_get(m->cgroup_unit, p); + if (u) + return u; + } +} + +Unit *manager_get_unit_by_pid_cgroup(Manager *m, pid_t pid) { + _cleanup_free_ char *cgroup = NULL; + int r; + + assert(m); + + if (pid <= 0) + return NULL; + + r = cg_pid_get_path(SYSTEMD_CGROUP_CONTROLLER, pid, &cgroup); + if (r < 0) + return NULL; + + return manager_get_unit_by_cgroup(m, cgroup); +} + +Unit *manager_get_unit_by_pid(Manager *m, pid_t pid) { + Unit *u; + + assert(m); + + if (pid <= 0) + return NULL; + + if (pid == 1) + return hashmap_get(m->units, SPECIAL_INIT_SCOPE); + + u = hashmap_get(m->watch_pids1, PID_TO_PTR(pid)); + if (u) + return u; + + u = hashmap_get(m->watch_pids2, PID_TO_PTR(pid)); + if (u) + return u; + + return manager_get_unit_by_pid_cgroup(m, pid); +} + +int manager_notify_cgroup_empty(Manager *m, const char *cgroup) { + Unit *u; + + assert(m); + assert(cgroup); + + log_debug("Got cgroup empty notification for: %s", cgroup); + + u = manager_get_unit_by_cgroup(m, cgroup); + if (!u) + return 0; + + return unit_notify_cgroup_empty(u); +} + +int unit_get_memory_current(Unit *u, uint64_t *ret) { + _cleanup_free_ char *v = NULL; + int r; + + assert(u); + assert(ret); + + if (!u->cgroup_path) + return -ENODATA; + + if ((u->cgroup_realized_mask & CGROUP_MASK_MEMORY) == 0) + return -ENODATA; + + if (cg_unified() <= 0) + r = cg_get_attribute("memory", u->cgroup_path, "memory.usage_in_bytes", &v); + else + r = cg_get_attribute("memory", u->cgroup_path, "memory.current", &v); + if (r == -ENOENT) + return -ENODATA; + if (r < 0) + return r; + + return safe_atou64(v, ret); +} + +int unit_get_tasks_current(Unit *u, uint64_t *ret) { + _cleanup_free_ char *v = NULL; + int r; + + assert(u); + assert(ret); + + if (!u->cgroup_path) + return -ENODATA; + + if ((u->cgroup_realized_mask & CGROUP_MASK_PIDS) == 0) + return -ENODATA; + + r = cg_get_attribute("pids", u->cgroup_path, "pids.current", &v); + if (r == -ENOENT) + return -ENODATA; + if (r < 0) + return r; + + return safe_atou64(v, ret); +} + +static int unit_get_cpu_usage_raw(Unit *u, nsec_t *ret) { + _cleanup_free_ char *v = NULL; + uint64_t ns; + int r; + + assert(u); + assert(ret); + + if (!u->cgroup_path) + return -ENODATA; + + if ((u->cgroup_realized_mask & CGROUP_MASK_CPUACCT) == 0) + return -ENODATA; + + r = cg_get_attribute("cpuacct", u->cgroup_path, "cpuacct.usage", &v); + if (r == -ENOENT) + return -ENODATA; + if (r < 0) + return r; + + r = safe_atou64(v, &ns); + if (r < 0) + return r; + + *ret = ns; + return 0; +} + +int unit_get_cpu_usage(Unit *u, nsec_t *ret) { + nsec_t ns; + int r; + + r = unit_get_cpu_usage_raw(u, &ns); + if (r < 0) + return r; + + if (ns > u->cpuacct_usage_base) + ns -= u->cpuacct_usage_base; + else + ns = 0; + + *ret = ns; + return 0; +} + +int unit_reset_cpu_usage(Unit *u) { + nsec_t ns; + int r; + + assert(u); + + r = unit_get_cpu_usage_raw(u, &ns); + if (r < 0) { + u->cpuacct_usage_base = 0; + return r; + } + + u->cpuacct_usage_base = ns; + return 0; +} + +bool unit_cgroup_delegate(Unit *u) { + CGroupContext *c; + + assert(u); + + c = unit_get_cgroup_context(u); + if (!c) + return false; + + return c->delegate; +} + +void unit_invalidate_cgroup(Unit *u, CGroupMask m) { + assert(u); + + if (!UNIT_HAS_CGROUP_CONTEXT(u)) + return; + + if (m == 0) + return; + + /* always invalidate compat pairs together */ + if (m & (CGROUP_MASK_IO | CGROUP_MASK_BLKIO)) + m |= CGROUP_MASK_IO | CGROUP_MASK_BLKIO; + + if ((u->cgroup_realized_mask & m) == 0) + return; + + u->cgroup_realized_mask &= ~m; + unit_add_to_cgroup_queue(u); +} + +void manager_invalidate_startup_units(Manager *m) { + Iterator i; + Unit *u; + + assert(m); + + SET_FOREACH(u, m->startup_units, i) + unit_invalidate_cgroup(u, CGROUP_MASK_CPU|CGROUP_MASK_IO|CGROUP_MASK_BLKIO); +} + +static const char* const cgroup_device_policy_table[_CGROUP_DEVICE_POLICY_MAX] = { + [CGROUP_AUTO] = "auto", + [CGROUP_CLOSED] = "closed", + [CGROUP_STRICT] = "strict", +}; + +DEFINE_STRING_TABLE_LOOKUP(cgroup_device_policy, CGroupDevicePolicy); diff --git a/src/grp-system/libcore/src/dbus-automount.c b/src/grp-system/libcore/src/dbus-automount.c new file mode 100644 index 0000000000..500dcee502 --- /dev/null +++ b/src/grp-system/libcore/src/dbus-automount.c @@ -0,0 +1,35 @@ +/*** + This file is part of systemd. + + Copyright 2010 Lennart Poettering + + systemd is free software; you can redistribute it and/or modify it + under the terms of the GNU Lesser General Public License as published by + the Free Software Foundation; either version 2.1 of the License, or + (at your option) any later version. + + systemd is distributed in the hope that it will be useful, but + WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + Lesser General Public License for more details. + + You should have received a copy of the GNU Lesser General Public License + along with systemd; If not, see . +***/ + +#include "core/automount.h" +#include "sd-bus/bus-util.h" +#include "systemd-basic/string-util.h" + +#include "dbus-automount.h" + +static BUS_DEFINE_PROPERTY_GET_ENUM(property_get_result, automount_result, AutomountResult); + +const sd_bus_vtable bus_automount_vtable[] = { + SD_BUS_VTABLE_START(0), + SD_BUS_PROPERTY("Where", "s", NULL, offsetof(Automount, where), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("DirectoryMode", "u", bus_property_get_mode, offsetof(Automount, directory_mode), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("Result", "s", property_get_result, offsetof(Automount, result), SD_BUS_VTABLE_PROPERTY_EMITS_CHANGE), + SD_BUS_PROPERTY("TimeoutIdleUSec", "t", bus_property_get_usec, offsetof(Automount, timeout_idle_usec), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_VTABLE_END +}; diff --git a/src/grp-system/libcore/src/dbus-automount.h b/src/grp-system/libcore/src/dbus-automount.h new file mode 100644 index 0000000000..7b51eb973a --- /dev/null +++ b/src/grp-system/libcore/src/dbus-automount.h @@ -0,0 +1,23 @@ +#pragma once + +/*** + This file is part of systemd. + + Copyright 2010 Lennart Poettering + + systemd is free software; you can redistribute it and/or modify it + under the terms of the GNU Lesser General Public License as published by + the Free Software Foundation; either version 2.1 of the License, or + (at your option) any later version. + + systemd is distributed in the hope that it will be useful, but + WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + Lesser General Public License for more details. + + You should have received a copy of the GNU Lesser General Public License + along with systemd; If not, see . +***/ + + +extern const sd_bus_vtable bus_automount_vtable[]; diff --git a/src/grp-system/libcore/src/dbus-busname.c b/src/grp-system/libcore/src/dbus-busname.c new file mode 100644 index 0000000000..d1324e66bd --- /dev/null +++ b/src/grp-system/libcore/src/dbus-busname.c @@ -0,0 +1,38 @@ +/*** + This file is part of systemd. + + Copyright 2013 Lennart Poettering + + systemd is free software; you can redistribute it and/or modify it + under the terms of the GNU Lesser General Public License as published by + the Free Software Foundation; either version 2.1 of the License, or + (at your option) any later version. + + systemd is distributed in the hope that it will be useful, but + WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + Lesser General Public License for more details. + + You should have received a copy of the GNU Lesser General Public License + along with systemd; If not, see . +***/ + +#include "core/busname.h" +#include "core/unit.h" +#include "sd-bus/bus-util.h" +#include "systemd-basic/string-util.h" + +#include "dbus-busname.h" + +static BUS_DEFINE_PROPERTY_GET_ENUM(property_get_result, busname_result, BusNameResult); + +const sd_bus_vtable bus_busname_vtable[] = { + SD_BUS_VTABLE_START(0), + SD_BUS_PROPERTY("Name", "s", NULL, offsetof(BusName, name), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("TimeoutUSec", "t", bus_property_get_usec, offsetof(BusName, timeout_usec), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("ControlPID", "u", bus_property_get_pid, offsetof(BusName, control_pid), SD_BUS_VTABLE_PROPERTY_EMITS_CHANGE), + SD_BUS_PROPERTY("Result", "s", property_get_result, offsetof(BusName, result), SD_BUS_VTABLE_PROPERTY_EMITS_CHANGE), + SD_BUS_PROPERTY("Activating", "b", bus_property_get_bool, offsetof(BusName, activating), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("AcceptFileDescriptors", "b", bus_property_get_bool, offsetof(BusName, accept_fd), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_VTABLE_END +}; diff --git a/src/grp-system/libcore/src/dbus-busname.h b/src/grp-system/libcore/src/dbus-busname.h new file mode 100644 index 0000000000..8643d1a404 --- /dev/null +++ b/src/grp-system/libcore/src/dbus-busname.h @@ -0,0 +1,23 @@ +#pragma once + +/*** + This file is part of systemd. + + Copyright 2013 Lennart Poettering + + systemd is free software; you can redistribute it and/or modify it + under the terms of the GNU Lesser General Public License as published by + the Free Software Foundation; either version 2.1 of the License, or + (at your option) any later version. + + systemd is distributed in the hope that it will be useful, but + WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + Lesser General Public License for more details. + + You should have received a copy of the GNU Lesser General Public License + along with systemd; If not, see . +***/ + + +extern const sd_bus_vtable bus_busname_vtable[]; diff --git a/src/grp-system/libcore/src/dbus-cgroup.c b/src/grp-system/libcore/src/dbus-cgroup.c new file mode 100644 index 0000000000..e531f780fd --- /dev/null +++ b/src/grp-system/libcore/src/dbus-cgroup.c @@ -0,0 +1,1110 @@ +/*** + This file is part of systemd. + + Copyright 2013 Lennart Poettering + + systemd is free software; you can redistribute it and/or modify it + under the terms of the GNU Lesser General Public License as published by + the Free Software Foundation; either version 2.1 of the License, or + (at your option) any later version. + + systemd is distributed in the hope that it will be useful, but + WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + Lesser General Public License for more details. + + You should have received a copy of the GNU Lesser General Public License + along with systemd; If not, see . +***/ + +#include "core/cgroup.h" +#include "sd-bus/bus-util.h" +#include "systemd-basic/alloc-util.h" +#include "systemd-basic/cgroup-util.h" +#include "systemd-basic/fd-util.h" +#include "systemd-basic/fileio.h" +#include "systemd-basic/path-util.h" + +#include "dbus-cgroup.h" + +static BUS_DEFINE_PROPERTY_GET_ENUM(property_get_cgroup_device_policy, cgroup_device_policy, CGroupDevicePolicy); + +static int property_get_io_device_weight( + sd_bus *bus, + const char *path, + const char *interface, + const char *property, + sd_bus_message *reply, + void *userdata, + sd_bus_error *error) { + + CGroupContext *c = userdata; + CGroupIODeviceWeight *w; + int r; + + assert(bus); + assert(reply); + assert(c); + + r = sd_bus_message_open_container(reply, 'a', "(st)"); + if (r < 0) + return r; + + LIST_FOREACH(device_weights, w, c->io_device_weights) { + r = sd_bus_message_append(reply, "(st)", w->path, w->weight); + if (r < 0) + return r; + } + + return sd_bus_message_close_container(reply); +} + +static int property_get_io_device_limits( + sd_bus *bus, + const char *path, + const char *interface, + const char *property, + sd_bus_message *reply, + void *userdata, + sd_bus_error *error) { + + CGroupContext *c = userdata; + CGroupIODeviceLimit *l; + int r; + + assert(bus); + assert(reply); + assert(c); + + r = sd_bus_message_open_container(reply, 'a', "(st)"); + if (r < 0) + return r; + + LIST_FOREACH(device_limits, l, c->io_device_limits) { + CGroupIOLimitType type; + + type = cgroup_io_limit_type_from_string(property); + if (type < 0 || l->limits[type] == cgroup_io_limit_defaults[type]) + continue; + + r = sd_bus_message_append(reply, "(st)", l->path, l->limits[type]); + if (r < 0) + return r; + } + + return sd_bus_message_close_container(reply); +} + +static int property_get_blockio_device_weight( + sd_bus *bus, + const char *path, + const char *interface, + const char *property, + sd_bus_message *reply, + void *userdata, + sd_bus_error *error) { + + CGroupContext *c = userdata; + CGroupBlockIODeviceWeight *w; + int r; + + assert(bus); + assert(reply); + assert(c); + + r = sd_bus_message_open_container(reply, 'a', "(st)"); + if (r < 0) + return r; + + LIST_FOREACH(device_weights, w, c->blockio_device_weights) { + r = sd_bus_message_append(reply, "(st)", w->path, w->weight); + if (r < 0) + return r; + } + + return sd_bus_message_close_container(reply); +} + +static int property_get_blockio_device_bandwidths( + sd_bus *bus, + const char *path, + const char *interface, + const char *property, + sd_bus_message *reply, + void *userdata, + sd_bus_error *error) { + + CGroupContext *c = userdata; + CGroupBlockIODeviceBandwidth *b; + int r; + + assert(bus); + assert(reply); + assert(c); + + r = sd_bus_message_open_container(reply, 'a', "(st)"); + if (r < 0) + return r; + + LIST_FOREACH(device_bandwidths, b, c->blockio_device_bandwidths) { + uint64_t v; + + if (streq(property, "BlockIOReadBandwidth")) + v = b->rbps; + else + v = b->wbps; + + if (v == CGROUP_LIMIT_MAX) + continue; + + r = sd_bus_message_append(reply, "(st)", b->path, v); + if (r < 0) + return r; + } + + return sd_bus_message_close_container(reply); +} + +static int property_get_device_allow( + sd_bus *bus, + const char *path, + const char *interface, + const char *property, + sd_bus_message *reply, + void *userdata, + sd_bus_error *error) { + + CGroupContext *c = userdata; + CGroupDeviceAllow *a; + int r; + + assert(bus); + assert(reply); + assert(c); + + r = sd_bus_message_open_container(reply, 'a', "(ss)"); + if (r < 0) + return r; + + LIST_FOREACH(device_allow, a, c->device_allow) { + unsigned k = 0; + char rwm[4]; + + if (a->r) + rwm[k++] = 'r'; + if (a->w) + rwm[k++] = 'w'; + if (a->m) + rwm[k++] = 'm'; + + rwm[k] = 0; + + r = sd_bus_message_append(reply, "(ss)", a->path, rwm); + if (r < 0) + return r; + } + + return sd_bus_message_close_container(reply); +} + +const sd_bus_vtable bus_cgroup_vtable[] = { + SD_BUS_VTABLE_START(0), + SD_BUS_PROPERTY("Delegate", "b", bus_property_get_bool, offsetof(CGroupContext, delegate), 0), + SD_BUS_PROPERTY("CPUAccounting", "b", bus_property_get_bool, offsetof(CGroupContext, cpu_accounting), 0), + SD_BUS_PROPERTY("CPUShares", "t", NULL, offsetof(CGroupContext, cpu_shares), 0), + SD_BUS_PROPERTY("StartupCPUShares", "t", NULL, offsetof(CGroupContext, startup_cpu_shares), 0), + SD_BUS_PROPERTY("CPUQuotaPerSecUSec", "t", bus_property_get_usec, offsetof(CGroupContext, cpu_quota_per_sec_usec), 0), + SD_BUS_PROPERTY("IOAccounting", "b", bus_property_get_bool, offsetof(CGroupContext, io_accounting), 0), + SD_BUS_PROPERTY("IOWeight", "t", NULL, offsetof(CGroupContext, io_weight), 0), + SD_BUS_PROPERTY("StartupIOWeight", "t", NULL, offsetof(CGroupContext, startup_io_weight), 0), + SD_BUS_PROPERTY("IODeviceWeight", "a(st)", property_get_io_device_weight, 0, 0), + SD_BUS_PROPERTY("IOReadBandwidthMax", "a(st)", property_get_io_device_limits, 0, 0), + SD_BUS_PROPERTY("IOWriteBandwidthMax", "a(st)", property_get_io_device_limits, 0, 0), + SD_BUS_PROPERTY("IOReadIOPSMax", "a(st)", property_get_io_device_limits, 0, 0), + SD_BUS_PROPERTY("IOWriteIOPSMax", "a(st)", property_get_io_device_limits, 0, 0), + SD_BUS_PROPERTY("BlockIOAccounting", "b", bus_property_get_bool, offsetof(CGroupContext, blockio_accounting), 0), + SD_BUS_PROPERTY("BlockIOWeight", "t", NULL, offsetof(CGroupContext, blockio_weight), 0), + SD_BUS_PROPERTY("StartupBlockIOWeight", "t", NULL, offsetof(CGroupContext, startup_blockio_weight), 0), + SD_BUS_PROPERTY("BlockIODeviceWeight", "a(st)", property_get_blockio_device_weight, 0, 0), + SD_BUS_PROPERTY("BlockIOReadBandwidth", "a(st)", property_get_blockio_device_bandwidths, 0, 0), + SD_BUS_PROPERTY("BlockIOWriteBandwidth", "a(st)", property_get_blockio_device_bandwidths, 0, 0), + SD_BUS_PROPERTY("MemoryAccounting", "b", bus_property_get_bool, offsetof(CGroupContext, memory_accounting), 0), + SD_BUS_PROPERTY("MemoryLow", "t", NULL, offsetof(CGroupContext, memory_low), 0), + SD_BUS_PROPERTY("MemoryHigh", "t", NULL, offsetof(CGroupContext, memory_high), 0), + SD_BUS_PROPERTY("MemoryMax", "t", NULL, offsetof(CGroupContext, memory_max), 0), + SD_BUS_PROPERTY("MemoryLimit", "t", NULL, offsetof(CGroupContext, memory_limit), 0), + SD_BUS_PROPERTY("DevicePolicy", "s", property_get_cgroup_device_policy, offsetof(CGroupContext, device_policy), 0), + SD_BUS_PROPERTY("DeviceAllow", "a(ss)", property_get_device_allow, 0, 0), + SD_BUS_PROPERTY("TasksAccounting", "b", bus_property_get_bool, offsetof(CGroupContext, tasks_accounting), 0), + SD_BUS_PROPERTY("TasksMax", "t", NULL, offsetof(CGroupContext, tasks_max), 0), + SD_BUS_VTABLE_END +}; + +static int bus_cgroup_set_transient_property( + Unit *u, + CGroupContext *c, + const char *name, + sd_bus_message *message, + UnitSetPropertiesMode mode, + sd_bus_error *error) { + + int r; + + assert(u); + assert(c); + assert(name); + assert(message); + + if (streq(name, "Delegate")) { + int b; + + r = sd_bus_message_read(message, "b", &b); + if (r < 0) + return r; + + if (mode != UNIT_CHECK) { + c->delegate = b; + unit_write_drop_in_private(u, mode, name, b ? "Delegate=yes" : "Delegate=no"); + } + + return 1; + } + + return 0; +} + +int bus_cgroup_set_property( + Unit *u, + CGroupContext *c, + const char *name, + sd_bus_message *message, + UnitSetPropertiesMode mode, + sd_bus_error *error) { + + CGroupIOLimitType iol_type; + int r; + + assert(u); + assert(c); + assert(name); + assert(message); + + if (streq(name, "CPUAccounting")) { + int b; + + r = sd_bus_message_read(message, "b", &b); + if (r < 0) + return r; + + if (mode != UNIT_CHECK) { + c->cpu_accounting = b; + unit_invalidate_cgroup(u, CGROUP_MASK_CPUACCT|CGROUP_MASK_CPU); + unit_write_drop_in_private(u, mode, name, b ? "CPUAccounting=yes" : "CPUAccounting=no"); + } + + return 1; + + } else if (streq(name, "CPUShares")) { + uint64_t shares; + + r = sd_bus_message_read(message, "t", &shares); + if (r < 0) + return r; + + if (!CGROUP_CPU_SHARES_IS_OK(shares)) + return sd_bus_error_set_errnof(error, EINVAL, "CPUShares value out of range"); + + if (mode != UNIT_CHECK) { + c->cpu_shares = shares; + unit_invalidate_cgroup(u, CGROUP_MASK_CPU); + + if (shares == CGROUP_CPU_SHARES_INVALID) + unit_write_drop_in_private(u, mode, name, "CPUShares="); + else + unit_write_drop_in_private_format(u, mode, name, "CPUShares=%" PRIu64, shares); + } + + return 1; + + } else if (streq(name, "StartupCPUShares")) { + uint64_t shares; + + r = sd_bus_message_read(message, "t", &shares); + if (r < 0) + return r; + + if (!CGROUP_CPU_SHARES_IS_OK(shares)) + return sd_bus_error_set_errnof(error, EINVAL, "StartupCPUShares value out of range"); + + if (mode != UNIT_CHECK) { + c->startup_cpu_shares = shares; + unit_invalidate_cgroup(u, CGROUP_MASK_CPU); + + if (shares == CGROUP_CPU_SHARES_INVALID) + unit_write_drop_in_private(u, mode, name, "StartupCPUShares="); + else + unit_write_drop_in_private_format(u, mode, name, "StartupCPUShares=%" PRIu64, shares); + } + + return 1; + + } else if (streq(name, "CPUQuotaPerSecUSec")) { + uint64_t u64; + + r = sd_bus_message_read(message, "t", &u64); + if (r < 0) + return r; + + if (u64 <= 0) + return sd_bus_error_set_errnof(error, EINVAL, "CPUQuotaPerSecUSec value out of range"); + + if (mode != UNIT_CHECK) { + c->cpu_quota_per_sec_usec = u64; + unit_invalidate_cgroup(u, CGROUP_MASK_CPU); + unit_write_drop_in_private_format(u, mode, "CPUQuota", "CPUQuota=%0.f%%", (double) (c->cpu_quota_per_sec_usec / 10000)); + } + + return 1; + + } else if (streq(name, "IOAccounting")) { + int b; + + r = sd_bus_message_read(message, "b", &b); + if (r < 0) + return r; + + if (mode != UNIT_CHECK) { + c->io_accounting = b; + unit_invalidate_cgroup(u, CGROUP_MASK_IO); + unit_write_drop_in_private(u, mode, name, b ? "IOAccounting=yes" : "IOAccounting=no"); + } + + return 1; + + } else if (streq(name, "IOWeight")) { + uint64_t weight; + + r = sd_bus_message_read(message, "t", &weight); + if (r < 0) + return r; + + if (!CGROUP_WEIGHT_IS_OK(weight)) + return sd_bus_error_set_errnof(error, EINVAL, "IOWeight value out of range"); + + if (mode != UNIT_CHECK) { + c->io_weight = weight; + unit_invalidate_cgroup(u, CGROUP_MASK_IO); + + if (weight == CGROUP_WEIGHT_INVALID) + unit_write_drop_in_private(u, mode, name, "IOWeight="); + else + unit_write_drop_in_private_format(u, mode, name, "IOWeight=%" PRIu64, weight); + } + + return 1; + + } else if (streq(name, "StartupIOWeight")) { + uint64_t weight; + + r = sd_bus_message_read(message, "t", &weight); + if (r < 0) + return r; + + if (CGROUP_WEIGHT_IS_OK(weight)) + return sd_bus_error_set_errnof(error, EINVAL, "StartupIOWeight value out of range"); + + if (mode != UNIT_CHECK) { + c->startup_io_weight = weight; + unit_invalidate_cgroup(u, CGROUP_MASK_IO); + + if (weight == CGROUP_WEIGHT_INVALID) + unit_write_drop_in_private(u, mode, name, "StartupIOWeight="); + else + unit_write_drop_in_private_format(u, mode, name, "StartupIOWeight=%" PRIu64, weight); + } + + return 1; + + } else if ((iol_type = cgroup_io_limit_type_from_string(name)) >= 0) { + const char *path; + unsigned n = 0; + uint64_t u64; + + r = sd_bus_message_enter_container(message, 'a', "(st)"); + if (r < 0) + return r; + + while ((r = sd_bus_message_read(message, "(st)", &path, &u64)) > 0) { + + if (mode != UNIT_CHECK) { + CGroupIODeviceLimit *a = NULL, *b; + + LIST_FOREACH(device_limits, b, c->io_device_limits) { + if (path_equal(path, b->path)) { + a = b; + break; + } + } + + if (!a) { + CGroupIOLimitType type; + + a = new0(CGroupIODeviceLimit, 1); + if (!a) + return -ENOMEM; + + a->path = strdup(path); + if (!a->path) { + free(a); + return -ENOMEM; + } + + for (type = 0; type < _CGROUP_IO_LIMIT_TYPE_MAX; type++) + a->limits[type] = cgroup_io_limit_defaults[type]; + + LIST_PREPEND(device_limits, c->io_device_limits, a); + } + + a->limits[iol_type] = u64; + } + + n++; + } + if (r < 0) + return r; + + r = sd_bus_message_exit_container(message); + if (r < 0) + return r; + + if (mode != UNIT_CHECK) { + CGroupIODeviceLimit *a; + _cleanup_free_ char *buf = NULL; + _cleanup_fclose_ FILE *f = NULL; + size_t size = 0; + + if (n == 0) { + LIST_FOREACH(device_limits, a, c->io_device_limits) + a->limits[iol_type] = cgroup_io_limit_defaults[iol_type]; + } + + unit_invalidate_cgroup(u, CGROUP_MASK_IO); + + f = open_memstream(&buf, &size); + if (!f) + return -ENOMEM; + + fprintf(f, "%s=\n", name); + LIST_FOREACH(device_limits, a, c->io_device_limits) + if (a->limits[iol_type] != cgroup_io_limit_defaults[iol_type]) + fprintf(f, "%s=%s %" PRIu64 "\n", name, a->path, a->limits[iol_type]); + + r = fflush_and_check(f); + if (r < 0) + return r; + unit_write_drop_in_private(u, mode, name, buf); + } + + return 1; + + } else if (streq(name, "IODeviceWeight")) { + const char *path; + uint64_t weight; + unsigned n = 0; + + r = sd_bus_message_enter_container(message, 'a', "(st)"); + if (r < 0) + return r; + + while ((r = sd_bus_message_read(message, "(st)", &path, &weight)) > 0) { + + if (!CGROUP_WEIGHT_IS_OK(weight) || weight == CGROUP_WEIGHT_INVALID) + return sd_bus_error_set_errnof(error, EINVAL, "IODeviceWeight out of range"); + + if (mode != UNIT_CHECK) { + CGroupIODeviceWeight *a = NULL, *b; + + LIST_FOREACH(device_weights, b, c->io_device_weights) { + if (path_equal(b->path, path)) { + a = b; + break; + } + } + + if (!a) { + a = new0(CGroupIODeviceWeight, 1); + if (!a) + return -ENOMEM; + + a->path = strdup(path); + if (!a->path) { + free(a); + return -ENOMEM; + } + LIST_PREPEND(device_weights,c->io_device_weights, a); + } + + a->weight = weight; + } + + n++; + } + + r = sd_bus_message_exit_container(message); + if (r < 0) + return r; + + if (mode != UNIT_CHECK) { + _cleanup_free_ char *buf = NULL; + _cleanup_fclose_ FILE *f = NULL; + CGroupIODeviceWeight *a; + size_t size = 0; + + if (n == 0) { + while (c->io_device_weights) + cgroup_context_free_io_device_weight(c, c->io_device_weights); + } + + unit_invalidate_cgroup(u, CGROUP_MASK_IO); + + f = open_memstream(&buf, &size); + if (!f) + return -ENOMEM; + + fputs("IODeviceWeight=\n", f); + LIST_FOREACH(device_weights, a, c->io_device_weights) + fprintf(f, "IODeviceWeight=%s %" PRIu64 "\n", a->path, a->weight); + + r = fflush_and_check(f); + if (r < 0) + return r; + unit_write_drop_in_private(u, mode, name, buf); + } + + return 1; + + } else if (streq(name, "BlockIOAccounting")) { + int b; + + r = sd_bus_message_read(message, "b", &b); + if (r < 0) + return r; + + if (mode != UNIT_CHECK) { + c->blockio_accounting = b; + unit_invalidate_cgroup(u, CGROUP_MASK_BLKIO); + unit_write_drop_in_private(u, mode, name, b ? "BlockIOAccounting=yes" : "BlockIOAccounting=no"); + } + + return 1; + + } else if (streq(name, "BlockIOWeight")) { + uint64_t weight; + + r = sd_bus_message_read(message, "t", &weight); + if (r < 0) + return r; + + if (!CGROUP_BLKIO_WEIGHT_IS_OK(weight)) + return sd_bus_error_set_errnof(error, EINVAL, "BlockIOWeight value out of range"); + + if (mode != UNIT_CHECK) { + c->blockio_weight = weight; + unit_invalidate_cgroup(u, CGROUP_MASK_BLKIO); + + if (weight == CGROUP_BLKIO_WEIGHT_INVALID) + unit_write_drop_in_private(u, mode, name, "BlockIOWeight="); + else + unit_write_drop_in_private_format(u, mode, name, "BlockIOWeight=%" PRIu64, weight); + } + + return 1; + + } else if (streq(name, "StartupBlockIOWeight")) { + uint64_t weight; + + r = sd_bus_message_read(message, "t", &weight); + if (r < 0) + return r; + + if (!CGROUP_BLKIO_WEIGHT_IS_OK(weight)) + return sd_bus_error_set_errnof(error, EINVAL, "StartupBlockIOWeight value out of range"); + + if (mode != UNIT_CHECK) { + c->startup_blockio_weight = weight; + unit_invalidate_cgroup(u, CGROUP_MASK_BLKIO); + + if (weight == CGROUP_BLKIO_WEIGHT_INVALID) + unit_write_drop_in_private(u, mode, name, "StartupBlockIOWeight="); + else + unit_write_drop_in_private_format(u, mode, name, "StartupBlockIOWeight=%" PRIu64, weight); + } + + return 1; + + } else if (STR_IN_SET(name, "BlockIOReadBandwidth", "BlockIOWriteBandwidth")) { + const char *path; + bool read = true; + unsigned n = 0; + uint64_t u64; + + if (streq(name, "BlockIOWriteBandwidth")) + read = false; + + r = sd_bus_message_enter_container(message, 'a', "(st)"); + if (r < 0) + return r; + + while ((r = sd_bus_message_read(message, "(st)", &path, &u64)) > 0) { + + if (mode != UNIT_CHECK) { + CGroupBlockIODeviceBandwidth *a = NULL, *b; + + LIST_FOREACH(device_bandwidths, b, c->blockio_device_bandwidths) { + if (path_equal(path, b->path)) { + a = b; + break; + } + } + + if (!a) { + a = new0(CGroupBlockIODeviceBandwidth, 1); + if (!a) + return -ENOMEM; + + a->rbps = CGROUP_LIMIT_MAX; + a->wbps = CGROUP_LIMIT_MAX; + a->path = strdup(path); + if (!a->path) { + free(a); + return -ENOMEM; + } + + LIST_PREPEND(device_bandwidths, c->blockio_device_bandwidths, a); + } + + if (read) + a->rbps = u64; + else + a->wbps = u64; + } + + n++; + } + if (r < 0) + return r; + + r = sd_bus_message_exit_container(message); + if (r < 0) + return r; + + if (mode != UNIT_CHECK) { + CGroupBlockIODeviceBandwidth *a; + _cleanup_free_ char *buf = NULL; + _cleanup_fclose_ FILE *f = NULL; + size_t size = 0; + + if (n == 0) { + LIST_FOREACH(device_bandwidths, a, c->blockio_device_bandwidths) { + if (read) + a->rbps = CGROUP_LIMIT_MAX; + else + a->wbps = CGROUP_LIMIT_MAX; + } + } + + unit_invalidate_cgroup(u, CGROUP_MASK_BLKIO); + + f = open_memstream(&buf, &size); + if (!f) + return -ENOMEM; + + if (read) { + fputs("BlockIOReadBandwidth=\n", f); + LIST_FOREACH(device_bandwidths, a, c->blockio_device_bandwidths) + if (a->rbps != CGROUP_LIMIT_MAX) + fprintf(f, "BlockIOReadBandwidth=%s %" PRIu64 "\n", a->path, a->rbps); + } else { + fputs("BlockIOWriteBandwidth=\n", f); + LIST_FOREACH(device_bandwidths, a, c->blockio_device_bandwidths) + if (a->wbps != CGROUP_LIMIT_MAX) + fprintf(f, "BlockIOWriteBandwidth=%s %" PRIu64 "\n", a->path, a->wbps); + } + + r = fflush_and_check(f); + if (r < 0) + return r; + unit_write_drop_in_private(u, mode, name, buf); + } + + return 1; + + } else if (streq(name, "BlockIODeviceWeight")) { + const char *path; + uint64_t weight; + unsigned n = 0; + + r = sd_bus_message_enter_container(message, 'a', "(st)"); + if (r < 0) + return r; + + while ((r = sd_bus_message_read(message, "(st)", &path, &weight)) > 0) { + + if (!CGROUP_BLKIO_WEIGHT_IS_OK(weight) || weight == CGROUP_BLKIO_WEIGHT_INVALID) + return sd_bus_error_set_errnof(error, EINVAL, "BlockIODeviceWeight out of range"); + + if (mode != UNIT_CHECK) { + CGroupBlockIODeviceWeight *a = NULL, *b; + + LIST_FOREACH(device_weights, b, c->blockio_device_weights) { + if (path_equal(b->path, path)) { + a = b; + break; + } + } + + if (!a) { + a = new0(CGroupBlockIODeviceWeight, 1); + if (!a) + return -ENOMEM; + + a->path = strdup(path); + if (!a->path) { + free(a); + return -ENOMEM; + } + LIST_PREPEND(device_weights,c->blockio_device_weights, a); + } + + a->weight = weight; + } + + n++; + } + + r = sd_bus_message_exit_container(message); + if (r < 0) + return r; + + if (mode != UNIT_CHECK) { + _cleanup_free_ char *buf = NULL; + _cleanup_fclose_ FILE *f = NULL; + CGroupBlockIODeviceWeight *a; + size_t size = 0; + + if (n == 0) { + while (c->blockio_device_weights) + cgroup_context_free_blockio_device_weight(c, c->blockio_device_weights); + } + + unit_invalidate_cgroup(u, CGROUP_MASK_BLKIO); + + f = open_memstream(&buf, &size); + if (!f) + return -ENOMEM; + + fputs("BlockIODeviceWeight=\n", f); + LIST_FOREACH(device_weights, a, c->blockio_device_weights) + fprintf(f, "BlockIODeviceWeight=%s %" PRIu64 "\n", a->path, a->weight); + + r = fflush_and_check(f); + if (r < 0) + return r; + unit_write_drop_in_private(u, mode, name, buf); + } + + return 1; + + } else if (streq(name, "MemoryAccounting")) { + int b; + + r = sd_bus_message_read(message, "b", &b); + if (r < 0) + return r; + + if (mode != UNIT_CHECK) { + c->memory_accounting = b; + unit_invalidate_cgroup(u, CGROUP_MASK_MEMORY); + unit_write_drop_in_private(u, mode, name, b ? "MemoryAccounting=yes" : "MemoryAccounting=no"); + } + + return 1; + + } else if (STR_IN_SET(name, "MemoryLow", "MemoryHigh", "MemoryMax")) { + uint64_t v; + + r = sd_bus_message_read(message, "t", &v); + if (r < 0) + return r; + if (v <= 0) + return sd_bus_error_set_errnof(error, EINVAL, "%s= is too small", name); + + if (mode != UNIT_CHECK) { + if (streq(name, "MemoryLow")) + c->memory_low = v; + else if (streq(name, "MemoryHigh")) + c->memory_high = v; + else + c->memory_max = v; + + unit_invalidate_cgroup(u, CGROUP_MASK_MEMORY); + + if (v == CGROUP_LIMIT_MAX) + unit_write_drop_in_private_format(u, mode, name, "%s=infinity", name); + else + unit_write_drop_in_private_format(u, mode, name, "%s=%" PRIu64, name, v); + } + + return 1; + + } else if (STR_IN_SET(name, "MemoryLowScale", "MemoryHighScale", "MemoryMaxScale")) { + uint32_t raw; + uint64_t v; + + r = sd_bus_message_read(message, "u", &raw); + if (r < 0) + return r; + + v = physical_memory_scale(raw, UINT32_MAX); + if (v <= 0 || v == UINT64_MAX) + return sd_bus_error_set_errnof(error, EINVAL, "%s= is out of range", name); + + if (mode != UNIT_CHECK) { + const char *e; + + /* Chop off suffix */ + assert_se(e = endswith(name, "Scale")); + name = strndupa(name, e - name); + + if (streq(name, "MemoryLow")) + c->memory_low = v; + else if (streq(name, "MemoryHigh")) + c->memory_high = v; + else + c->memory_max = v; + + unit_invalidate_cgroup(u, CGROUP_MASK_MEMORY); + unit_write_drop_in_private_format(u, mode, name, "%s=%" PRIu32 "%%", name, + (uint32_t) (DIV_ROUND_UP((uint64_t) raw * 100U, (uint64_t) UINT32_MAX))); + } + + return 1; + + } else if (streq(name, "MemoryLimit")) { + uint64_t limit; + + r = sd_bus_message_read(message, "t", &limit); + if (r < 0) + return r; + if (limit <= 0) + return sd_bus_error_set_errnof(error, EINVAL, "%s= is too small", name); + + if (mode != UNIT_CHECK) { + c->memory_limit = limit; + unit_invalidate_cgroup(u, CGROUP_MASK_MEMORY); + + if (limit == (uint64_t) -1) + unit_write_drop_in_private(u, mode, name, "MemoryLimit=infinity"); + else + unit_write_drop_in_private_format(u, mode, name, "MemoryLimit=%" PRIu64, limit); + } + + return 1; + + } else if (streq(name, "MemoryLimitScale")) { + uint64_t limit; + uint32_t raw; + + r = sd_bus_message_read(message, "u", &raw); + if (r < 0) + return r; + + limit = physical_memory_scale(raw, UINT32_MAX); + if (limit <= 0 || limit == UINT64_MAX) + return sd_bus_error_set_errnof(error, EINVAL, "%s= is out of range", name); + + if (mode != UNIT_CHECK) { + c->memory_limit = limit; + unit_invalidate_cgroup(u, CGROUP_MASK_MEMORY); + unit_write_drop_in_private_format(u, mode, "MemoryLimit", "MemoryLimit=%" PRIu32 "%%", + (uint32_t) (DIV_ROUND_UP((uint64_t) raw * 100U, (uint64_t) UINT32_MAX))); + } + + return 1; + + } else if (streq(name, "DevicePolicy")) { + const char *policy; + CGroupDevicePolicy p; + + r = sd_bus_message_read(message, "s", &policy); + if (r < 0) + return r; + + p = cgroup_device_policy_from_string(policy); + if (p < 0) + return -EINVAL; + + if (mode != UNIT_CHECK) { + c->device_policy = p; + unit_invalidate_cgroup(u, CGROUP_MASK_DEVICES); + unit_write_drop_in_private_format(u, mode, name, "DevicePolicy=%s", policy); + } + + return 1; + + } else if (streq(name, "DeviceAllow")) { + const char *path, *rwm; + unsigned n = 0; + + r = sd_bus_message_enter_container(message, 'a', "(ss)"); + if (r < 0) + return r; + + while ((r = sd_bus_message_read(message, "(ss)", &path, &rwm)) > 0) { + + if ((!startswith(path, "/dev/") && + !startswith(path, "/run/systemd/inaccessible/") && + !startswith(path, "block-") && + !startswith(path, "char-")) || + strpbrk(path, WHITESPACE)) + return sd_bus_error_set_errnof(error, EINVAL, "DeviceAllow= requires device node"); + + if (isempty(rwm)) + rwm = "rwm"; + + if (!in_charset(rwm, "rwm")) + return sd_bus_error_set_errnof(error, EINVAL, "DeviceAllow= requires combination of rwm flags"); + + if (mode != UNIT_CHECK) { + CGroupDeviceAllow *a = NULL, *b; + + LIST_FOREACH(device_allow, b, c->device_allow) { + if (path_equal(b->path, path)) { + a = b; + break; + } + } + + if (!a) { + a = new0(CGroupDeviceAllow, 1); + if (!a) + return -ENOMEM; + + a->path = strdup(path); + if (!a->path) { + free(a); + return -ENOMEM; + } + + LIST_PREPEND(device_allow, c->device_allow, a); + } + + a->r = !!strchr(rwm, 'r'); + a->w = !!strchr(rwm, 'w'); + a->m = !!strchr(rwm, 'm'); + } + + n++; + } + if (r < 0) + return r; + + r = sd_bus_message_exit_container(message); + if (r < 0) + return r; + + if (mode != UNIT_CHECK) { + _cleanup_free_ char *buf = NULL; + _cleanup_fclose_ FILE *f = NULL; + CGroupDeviceAllow *a; + size_t size = 0; + + if (n == 0) { + while (c->device_allow) + cgroup_context_free_device_allow(c, c->device_allow); + } + + unit_invalidate_cgroup(u, CGROUP_MASK_DEVICES); + + f = open_memstream(&buf, &size); + if (!f) + return -ENOMEM; + + fputs("DeviceAllow=\n", f); + LIST_FOREACH(device_allow, a, c->device_allow) + fprintf(f, "DeviceAllow=%s %s%s%s\n", a->path, a->r ? "r" : "", a->w ? "w" : "", a->m ? "m" : ""); + + r = fflush_and_check(f); + if (r < 0) + return r; + unit_write_drop_in_private(u, mode, name, buf); + } + + return 1; + + } else if (streq(name, "TasksAccounting")) { + int b; + + r = sd_bus_message_read(message, "b", &b); + if (r < 0) + return r; + + if (mode != UNIT_CHECK) { + c->tasks_accounting = b; + unit_invalidate_cgroup(u, CGROUP_MASK_PIDS); + unit_write_drop_in_private(u, mode, name, b ? "TasksAccounting=yes" : "TasksAccounting=no"); + } + + return 1; + + } else if (streq(name, "TasksMax")) { + uint64_t limit; + + r = sd_bus_message_read(message, "t", &limit); + if (r < 0) + return r; + if (limit <= 0) + return sd_bus_error_set_errnof(error, EINVAL, "%s= is too small", name); + + if (mode != UNIT_CHECK) { + c->tasks_max = limit; + unit_invalidate_cgroup(u, CGROUP_MASK_PIDS); + + if (limit == (uint64_t) -1) + unit_write_drop_in_private(u, mode, name, "TasksMax=infinity"); + else + unit_write_drop_in_private_format(u, mode, name, "TasksMax=%" PRIu64, limit); + } + + return 1; + } else if (streq(name, "TasksMaxScale")) { + uint64_t limit; + uint32_t raw; + + r = sd_bus_message_read(message, "u", &raw); + if (r < 0) + return r; + + limit = system_tasks_max_scale(raw, UINT32_MAX); + if (limit <= 0 || limit >= UINT64_MAX) + return sd_bus_error_set_errnof(error, EINVAL, "%s= is out of range", name); + + if (mode != UNIT_CHECK) { + c->tasks_max = limit; + unit_invalidate_cgroup(u, CGROUP_MASK_PIDS); + unit_write_drop_in_private_format(u, mode, name, "TasksMax=%" PRIu32 "%%", + (uint32_t) (DIV_ROUND_UP((uint64_t) raw * 100U, (uint64_t) UINT32_MAX))); + } + + return 1; + } + + if (u->transient && u->load_state == UNIT_STUB) { + r = bus_cgroup_set_transient_property(u, c, name, message, mode, error); + if (r != 0) + return r; + + } + + return 0; +} diff --git a/src/grp-system/libcore/src/dbus-cgroup.h b/src/grp-system/libcore/src/dbus-cgroup.h new file mode 100644 index 0000000000..b61ca2fcef --- /dev/null +++ b/src/grp-system/libcore/src/dbus-cgroup.h @@ -0,0 +1,28 @@ +#pragma once + +/*** + This file is part of systemd. + + Copyright 2013 Lennart Poettering + + systemd is free software; you can redistribute it and/or modify it + under the terms of the GNU Lesser General Public License as published by + the Free Software Foundation; either version 2.1 of the License, or + (at your option) any later version. + + systemd is distributed in the hope that it will be useful, but + WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + Lesser General Public License for more details. + + You should have received a copy of the GNU Lesser General Public License + along with systemd; If not, see . +***/ + +#include + +#include "core/cgroup.h" + +extern const sd_bus_vtable bus_cgroup_vtable[]; + +int bus_cgroup_set_property(Unit *u, CGroupContext *c, const char *name, sd_bus_message *message, UnitSetPropertiesMode mode, sd_bus_error *error); diff --git a/src/grp-system/libcore/src/dbus-device.c b/src/grp-system/libcore/src/dbus-device.c new file mode 100644 index 0000000000..75e9beb55e --- /dev/null +++ b/src/grp-system/libcore/src/dbus-device.c @@ -0,0 +1,29 @@ +/*** + This file is part of systemd. + + Copyright 2010 Lennart Poettering + + systemd is free software; you can redistribute it and/or modify it + under the terms of the GNU Lesser General Public License as published by + the Free Software Foundation; either version 2.1 of the License, or + (at your option) any later version. + + systemd is distributed in the hope that it will be useful, but + WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + Lesser General Public License for more details. + + You should have received a copy of the GNU Lesser General Public License + along with systemd; If not, see . +***/ + +#include "core/device.h" +#include "core/unit.h" + +#include "dbus-device.h" + +const sd_bus_vtable bus_device_vtable[] = { + SD_BUS_VTABLE_START(0), + SD_BUS_PROPERTY("SysFSPath", "s", NULL, offsetof(Device, sysfs), SD_BUS_VTABLE_PROPERTY_EMITS_CHANGE), + SD_BUS_VTABLE_END +}; diff --git a/src/grp-system/libcore/src/dbus-device.h b/src/grp-system/libcore/src/dbus-device.h new file mode 100644 index 0000000000..84b7ed56ec --- /dev/null +++ b/src/grp-system/libcore/src/dbus-device.h @@ -0,0 +1,24 @@ +#pragma once + +/*** + This file is part of systemd. + + Copyright 2010 Lennart Poettering + + systemd is free software; you can redistribute it and/or modify it + under the terms of the GNU Lesser General Public License as published by + the Free Software Foundation; either version 2.1 of the License, or + (at your option) any later version. + + systemd is distributed in the hope that it will be useful, but + WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + Lesser General Public License for more details. + + You should have received a copy of the GNU Lesser General Public License + along with systemd; If not, see . +***/ + +#include "core/unit.h" + +extern const sd_bus_vtable bus_device_vtable[]; diff --git a/src/grp-system/libcore/src/dbus-execute.c b/src/grp-system/libcore/src/dbus-execute.c new file mode 100644 index 0000000000..8029497c3c --- /dev/null +++ b/src/grp-system/libcore/src/dbus-execute.c @@ -0,0 +1,1555 @@ +/*** + This file is part of systemd. + + Copyright 2010 Lennart Poettering + + systemd is free software; you can redistribute it and/or modify it + under the terms of the GNU Lesser General Public License as published by + the Free Software Foundation; either version 2.1 of the License, or + (at your option) any later version. + + systemd is distributed in the hope that it will be useful, but + WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + Lesser General Public License for more details. + + You should have received a copy of the GNU Lesser General Public License + along with systemd; If not, see . +***/ + +#include + +#ifdef HAVE_SECCOMP +#include +#endif + +#include "core/execute.h" +#include "core/namespace.h" +#include "sd-bus/bus-util.h" +#include "systemd-basic/af-list.h" +#include "systemd-basic/alloc-util.h" +#include "systemd-basic/capability-util.h" +#include "systemd-basic/env-util.h" +#include "systemd-basic/fd-util.h" +#include "systemd-basic/fileio.h" +#include "systemd-basic/ioprio.h" +#include "systemd-basic/missing.h" +#include "systemd-basic/parse-util.h" +#include "systemd-basic/path-util.h" +#include "systemd-basic/process-util.h" +#include "systemd-basic/rlimit-util.h" + +#include "dbus-execute.h" +#ifdef HAVE_SECCOMP +#include "systemd-shared/seccomp-util.h" +#endif +#include "systemd-basic/strv.h" +#include "systemd-basic/syslog-util.h" +#include "systemd-basic/utf8.h" + +BUS_DEFINE_PROPERTY_GET_ENUM(bus_property_get_exec_output, exec_output, ExecOutput); + +static BUS_DEFINE_PROPERTY_GET_ENUM(property_get_exec_input, exec_input, ExecInput); + +static BUS_DEFINE_PROPERTY_GET_ENUM(property_get_exec_utmp_mode, exec_utmp_mode, ExecUtmpMode); + +static BUS_DEFINE_PROPERTY_GET_ENUM(bus_property_get_protect_home, protect_home, ProtectHome); +static BUS_DEFINE_PROPERTY_GET_ENUM(bus_property_get_protect_system, protect_system, ProtectSystem); + +static int property_get_environment_files( + sd_bus *bus, + const char *path, + const char *interface, + const char *property, + sd_bus_message *reply, + void *userdata, + sd_bus_error *error) { + + ExecContext *c = userdata; + char **j; + int r; + + assert(bus); + assert(reply); + assert(c); + + r = sd_bus_message_open_container(reply, 'a', "(sb)"); + if (r < 0) + return r; + + STRV_FOREACH(j, c->environment_files) { + const char *fn = *j; + + r = sd_bus_message_append(reply, "(sb)", fn[0] == '-' ? fn + 1 : fn, fn[0] == '-'); + if (r < 0) + return r; + } + + return sd_bus_message_close_container(reply); +} + +static int property_get_oom_score_adjust( + sd_bus *bus, + const char *path, + const char *interface, + const char *property, + sd_bus_message *reply, + void *userdata, + sd_bus_error *error) { + + + ExecContext *c = userdata; + int32_t n; + + assert(bus); + assert(reply); + assert(c); + + if (c->oom_score_adjust_set) + n = c->oom_score_adjust; + else { + _cleanup_free_ char *t = NULL; + + n = 0; + if (read_one_line_file("/proc/self/oom_score_adj", &t) >= 0) + safe_atoi32(t, &n); + } + + return sd_bus_message_append(reply, "i", n); +} + +static int property_get_nice( + sd_bus *bus, + const char *path, + const char *interface, + const char *property, + sd_bus_message *reply, + void *userdata, + sd_bus_error *error) { + + + ExecContext *c = userdata; + int32_t n; + + assert(bus); + assert(reply); + assert(c); + + if (c->nice_set) + n = c->nice; + else { + errno = 0; + n = getpriority(PRIO_PROCESS, 0); + if (errno > 0) + n = 0; + } + + return sd_bus_message_append(reply, "i", n); +} + +static int property_get_ioprio( + sd_bus *bus, + const char *path, + const char *interface, + const char *property, + sd_bus_message *reply, + void *userdata, + sd_bus_error *error) { + + + ExecContext *c = userdata; + int32_t n; + + assert(bus); + assert(reply); + assert(c); + + if (c->ioprio_set) + n = c->ioprio; + else { + n = ioprio_get(IOPRIO_WHO_PROCESS, 0); + if (n < 0) + n = IOPRIO_PRIO_VALUE(IOPRIO_CLASS_BE, 4); + } + + return sd_bus_message_append(reply, "i", n); +} + +static int property_get_cpu_sched_policy( + sd_bus *bus, + const char *path, + const char *interface, + const char *property, + sd_bus_message *reply, + void *userdata, + sd_bus_error *error) { + + ExecContext *c = userdata; + int32_t n; + + assert(bus); + assert(reply); + assert(c); + + if (c->cpu_sched_set) + n = c->cpu_sched_policy; + else { + n = sched_getscheduler(0); + if (n < 0) + n = SCHED_OTHER; + } + + return sd_bus_message_append(reply, "i", n); +} + +static int property_get_cpu_sched_priority( + sd_bus *bus, + const char *path, + const char *interface, + const char *property, + sd_bus_message *reply, + void *userdata, + sd_bus_error *error) { + + ExecContext *c = userdata; + int32_t n; + + assert(bus); + assert(reply); + assert(c); + + if (c->cpu_sched_set) + n = c->cpu_sched_priority; + else { + struct sched_param p = {}; + + if (sched_getparam(0, &p) >= 0) + n = p.sched_priority; + else + n = 0; + } + + return sd_bus_message_append(reply, "i", n); +} + +static int property_get_cpu_affinity( + sd_bus *bus, + const char *path, + const char *interface, + const char *property, + sd_bus_message *reply, + void *userdata, + sd_bus_error *error) { + + ExecContext *c = userdata; + + assert(bus); + assert(reply); + assert(c); + + if (c->cpuset) + return sd_bus_message_append_array(reply, 'y', c->cpuset, CPU_ALLOC_SIZE(c->cpuset_ncpus)); + else + return sd_bus_message_append_array(reply, 'y', NULL, 0); +} + +static int property_get_timer_slack_nsec( + sd_bus *bus, + const char *path, + const char *interface, + const char *property, + sd_bus_message *reply, + void *userdata, + sd_bus_error *error) { + + ExecContext *c = userdata; + uint64_t u; + + assert(bus); + assert(reply); + assert(c); + + if (c->timer_slack_nsec != NSEC_INFINITY) + u = (uint64_t) c->timer_slack_nsec; + else + u = (uint64_t) prctl(PR_GET_TIMERSLACK); + + return sd_bus_message_append(reply, "t", u); +} + +static int property_get_capability_bounding_set( + sd_bus *bus, + const char *path, + const char *interface, + const char *property, + sd_bus_message *reply, + void *userdata, + sd_bus_error *error) { + + ExecContext *c = userdata; + + assert(bus); + assert(reply); + assert(c); + + return sd_bus_message_append(reply, "t", c->capability_bounding_set); +} + +static int property_get_ambient_capabilities( + sd_bus *bus, + const char *path, + const char *interface, + const char *property, + sd_bus_message *reply, + void *userdata, + sd_bus_error *error) { + + ExecContext *c = userdata; + + assert(bus); + assert(reply); + assert(c); + + return sd_bus_message_append(reply, "t", c->capability_ambient_set); +} + +static int property_get_empty_string( + sd_bus *bus, + const char *path, + const char *interface, + const char *property, + sd_bus_message *reply, + void *userdata, + sd_bus_error *error) { + + assert(bus); + assert(reply); + + return sd_bus_message_append(reply, "s", ""); +} + +static int property_get_syscall_filter( + sd_bus *bus, + const char *path, + const char *interface, + const char *property, + sd_bus_message *reply, + void *userdata, + sd_bus_error *error) { + + ExecContext *c = userdata; + _cleanup_strv_free_ char **l = NULL; + int r; + +#ifdef HAVE_SECCOMP + Iterator i; + void *id; +#endif + + assert(bus); + assert(reply); + assert(c); + + r = sd_bus_message_open_container(reply, 'r', "bas"); + if (r < 0) + return r; + + r = sd_bus_message_append(reply, "b", c->syscall_whitelist); + if (r < 0) + return r; + +#ifdef HAVE_SECCOMP + SET_FOREACH(id, c->syscall_filter, i) { + char *name; + + name = seccomp_syscall_resolve_num_arch(SCMP_ARCH_NATIVE, PTR_TO_INT(id) - 1); + if (!name) + continue; + + r = strv_consume(&l, name); + if (r < 0) + return r; + } +#endif + + strv_sort(l); + + r = sd_bus_message_append_strv(reply, l); + if (r < 0) + return r; + + return sd_bus_message_close_container(reply); +} + +static int property_get_syscall_archs( + sd_bus *bus, + const char *path, + const char *interface, + const char *property, + sd_bus_message *reply, + void *userdata, + sd_bus_error *error) { + + ExecContext *c = userdata; + _cleanup_strv_free_ char **l = NULL; + int r; + +#ifdef HAVE_SECCOMP + Iterator i; + void *id; +#endif + + assert(bus); + assert(reply); + assert(c); + +#ifdef HAVE_SECCOMP + SET_FOREACH(id, c->syscall_archs, i) { + const char *name; + + name = seccomp_arch_to_string(PTR_TO_UINT32(id) - 1); + if (!name) + continue; + + r = strv_extend(&l, name); + if (r < 0) + return -ENOMEM; + } +#endif + + strv_sort(l); + + r = sd_bus_message_append_strv(reply, l); + if (r < 0) + return r; + + return 0; +} + +static int property_get_syscall_errno( + sd_bus *bus, + const char *path, + const char *interface, + const char *property, + sd_bus_message *reply, + void *userdata, + sd_bus_error *error) { + + ExecContext *c = userdata; + + assert(bus); + assert(reply); + assert(c); + + return sd_bus_message_append(reply, "i", (int32_t) c->syscall_errno); +} + +static int property_get_selinux_context( + sd_bus *bus, + const char *path, + const char *interface, + const char *property, + sd_bus_message *reply, + void *userdata, + sd_bus_error *error) { + + ExecContext *c = userdata; + + assert(bus); + assert(reply); + assert(c); + + return sd_bus_message_append(reply, "(bs)", c->selinux_context_ignore, c->selinux_context); +} + +static int property_get_apparmor_profile( + sd_bus *bus, + const char *path, + const char *interface, + const char *property, + sd_bus_message *reply, + void *userdata, + sd_bus_error *error) { + + ExecContext *c = userdata; + + assert(bus); + assert(reply); + assert(c); + + return sd_bus_message_append(reply, "(bs)", c->apparmor_profile_ignore, c->apparmor_profile); +} + +static int property_get_smack_process_label( + sd_bus *bus, + const char *path, + const char *interface, + const char *property, + sd_bus_message *reply, + void *userdata, + sd_bus_error *error) { + + ExecContext *c = userdata; + + assert(bus); + assert(reply); + assert(c); + + return sd_bus_message_append(reply, "(bs)", c->smack_process_label_ignore, c->smack_process_label); +} + +static int property_get_personality( + sd_bus *bus, + const char *path, + const char *interface, + const char *property, + sd_bus_message *reply, + void *userdata, + sd_bus_error *error) { + + ExecContext *c = userdata; + + assert(bus); + assert(reply); + assert(c); + + return sd_bus_message_append(reply, "s", personality_to_string(c->personality)); +} + +static int property_get_address_families( + sd_bus *bus, + const char *path, + const char *interface, + const char *property, + sd_bus_message *reply, + void *userdata, + sd_bus_error *error) { + + ExecContext *c = userdata; + _cleanup_strv_free_ char **l = NULL; + Iterator i; + void *af; + int r; + + assert(bus); + assert(reply); + assert(c); + + r = sd_bus_message_open_container(reply, 'r', "bas"); + if (r < 0) + return r; + + r = sd_bus_message_append(reply, "b", c->address_families_whitelist); + if (r < 0) + return r; + + SET_FOREACH(af, c->address_families, i) { + const char *name; + + name = af_to_name(PTR_TO_INT(af)); + if (!name) + continue; + + r = strv_extend(&l, name); + if (r < 0) + return -ENOMEM; + } + + strv_sort(l); + + r = sd_bus_message_append_strv(reply, l); + if (r < 0) + return r; + + return sd_bus_message_close_container(reply); +} + +static int property_get_working_directory( + sd_bus *bus, + const char *path, + const char *interface, + const char *property, + sd_bus_message *reply, + void *userdata, + sd_bus_error *error) { + + ExecContext *c = userdata; + const char *wd; + + assert(bus); + assert(reply); + assert(c); + + if (c->working_directory_home) + wd = "~"; + else + wd = c->working_directory; + + if (c->working_directory_missing_ok) + wd = strjoina("!", wd); + + return sd_bus_message_append(reply, "s", wd); +} + +static int property_get_syslog_level( + sd_bus *bus, + const char *path, + const char *interface, + const char *property, + sd_bus_message *reply, + void *userdata, + sd_bus_error *error) { + + ExecContext *c = userdata; + + assert(bus); + assert(reply); + assert(c); + + return sd_bus_message_append(reply, "i", LOG_PRI(c->syslog_priority)); +} + +static int property_get_syslog_facility( + sd_bus *bus, + const char *path, + const char *interface, + const char *property, + sd_bus_message *reply, + void *userdata, + sd_bus_error *error) { + + ExecContext *c = userdata; + + assert(bus); + assert(reply); + assert(c); + + return sd_bus_message_append(reply, "i", LOG_FAC(c->syslog_priority)); +} + +const sd_bus_vtable bus_exec_vtable[] = { + SD_BUS_VTABLE_START(0), + SD_BUS_PROPERTY("Environment", "as", NULL, offsetof(ExecContext, environment), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("EnvironmentFiles", "a(sb)", property_get_environment_files, 0, SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("PassEnvironment", "as", NULL, offsetof(ExecContext, pass_environment), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("UMask", "u", bus_property_get_mode, offsetof(ExecContext, umask), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("LimitCPU", "t", bus_property_get_rlimit, offsetof(ExecContext, rlimit[RLIMIT_CPU]), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("LimitCPUSoft", "t", bus_property_get_rlimit, offsetof(ExecContext, rlimit[RLIMIT_CPU]), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("LimitFSIZE", "t", bus_property_get_rlimit, offsetof(ExecContext, rlimit[RLIMIT_FSIZE]), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("LimitFSIZESoft", "t", bus_property_get_rlimit, offsetof(ExecContext, rlimit[RLIMIT_FSIZE]), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("LimitDATA", "t", bus_property_get_rlimit, offsetof(ExecContext, rlimit[RLIMIT_DATA]), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("LimitDATASoft", "t", bus_property_get_rlimit, offsetof(ExecContext, rlimit[RLIMIT_DATA]), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("LimitSTACK", "t", bus_property_get_rlimit, offsetof(ExecContext, rlimit[RLIMIT_STACK]), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("LimitSTACKSoft", "t", bus_property_get_rlimit, offsetof(ExecContext, rlimit[RLIMIT_STACK]), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("LimitCORE", "t", bus_property_get_rlimit, offsetof(ExecContext, rlimit[RLIMIT_CORE]), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("LimitCORESoft", "t", bus_property_get_rlimit, offsetof(ExecContext, rlimit[RLIMIT_CORE]), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("LimitRSS", "t", bus_property_get_rlimit, offsetof(ExecContext, rlimit[RLIMIT_RSS]), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("LimitRSSSoft", "t", bus_property_get_rlimit, offsetof(ExecContext, rlimit[RLIMIT_RSS]), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("LimitNOFILE", "t", bus_property_get_rlimit, offsetof(ExecContext, rlimit[RLIMIT_NOFILE]), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("LimitNOFILESoft", "t", bus_property_get_rlimit, offsetof(ExecContext, rlimit[RLIMIT_NOFILE]), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("LimitAS", "t", bus_property_get_rlimit, offsetof(ExecContext, rlimit[RLIMIT_AS]), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("LimitASSoft", "t", bus_property_get_rlimit, offsetof(ExecContext, rlimit[RLIMIT_AS]), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("LimitNPROC", "t", bus_property_get_rlimit, offsetof(ExecContext, rlimit[RLIMIT_NPROC]), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("LimitNPROCSoft", "t", bus_property_get_rlimit, offsetof(ExecContext, rlimit[RLIMIT_NPROC]), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("LimitMEMLOCK", "t", bus_property_get_rlimit, offsetof(ExecContext, rlimit[RLIMIT_MEMLOCK]), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("LimitMEMLOCKSoft", "t", bus_property_get_rlimit, offsetof(ExecContext, rlimit[RLIMIT_MEMLOCK]), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("LimitLOCKS", "t", bus_property_get_rlimit, offsetof(ExecContext, rlimit[RLIMIT_LOCKS]), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("LimitLOCKSSoft", "t", bus_property_get_rlimit, offsetof(ExecContext, rlimit[RLIMIT_LOCKS]), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("LimitSIGPENDING", "t", bus_property_get_rlimit, offsetof(ExecContext, rlimit[RLIMIT_SIGPENDING]), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("LimitSIGPENDINGSoft", "t", bus_property_get_rlimit, offsetof(ExecContext, rlimit[RLIMIT_SIGPENDING]), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("LimitMSGQUEUE", "t", bus_property_get_rlimit, offsetof(ExecContext, rlimit[RLIMIT_MSGQUEUE]), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("LimitMSGQUEUESoft", "t", bus_property_get_rlimit, offsetof(ExecContext, rlimit[RLIMIT_MSGQUEUE]), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("LimitNICE", "t", bus_property_get_rlimit, offsetof(ExecContext, rlimit[RLIMIT_NICE]), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("LimitNICESoft", "t", bus_property_get_rlimit, offsetof(ExecContext, rlimit[RLIMIT_NICE]), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("LimitRTPRIO", "t", bus_property_get_rlimit, offsetof(ExecContext, rlimit[RLIMIT_RTPRIO]), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("LimitRTPRIOSoft", "t", bus_property_get_rlimit, offsetof(ExecContext, rlimit[RLIMIT_RTPRIO]), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("LimitRTTIME", "t", bus_property_get_rlimit, offsetof(ExecContext, rlimit[RLIMIT_RTTIME]), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("LimitRTTIMESoft", "t", bus_property_get_rlimit, offsetof(ExecContext, rlimit[RLIMIT_RTTIME]), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("WorkingDirectory", "s", property_get_working_directory, 0, SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("RootDirectory", "s", NULL, offsetof(ExecContext, root_directory), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("OOMScoreAdjust", "i", property_get_oom_score_adjust, 0, SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("Nice", "i", property_get_nice, 0, SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("IOScheduling", "i", property_get_ioprio, 0, SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("CPUSchedulingPolicy", "i", property_get_cpu_sched_policy, 0, SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("CPUSchedulingPriority", "i", property_get_cpu_sched_priority, 0, SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("CPUAffinity", "ay", property_get_cpu_affinity, 0, SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("TimerSlackNSec", "t", property_get_timer_slack_nsec, 0, SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("CPUSchedulingResetOnFork", "b", bus_property_get_bool, offsetof(ExecContext, cpu_sched_reset_on_fork), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("NonBlocking", "b", bus_property_get_bool, offsetof(ExecContext, non_blocking), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("StandardInput", "s", property_get_exec_input, offsetof(ExecContext, std_input), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("StandardOutput", "s", bus_property_get_exec_output, offsetof(ExecContext, std_output), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("StandardError", "s", bus_property_get_exec_output, offsetof(ExecContext, std_error), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("TTYPath", "s", NULL, offsetof(ExecContext, tty_path), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("TTYReset", "b", bus_property_get_bool, offsetof(ExecContext, tty_reset), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("TTYVHangup", "b", bus_property_get_bool, offsetof(ExecContext, tty_vhangup), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("TTYVTDisallocate", "b", bus_property_get_bool, offsetof(ExecContext, tty_vt_disallocate), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("SyslogPriority", "i", bus_property_get_int, offsetof(ExecContext, syslog_priority), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("SyslogIdentifier", "s", NULL, offsetof(ExecContext, syslog_identifier), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("SyslogLevelPrefix", "b", bus_property_get_bool, offsetof(ExecContext, syslog_level_prefix), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("SyslogLevel", "i", property_get_syslog_level, 0, SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("SyslogFacility", "i", property_get_syslog_facility, 0, SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("Capabilities", "s", property_get_empty_string, 0, SD_BUS_VTABLE_PROPERTY_CONST|SD_BUS_VTABLE_HIDDEN), + SD_BUS_PROPERTY("SecureBits", "i", bus_property_get_int, offsetof(ExecContext, secure_bits), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("CapabilityBoundingSet", "t", property_get_capability_bounding_set, 0, SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("AmbientCapabilities", "t", property_get_ambient_capabilities, 0, SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("User", "s", NULL, offsetof(ExecContext, user), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("Group", "s", NULL, offsetof(ExecContext, group), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("SupplementaryGroups", "as", NULL, offsetof(ExecContext, supplementary_groups), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("PAMName", "s", NULL, offsetof(ExecContext, pam_name), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("ReadWriteDirectories", "as", NULL, offsetof(ExecContext, read_write_paths), SD_BUS_VTABLE_PROPERTY_CONST|SD_BUS_VTABLE_HIDDEN), + SD_BUS_PROPERTY("ReadOnlyDirectories", "as", NULL, offsetof(ExecContext, read_only_paths), SD_BUS_VTABLE_PROPERTY_CONST|SD_BUS_VTABLE_HIDDEN), + SD_BUS_PROPERTY("InaccessibleDirectories", "as", NULL, offsetof(ExecContext, inaccessible_paths), SD_BUS_VTABLE_PROPERTY_CONST|SD_BUS_VTABLE_HIDDEN), + SD_BUS_PROPERTY("ReadWritePaths", "as", NULL, offsetof(ExecContext, read_write_paths), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("ReadOnlyPaths", "as", NULL, offsetof(ExecContext, read_only_paths), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("InaccessiblePaths", "as", NULL, offsetof(ExecContext, inaccessible_paths), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("MountFlags", "t", bus_property_get_ulong, offsetof(ExecContext, mount_flags), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("PrivateTmp", "b", bus_property_get_bool, offsetof(ExecContext, private_tmp), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("PrivateNetwork", "b", bus_property_get_bool, offsetof(ExecContext, private_network), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("PrivateDevices", "b", bus_property_get_bool, offsetof(ExecContext, private_devices), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("ProtectHome", "s", bus_property_get_protect_home, offsetof(ExecContext, protect_home), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("ProtectSystem", "s", bus_property_get_protect_system, offsetof(ExecContext, protect_system), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("SameProcessGroup", "b", bus_property_get_bool, offsetof(ExecContext, same_pgrp), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("UtmpIdentifier", "s", NULL, offsetof(ExecContext, utmp_id), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("UtmpMode", "s", property_get_exec_utmp_mode, offsetof(ExecContext, utmp_mode), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("SELinuxContext", "(bs)", property_get_selinux_context, 0, SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("AppArmorProfile", "(bs)", property_get_apparmor_profile, 0, SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("SmackProcessLabel", "(bs)", property_get_smack_process_label, 0, SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("IgnoreSIGPIPE", "b", bus_property_get_bool, offsetof(ExecContext, ignore_sigpipe), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("NoNewPrivileges", "b", bus_property_get_bool, offsetof(ExecContext, no_new_privileges), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("SystemCallFilter", "(bas)", property_get_syscall_filter, 0, SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("SystemCallArchitectures", "as", property_get_syscall_archs, 0, SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("SystemCallErrorNumber", "i", property_get_syscall_errno, 0, SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("Personality", "s", property_get_personality, 0, SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("RestrictAddressFamilies", "(bas)", property_get_address_families, 0, SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("RuntimeDirectoryMode", "u", bus_property_get_mode, offsetof(ExecContext, runtime_directory_mode), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("RuntimeDirectory", "as", NULL, offsetof(ExecContext, runtime_directory), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("MemoryDenyWriteExecute", "b", bus_property_get_bool, offsetof(ExecContext, memory_deny_write_execute), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("RestrictRealtime", "b", bus_property_get_bool, offsetof(ExecContext, restrict_realtime), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_VTABLE_END +}; + +static int append_exec_command(sd_bus_message *reply, ExecCommand *c) { + int r; + + assert(reply); + assert(c); + + if (!c->path) + return 0; + + r = sd_bus_message_open_container(reply, 'r', "sasbttttuii"); + if (r < 0) + return r; + + r = sd_bus_message_append(reply, "s", c->path); + if (r < 0) + return r; + + r = sd_bus_message_append_strv(reply, c->argv); + if (r < 0) + return r; + + r = sd_bus_message_append(reply, "bttttuii", + c->ignore, + c->exec_status.start_timestamp.realtime, + c->exec_status.start_timestamp.monotonic, + c->exec_status.exit_timestamp.realtime, + c->exec_status.exit_timestamp.monotonic, + (uint32_t) c->exec_status.pid, + (int32_t) c->exec_status.code, + (int32_t) c->exec_status.status); + if (r < 0) + return r; + + return sd_bus_message_close_container(reply); +} + +int bus_property_get_exec_command( + sd_bus *bus, + const char *path, + const char *interface, + const char *property, + sd_bus_message *reply, + void *userdata, + sd_bus_error *ret_error) { + + ExecCommand *c = (ExecCommand*) userdata; + int r; + + assert(bus); + assert(reply); + + r = sd_bus_message_open_container(reply, 'a', "(sasbttttuii)"); + if (r < 0) + return r; + + r = append_exec_command(reply, c); + if (r < 0) + return r; + + return sd_bus_message_close_container(reply); +} + +int bus_property_get_exec_command_list( + sd_bus *bus, + const char *path, + const char *interface, + const char *property, + sd_bus_message *reply, + void *userdata, + sd_bus_error *ret_error) { + + ExecCommand *c = *(ExecCommand**) userdata; + int r; + + assert(bus); + assert(reply); + + r = sd_bus_message_open_container(reply, 'a', "(sasbttttuii)"); + if (r < 0) + return r; + + LIST_FOREACH(command, c, c) { + r = append_exec_command(reply, c); + if (r < 0) + return r; + } + + return sd_bus_message_close_container(reply); +} + +int bus_exec_context_set_transient_property( + Unit *u, + ExecContext *c, + const char *name, + sd_bus_message *message, + UnitSetPropertiesMode mode, + sd_bus_error *error) { + + const char *soft = NULL; + int r, ri; + + assert(u); + assert(c); + assert(name); + assert(message); + + if (streq(name, "User")) { + const char *uu; + + r = sd_bus_message_read(message, "s", &uu); + if (r < 0) + return r; + + if (mode != UNIT_CHECK) { + + if (isempty(uu)) + c->user = mfree(c->user); + else if (free_and_strdup(&c->user, uu) < 0) + return -ENOMEM; + + unit_write_drop_in_private_format(u, mode, name, "User=%s", uu); + } + + return 1; + + } else if (streq(name, "Group")) { + const char *gg; + + r = sd_bus_message_read(message, "s", &gg); + if (r < 0) + return r; + + if (mode != UNIT_CHECK) { + + if (isempty(gg)) + c->group = mfree(c->group); + else if (free_and_strdup(&c->group, gg) < 0) + return -ENOMEM; + + unit_write_drop_in_private_format(u, mode, name, "Group=%s", gg); + } + + return 1; + } else if (streq(name, "SyslogIdentifier")) { + const char *id; + + r = sd_bus_message_read(message, "s", &id); + if (r < 0) + return r; + + if (mode != UNIT_CHECK) { + + if (isempty(id)) + c->syslog_identifier = mfree(c->syslog_identifier); + else if (free_and_strdup(&c->syslog_identifier, id) < 0) + return -ENOMEM; + + unit_write_drop_in_private_format(u, mode, name, "SyslogIdentifier=%s", id); + } + + return 1; + } else if (streq(name, "SyslogLevel")) { + int level; + + r = sd_bus_message_read(message, "i", &level); + if (r < 0) + return r; + + if (!log_level_is_valid(level)) + return sd_bus_error_setf(error, SD_BUS_ERROR_INVALID_ARGS, "Log level value out of range"); + + if (mode != UNIT_CHECK) { + c->syslog_priority = (c->syslog_priority & LOG_FACMASK) | level; + unit_write_drop_in_private_format(u, mode, name, "SyslogLevel=%i", level); + } + + return 1; + } else if (streq(name, "SyslogFacility")) { + int facility; + + r = sd_bus_message_read(message, "i", &facility); + if (r < 0) + return r; + + if (!log_facility_unshifted_is_valid(facility)) + return sd_bus_error_setf(error, SD_BUS_ERROR_INVALID_ARGS, "Log facility value out of range"); + + if (mode != UNIT_CHECK) { + c->syslog_priority = (facility << 3) | LOG_PRI(c->syslog_priority); + unit_write_drop_in_private_format(u, mode, name, "SyslogFacility=%i", facility); + } + + return 1; + } else if (streq(name, "Nice")) { + int n; + + r = sd_bus_message_read(message, "i", &n); + if (r < 0) + return r; + + if (n < PRIO_MIN || n >= PRIO_MAX) + return sd_bus_error_setf(error, SD_BUS_ERROR_INVALID_ARGS, "Nice value out of range"); + + if (mode != UNIT_CHECK) { + c->nice = n; + unit_write_drop_in_private_format(u, mode, name, "Nice=%i", n); + } + + return 1; + + } else if (STR_IN_SET(name, "TTYPath", "RootDirectory")) { + const char *s; + + r = sd_bus_message_read(message, "s", &s); + if (r < 0) + return r; + + if (!path_is_absolute(s)) + return sd_bus_error_setf(error, SD_BUS_ERROR_INVALID_ARGS, "%s takes an absolute path", name); + + if (mode != UNIT_CHECK) { + if (streq(name, "TTYPath")) + r = free_and_strdup(&c->tty_path, s); + else { + assert(streq(name, "RootDirectory")); + r = free_and_strdup(&c->root_directory, s); + } + if (r < 0) + return r; + + unit_write_drop_in_private_format(u, mode, name, "%s=%s", name, s); + } + + return 1; + + } else if (streq(name, "WorkingDirectory")) { + const char *s; + bool missing_ok; + + r = sd_bus_message_read(message, "s", &s); + if (r < 0) + return r; + + if (s[0] == '-') { + missing_ok = true; + s++; + } else + missing_ok = false; + + if (!streq(s, "~") && !path_is_absolute(s)) + return sd_bus_error_setf(error, SD_BUS_ERROR_INVALID_ARGS, "WorkingDirectory= expects an absolute path or '~'"); + + if (mode != UNIT_CHECK) { + if (streq(s, "~")) { + c->working_directory = mfree(c->working_directory); + c->working_directory_home = true; + } else { + r = free_and_strdup(&c->working_directory, s); + if (r < 0) + return r; + + c->working_directory_home = false; + } + + c->working_directory_missing_ok = missing_ok; + unit_write_drop_in_private_format(u, mode, name, "WorkingDirectory=%s%s", missing_ok ? "-" : "", s); + } + + return 1; + + } else if (streq(name, "StandardInput")) { + const char *s; + ExecInput p; + + r = sd_bus_message_read(message, "s", &s); + if (r < 0) + return r; + + p = exec_input_from_string(s); + if (p < 0) + return sd_bus_error_setf(error, SD_BUS_ERROR_INVALID_ARGS, "Invalid standard input name"); + + if (mode != UNIT_CHECK) { + c->std_input = p; + + unit_write_drop_in_private_format(u, mode, name, "StandardInput=%s", exec_input_to_string(p)); + } + + return 1; + + + } else if (streq(name, "StandardOutput")) { + const char *s; + ExecOutput p; + + r = sd_bus_message_read(message, "s", &s); + if (r < 0) + return r; + + p = exec_output_from_string(s); + if (p < 0) + return sd_bus_error_setf(error, SD_BUS_ERROR_INVALID_ARGS, "Invalid standard output name"); + + if (mode != UNIT_CHECK) { + c->std_output = p; + + unit_write_drop_in_private_format(u, mode, name, "StandardOutput=%s", exec_output_to_string(p)); + } + + return 1; + + } else if (streq(name, "StandardError")) { + const char *s; + ExecOutput p; + + r = sd_bus_message_read(message, "s", &s); + if (r < 0) + return r; + + p = exec_output_from_string(s); + if (p < 0) + return sd_bus_error_setf(error, SD_BUS_ERROR_INVALID_ARGS, "Invalid standard error name"); + + if (mode != UNIT_CHECK) { + c->std_error = p; + + unit_write_drop_in_private_format(u, mode, name, "StandardError=%s", exec_output_to_string(p)); + } + + return 1; + + } else if (STR_IN_SET(name, + "IgnoreSIGPIPE", "TTYVHangup", "TTYReset", + "PrivateTmp", "PrivateDevices", "PrivateNetwork", + "NoNewPrivileges", "SyslogLevelPrefix", "MemoryDenyWriteExecute", "RestrictRealtime")) { + int b; + + r = sd_bus_message_read(message, "b", &b); + if (r < 0) + return r; + + if (mode != UNIT_CHECK) { + if (streq(name, "IgnoreSIGPIPE")) + c->ignore_sigpipe = b; + else if (streq(name, "TTYVHangup")) + c->tty_vhangup = b; + else if (streq(name, "TTYReset")) + c->tty_reset = b; + else if (streq(name, "PrivateTmp")) + c->private_tmp = b; + else if (streq(name, "PrivateDevices")) + c->private_devices = b; + else if (streq(name, "PrivateNetwork")) + c->private_network = b; + else if (streq(name, "NoNewPrivileges")) + c->no_new_privileges = b; + else if (streq(name, "SyslogLevelPrefix")) + c->syslog_level_prefix = b; + else if (streq(name, "MemoryDenyWriteExecute")) + c->memory_deny_write_execute = b; + else if (streq(name, "RestrictRealtime")) + c->restrict_realtime = b; + + unit_write_drop_in_private_format(u, mode, name, "%s=%s", name, yes_no(b)); + } + + return 1; + + } else if (streq(name, "UtmpIdentifier")) { + const char *id; + + r = sd_bus_message_read(message, "s", &id); + if (r < 0) + return r; + + if (mode != UNIT_CHECK) { + if (isempty(id)) + c->utmp_id = mfree(c->utmp_id); + else if (free_and_strdup(&c->utmp_id, id) < 0) + return -ENOMEM; + + unit_write_drop_in_private_format(u, mode, name, "UtmpIdentifier=%s", strempty(id)); + } + + return 1; + + } else if (streq(name, "UtmpMode")) { + const char *s; + ExecUtmpMode m; + + r = sd_bus_message_read(message, "s", &s); + if (r < 0) + return r; + + m = exec_utmp_mode_from_string(s); + if (m < 0) + return sd_bus_error_setf(error, SD_BUS_ERROR_INVALID_ARGS, "Invalid utmp mode"); + + if (mode != UNIT_CHECK) { + c->utmp_mode = m; + + unit_write_drop_in_private_format(u, mode, name, "UtmpMode=%s", exec_utmp_mode_to_string(m)); + } + + return 1; + + } else if (streq(name, "PAMName")) { + const char *n; + + r = sd_bus_message_read(message, "s", &n); + if (r < 0) + return r; + + if (mode != UNIT_CHECK) { + if (isempty(n)) + c->pam_name = mfree(c->pam_name); + else if (free_and_strdup(&c->pam_name, n) < 0) + return -ENOMEM; + + unit_write_drop_in_private_format(u, mode, name, "PAMName=%s", strempty(n)); + } + + return 1; + + } else if (streq(name, "Environment")) { + + _cleanup_strv_free_ char **l = NULL; + + r = sd_bus_message_read_strv(message, &l); + if (r < 0) + return r; + + if (!strv_env_is_valid(l)) + return sd_bus_error_setf(error, SD_BUS_ERROR_INVALID_ARGS, "Invalid environment block."); + + if (mode != UNIT_CHECK) { + _cleanup_free_ char *joined = NULL; + char **e; + + if (strv_length(l) == 0) { + c->environment = strv_free(c->environment); + unit_write_drop_in_private_format(u, mode, name, "Environment="); + } else { + e = strv_env_merge(2, c->environment, l); + if (!e) + return -ENOMEM; + + strv_free(c->environment); + c->environment = e; + + joined = strv_join_quoted(c->environment); + if (!joined) + return -ENOMEM; + + unit_write_drop_in_private_format(u, mode, name, "Environment=%s", joined); + } + } + + return 1; + + } else if (streq(name, "TimerSlackNSec")) { + + nsec_t n; + + r = sd_bus_message_read(message, "t", &n); + if (r < 0) + return r; + + if (mode != UNIT_CHECK) { + c->timer_slack_nsec = n; + unit_write_drop_in_private_format(u, mode, name, "TimerSlackNSec=" NSEC_FMT, n); + } + + return 1; + + } else if (streq(name, "OOMScoreAdjust")) { + int oa; + + r = sd_bus_message_read(message, "i", &oa); + if (r < 0) + return r; + + if (!oom_score_adjust_is_valid(oa)) + return sd_bus_error_setf(error, SD_BUS_ERROR_INVALID_ARGS, "OOM score adjust value out of range"); + + if (mode != UNIT_CHECK) { + c->oom_score_adjust = oa; + c->oom_score_adjust_set = true; + unit_write_drop_in_private_format(u, mode, name, "OOMScoreAdjust=%i", oa); + } + + return 1; + + } else if (streq(name, "EnvironmentFiles")) { + + _cleanup_free_ char *joined = NULL; + _cleanup_fclose_ FILE *f = NULL; + _cleanup_free_ char **l = NULL; + size_t size = 0; + char **i; + + r = sd_bus_message_enter_container(message, 'a', "(sb)"); + if (r < 0) + return r; + + f = open_memstream(&joined, &size); + if (!f) + return -ENOMEM; + + STRV_FOREACH(i, c->environment_files) + fprintf(f, "EnvironmentFile=%s", *i); + + while ((r = sd_bus_message_enter_container(message, 'r', "sb")) > 0) { + const char *path; + int b; + + r = sd_bus_message_read(message, "sb", &path, &b); + if (r < 0) + return r; + + r = sd_bus_message_exit_container(message); + if (r < 0) + return r; + + if (!isempty(path) && !path_is_absolute(path)) + return sd_bus_error_set_errnof(error, EINVAL, "Path %s is not absolute.", path); + + if (mode != UNIT_CHECK) { + char *buf = NULL; + + buf = strjoin(b ? "-" : "", path, NULL); + if (!buf) + return -ENOMEM; + + fprintf(f, "EnvironmentFile=%s", buf); + + r = strv_consume(&l, buf); + if (r < 0) + return r; + } + } + if (r < 0) + return r; + + r = sd_bus_message_exit_container(message); + if (r < 0) + return r; + + r = fflush_and_check(f); + if (r < 0) + return r; + + if (mode != UNIT_CHECK) { + if (strv_isempty(l)) { + c->environment_files = strv_free(c->environment_files); + unit_write_drop_in_private(u, mode, name, "EnvironmentFile="); + } else { + r = strv_extend_strv(&c->environment_files, l, true); + if (r < 0) + return r; + + unit_write_drop_in_private(u, mode, name, joined); + } + } + + return 1; + + } else if (streq(name, "PassEnvironment")) { + + _cleanup_strv_free_ char **l = NULL; + + r = sd_bus_message_read_strv(message, &l); + if (r < 0) + return r; + + if (!strv_env_name_is_valid(l)) + return sd_bus_error_setf(error, SD_BUS_ERROR_INVALID_ARGS, "Invalid PassEnvironment block."); + + if (mode != UNIT_CHECK) { + if (strv_isempty(l)) { + c->pass_environment = strv_free(c->pass_environment); + unit_write_drop_in_private_format(u, mode, name, "PassEnvironment="); + } else { + _cleanup_free_ char *joined = NULL; + + r = strv_extend_strv(&c->pass_environment, l, true); + if (r < 0) + return r; + + joined = strv_join_quoted(c->pass_environment); + if (!joined) + return -ENOMEM; + + unit_write_drop_in_private_format(u, mode, name, "PassEnvironment=%s", joined); + } + } + + return 1; + + } else if (STR_IN_SET(name, "ReadWriteDirectories", "ReadOnlyDirectories", "InaccessibleDirectories", + "ReadWritePaths", "ReadOnlyPaths", "InaccessiblePaths")) { + _cleanup_strv_free_ char **l = NULL; + char ***dirs; + char **p; + + r = sd_bus_message_read_strv(message, &l); + if (r < 0) + return r; + + STRV_FOREACH(p, l) { + int offset; + if (!utf8_is_valid(*p)) + return sd_bus_error_setf(error, SD_BUS_ERROR_INVALID_ARGS, "Invalid %s", name); + + offset = **p == '-'; + if (!path_is_absolute(*p + offset)) + return sd_bus_error_setf(error, SD_BUS_ERROR_INVALID_ARGS, "Invalid %s", name); + } + + if (mode != UNIT_CHECK) { + _cleanup_free_ char *joined = NULL; + + if (STR_IN_SET(name, "ReadWriteDirectories", "ReadWritePaths")) + dirs = &c->read_write_paths; + else if (STR_IN_SET(name, "ReadOnlyDirectories", "ReadOnlyPaths")) + dirs = &c->read_only_paths; + else /* "InaccessiblePaths" */ + dirs = &c->inaccessible_paths; + + if (strv_length(l) == 0) { + *dirs = strv_free(*dirs); + unit_write_drop_in_private_format(u, mode, name, "%s=", name); + } else { + r = strv_extend_strv(dirs, l, true); + + if (r < 0) + return -ENOMEM; + + joined = strv_join_quoted(*dirs); + if (!joined) + return -ENOMEM; + + unit_write_drop_in_private_format(u, mode, name, "%s=%s", name, joined); + } + + } + + return 1; + + } else if (streq(name, "ProtectSystem")) { + const char *s; + ProtectSystem ps; + + r = sd_bus_message_read(message, "s", &s); + if (r < 0) + return r; + + r = parse_boolean(s); + if (r > 0) + ps = PROTECT_SYSTEM_YES; + else if (r == 0) + ps = PROTECT_SYSTEM_NO; + else { + ps = protect_system_from_string(s); + if (ps < 0) + return sd_bus_error_setf(error, SD_BUS_ERROR_INVALID_ARGS, "Failed to parse protect system value"); + } + + if (mode != UNIT_CHECK) { + c->protect_system = ps; + unit_write_drop_in_private_format(u, mode, name, "%s=%s", name, s); + } + + return 1; + + } else if (streq(name, "ProtectHome")) { + const char *s; + ProtectHome ph; + + r = sd_bus_message_read(message, "s", &s); + if (r < 0) + return r; + + r = parse_boolean(s); + if (r > 0) + ph = PROTECT_HOME_YES; + else if (r == 0) + ph = PROTECT_HOME_NO; + else { + ph = protect_home_from_string(s); + if (ph < 0) + return sd_bus_error_setf(error, SD_BUS_ERROR_INVALID_ARGS, "Failed to parse protect home value"); + } + + if (mode != UNIT_CHECK) { + c->protect_home = ph; + unit_write_drop_in_private_format(u, mode, name, "%s=%s", name, s); + } + + return 1; + + } else if (streq(name, "RuntimeDirectory")) { + _cleanup_strv_free_ char **l = NULL; + char **p; + + r = sd_bus_message_read_strv(message, &l); + if (r < 0) + return r; + + STRV_FOREACH(p, l) { + if (!filename_is_valid(*p)) + return sd_bus_error_setf(error, SD_BUS_ERROR_INVALID_ARGS, "Runtime directory is not valid %s", *p); + } + + if (mode != UNIT_CHECK) { + _cleanup_free_ char *joined = NULL; + + if (strv_isempty(l)) { + c->runtime_directory = strv_free(c->runtime_directory); + unit_write_drop_in_private_format(u, mode, name, "%s=", name); + } else { + r = strv_extend_strv(&c->runtime_directory, l, true); + + if (r < 0) + return -ENOMEM; + + joined = strv_join_quoted(c->runtime_directory); + if (!joined) + return -ENOMEM; + + unit_write_drop_in_private_format(u, mode, name, "%s=%s", name, joined); + } + } + + return 1; + + } else if (streq(name, "SELinuxContext")) { + const char *s; + + r = sd_bus_message_read(message, "s", &s); + if (r < 0) + return r; + + if (mode != UNIT_CHECK) { + if (isempty(s)) + c->selinux_context = mfree(c->selinux_context); + else if (free_and_strdup(&c->selinux_context, s) < 0) + return -ENOMEM; + + unit_write_drop_in_private_format(u, mode, name, "%s=%s", name, strempty(s)); + } + + return 1; + + } + + ri = rlimit_from_string(name); + if (ri < 0) { + soft = endswith(name, "Soft"); + if (soft) { + const char *n; + + n = strndupa(name, soft - name); + ri = rlimit_from_string(n); + if (ri >= 0) + name = n; + + } + } + + if (ri >= 0) { + uint64_t rl; + rlim_t x; + + r = sd_bus_message_read(message, "t", &rl); + if (r < 0) + return r; + + if (rl == (uint64_t) -1) + x = RLIM_INFINITY; + else { + x = (rlim_t) rl; + + if ((uint64_t) x != rl) + return -ERANGE; + } + + if (mode != UNIT_CHECK) { + _cleanup_free_ char *f = NULL; + struct rlimit nl; + + if (c->rlimit[ri]) { + nl = *c->rlimit[ri]; + + if (soft) + nl.rlim_cur = x; + else + nl.rlim_max = x; + } else + /* When the resource limit is not initialized yet, then assign the value to both fields */ + nl = (struct rlimit) { + .rlim_cur = x, + .rlim_max = x, + }; + + r = rlimit_format(&nl, &f); + if (r < 0) + return r; + + if (c->rlimit[ri]) + *c->rlimit[ri] = nl; + else { + c->rlimit[ri] = newdup(struct rlimit, &nl, 1); + if (!c->rlimit[ri]) + return -ENOMEM; + } + + unit_write_drop_in_private_format(u, mode, name, "%s=%s", name, f); + } + + return 1; + } + + return 0; +} diff --git a/src/grp-system/libcore/src/dbus-execute.h b/src/grp-system/libcore/src/dbus-execute.h new file mode 100644 index 0000000000..25137d9516 --- /dev/null +++ b/src/grp-system/libcore/src/dbus-execute.h @@ -0,0 +1,45 @@ +#pragma once + +/*** + This file is part of systemd. + + Copyright 2010 Lennart Poettering + + systemd is free software; you can redistribute it and/or modify it + under the terms of the GNU Lesser General Public License as published by + the Free Software Foundation; either version 2.1 of the License, or + (at your option) any later version. + + systemd is distributed in the hope that it will be useful, but + WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + Lesser General Public License for more details. + + You should have received a copy of the GNU Lesser General Public License + along with systemd; If not, see . +***/ + +#include + +#include "core/execute.h" + +#define BUS_EXEC_STATUS_VTABLE(prefix, offset, flags) \ + BUS_PROPERTY_DUAL_TIMESTAMP(prefix "StartTimestamp", (offset) + offsetof(ExecStatus, start_timestamp), flags), \ + BUS_PROPERTY_DUAL_TIMESTAMP(prefix "ExitTimestamp", (offset) + offsetof(ExecStatus, exit_timestamp), flags), \ + SD_BUS_PROPERTY(prefix "PID", "u", bus_property_get_pid, (offset) + offsetof(ExecStatus, pid), flags), \ + SD_BUS_PROPERTY(prefix "Code", "i", bus_property_get_int, (offset) + offsetof(ExecStatus, code), flags), \ + SD_BUS_PROPERTY(prefix "Status", "i", bus_property_get_int, (offset) + offsetof(ExecStatus, status), flags) + +#define BUS_EXEC_COMMAND_VTABLE(name, offset, flags) \ + SD_BUS_PROPERTY(name, "a(sasbttttuii)", bus_property_get_exec_command, offset, flags) + +#define BUS_EXEC_COMMAND_LIST_VTABLE(name, offset, flags) \ + SD_BUS_PROPERTY(name, "a(sasbttttuii)", bus_property_get_exec_command_list, offset, flags) + +extern const sd_bus_vtable bus_exec_vtable[]; + +int bus_property_get_exec_output(sd_bus *bus, const char *path, const char *interface, const char *property, sd_bus_message *reply, void *userdata, sd_bus_error *ret_error); +int bus_property_get_exec_command(sd_bus *bus, const char *path, const char *interface, const char *property, sd_bus_message *reply, void *userdata, sd_bus_error *ret_error); +int bus_property_get_exec_command_list(sd_bus *bus, const char *path, const char *interface, const char *property, sd_bus_message *reply, void *userdata, sd_bus_error *ret_error); + +int bus_exec_context_set_transient_property(Unit *u, ExecContext *c, const char *name, sd_bus_message *message, UnitSetPropertiesMode mode, sd_bus_error *error); diff --git a/src/grp-system/libcore/src/dbus-job.c b/src/grp-system/libcore/src/dbus-job.c new file mode 100644 index 0000000000..0eddfe60d7 --- /dev/null +++ b/src/grp-system/libcore/src/dbus-job.c @@ -0,0 +1,194 @@ +/*** + This file is part of systemd. + + Copyright 2010 Lennart Poettering + + systemd is free software; you can redistribute it and/or modify it + under the terms of the GNU Lesser General Public License as published by + the Free Software Foundation; either version 2.1 of the License, or + (at your option) any later version. + + systemd is distributed in the hope that it will be useful, but + WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + Lesser General Public License for more details. + + You should have received a copy of the GNU Lesser General Public License + along with systemd; If not, see . +***/ + +#include + +#include "core/job.h" +#include "systemd-basic/alloc-util.h" +#include "systemd-basic/log.h" +#include "systemd-basic/string-util.h" + +#include "dbus-job.h" +#include "dbus.h" +#include "selinux-access.h" + +static BUS_DEFINE_PROPERTY_GET_ENUM(property_get_type, job_type, JobType); +static BUS_DEFINE_PROPERTY_GET_ENUM(property_get_state, job_state, JobState); + +static int property_get_unit( + sd_bus *bus, + const char *path, + const char *interface, + const char *property, + sd_bus_message *reply, + void *userdata, + sd_bus_error *error) { + + _cleanup_free_ char *p = NULL; + Job *j = userdata; + + assert(bus); + assert(reply); + assert(j); + + p = unit_dbus_path(j->unit); + if (!p) + return -ENOMEM; + + return sd_bus_message_append(reply, "(so)", j->unit->id, p); +} + +int bus_job_method_cancel(sd_bus_message *message, void *userdata, sd_bus_error *error) { + Job *j = userdata; + int r; + + assert(message); + assert(j); + + r = mac_selinux_unit_access_check(j->unit, message, "stop", error); + if (r < 0) + return r; + + /* Access is granted to the job owner */ + if (!sd_bus_track_contains(j->clients, sd_bus_message_get_sender(message))) { + + /* And for everybody else consult PolicyKit */ + r = bus_verify_manage_units_async(j->unit->manager, message, error); + if (r < 0) + return r; + if (r == 0) + return 1; /* No authorization for now, but the async polkit stuff will call us again when it has it */ + } + + job_finish_and_invalidate(j, JOB_CANCELED, true, false); + + return sd_bus_reply_method_return(message, NULL); +} + +const sd_bus_vtable bus_job_vtable[] = { + SD_BUS_VTABLE_START(0), + SD_BUS_METHOD("Cancel", NULL, NULL, bus_job_method_cancel, SD_BUS_VTABLE_UNPRIVILEGED), + SD_BUS_PROPERTY("Id", "u", NULL, offsetof(Job, id), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("Unit", "(so)", property_get_unit, 0, SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("JobType", "s", property_get_type, offsetof(Job, type), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("State", "s", property_get_state, offsetof(Job, state), SD_BUS_VTABLE_PROPERTY_EMITS_CHANGE), + SD_BUS_VTABLE_END +}; + +static int send_new_signal(sd_bus *bus, void *userdata) { + _cleanup_(sd_bus_message_unrefp) sd_bus_message *m = NULL; + _cleanup_free_ char *p = NULL; + Job *j = userdata; + int r; + + assert(bus); + assert(j); + + p = job_dbus_path(j); + if (!p) + return -ENOMEM; + + r = sd_bus_message_new_signal( + bus, + &m, + "/org/freedesktop/systemd1", + "org.freedesktop.systemd1.Manager", + "JobNew"); + if (r < 0) + return r; + + r = sd_bus_message_append(m, "uos", j->id, p, j->unit->id); + if (r < 0) + return r; + + return sd_bus_send(bus, m, NULL); +} + +static int send_changed_signal(sd_bus *bus, void *userdata) { + _cleanup_free_ char *p = NULL; + Job *j = userdata; + + assert(bus); + assert(j); + + p = job_dbus_path(j); + if (!p) + return -ENOMEM; + + return sd_bus_emit_properties_changed(bus, p, "org.freedesktop.systemd1.Job", "State", NULL); +} + +void bus_job_send_change_signal(Job *j) { + int r; + + assert(j); + + if (j->in_dbus_queue) { + LIST_REMOVE(dbus_queue, j->manager->dbus_job_queue, j); + j->in_dbus_queue = false; + } + + r = bus_foreach_bus(j->manager, j->clients, j->sent_dbus_new_signal ? send_changed_signal : send_new_signal, j); + if (r < 0) + log_debug_errno(r, "Failed to send job change signal for %u: %m", j->id); + + j->sent_dbus_new_signal = true; +} + +static int send_removed_signal(sd_bus *bus, void *userdata) { + _cleanup_(sd_bus_message_unrefp) sd_bus_message *m = NULL; + _cleanup_free_ char *p = NULL; + Job *j = userdata; + int r; + + assert(bus); + assert(j); + + p = job_dbus_path(j); + if (!p) + return -ENOMEM; + + r = sd_bus_message_new_signal( + bus, + &m, + "/org/freedesktop/systemd1", + "org.freedesktop.systemd1.Manager", + "JobRemoved"); + if (r < 0) + return r; + + r = sd_bus_message_append(m, "uoss", j->id, p, j->unit->id, job_result_to_string(j->result)); + if (r < 0) + return r; + + return sd_bus_send(bus, m, NULL); +} + +void bus_job_send_removed_signal(Job *j) { + int r; + + assert(j); + + if (!j->sent_dbus_new_signal) + bus_job_send_change_signal(j); + + r = bus_foreach_bus(j->manager, j->clients, send_removed_signal, j); + if (r < 0) + log_debug_errno(r, "Failed to send job remove signal for %u: %m", j->id); +} diff --git a/src/grp-system/libcore/src/dbus-job.h b/src/grp-system/libcore/src/dbus-job.h new file mode 100644 index 0000000000..4e2d4f086d --- /dev/null +++ b/src/grp-system/libcore/src/dbus-job.h @@ -0,0 +1,31 @@ +#pragma once + +/*** + This file is part of systemd. + + Copyright 2010 Lennart Poettering + + systemd is free software; you can redistribute it and/or modify it + under the terms of the GNU Lesser General Public License as published by + the Free Software Foundation; either version 2.1 of the License, or + (at your option) any later version. + + systemd is distributed in the hope that it will be useful, but + WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + Lesser General Public License for more details. + + You should have received a copy of the GNU Lesser General Public License + along with systemd; If not, see . +***/ + +#include + +#include "core/job.h" + +extern const sd_bus_vtable bus_job_vtable[]; + +int bus_job_method_cancel(sd_bus_message *message, void *job, sd_bus_error *error); + +void bus_job_send_change_signal(Job *j); +void bus_job_send_removed_signal(Job *j); diff --git a/src/grp-system/libcore/src/dbus-kill.c b/src/grp-system/libcore/src/dbus-kill.c new file mode 100644 index 0000000000..71337306a6 --- /dev/null +++ b/src/grp-system/libcore/src/dbus-kill.c @@ -0,0 +1,123 @@ +/*** + This file is part of systemd. + + Copyright 2012 Lennart Poettering + + systemd is free software; you can redistribute it and/or modify it + under the terms of the GNU Lesser General Public License as published by + the Free Software Foundation; either version 2.1 of the License, or + (at your option) any later version. + + systemd is distributed in the hope that it will be useful, but + WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + Lesser General Public License for more details. + + You should have received a copy of the GNU Lesser General Public License + along with systemd; If not, see . +***/ + +#include "core/kill.h" +#include "sd-bus/bus-util.h" +#include "systemd-basic/signal-util.h" + +#include "dbus-kill.h" + +static BUS_DEFINE_PROPERTY_GET_ENUM(property_get_kill_mode, kill_mode, KillMode); + +const sd_bus_vtable bus_kill_vtable[] = { + SD_BUS_VTABLE_START(0), + SD_BUS_PROPERTY("KillMode", "s", property_get_kill_mode, offsetof(KillContext, kill_mode), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("KillSignal", "i", bus_property_get_int, offsetof(KillContext, kill_signal), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("SendSIGKILL", "b", bus_property_get_bool, offsetof(KillContext, send_sigkill), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("SendSIGHUP", "b", bus_property_get_bool, offsetof(KillContext, send_sighup), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_VTABLE_END +}; + +int bus_kill_context_set_transient_property( + Unit *u, + KillContext *c, + const char *name, + sd_bus_message *message, + UnitSetPropertiesMode mode, + sd_bus_error *error) { + + int r; + + assert(u); + assert(c); + assert(name); + assert(message); + + if (streq(name, "KillMode")) { + const char *m; + KillMode k; + + r = sd_bus_message_read(message, "s", &m); + if (r < 0) + return r; + + k = kill_mode_from_string(m); + if (k < 0) + return sd_bus_error_setf(error, SD_BUS_ERROR_INVALID_ARGS, "Kill mode '%s' not known.", m); + + if (mode != UNIT_CHECK) { + c->kill_mode = k; + + unit_write_drop_in_private_format(u, mode, name, "KillMode=%s", kill_mode_to_string(k)); + } + + return 1; + + } else if (streq(name, "KillSignal")) { + int sig; + + r = sd_bus_message_read(message, "i", &sig); + if (r < 0) + return r; + + if (!SIGNAL_VALID(sig)) + return sd_bus_error_setf(error, SD_BUS_ERROR_INVALID_ARGS, "Signal %i out of range", sig); + + if (mode != UNIT_CHECK) { + c->kill_signal = sig; + + unit_write_drop_in_private_format(u, mode, name, "KillSignal=%s", signal_to_string(sig)); + } + + return 1; + + } else if (streq(name, "SendSIGHUP")) { + int b; + + r = sd_bus_message_read(message, "b", &b); + if (r < 0) + return r; + + if (mode != UNIT_CHECK) { + c->send_sighup = b; + + unit_write_drop_in_private_format(u, mode, name, "SendSIGHUP=%s", yes_no(b)); + } + + return 1; + + } else if (streq(name, "SendSIGKILL")) { + int b; + + r = sd_bus_message_read(message, "b", &b); + if (r < 0) + return r; + + if (mode != UNIT_CHECK) { + c->send_sigkill = b; + + unit_write_drop_in_private_format(u, mode, name, "SendSIGKILL=%s", yes_no(b)); + } + + return 1; + + } + + return 0; +} diff --git a/src/grp-system/libcore/src/dbus-kill.h b/src/grp-system/libcore/src/dbus-kill.h new file mode 100644 index 0000000000..3f908c40d9 --- /dev/null +++ b/src/grp-system/libcore/src/dbus-kill.h @@ -0,0 +1,29 @@ +#pragma once + +/*** + This file is part of systemd. + + Copyright 2012 Lennart Poettering + + systemd is free software; you can redistribute it and/or modify it + under the terms of the GNU Lesser General Public License as published by + the Free Software Foundation; either version 2.1 of the License, or + (at your option) any later version. + + systemd is distributed in the hope that it will be useful, but + WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + Lesser General Public License for more details. + + You should have received a copy of the GNU Lesser General Public License + along with systemd; If not, see . +***/ + +#include + +#include "core/kill.h" +#include "core/unit.h" + +extern const sd_bus_vtable bus_kill_vtable[]; + +int bus_kill_context_set_transient_property(Unit *u, KillContext *c, const char *name, sd_bus_message *message, UnitSetPropertiesMode mode, sd_bus_error *error); diff --git a/src/grp-system/libcore/src/dbus-manager.c b/src/grp-system/libcore/src/dbus-manager.c new file mode 100644 index 0000000000..8ae359f2d4 --- /dev/null +++ b/src/grp-system/libcore/src/dbus-manager.c @@ -0,0 +1,2307 @@ +/*** + This file is part of systemd. + + Copyright 2010 Lennart Poettering + + systemd is free software; you can redistribute it and/or modify it + under the terms of the GNU Lesser General Public License as published by + the Free Software Foundation; either version 2.1 of the License, or + (at your option) any later version. + + systemd is distributed in the hope that it will be useful, but + WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + Lesser General Public License for more details. + + You should have received a copy of the GNU Lesser General Public License + along with systemd; If not, see . +***/ + +#include +#include +#include + +#include "core/dbus-manager.h" +#include "sd-bus/bus-common-errors.h" +#include "systemd-basic/alloc-util.h" +#include "systemd-basic/architecture.h" +#include "systemd-basic/build.h" +#include "systemd-basic/clock-util.h" +#include "systemd-basic/env-util.h" +#include "systemd-basic/fd-util.h" +#include "systemd-basic/fileio.h" +#include "systemd-basic/formats-util.h" +#include "systemd-basic/log.h" +#include "systemd-basic/path-util.h" +#include "systemd-basic/stat-util.h" +#include "systemd-basic/string-util.h" +#include "systemd-basic/strv.h" +#include "systemd-basic/syslog-util.h" +#include "systemd-basic/virt.h" +#include "systemd-shared/install.h" +#include "systemd-shared/watchdog.h" + +#include "dbus-execute.h" +#include "dbus-job.h" +#include "dbus-unit.h" +#include "dbus.h" +#include "selinux-access.h" + +static int property_get_version( + sd_bus *bus, + const char *path, + const char *interface, + const char *property, + sd_bus_message *reply, + void *userdata, + sd_bus_error *error) { + + assert(bus); + assert(reply); + + return sd_bus_message_append(reply, "s", PACKAGE_VERSION); +} + +static int property_get_features( + sd_bus *bus, + const char *path, + const char *interface, + const char *property, + sd_bus_message *reply, + void *userdata, + sd_bus_error *error) { + + assert(bus); + assert(reply); + + return sd_bus_message_append(reply, "s", SYSTEMD_FEATURES); +} + +static int property_get_virtualization( + sd_bus *bus, + const char *path, + const char *interface, + const char *property, + sd_bus_message *reply, + void *userdata, + sd_bus_error *error) { + + int v; + + assert(bus); + assert(reply); + + v = detect_virtualization(); + + /* Make sure to return the empty string when we detect no virtualization, as that is the API. + * + * https://github.com/systemd/systemd/issues/1423 + */ + + return sd_bus_message_append( + reply, "s", + v == VIRTUALIZATION_NONE ? "" : virtualization_to_string(v)); +} + +static int property_get_architecture( + sd_bus *bus, + const char *path, + const char *interface, + const char *property, + sd_bus_message *reply, + void *userdata, + sd_bus_error *error) { + + assert(bus); + assert(reply); + + return sd_bus_message_append(reply, "s", architecture_to_string(uname_architecture())); +} + +static int property_get_tainted( + sd_bus *bus, + const char *path, + const char *interface, + const char *property, + sd_bus_message *reply, + void *userdata, + sd_bus_error *error) { + + char buf[sizeof("split-usr:cgroups-missing:local-hwclock:")] = "", *e = buf; + Manager *m = userdata; + + assert(bus); + assert(reply); + assert(m); + + if (m->taint_usr) + e = stpcpy(e, "split-usr:"); + + if (access("/proc/cgroups", F_OK) < 0) + e = stpcpy(e, "cgroups-missing:"); + + if (clock_is_localtime(NULL) > 0) + e = stpcpy(e, "local-hwclock:"); + + /* remove the last ':' */ + if (e != buf) + e[-1] = 0; + + return sd_bus_message_append(reply, "s", buf); +} + +static int property_get_log_target( + sd_bus *bus, + const char *path, + const char *interface, + const char *property, + sd_bus_message *reply, + void *userdata, + sd_bus_error *error) { + + assert(bus); + assert(reply); + + return sd_bus_message_append(reply, "s", log_target_to_string(log_get_target())); +} + +static int property_set_log_target( + sd_bus *bus, + const char *path, + const char *interface, + const char *property, + sd_bus_message *value, + void *userdata, + sd_bus_error *error) { + + const char *t; + int r; + + assert(bus); + assert(value); + + r = sd_bus_message_read(value, "s", &t); + if (r < 0) + return r; + + return log_set_target_from_string(t); +} + +static int property_get_log_level( + sd_bus *bus, + const char *path, + const char *interface, + const char *property, + sd_bus_message *reply, + void *userdata, + sd_bus_error *error) { + + _cleanup_free_ char *t = NULL; + int r; + + assert(bus); + assert(reply); + + r = log_level_to_string_alloc(log_get_max_level(), &t); + if (r < 0) + return r; + + return sd_bus_message_append(reply, "s", t); +} + +static int property_set_log_level( + sd_bus *bus, + const char *path, + const char *interface, + const char *property, + sd_bus_message *value, + void *userdata, + sd_bus_error *error) { + + const char *t; + int r; + + assert(bus); + assert(value); + + r = sd_bus_message_read(value, "s", &t); + if (r < 0) + return r; + + r = log_set_max_level_from_string(t); + if (r == 0) + log_info("Setting log level to %s.", t); + return r; +} + +static int property_get_n_names( + sd_bus *bus, + const char *path, + const char *interface, + const char *property, + sd_bus_message *reply, + void *userdata, + sd_bus_error *error) { + + Manager *m = userdata; + + assert(bus); + assert(reply); + assert(m); + + return sd_bus_message_append(reply, "u", (uint32_t) hashmap_size(m->units)); +} + +static int property_get_n_failed_units( + sd_bus *bus, + const char *path, + const char *interface, + const char *property, + sd_bus_message *reply, + void *userdata, + sd_bus_error *error) { + + Manager *m = userdata; + + assert(bus); + assert(reply); + assert(m); + + return sd_bus_message_append(reply, "u", (uint32_t) set_size(m->failed_units)); +} + +static int property_get_n_jobs( + sd_bus *bus, + const char *path, + const char *interface, + const char *property, + sd_bus_message *reply, + void *userdata, + sd_bus_error *error) { + + Manager *m = userdata; + + assert(bus); + assert(reply); + assert(m); + + return sd_bus_message_append(reply, "u", (uint32_t) hashmap_size(m->jobs)); +} + +static int property_get_progress( + sd_bus *bus, + const char *path, + const char *interface, + const char *property, + sd_bus_message *reply, + void *userdata, + sd_bus_error *error) { + + Manager *m = userdata; + double d; + + assert(bus); + assert(reply); + assert(m); + + if (dual_timestamp_is_set(&m->finish_timestamp)) + d = 1.0; + else + d = 1.0 - ((double) hashmap_size(m->jobs) / (double) m->n_installed_jobs); + + return sd_bus_message_append(reply, "d", d); +} + +static int property_get_system_state( + sd_bus *bus, + const char *path, + const char *interface, + const char *property, + sd_bus_message *reply, + void *userdata, + sd_bus_error *error) { + + Manager *m = userdata; + + assert(bus); + assert(reply); + assert(m); + + return sd_bus_message_append(reply, "s", manager_state_to_string(manager_state(m))); +} + +static int property_set_runtime_watchdog( + sd_bus *bus, + const char *path, + const char *interface, + const char *property, + sd_bus_message *value, + void *userdata, + sd_bus_error *error) { + + usec_t *t = userdata; + int r; + + assert(bus); + assert(value); + + assert_cc(sizeof(usec_t) == sizeof(uint64_t)); + + r = sd_bus_message_read(value, "t", t); + if (r < 0) + return r; + + return watchdog_set_timeout(t); +} + +static int property_get_timer_slack_nsec( + sd_bus *bus, + const char *path, + const char *interface, + const char *property, + sd_bus_message *reply, + void *userdata, + sd_bus_error *error) { + + assert(bus); + assert(reply); + + return sd_bus_message_append(reply, "t", (uint64_t) prctl(PR_GET_TIMERSLACK)); +} + +static int method_get_unit(sd_bus_message *message, void *userdata, sd_bus_error *error) { + _cleanup_free_ char *path = NULL; + Manager *m = userdata; + const char *name; + Unit *u; + int r; + + assert(message); + assert(m); + + /* Anyone can call this method */ + + r = sd_bus_message_read(message, "s", &name); + if (r < 0) + return r; + + if (isempty(name)) { + _cleanup_(sd_bus_creds_unrefp) sd_bus_creds *creds = NULL; + pid_t pid; + + r = sd_bus_query_sender_creds(message, SD_BUS_CREDS_PID, &creds); + if (r < 0) + return r; + + r = sd_bus_creds_get_pid(creds, &pid); + if (r < 0) + return r; + + u = manager_get_unit_by_pid(m, pid); + if (!u) + return sd_bus_error_setf(error, BUS_ERROR_NO_SUCH_UNIT, "Client not member of any unit."); + } else { + u = manager_get_unit(m, name); + if (!u) + return sd_bus_error_setf(error, BUS_ERROR_NO_SUCH_UNIT, "Unit %s not loaded.", name); + } + + r = mac_selinux_unit_access_check(u, message, "status", error); + if (r < 0) + return r; + + path = unit_dbus_path(u); + if (!path) + return -ENOMEM; + + return sd_bus_reply_method_return(message, "o", path); +} + +static int method_get_unit_by_pid(sd_bus_message *message, void *userdata, sd_bus_error *error) { + _cleanup_free_ char *path = NULL; + Manager *m = userdata; + pid_t pid; + Unit *u; + int r; + + assert(message); + assert(m); + + assert_cc(sizeof(pid_t) == sizeof(uint32_t)); + + /* Anyone can call this method */ + + r = sd_bus_message_read(message, "u", &pid); + if (r < 0) + return r; + if (pid < 0) + return sd_bus_error_setf(error, SD_BUS_ERROR_INVALID_ARGS, "Invalid PID " PID_FMT, pid); + + if (pid == 0) { + _cleanup_(sd_bus_creds_unrefp) sd_bus_creds *creds = NULL; + + r = sd_bus_query_sender_creds(message, SD_BUS_CREDS_PID, &creds); + if (r < 0) + return r; + + r = sd_bus_creds_get_pid(creds, &pid); + if (r < 0) + return r; + } + + u = manager_get_unit_by_pid(m, pid); + if (!u) + return sd_bus_error_setf(error, BUS_ERROR_NO_UNIT_FOR_PID, "PID "PID_FMT" does not belong to any loaded unit.", pid); + + r = mac_selinux_unit_access_check(u, message, "status", error); + if (r < 0) + return r; + + path = unit_dbus_path(u); + if (!path) + return -ENOMEM; + + return sd_bus_reply_method_return(message, "o", path); +} + +static int method_load_unit(sd_bus_message *message, void *userdata, sd_bus_error *error) { + _cleanup_free_ char *path = NULL; + Manager *m = userdata; + const char *name; + Unit *u; + int r; + + assert(message); + assert(m); + + /* Anyone can call this method */ + + r = sd_bus_message_read(message, "s", &name); + if (r < 0) + return r; + + if (isempty(name)) { + _cleanup_(sd_bus_creds_unrefp) sd_bus_creds *creds = NULL; + pid_t pid; + + r = sd_bus_query_sender_creds(message, SD_BUS_CREDS_PID, &creds); + if (r < 0) + return r; + + r = sd_bus_creds_get_pid(creds, &pid); + if (r < 0) + return r; + + u = manager_get_unit_by_pid(m, pid); + if (!u) + return sd_bus_error_setf(error, BUS_ERROR_NO_SUCH_UNIT, "Client not member of any unit."); + } else { + r = manager_load_unit(m, name, NULL, error, &u); + if (r < 0) + return r; + } + + r = mac_selinux_unit_access_check(u, message, "status", error); + if (r < 0) + return r; + + path = unit_dbus_path(u); + if (!path) + return -ENOMEM; + + return sd_bus_reply_method_return(message, "o", path); +} + +static int method_start_unit_generic(sd_bus_message *message, Manager *m, JobType job_type, bool reload_if_possible, sd_bus_error *error) { + const char *name; + Unit *u; + int r; + + assert(message); + assert(m); + + r = sd_bus_message_read(message, "s", &name); + if (r < 0) + return r; + + r = manager_load_unit(m, name, NULL, error, &u); + if (r < 0) + return r; + + return bus_unit_method_start_generic(message, u, job_type, reload_if_possible, error); +} + +static int method_start_unit(sd_bus_message *message, void *userdata, sd_bus_error *error) { + return method_start_unit_generic(message, userdata, JOB_START, false, error); +} + +static int method_stop_unit(sd_bus_message *message, void *userdata, sd_bus_error *error) { + return method_start_unit_generic(message, userdata, JOB_STOP, false, error); +} + +static int method_reload_unit(sd_bus_message *message, void *userdata, sd_bus_error *error) { + return method_start_unit_generic(message, userdata, JOB_RELOAD, false, error); +} + +static int method_restart_unit(sd_bus_message *message, void *userdata, sd_bus_error *error) { + return method_start_unit_generic(message, userdata, JOB_RESTART, false, error); +} + +static int method_try_restart_unit(sd_bus_message *message, void *userdata, sd_bus_error *error) { + return method_start_unit_generic(message, userdata, JOB_TRY_RESTART, false, error); +} + +static int method_reload_or_restart_unit(sd_bus_message *message, void *userdata, sd_bus_error *error) { + return method_start_unit_generic(message, userdata, JOB_RESTART, true, error); +} + +static int method_reload_or_try_restart_unit(sd_bus_message *message, void *userdata, sd_bus_error *error) { + return method_start_unit_generic(message, userdata, JOB_TRY_RESTART, true, error); +} + +static int method_start_unit_replace(sd_bus_message *message, void *userdata, sd_bus_error *error) { + Manager *m = userdata; + const char *old_name; + Unit *u; + int r; + + assert(message); + assert(m); + + r = sd_bus_message_read(message, "s", &old_name); + if (r < 0) + return r; + + u = manager_get_unit(m, old_name); + if (!u || !u->job || u->job->type != JOB_START) + return sd_bus_error_setf(error, BUS_ERROR_NO_SUCH_JOB, "No job queued for unit %s", old_name); + + return method_start_unit_generic(message, m, JOB_START, false, error); +} + +static int method_kill_unit(sd_bus_message *message, void *userdata, sd_bus_error *error) { + Manager *m = userdata; + const char *name; + Unit *u; + int r; + + assert(message); + assert(m); + + r = sd_bus_message_read(message, "s", &name); + if (r < 0) + return r; + + u = manager_get_unit(m, name); + if (!u) + return sd_bus_error_setf(error, BUS_ERROR_NO_SUCH_UNIT, "Unit %s is not loaded.", name); + + return bus_unit_method_kill(message, u, error); +} + +static int method_reset_failed_unit(sd_bus_message *message, void *userdata, sd_bus_error *error) { + Manager *m = userdata; + const char *name; + Unit *u; + int r; + + assert(message); + assert(m); + + r = sd_bus_message_read(message, "s", &name); + if (r < 0) + return r; + + u = manager_get_unit(m, name); + if (!u) + return sd_bus_error_setf(error, BUS_ERROR_NO_SUCH_UNIT, "Unit %s is not loaded.", name); + + return bus_unit_method_reset_failed(message, u, error); +} + +static int method_set_unit_properties(sd_bus_message *message, void *userdata, sd_bus_error *error) { + Manager *m = userdata; + const char *name; + Unit *u; + int r; + + assert(message); + assert(m); + + r = sd_bus_message_read(message, "s", &name); + if (r < 0) + return r; + + r = manager_load_unit(m, name, NULL, error, &u); + if (r < 0) + return r; + + r = bus_unit_check_load_state(u, error); + if (r < 0) + return r; + + return bus_unit_method_set_properties(message, u, error); +} + +static int reply_unit_info(sd_bus_message *reply, Unit *u) { + _cleanup_free_ char *unit_path = NULL, *job_path = NULL; + Unit *following; + + following = unit_following(u); + + unit_path = unit_dbus_path(u); + if (!unit_path) + return -ENOMEM; + + if (u->job) { + job_path = job_dbus_path(u->job); + if (!job_path) + return -ENOMEM; + } + + return sd_bus_message_append( + reply, "(ssssssouso)", + u->id, + unit_description(u), + unit_load_state_to_string(u->load_state), + unit_active_state_to_string(unit_active_state(u)), + unit_sub_state_to_string(u), + following ? following->id : "", + unit_path, + u->job ? u->job->id : 0, + u->job ? job_type_to_string(u->job->type) : "", + job_path ? job_path : "/"); +} + +static int method_list_units_by_names(sd_bus_message *message, void *userdata, sd_bus_error *error) { + _cleanup_(sd_bus_message_unrefp) sd_bus_message *reply = NULL; + Manager *m = userdata; + int r; + char **unit; + _cleanup_strv_free_ char **units = NULL; + + assert(message); + assert(m); + + r = sd_bus_message_read_strv(message, &units); + if (r < 0) + return r; + + r = sd_bus_message_new_method_return(message, &reply); + if (r < 0) + return r; + + r = sd_bus_message_open_container(reply, 'a', "(ssssssouso)"); + if (r < 0) + return r; + + STRV_FOREACH(unit, units) { + Unit *u; + + if (!unit_name_is_valid(*unit, UNIT_NAME_ANY)) + continue; + + r = manager_load_unit(m, *unit, NULL, error, &u); + if (r < 0) + return r; + + r = reply_unit_info(reply, u); + if (r < 0) + return r; + } + + r = sd_bus_message_close_container(reply); + if (r < 0) + return r; + + return sd_bus_send(NULL, reply, NULL); +} + +static int method_get_unit_processes(sd_bus_message *message, void *userdata, sd_bus_error *error) { + Manager *m = userdata; + const char *name; + Unit *u; + int r; + + assert(message); + assert(m); + + r = sd_bus_message_read(message, "s", &name); + if (r < 0) + return r; + + r = manager_load_unit(m, name, NULL, error, &u); + if (r < 0) + return r; + + r = bus_unit_check_load_state(u, error); + if (r < 0) + return r; + + return bus_unit_method_get_processes(message, u, error); +} + +static int transient_unit_from_message( + Manager *m, + sd_bus_message *message, + const char *name, + Unit **unit, + sd_bus_error *error) { + + UnitType t; + Unit *u; + int r; + + assert(m); + assert(message); + assert(name); + + t = unit_name_to_type(name); + if (t < 0) + return sd_bus_error_setf(error, SD_BUS_ERROR_INVALID_ARGS, "Invalid unit name or type."); + + if (!unit_vtable[t]->can_transient) + return sd_bus_error_setf(error, SD_BUS_ERROR_INVALID_ARGS, "Unit type %s does not support transient units.", unit_type_to_string(t)); + + r = manager_load_unit(m, name, NULL, error, &u); + if (r < 0) + return r; + + if (!unit_is_pristine(u)) + return sd_bus_error_setf(error, BUS_ERROR_UNIT_EXISTS, "Unit %s already exists.", name); + + /* OK, the unit failed to load and is unreferenced, now let's + * fill in the transient data instead */ + r = unit_make_transient(u); + if (r < 0) + return r; + + /* Set our properties */ + r = bus_unit_set_properties(u, message, UNIT_RUNTIME, false, error); + if (r < 0) + return r; + + /* Now load the missing bits of the unit we just created */ + unit_add_to_load_queue(u); + manager_dispatch_load_queue(m); + + *unit = u; + + return 0; +} + +static int transient_aux_units_from_message( + Manager *m, + sd_bus_message *message, + sd_bus_error *error) { + + int r; + + assert(m); + assert(message); + + r = sd_bus_message_enter_container(message, 'a', "(sa(sv))"); + if (r < 0) + return r; + + while ((r = sd_bus_message_enter_container(message, 'r', "sa(sv)")) > 0) { + const char *name = NULL; + Unit *u; + + r = sd_bus_message_read(message, "s", &name); + if (r < 0) + return r; + + r = transient_unit_from_message(m, message, name, &u, error); + if (r < 0) + return r; + + r = sd_bus_message_exit_container(message); + if (r < 0) + return r; + } + if (r < 0) + return r; + + r = sd_bus_message_exit_container(message); + if (r < 0) + return r; + + return 0; +} + +static int method_start_transient_unit(sd_bus_message *message, void *userdata, sd_bus_error *error) { + const char *name, *smode; + Manager *m = userdata; + JobMode mode; + Unit *u; + int r; + + assert(message); + assert(m); + + r = mac_selinux_access_check(message, "start", error); + if (r < 0) + return r; + + r = sd_bus_message_read(message, "ss", &name, &smode); + if (r < 0) + return r; + + mode = job_mode_from_string(smode); + if (mode < 0) + return sd_bus_error_setf(error, SD_BUS_ERROR_INVALID_ARGS, "Job mode %s is invalid.", smode); + + r = bus_verify_manage_units_async(m, message, error); + if (r < 0) + return r; + if (r == 0) + return 1; /* No authorization for now, but the async polkit stuff will call us again when it has it */ + + r = transient_unit_from_message(m, message, name, &u, error); + if (r < 0) + return r; + + r = transient_aux_units_from_message(m, message, error); + if (r < 0) + return r; + + /* Finally, start it */ + return bus_unit_queue_job(message, u, JOB_START, mode, false, error); +} + +static int method_get_job(sd_bus_message *message, void *userdata, sd_bus_error *error) { + _cleanup_free_ char *path = NULL; + Manager *m = userdata; + uint32_t id; + Job *j; + int r; + + assert(message); + assert(m); + + /* Anyone can call this method */ + + r = sd_bus_message_read(message, "u", &id); + if (r < 0) + return r; + + j = manager_get_job(m, id); + if (!j) + return sd_bus_error_setf(error, BUS_ERROR_NO_SUCH_JOB, "Job %u does not exist.", (unsigned) id); + + r = mac_selinux_unit_access_check(j->unit, message, "status", error); + if (r < 0) + return r; + + path = job_dbus_path(j); + if (!path) + return -ENOMEM; + + return sd_bus_reply_method_return(message, "o", path); +} + +static int method_cancel_job(sd_bus_message *message, void *userdata, sd_bus_error *error) { + Manager *m = userdata; + uint32_t id; + Job *j; + int r; + + assert(message); + assert(m); + + r = sd_bus_message_read(message, "u", &id); + if (r < 0) + return r; + + j = manager_get_job(m, id); + if (!j) + return sd_bus_error_setf(error, BUS_ERROR_NO_SUCH_JOB, "Job %u does not exist.", (unsigned) id); + + return bus_job_method_cancel(message, j, error); +} + +static int method_clear_jobs(sd_bus_message *message, void *userdata, sd_bus_error *error) { + Manager *m = userdata; + int r; + + assert(message); + assert(m); + + r = mac_selinux_access_check(message, "reload", error); + if (r < 0) + return r; + + r = bus_verify_manage_units_async(m, message, error); + if (r < 0) + return r; + if (r == 0) + return 1; /* No authorization for now, but the async polkit stuff will call us again when it has it */ + + manager_clear_jobs(m); + + return sd_bus_reply_method_return(message, NULL); +} + +static int method_reset_failed(sd_bus_message *message, void *userdata, sd_bus_error *error) { + Manager *m = userdata; + int r; + + assert(message); + assert(m); + + r = mac_selinux_access_check(message, "reload", error); + if (r < 0) + return r; + + r = bus_verify_manage_units_async(m, message, error); + if (r < 0) + return r; + if (r == 0) + return 1; /* No authorization for now, but the async polkit stuff will call us again when it has it */ + + manager_reset_failed(m); + + return sd_bus_reply_method_return(message, NULL); +} + +static int list_units_filtered(sd_bus_message *message, void *userdata, sd_bus_error *error, char **states, char **patterns) { + _cleanup_(sd_bus_message_unrefp) sd_bus_message *reply = NULL; + Manager *m = userdata; + const char *k; + Iterator i; + Unit *u; + int r; + + assert(message); + assert(m); + + /* Anyone can call this method */ + + r = mac_selinux_access_check(message, "status", error); + if (r < 0) + return r; + + r = sd_bus_message_new_method_return(message, &reply); + if (r < 0) + return r; + + r = sd_bus_message_open_container(reply, 'a', "(ssssssouso)"); + if (r < 0) + return r; + + HASHMAP_FOREACH_KEY(u, k, m->units, i) { + if (k != u->id) + continue; + + if (!strv_isempty(states) && + !strv_contains(states, unit_load_state_to_string(u->load_state)) && + !strv_contains(states, unit_active_state_to_string(unit_active_state(u))) && + !strv_contains(states, unit_sub_state_to_string(u))) + continue; + + if (!strv_isempty(patterns) && + !strv_fnmatch_or_empty(patterns, u->id, FNM_NOESCAPE)) + continue; + + r = reply_unit_info(reply, u); + if (r < 0) + return r; + } + + r = sd_bus_message_close_container(reply); + if (r < 0) + return r; + + return sd_bus_send(NULL, reply, NULL); +} + +static int method_list_units(sd_bus_message *message, void *userdata, sd_bus_error *error) { + return list_units_filtered(message, userdata, error, NULL, NULL); +} + +static int method_list_units_filtered(sd_bus_message *message, void *userdata, sd_bus_error *error) { + _cleanup_strv_free_ char **states = NULL; + int r; + + r = sd_bus_message_read_strv(message, &states); + if (r < 0) + return r; + + return list_units_filtered(message, userdata, error, states, NULL); +} + +static int method_list_units_by_patterns(sd_bus_message *message, void *userdata, sd_bus_error *error) { + _cleanup_strv_free_ char **states = NULL; + _cleanup_strv_free_ char **patterns = NULL; + int r; + + r = sd_bus_message_read_strv(message, &states); + if (r < 0) + return r; + + r = sd_bus_message_read_strv(message, &patterns); + if (r < 0) + return r; + + return list_units_filtered(message, userdata, error, states, patterns); +} + +static int method_list_jobs(sd_bus_message *message, void *userdata, sd_bus_error *error) { + _cleanup_(sd_bus_message_unrefp) sd_bus_message *reply = NULL; + Manager *m = userdata; + Iterator i; + Job *j; + int r; + + assert(message); + assert(m); + + /* Anyone can call this method */ + + r = mac_selinux_access_check(message, "status", error); + if (r < 0) + return r; + + r = sd_bus_message_new_method_return(message, &reply); + if (r < 0) + return r; + + r = sd_bus_message_open_container(reply, 'a', "(usssoo)"); + if (r < 0) + return r; + + HASHMAP_FOREACH(j, m->jobs, i) { + _cleanup_free_ char *unit_path = NULL, *job_path = NULL; + + job_path = job_dbus_path(j); + if (!job_path) + return -ENOMEM; + + unit_path = unit_dbus_path(j->unit); + if (!unit_path) + return -ENOMEM; + + r = sd_bus_message_append( + reply, "(usssoo)", + j->id, + j->unit->id, + job_type_to_string(j->type), + job_state_to_string(j->state), + job_path, + unit_path); + if (r < 0) + return r; + } + + r = sd_bus_message_close_container(reply); + if (r < 0) + return r; + + return sd_bus_send(NULL, reply, NULL); +} + +static int method_subscribe(sd_bus_message *message, void *userdata, sd_bus_error *error) { + Manager *m = userdata; + int r; + + assert(message); + assert(m); + + /* Anyone can call this method */ + + r = mac_selinux_access_check(message, "status", error); + if (r < 0) + return r; + + if (sd_bus_message_get_bus(message) == m->api_bus) { + + /* Note that direct bus connection subscribe by + * default, we only track peers on the API bus here */ + + if (!m->subscribed) { + r = sd_bus_track_new(sd_bus_message_get_bus(message), &m->subscribed, NULL, NULL); + if (r < 0) + return r; + } + + r = sd_bus_track_add_sender(m->subscribed, message); + if (r < 0) + return r; + if (r == 0) + return sd_bus_error_setf(error, BUS_ERROR_ALREADY_SUBSCRIBED, "Client is already subscribed."); + } + + return sd_bus_reply_method_return(message, NULL); +} + +static int method_unsubscribe(sd_bus_message *message, void *userdata, sd_bus_error *error) { + Manager *m = userdata; + int r; + + assert(message); + assert(m); + + /* Anyone can call this method */ + + r = mac_selinux_access_check(message, "status", error); + if (r < 0) + return r; + + if (sd_bus_message_get_bus(message) == m->api_bus) { + r = sd_bus_track_remove_sender(m->subscribed, message); + if (r < 0) + return r; + if (r == 0) + return sd_bus_error_setf(error, BUS_ERROR_NOT_SUBSCRIBED, "Client is not subscribed."); + } + + return sd_bus_reply_method_return(message, NULL); +} + +static int method_dump(sd_bus_message *message, void *userdata, sd_bus_error *error) { + _cleanup_free_ char *dump = NULL; + _cleanup_fclose_ FILE *f = NULL; + Manager *m = userdata; + size_t size; + int r; + + assert(message); + assert(m); + + /* Anyone can call this method */ + + r = mac_selinux_access_check(message, "status", error); + if (r < 0) + return r; + + f = open_memstream(&dump, &size); + if (!f) + return -ENOMEM; + + manager_dump_units(m, f, NULL); + manager_dump_jobs(m, f, NULL); + + r = fflush_and_check(f); + if (r < 0) + return r; + + return sd_bus_reply_method_return(message, "s", dump); +} + +static int method_refuse_snapshot(sd_bus_message *message, void *userdata, sd_bus_error *error) { + return sd_bus_error_setf(error, SD_BUS_ERROR_NOT_SUPPORTED, "Support for snapshots has been removed."); +} + +static int method_reload(sd_bus_message *message, void *userdata, sd_bus_error *error) { + Manager *m = userdata; + int r; + + assert(message); + assert(m); + + r = mac_selinux_access_check(message, "reload", error); + if (r < 0) + return r; + + r = bus_verify_reload_daemon_async(m, message, error); + if (r < 0) + return r; + if (r == 0) + return 1; /* No authorization for now, but the async polkit stuff will call us again when it has it */ + + /* Instead of sending the reply back right away, we just + * remember that we need to and then send it after the reload + * is finished. That way the caller knows when the reload + * finished. */ + + assert(!m->queued_message); + r = sd_bus_message_new_method_return(message, &m->queued_message); + if (r < 0) + return r; + + m->exit_code = MANAGER_RELOAD; + + return 1; +} + +static int method_reexecute(sd_bus_message *message, void *userdata, sd_bus_error *error) { + Manager *m = userdata; + int r; + + assert(message); + assert(m); + + r = mac_selinux_access_check(message, "reload", error); + if (r < 0) + return r; + + r = bus_verify_reload_daemon_async(m, message, error); + if (r < 0) + return r; + if (r == 0) + return 1; /* No authorization for now, but the async polkit stuff will call us again when it has it */ + + /* We don't send a reply back here, the client should + * just wait for us disconnecting. */ + + m->exit_code = MANAGER_REEXECUTE; + return 1; +} + +static int method_exit(sd_bus_message *message, void *userdata, sd_bus_error *error) { + Manager *m = userdata; + int r; + + assert(message); + assert(m); + + r = mac_selinux_access_check(message, "halt", error); + if (r < 0) + return r; + + /* Exit() (in contrast to SetExitCode()) is actually allowed even if + * we are running on the host. It will fall back on reboot() in + * systemd-shutdown if it cannot do the exit() because it isn't a + * container. */ + + m->exit_code = MANAGER_EXIT; + + return sd_bus_reply_method_return(message, NULL); +} + +static int method_reboot(sd_bus_message *message, void *userdata, sd_bus_error *error) { + Manager *m = userdata; + int r; + + assert(message); + assert(m); + + r = mac_selinux_access_check(message, "reboot", error); + if (r < 0) + return r; + + if (!MANAGER_IS_SYSTEM(m)) + return sd_bus_error_setf(error, SD_BUS_ERROR_NOT_SUPPORTED, "Reboot is only supported for system managers."); + + m->exit_code = MANAGER_REBOOT; + + return sd_bus_reply_method_return(message, NULL); +} + +static int method_poweroff(sd_bus_message *message, void *userdata, sd_bus_error *error) { + Manager *m = userdata; + int r; + + assert(message); + assert(m); + + r = mac_selinux_access_check(message, "halt", error); + if (r < 0) + return r; + + if (!MANAGER_IS_SYSTEM(m)) + return sd_bus_error_setf(error, SD_BUS_ERROR_NOT_SUPPORTED, "Powering off is only supported for system managers."); + + m->exit_code = MANAGER_POWEROFF; + + return sd_bus_reply_method_return(message, NULL); +} + +static int method_halt(sd_bus_message *message, void *userdata, sd_bus_error *error) { + Manager *m = userdata; + int r; + + assert(message); + assert(m); + + r = mac_selinux_access_check(message, "halt", error); + if (r < 0) + return r; + + if (!MANAGER_IS_SYSTEM(m)) + return sd_bus_error_setf(error, SD_BUS_ERROR_NOT_SUPPORTED, "Halt is only supported for system managers."); + + m->exit_code = MANAGER_HALT; + + return sd_bus_reply_method_return(message, NULL); +} + +static int method_kexec(sd_bus_message *message, void *userdata, sd_bus_error *error) { + Manager *m = userdata; + int r; + + assert(message); + assert(m); + + r = mac_selinux_access_check(message, "reboot", error); + if (r < 0) + return r; + + if (!MANAGER_IS_SYSTEM(m)) + return sd_bus_error_setf(error, SD_BUS_ERROR_NOT_SUPPORTED, "KExec is only supported for system managers."); + + m->exit_code = MANAGER_KEXEC; + + return sd_bus_reply_method_return(message, NULL); +} + +static int method_switch_root(sd_bus_message *message, void *userdata, sd_bus_error *error) { + char *ri = NULL, *rt = NULL; + const char *root, *init; + Manager *m = userdata; + int r; + + assert(message); + assert(m); + + r = mac_selinux_access_check(message, "reboot", error); + if (r < 0) + return r; + + if (!MANAGER_IS_SYSTEM(m)) + return sd_bus_error_setf(error, SD_BUS_ERROR_NOT_SUPPORTED, "Root switching is only supported by system manager."); + + r = sd_bus_message_read(message, "ss", &root, &init); + if (r < 0) + return r; + + if (path_equal(root, "/") || !path_is_absolute(root)) + return sd_bus_error_setf(error, SD_BUS_ERROR_INVALID_ARGS, "Invalid switch root path %s", root); + + /* Safety check */ + if (isempty(init)) { + if (!path_is_os_tree(root)) + return sd_bus_error_setf(error, SD_BUS_ERROR_INVALID_ARGS, "Specified switch root path %s does not seem to be an OS tree. os-release file is missing.", root); + } else { + _cleanup_free_ char *p = NULL; + + if (!path_is_absolute(init)) + return sd_bus_error_setf(error, SD_BUS_ERROR_INVALID_ARGS, "Invalid init path %s", init); + + p = strappend(root, init); + if (!p) + return -ENOMEM; + + if (access(p, X_OK) < 0) + return sd_bus_error_setf(error, SD_BUS_ERROR_INVALID_ARGS, "Specified init binary %s does not exist.", p); + } + + rt = strdup(root); + if (!rt) + return -ENOMEM; + + if (!isempty(init)) { + ri = strdup(init); + if (!ri) { + free(rt); + return -ENOMEM; + } + } + + free(m->switch_root); + m->switch_root = rt; + + free(m->switch_root_init); + m->switch_root_init = ri; + + m->exit_code = MANAGER_SWITCH_ROOT; + + return sd_bus_reply_method_return(message, NULL); +} + +static int method_set_environment(sd_bus_message *message, void *userdata, sd_bus_error *error) { + _cleanup_strv_free_ char **plus = NULL; + Manager *m = userdata; + int r; + + assert(message); + assert(m); + + r = mac_selinux_access_check(message, "reload", error); + if (r < 0) + return r; + + r = sd_bus_message_read_strv(message, &plus); + if (r < 0) + return r; + if (!strv_env_is_valid(plus)) + return sd_bus_error_setf(error, SD_BUS_ERROR_INVALID_ARGS, "Invalid environment assignments"); + + r = bus_verify_set_environment_async(m, message, error); + if (r < 0) + return r; + if (r == 0) + return 1; /* No authorization for now, but the async polkit stuff will call us again when it has it */ + + r = manager_environment_add(m, NULL, plus); + if (r < 0) + return r; + + return sd_bus_reply_method_return(message, NULL); +} + +static int method_unset_environment(sd_bus_message *message, void *userdata, sd_bus_error *error) { + _cleanup_strv_free_ char **minus = NULL; + Manager *m = userdata; + int r; + + assert(message); + assert(m); + + r = mac_selinux_access_check(message, "reload", error); + if (r < 0) + return r; + + r = sd_bus_message_read_strv(message, &minus); + if (r < 0) + return r; + + if (!strv_env_name_or_assignment_is_valid(minus)) + return sd_bus_error_setf(error, SD_BUS_ERROR_INVALID_ARGS, "Invalid environment variable names or assignments"); + + r = bus_verify_set_environment_async(m, message, error); + if (r < 0) + return r; + if (r == 0) + return 1; /* No authorization for now, but the async polkit stuff will call us again when it has it */ + + r = manager_environment_add(m, minus, NULL); + if (r < 0) + return r; + + return sd_bus_reply_method_return(message, NULL); +} + +static int method_unset_and_set_environment(sd_bus_message *message, void *userdata, sd_bus_error *error) { + _cleanup_strv_free_ char **minus = NULL, **plus = NULL; + Manager *m = userdata; + int r; + + assert(message); + assert(m); + + r = mac_selinux_access_check(message, "reload", error); + if (r < 0) + return r; + + r = sd_bus_message_read_strv(message, &minus); + if (r < 0) + return r; + + r = sd_bus_message_read_strv(message, &plus); + if (r < 0) + return r; + + if (!strv_env_name_or_assignment_is_valid(minus)) + return sd_bus_error_setf(error, SD_BUS_ERROR_INVALID_ARGS, "Invalid environment variable names or assignments"); + if (!strv_env_is_valid(plus)) + return sd_bus_error_setf(error, SD_BUS_ERROR_INVALID_ARGS, "Invalid environment assignments"); + + r = bus_verify_set_environment_async(m, message, error); + if (r < 0) + return r; + if (r == 0) + return 1; /* No authorization for now, but the async polkit stuff will call us again when it has it */ + + r = manager_environment_add(m, minus, plus); + if (r < 0) + return r; + + return sd_bus_reply_method_return(message, NULL); +} + +static int method_set_exit_code(sd_bus_message *message, void *userdata, sd_bus_error *error) { + uint8_t code; + Manager *m = userdata; + int r; + + assert(message); + assert(m); + + r = mac_selinux_access_check(message, "exit", error); + if (r < 0) + return r; + + r = sd_bus_message_read_basic(message, 'y', &code); + if (r < 0) + return r; + + if (MANAGER_IS_SYSTEM(m) && detect_container() <= 0) + return sd_bus_error_setf(error, SD_BUS_ERROR_NOT_SUPPORTED, "ExitCode can only be set for user service managers or in containers."); + + m->return_value = code; + + return sd_bus_reply_method_return(message, NULL); +} + +static int list_unit_files_by_patterns(sd_bus_message *message, void *userdata, sd_bus_error *error, char **states, char **patterns) { + _cleanup_(sd_bus_message_unrefp) sd_bus_message *reply = NULL; + Manager *m = userdata; + UnitFileList *item; + Hashmap *h; + Iterator i; + int r; + + assert(message); + assert(m); + + /* Anyone can call this method */ + + r = mac_selinux_access_check(message, "status", error); + if (r < 0) + return r; + + r = sd_bus_message_new_method_return(message, &reply); + if (r < 0) + return r; + + h = hashmap_new(&string_hash_ops); + if (!h) + return -ENOMEM; + + r = unit_file_get_list(m->unit_file_scope, NULL, h, states, patterns); + if (r < 0) + goto fail; + + r = sd_bus_message_open_container(reply, 'a', "(ss)"); + if (r < 0) + goto fail; + + HASHMAP_FOREACH(item, h, i) { + + r = sd_bus_message_append(reply, "(ss)", item->path, unit_file_state_to_string(item->state)); + if (r < 0) + goto fail; + } + + unit_file_list_free(h); + + r = sd_bus_message_close_container(reply); + if (r < 0) + return r; + + return sd_bus_send(NULL, reply, NULL); + +fail: + unit_file_list_free(h); + return r; +} + +static int method_list_unit_files(sd_bus_message *message, void *userdata, sd_bus_error *error) { + return list_unit_files_by_patterns(message, userdata, error, NULL, NULL); +} + +static int method_list_unit_files_by_patterns(sd_bus_message *message, void *userdata, sd_bus_error *error) { + _cleanup_strv_free_ char **states = NULL; + _cleanup_strv_free_ char **patterns = NULL; + int r; + + r = sd_bus_message_read_strv(message, &states); + if (r < 0) + return r; + + r = sd_bus_message_read_strv(message, &patterns); + if (r < 0) + return r; + + return list_unit_files_by_patterns(message, userdata, error, states, patterns); +} + +static int method_get_unit_file_state(sd_bus_message *message, void *userdata, sd_bus_error *error) { + Manager *m = userdata; + const char *name; + UnitFileState state; + int r; + + assert(message); + assert(m); + + /* Anyone can call this method */ + + r = mac_selinux_access_check(message, "status", error); + if (r < 0) + return r; + + r = sd_bus_message_read(message, "s", &name); + if (r < 0) + return r; + + r = unit_file_get_state(m->unit_file_scope, NULL, name, &state); + if (r < 0) + return r; + + return sd_bus_reply_method_return(message, "s", unit_file_state_to_string(state)); +} + +static int method_get_default_target(sd_bus_message *message, void *userdata, sd_bus_error *error) { + _cleanup_free_ char *default_target = NULL; + Manager *m = userdata; + int r; + + assert(message); + assert(m); + + /* Anyone can call this method */ + + r = mac_selinux_access_check(message, "status", error); + if (r < 0) + return r; + + r = unit_file_get_default(m->unit_file_scope, NULL, &default_target); + if (r < 0) + return r; + + return sd_bus_reply_method_return(message, "s", default_target); +} + +static int send_unit_files_changed(sd_bus *bus, void *userdata) { + _cleanup_(sd_bus_message_unrefp) sd_bus_message *message = NULL; + int r; + + assert(bus); + + r = sd_bus_message_new_signal(bus, &message, "/org/freedesktop/systemd1", "org.freedesktop.systemd1.Manager", "UnitFilesChanged"); + if (r < 0) + return r; + + return sd_bus_send(bus, message, NULL); +} + +static int reply_unit_file_changes_and_free( + Manager *m, + sd_bus_message *message, + int carries_install_info, + UnitFileChange *changes, + unsigned n_changes) { + + _cleanup_(sd_bus_message_unrefp) sd_bus_message *reply = NULL; + unsigned i; + int r; + + if (unit_file_changes_have_modification(changes, n_changes)) { + r = bus_foreach_bus(m, NULL, send_unit_files_changed, NULL); + if (r < 0) + log_debug_errno(r, "Failed to send UnitFilesChanged signal: %m"); + } + + r = sd_bus_message_new_method_return(message, &reply); + if (r < 0) + goto fail; + + if (carries_install_info >= 0) { + r = sd_bus_message_append(reply, "b", carries_install_info); + if (r < 0) + goto fail; + } + + r = sd_bus_message_open_container(reply, 'a', "(sss)"); + if (r < 0) + goto fail; + + for (i = 0; i < n_changes; i++) + if (changes[i].type >= 0) { + const char *change = unit_file_change_type_to_string(changes[i].type); + assert(change != NULL); + + r = sd_bus_message_append( + reply, "(sss)", + change, + changes[i].path, + changes[i].source); + if (r < 0) + goto fail; + } + + r = sd_bus_message_close_container(reply); + if (r < 0) + goto fail; + + unit_file_changes_free(changes, n_changes); + return sd_bus_send(NULL, reply, NULL); + +fail: + unit_file_changes_free(changes, n_changes); + return r; +} + +/* Create an error reply, using the error information from changes[] + * if possible, and fall back to generating an error from error code c. + * The error message only describes the first error. + * + * Coordinate with unit_file_dump_changes() in install.c. + */ +static int install_error( + sd_bus_error *error, + int c, + UnitFileChange *changes, + unsigned n_changes) { + int r; + unsigned i; + assert(c < 0); + + for (i = 0; i < n_changes; i++) + switch(changes[i].type) { + case 0 ... INT_MAX: + continue; + case -EEXIST: + if (changes[i].source) + r = sd_bus_error_setf(error, BUS_ERROR_UNIT_EXISTS, + "File %s already exists and is a symlink to %s.", + changes[i].path, changes[i].source); + else + r = sd_bus_error_setf(error, BUS_ERROR_UNIT_EXISTS, + "File %s already exists.", + changes[i].path); + goto found; + case -ERFKILL: + r = sd_bus_error_setf(error, BUS_ERROR_UNIT_MASKED, + "Unit file %s is masked.", changes[i].path); + goto found; + case -EADDRNOTAVAIL: + r = sd_bus_error_setf(error, BUS_ERROR_UNIT_GENERATED, + "Unit %s is transient or generated.", changes[i].path); + goto found; + case -ELOOP: + r = sd_bus_error_setf(error, BUS_ERROR_UNIT_LINKED, + "Refusing to operate on linked unit file %s", changes[i].path); + goto found; + default: + r = sd_bus_error_set_errnof(error, changes[i].type, "File %s: %m", changes[i].path); + goto found; + } + + r = c; + found: + unit_file_changes_free(changes, n_changes); + return r; +} + +static int method_enable_unit_files_generic( + sd_bus_message *message, + Manager *m, + int (*call)(UnitFileScope scope, bool runtime, const char *root_dir, char *files[], bool force, UnitFileChange **changes, unsigned *n_changes), + bool carries_install_info, + sd_bus_error *error) { + + _cleanup_strv_free_ char **l = NULL; + UnitFileChange *changes = NULL; + unsigned n_changes = 0; + int runtime, force, r; + + assert(message); + assert(m); + + r = sd_bus_message_read_strv(message, &l); + if (r < 0) + return r; + + r = sd_bus_message_read(message, "bb", &runtime, &force); + if (r < 0) + return r; + + r = bus_verify_manage_unit_files_async(m, message, error); + if (r < 0) + return r; + if (r == 0) + return 1; /* No authorization for now, but the async polkit stuff will call us again when it has it */ + + r = call(m->unit_file_scope, runtime, NULL, l, force, &changes, &n_changes); + if (r < 0) + return install_error(error, r, changes, n_changes); + + return reply_unit_file_changes_and_free(m, message, carries_install_info ? r : -1, changes, n_changes); +} + +static int method_enable_unit_files(sd_bus_message *message, void *userdata, sd_bus_error *error) { + return method_enable_unit_files_generic(message, userdata, unit_file_enable, true, error); +} + +static int method_reenable_unit_files(sd_bus_message *message, void *userdata, sd_bus_error *error) { + return method_enable_unit_files_generic(message, userdata, unit_file_reenable, true, error); +} + +static int method_link_unit_files(sd_bus_message *message, void *userdata, sd_bus_error *error) { + return method_enable_unit_files_generic(message, userdata, unit_file_link, false, error); +} + +static int unit_file_preset_without_mode(UnitFileScope scope, bool runtime, const char *root_dir, char **files, bool force, UnitFileChange **changes, unsigned *n_changes) { + return unit_file_preset(scope, runtime, root_dir, files, UNIT_FILE_PRESET_FULL, force, changes, n_changes); +} + +static int method_preset_unit_files(sd_bus_message *message, void *userdata, sd_bus_error *error) { + return method_enable_unit_files_generic(message, userdata, unit_file_preset_without_mode, true, error); +} + +static int method_mask_unit_files(sd_bus_message *message, void *userdata, sd_bus_error *error) { + return method_enable_unit_files_generic(message, userdata, unit_file_mask, false, error); +} + +static int method_preset_unit_files_with_mode(sd_bus_message *message, void *userdata, sd_bus_error *error) { + + _cleanup_strv_free_ char **l = NULL; + UnitFileChange *changes = NULL; + unsigned n_changes = 0; + Manager *m = userdata; + UnitFilePresetMode mm; + int runtime, force, r; + const char *mode; + + assert(message); + assert(m); + + r = sd_bus_message_read_strv(message, &l); + if (r < 0) + return r; + + r = sd_bus_message_read(message, "sbb", &mode, &runtime, &force); + if (r < 0) + return r; + + if (isempty(mode)) + mm = UNIT_FILE_PRESET_FULL; + else { + mm = unit_file_preset_mode_from_string(mode); + if (mm < 0) + return -EINVAL; + } + + r = bus_verify_manage_unit_files_async(m, message, error); + if (r < 0) + return r; + if (r == 0) + return 1; /* No authorization for now, but the async polkit stuff will call us again when it has it */ + + r = unit_file_preset(m->unit_file_scope, runtime, NULL, l, mm, force, &changes, &n_changes); + if (r < 0) + return install_error(error, r, changes, n_changes); + + return reply_unit_file_changes_and_free(m, message, r, changes, n_changes); +} + +static int method_disable_unit_files_generic( + sd_bus_message *message, + Manager *m, + int (*call)(UnitFileScope scope, bool runtime, const char *root_dir, char *files[], UnitFileChange **changes, unsigned *n_changes), + sd_bus_error *error) { + + _cleanup_strv_free_ char **l = NULL; + UnitFileChange *changes = NULL; + unsigned n_changes = 0; + int r, runtime; + + assert(message); + assert(m); + + r = sd_bus_message_read_strv(message, &l); + if (r < 0) + return r; + + r = sd_bus_message_read(message, "b", &runtime); + if (r < 0) + return r; + + r = bus_verify_manage_unit_files_async(m, message, error); + if (r < 0) + return r; + if (r == 0) + return 1; /* No authorization for now, but the async polkit stuff will call us again when it has it */ + + r = call(m->unit_file_scope, runtime, NULL, l, &changes, &n_changes); + if (r < 0) + return install_error(error, r, changes, n_changes); + + return reply_unit_file_changes_and_free(m, message, -1, changes, n_changes); +} + +static int method_disable_unit_files(sd_bus_message *message, void *userdata, sd_bus_error *error) { + return method_disable_unit_files_generic(message, userdata, unit_file_disable, error); +} + +static int method_unmask_unit_files(sd_bus_message *message, void *userdata, sd_bus_error *error) { + return method_disable_unit_files_generic(message, userdata, unit_file_unmask, error); +} + +static int method_revert_unit_files(sd_bus_message *message, void *userdata, sd_bus_error *error) { + _cleanup_strv_free_ char **l = NULL; + UnitFileChange *changes = NULL; + unsigned n_changes = 0; + Manager *m = userdata; + int r; + + assert(message); + assert(m); + + r = sd_bus_message_read_strv(message, &l); + if (r < 0) + return r; + + r = bus_verify_manage_unit_files_async(m, message, error); + if (r < 0) + return r; + if (r == 0) + return 1; /* No authorization for now, but the async polkit stuff will call us again when it has it */ + + r = unit_file_revert(m->unit_file_scope, NULL, l, &changes, &n_changes); + if (r < 0) + return install_error(error, r, changes, n_changes); + + return reply_unit_file_changes_and_free(m, message, -1, changes, n_changes); +} + +static int method_set_default_target(sd_bus_message *message, void *userdata, sd_bus_error *error) { + UnitFileChange *changes = NULL; + unsigned n_changes = 0; + Manager *m = userdata; + const char *name; + int force, r; + + assert(message); + assert(m); + + r = mac_selinux_access_check(message, "enable", error); + if (r < 0) + return r; + + r = sd_bus_message_read(message, "sb", &name, &force); + if (r < 0) + return r; + + r = bus_verify_manage_unit_files_async(m, message, error); + if (r < 0) + return r; + if (r == 0) + return 1; /* No authorization for now, but the async polkit stuff will call us again when it has it */ + + r = unit_file_set_default(m->unit_file_scope, NULL, name, force, &changes, &n_changes); + if (r < 0) + return install_error(error, r, changes, n_changes); + + return reply_unit_file_changes_and_free(m, message, -1, changes, n_changes); +} + +static int method_preset_all_unit_files(sd_bus_message *message, void *userdata, sd_bus_error *error) { + UnitFileChange *changes = NULL; + unsigned n_changes = 0; + Manager *m = userdata; + UnitFilePresetMode mm; + const char *mode; + int force, runtime, r; + + assert(message); + assert(m); + + r = mac_selinux_access_check(message, "enable", error); + if (r < 0) + return r; + + r = sd_bus_message_read(message, "sbb", &mode, &runtime, &force); + if (r < 0) + return r; + + if (isempty(mode)) + mm = UNIT_FILE_PRESET_FULL; + else { + mm = unit_file_preset_mode_from_string(mode); + if (mm < 0) + return -EINVAL; + } + + r = bus_verify_manage_unit_files_async(m, message, error); + if (r < 0) + return r; + if (r == 0) + return 1; /* No authorization for now, but the async polkit stuff will call us again when it has it */ + + r = unit_file_preset_all(m->unit_file_scope, runtime, NULL, mm, force, &changes, &n_changes); + if (r < 0) + return install_error(error, r, changes, n_changes); + + return reply_unit_file_changes_and_free(m, message, -1, changes, n_changes); +} + +static int method_add_dependency_unit_files(sd_bus_message *message, void *userdata, sd_bus_error *error) { + _cleanup_strv_free_ char **l = NULL; + Manager *m = userdata; + UnitFileChange *changes = NULL; + unsigned n_changes = 0; + int runtime, force, r; + char *target, *type; + UnitDependency dep; + + assert(message); + assert(m); + + r = bus_verify_manage_unit_files_async(m, message, error); + if (r < 0) + return r; + if (r == 0) + return 1; /* No authorization for now, but the async polkit stuff will call us again when it has it */ + + r = sd_bus_message_read_strv(message, &l); + if (r < 0) + return r; + + r = sd_bus_message_read(message, "ssbb", &target, &type, &runtime, &force); + if (r < 0) + return r; + + dep = unit_dependency_from_string(type); + if (dep < 0) + return -EINVAL; + + r = unit_file_add_dependency(m->unit_file_scope, runtime, NULL, l, target, dep, force, &changes, &n_changes); + if (r < 0) + return install_error(error, r, changes, n_changes); + + return reply_unit_file_changes_and_free(m, message, -1, changes, n_changes); +} + +const sd_bus_vtable bus_manager_vtable[] = { + SD_BUS_VTABLE_START(0), + + SD_BUS_PROPERTY("Version", "s", property_get_version, 0, SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("Features", "s", property_get_features, 0, SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("Virtualization", "s", property_get_virtualization, 0, SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("Architecture", "s", property_get_architecture, 0, SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("Tainted", "s", property_get_tainted, 0, SD_BUS_VTABLE_PROPERTY_CONST), + BUS_PROPERTY_DUAL_TIMESTAMP("FirmwareTimestamp", offsetof(Manager, firmware_timestamp), SD_BUS_VTABLE_PROPERTY_CONST), + BUS_PROPERTY_DUAL_TIMESTAMP("LoaderTimestamp", offsetof(Manager, loader_timestamp), SD_BUS_VTABLE_PROPERTY_CONST), + BUS_PROPERTY_DUAL_TIMESTAMP("KernelTimestamp", offsetof(Manager, kernel_timestamp), SD_BUS_VTABLE_PROPERTY_CONST), + BUS_PROPERTY_DUAL_TIMESTAMP("InitRDTimestamp", offsetof(Manager, initrd_timestamp), SD_BUS_VTABLE_PROPERTY_CONST), + BUS_PROPERTY_DUAL_TIMESTAMP("UserspaceTimestamp", offsetof(Manager, userspace_timestamp), SD_BUS_VTABLE_PROPERTY_CONST), + BUS_PROPERTY_DUAL_TIMESTAMP("FinishTimestamp", offsetof(Manager, finish_timestamp), SD_BUS_VTABLE_PROPERTY_CONST), + BUS_PROPERTY_DUAL_TIMESTAMP("SecurityStartTimestamp", offsetof(Manager, security_start_timestamp), SD_BUS_VTABLE_PROPERTY_CONST), + BUS_PROPERTY_DUAL_TIMESTAMP("SecurityFinishTimestamp", offsetof(Manager, security_finish_timestamp), SD_BUS_VTABLE_PROPERTY_CONST), + BUS_PROPERTY_DUAL_TIMESTAMP("GeneratorsStartTimestamp", offsetof(Manager, generators_start_timestamp), SD_BUS_VTABLE_PROPERTY_CONST), + BUS_PROPERTY_DUAL_TIMESTAMP("GeneratorsFinishTimestamp", offsetof(Manager, generators_finish_timestamp), SD_BUS_VTABLE_PROPERTY_CONST), + BUS_PROPERTY_DUAL_TIMESTAMP("UnitsLoadStartTimestamp", offsetof(Manager, units_load_start_timestamp), SD_BUS_VTABLE_PROPERTY_CONST), + BUS_PROPERTY_DUAL_TIMESTAMP("UnitsLoadFinishTimestamp", offsetof(Manager, units_load_finish_timestamp), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_WRITABLE_PROPERTY("LogLevel", "s", property_get_log_level, property_set_log_level, 0, 0), + SD_BUS_WRITABLE_PROPERTY("LogTarget", "s", property_get_log_target, property_set_log_target, 0, 0), + SD_BUS_PROPERTY("NNames", "u", property_get_n_names, 0, 0), + SD_BUS_PROPERTY("NFailedUnits", "u", property_get_n_failed_units, 0, SD_BUS_VTABLE_PROPERTY_EMITS_CHANGE), + SD_BUS_PROPERTY("NJobs", "u", property_get_n_jobs, 0, 0), + SD_BUS_PROPERTY("NInstalledJobs", "u", bus_property_get_unsigned, offsetof(Manager, n_installed_jobs), 0), + SD_BUS_PROPERTY("NFailedJobs", "u", bus_property_get_unsigned, offsetof(Manager, n_failed_jobs), 0), + SD_BUS_PROPERTY("Progress", "d", property_get_progress, 0, 0), + SD_BUS_PROPERTY("Environment", "as", NULL, offsetof(Manager, environment), 0), + SD_BUS_PROPERTY("ConfirmSpawn", "b", bus_property_get_bool, offsetof(Manager, confirm_spawn), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("ShowStatus", "b", bus_property_get_bool, offsetof(Manager, show_status), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("UnitPath", "as", NULL, offsetof(Manager, lookup_paths.search_path), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("DefaultStandardOutput", "s", bus_property_get_exec_output, offsetof(Manager, default_std_output), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("DefaultStandardError", "s", bus_property_get_exec_output, offsetof(Manager, default_std_output), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_WRITABLE_PROPERTY("RuntimeWatchdogUSec", "t", bus_property_get_usec, property_set_runtime_watchdog, offsetof(Manager, runtime_watchdog), 0), + SD_BUS_WRITABLE_PROPERTY("ShutdownWatchdogUSec", "t", bus_property_get_usec, bus_property_set_usec, offsetof(Manager, shutdown_watchdog), 0), + SD_BUS_PROPERTY("ControlGroup", "s", NULL, offsetof(Manager, cgroup_root), 0), + SD_BUS_PROPERTY("SystemState", "s", property_get_system_state, 0, 0), + SD_BUS_PROPERTY("ExitCode", "y", bus_property_get_unsigned, offsetof(Manager, return_value), 0), + SD_BUS_PROPERTY("DefaultTimerAccuracyUSec", "t", bus_property_get_usec, offsetof(Manager, default_timer_accuracy_usec), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("DefaultTimeoutStartUSec", "t", bus_property_get_usec, offsetof(Manager, default_timeout_start_usec), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("DefaultTimeoutStopUSec", "t", bus_property_get_usec, offsetof(Manager, default_timeout_stop_usec), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("DefaultRestartUSec", "t", bus_property_get_usec, offsetof(Manager, default_restart_usec), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("DefaultStartLimitIntervalSec", "t", bus_property_get_usec, offsetof(Manager, default_start_limit_interval), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("DefaultStartLimitInterval", "t", bus_property_get_usec, offsetof(Manager, default_start_limit_interval), SD_BUS_VTABLE_PROPERTY_CONST|SD_BUS_VTABLE_HIDDEN), /* obsolete alias name */ + SD_BUS_PROPERTY("DefaultStartLimitBurst", "u", bus_property_get_unsigned, offsetof(Manager, default_start_limit_burst), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("DefaultCPUAccounting", "b", bus_property_get_bool, offsetof(Manager, default_cpu_accounting), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("DefaultBlockIOAccounting", "b", bus_property_get_bool, offsetof(Manager, default_blockio_accounting), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("DefaultMemoryAccounting", "b", bus_property_get_bool, offsetof(Manager, default_memory_accounting), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("DefaultTasksAccounting", "b", bus_property_get_bool, offsetof(Manager, default_tasks_accounting), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("DefaultLimitCPU", "t", bus_property_get_rlimit, offsetof(Manager, rlimit[RLIMIT_CPU]), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("DefaultLimitCPUSoft", "t", bus_property_get_rlimit, offsetof(Manager, rlimit[RLIMIT_CPU]), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("DefaultLimitFSIZE", "t", bus_property_get_rlimit, offsetof(Manager, rlimit[RLIMIT_FSIZE]), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("DefaultLimitFSIZESoft", "t", bus_property_get_rlimit, offsetof(Manager, rlimit[RLIMIT_FSIZE]), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("DefaultLimitDATA", "t", bus_property_get_rlimit, offsetof(Manager, rlimit[RLIMIT_DATA]), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("DefaultLimitDATASoft", "t", bus_property_get_rlimit, offsetof(Manager, rlimit[RLIMIT_DATA]), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("DefaultLimitSTACK", "t", bus_property_get_rlimit, offsetof(Manager, rlimit[RLIMIT_STACK]), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("DefaultLimitSTACKSoft", "t", bus_property_get_rlimit, offsetof(Manager, rlimit[RLIMIT_STACK]), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("DefaultLimitCORE", "t", bus_property_get_rlimit, offsetof(Manager, rlimit[RLIMIT_CORE]), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("DefaultLimitCORESoft", "t", bus_property_get_rlimit, offsetof(Manager, rlimit[RLIMIT_CORE]), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("DefaultLimitRSS", "t", bus_property_get_rlimit, offsetof(Manager, rlimit[RLIMIT_RSS]), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("DefaultLimitRSSSoft", "t", bus_property_get_rlimit, offsetof(Manager, rlimit[RLIMIT_RSS]), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("DefaultLimitNOFILE", "t", bus_property_get_rlimit, offsetof(Manager, rlimit[RLIMIT_NOFILE]), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("DefaultLimitNOFILESoft", "t", bus_property_get_rlimit, offsetof(Manager, rlimit[RLIMIT_NOFILE]), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("DefaultLimitAS", "t", bus_property_get_rlimit, offsetof(Manager, rlimit[RLIMIT_AS]), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("DefaultLimitASSoft", "t", bus_property_get_rlimit, offsetof(Manager, rlimit[RLIMIT_AS]), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("DefaultLimitNPROC", "t", bus_property_get_rlimit, offsetof(Manager, rlimit[RLIMIT_NPROC]), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("DefaultLimitNPROCSoft", "t", bus_property_get_rlimit, offsetof(Manager, rlimit[RLIMIT_NPROC]), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("DefaultLimitMEMLOCK", "t", bus_property_get_rlimit, offsetof(Manager, rlimit[RLIMIT_MEMLOCK]), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("DefaultLimitMEMLOCKSoft", "t", bus_property_get_rlimit, offsetof(Manager, rlimit[RLIMIT_MEMLOCK]), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("DefaultLimitLOCKS", "t", bus_property_get_rlimit, offsetof(Manager, rlimit[RLIMIT_LOCKS]), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("DefaultLimitLOCKSSoft", "t", bus_property_get_rlimit, offsetof(Manager, rlimit[RLIMIT_LOCKS]), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("DefaultLimitSIGPENDING", "t", bus_property_get_rlimit, offsetof(Manager, rlimit[RLIMIT_SIGPENDING]), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("DefaultLimitSIGPENDINGSoft", "t", bus_property_get_rlimit, offsetof(Manager, rlimit[RLIMIT_SIGPENDING]), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("DefaultLimitMSGQUEUE", "t", bus_property_get_rlimit, offsetof(Manager, rlimit[RLIMIT_MSGQUEUE]), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("DefaultLimitMSGQUEUESoft", "t", bus_property_get_rlimit, offsetof(Manager, rlimit[RLIMIT_MSGQUEUE]), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("DefaultLimitNICE", "t", bus_property_get_rlimit, offsetof(Manager, rlimit[RLIMIT_NICE]), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("DefaultLimitNICESoft", "t", bus_property_get_rlimit, offsetof(Manager, rlimit[RLIMIT_NICE]), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("DefaultLimitRTPRIO", "t", bus_property_get_rlimit, offsetof(Manager, rlimit[RLIMIT_RTPRIO]), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("DefaultLimitRTPRIOSoft", "t", bus_property_get_rlimit, offsetof(Manager, rlimit[RLIMIT_RTPRIO]), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("DefaultLimitRTTIME", "t", bus_property_get_rlimit, offsetof(Manager, rlimit[RLIMIT_RTTIME]), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("DefaultLimitRTTIMESoft", "t", bus_property_get_rlimit, offsetof(Manager, rlimit[RLIMIT_RTTIME]), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("DefaultTasksMax", "t", NULL, offsetof(Manager, default_tasks_max), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("TimerSlackNSec", "t", property_get_timer_slack_nsec, 0, SD_BUS_VTABLE_PROPERTY_CONST), + + SD_BUS_METHOD("GetUnit", "s", "o", method_get_unit, SD_BUS_VTABLE_UNPRIVILEGED), + SD_BUS_METHOD("GetUnitByPID", "u", "o", method_get_unit_by_pid, SD_BUS_VTABLE_UNPRIVILEGED), + SD_BUS_METHOD("LoadUnit", "s", "o", method_load_unit, SD_BUS_VTABLE_UNPRIVILEGED), + SD_BUS_METHOD("StartUnit", "ss", "o", method_start_unit, SD_BUS_VTABLE_UNPRIVILEGED), + SD_BUS_METHOD("StartUnitReplace", "sss", "o", method_start_unit_replace, SD_BUS_VTABLE_UNPRIVILEGED), + SD_BUS_METHOD("StopUnit", "ss", "o", method_stop_unit, SD_BUS_VTABLE_UNPRIVILEGED), + SD_BUS_METHOD("ReloadUnit", "ss", "o", method_reload_unit, SD_BUS_VTABLE_UNPRIVILEGED), + SD_BUS_METHOD("RestartUnit", "ss", "o", method_restart_unit, SD_BUS_VTABLE_UNPRIVILEGED), + SD_BUS_METHOD("TryRestartUnit", "ss", "o", method_try_restart_unit, SD_BUS_VTABLE_UNPRIVILEGED), + SD_BUS_METHOD("ReloadOrRestartUnit", "ss", "o", method_reload_or_restart_unit, SD_BUS_VTABLE_UNPRIVILEGED), + SD_BUS_METHOD("ReloadOrTryRestartUnit", "ss", "o", method_reload_or_try_restart_unit, SD_BUS_VTABLE_UNPRIVILEGED), + SD_BUS_METHOD("KillUnit", "ssi", NULL, method_kill_unit, SD_BUS_VTABLE_UNPRIVILEGED), + SD_BUS_METHOD("ResetFailedUnit", "s", NULL, method_reset_failed_unit, SD_BUS_VTABLE_UNPRIVILEGED), + SD_BUS_METHOD("SetUnitProperties", "sba(sv)", NULL, method_set_unit_properties, SD_BUS_VTABLE_UNPRIVILEGED), + SD_BUS_METHOD("StartTransientUnit", "ssa(sv)a(sa(sv))", "o", method_start_transient_unit, SD_BUS_VTABLE_UNPRIVILEGED), + SD_BUS_METHOD("GetUnitProcesses", "s", "a(sus)", method_get_unit_processes, SD_BUS_VTABLE_UNPRIVILEGED), + SD_BUS_METHOD("GetJob", "u", "o", method_get_job, SD_BUS_VTABLE_UNPRIVILEGED), + SD_BUS_METHOD("CancelJob", "u", NULL, method_cancel_job, SD_BUS_VTABLE_UNPRIVILEGED), + SD_BUS_METHOD("ClearJobs", NULL, NULL, method_clear_jobs, SD_BUS_VTABLE_UNPRIVILEGED), + SD_BUS_METHOD("ResetFailed", NULL, NULL, method_reset_failed, SD_BUS_VTABLE_UNPRIVILEGED), + SD_BUS_METHOD("ListUnits", NULL, "a(ssssssouso)", method_list_units, SD_BUS_VTABLE_UNPRIVILEGED), + SD_BUS_METHOD("ListUnitsFiltered", "as", "a(ssssssouso)", method_list_units_filtered, SD_BUS_VTABLE_UNPRIVILEGED), + SD_BUS_METHOD("ListUnitsByPatterns", "asas", "a(ssssssouso)", method_list_units_by_patterns, SD_BUS_VTABLE_UNPRIVILEGED), + SD_BUS_METHOD("ListUnitsByNames", "as", "a(ssssssouso)", method_list_units_by_names, SD_BUS_VTABLE_UNPRIVILEGED), + SD_BUS_METHOD("ListJobs", NULL, "a(usssoo)", method_list_jobs, SD_BUS_VTABLE_UNPRIVILEGED), + SD_BUS_METHOD("Subscribe", NULL, NULL, method_subscribe, SD_BUS_VTABLE_UNPRIVILEGED), + SD_BUS_METHOD("Unsubscribe", NULL, NULL, method_unsubscribe, SD_BUS_VTABLE_UNPRIVILEGED), + SD_BUS_METHOD("Dump", NULL, "s", method_dump, SD_BUS_VTABLE_UNPRIVILEGED), + SD_BUS_METHOD("CreateSnapshot", "sb", "o", method_refuse_snapshot, SD_BUS_VTABLE_UNPRIVILEGED), + SD_BUS_METHOD("RemoveSnapshot", "s", NULL, method_refuse_snapshot, SD_BUS_VTABLE_UNPRIVILEGED), + SD_BUS_METHOD("Reload", NULL, NULL, method_reload, SD_BUS_VTABLE_UNPRIVILEGED), + SD_BUS_METHOD("Reexecute", NULL, NULL, method_reexecute, SD_BUS_VTABLE_UNPRIVILEGED), + SD_BUS_METHOD("Exit", NULL, NULL, method_exit, 0), + SD_BUS_METHOD("Reboot", NULL, NULL, method_reboot, SD_BUS_VTABLE_CAPABILITY(CAP_SYS_BOOT)), + SD_BUS_METHOD("PowerOff", NULL, NULL, method_poweroff, SD_BUS_VTABLE_CAPABILITY(CAP_SYS_BOOT)), + SD_BUS_METHOD("Halt", NULL, NULL, method_halt, SD_BUS_VTABLE_CAPABILITY(CAP_SYS_BOOT)), + SD_BUS_METHOD("KExec", NULL, NULL, method_kexec, SD_BUS_VTABLE_CAPABILITY(CAP_SYS_BOOT)), + SD_BUS_METHOD("SwitchRoot", "ss", NULL, method_switch_root, SD_BUS_VTABLE_CAPABILITY(CAP_SYS_BOOT)), + SD_BUS_METHOD("SetEnvironment", "as", NULL, method_set_environment, SD_BUS_VTABLE_UNPRIVILEGED), + SD_BUS_METHOD("UnsetEnvironment", "as", NULL, method_unset_environment, SD_BUS_VTABLE_UNPRIVILEGED), + SD_BUS_METHOD("UnsetAndSetEnvironment", "asas", NULL, method_unset_and_set_environment, SD_BUS_VTABLE_UNPRIVILEGED), + SD_BUS_METHOD("ListUnitFiles", NULL, "a(ss)", method_list_unit_files, SD_BUS_VTABLE_UNPRIVILEGED), + SD_BUS_METHOD("ListUnitFilesByPatterns", "asas", "a(ss)", method_list_unit_files_by_patterns, SD_BUS_VTABLE_UNPRIVILEGED), + SD_BUS_METHOD("GetUnitFileState", "s", "s", method_get_unit_file_state, SD_BUS_VTABLE_UNPRIVILEGED), + SD_BUS_METHOD("EnableUnitFiles", "asbb", "ba(sss)", method_enable_unit_files, SD_BUS_VTABLE_UNPRIVILEGED), + SD_BUS_METHOD("DisableUnitFiles", "asb", "a(sss)", method_disable_unit_files, SD_BUS_VTABLE_UNPRIVILEGED), + SD_BUS_METHOD("ReenableUnitFiles", "asbb", "ba(sss)", method_reenable_unit_files, SD_BUS_VTABLE_UNPRIVILEGED), + SD_BUS_METHOD("LinkUnitFiles", "asbb", "a(sss)", method_link_unit_files, SD_BUS_VTABLE_UNPRIVILEGED), + SD_BUS_METHOD("PresetUnitFiles", "asbb", "ba(sss)", method_preset_unit_files, SD_BUS_VTABLE_UNPRIVILEGED), + SD_BUS_METHOD("PresetUnitFilesWithMode", "assbb", "ba(sss)", method_preset_unit_files_with_mode, SD_BUS_VTABLE_UNPRIVILEGED), + SD_BUS_METHOD("MaskUnitFiles", "asbb", "a(sss)", method_mask_unit_files, SD_BUS_VTABLE_UNPRIVILEGED), + SD_BUS_METHOD("UnmaskUnitFiles", "asb", "a(sss)", method_unmask_unit_files, SD_BUS_VTABLE_UNPRIVILEGED), + SD_BUS_METHOD("RevertUnitFiles", "as", "a(sss)", method_revert_unit_files, SD_BUS_VTABLE_UNPRIVILEGED), + SD_BUS_METHOD("SetDefaultTarget", "sb", "a(sss)", method_set_default_target, SD_BUS_VTABLE_UNPRIVILEGED), + SD_BUS_METHOD("GetDefaultTarget", NULL, "s", method_get_default_target, SD_BUS_VTABLE_UNPRIVILEGED), + SD_BUS_METHOD("PresetAllUnitFiles", "sbb", "a(sss)", method_preset_all_unit_files, SD_BUS_VTABLE_UNPRIVILEGED), + SD_BUS_METHOD("AddDependencyUnitFiles", "asssbb", "a(sss)", method_add_dependency_unit_files, SD_BUS_VTABLE_UNPRIVILEGED), + SD_BUS_METHOD("SetExitCode", "y", NULL, method_set_exit_code, SD_BUS_VTABLE_UNPRIVILEGED), + + SD_BUS_SIGNAL("UnitNew", "so", 0), + SD_BUS_SIGNAL("UnitRemoved", "so", 0), + SD_BUS_SIGNAL("JobNew", "uos", 0), + SD_BUS_SIGNAL("JobRemoved", "uoss", 0), + SD_BUS_SIGNAL("StartupFinished", "tttttt", 0), + SD_BUS_SIGNAL("UnitFilesChanged", NULL, 0), + SD_BUS_SIGNAL("Reloading", "b", 0), + + SD_BUS_VTABLE_END +}; + +static int send_finished(sd_bus *bus, void *userdata) { + _cleanup_(sd_bus_message_unrefp) sd_bus_message *message = NULL; + usec_t *times = userdata; + int r; + + assert(bus); + assert(times); + + r = sd_bus_message_new_signal(bus, &message, "/org/freedesktop/systemd1", "org.freedesktop.systemd1.Manager", "StartupFinished"); + if (r < 0) + return r; + + r = sd_bus_message_append(message, "tttttt", times[0], times[1], times[2], times[3], times[4], times[5]); + if (r < 0) + return r; + + return sd_bus_send(bus, message, NULL); +} + +void bus_manager_send_finished( + Manager *m, + usec_t firmware_usec, + usec_t loader_usec, + usec_t kernel_usec, + usec_t initrd_usec, + usec_t userspace_usec, + usec_t total_usec) { + + int r; + + assert(m); + + r = bus_foreach_bus( + m, + NULL, + send_finished, + (usec_t[6]) { + firmware_usec, + loader_usec, + kernel_usec, + initrd_usec, + userspace_usec, + total_usec + }); + if (r < 0) + log_debug_errno(r, "Failed to send finished signal: %m"); +} + +static int send_reloading(sd_bus *bus, void *userdata) { + _cleanup_(sd_bus_message_unrefp) sd_bus_message *message = NULL; + int r; + + assert(bus); + + r = sd_bus_message_new_signal(bus, &message, "/org/freedesktop/systemd1", "org.freedesktop.systemd1.Manager", "Reloading"); + if (r < 0) + return r; + + r = sd_bus_message_append(message, "b", PTR_TO_INT(userdata)); + if (r < 0) + return r; + + return sd_bus_send(bus, message, NULL); +} + +void bus_manager_send_reloading(Manager *m, bool active) { + int r; + + assert(m); + + r = bus_foreach_bus(m, NULL, send_reloading, INT_TO_PTR(active)); + if (r < 0) + log_debug_errno(r, "Failed to send reloading signal: %m"); +} + +static int send_changed_signal(sd_bus *bus, void *userdata) { + assert(bus); + + return sd_bus_emit_properties_changed_strv(bus, + "/org/freedesktop/systemd1", + "org.freedesktop.systemd1.Manager", + NULL); +} + +void bus_manager_send_change_signal(Manager *m) { + int r; + + assert(m); + + r = bus_foreach_bus(m, NULL, send_changed_signal, NULL); + if (r < 0) + log_debug_errno(r, "Failed to send manager change signal: %m"); +} diff --git a/src/grp-system/libcore/src/dbus-mount.c b/src/grp-system/libcore/src/dbus-mount.c new file mode 100644 index 0000000000..ea2abb0e4e --- /dev/null +++ b/src/grp-system/libcore/src/dbus-mount.c @@ -0,0 +1,212 @@ +/*** + This file is part of systemd. + + Copyright 2010 Lennart Poettering + + systemd is free software; you can redistribute it and/or modify it + under the terms of the GNU Lesser General Public License as published by + the Free Software Foundation; either version 2.1 of the License, or + (at your option) any later version. + + systemd is distributed in the hope that it will be useful, but + WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + Lesser General Public License for more details. + + You should have received a copy of the GNU Lesser General Public License + along with systemd; If not, see . +***/ + +#include "core/mount.h" +#include "core/unit.h" +#include "sd-bus/bus-util.h" +#include "systemd-basic/string-util.h" + +#include "dbus-cgroup.h" +#include "dbus-execute.h" +#include "dbus-kill.h" +#include "dbus-mount.h" + +static int property_get_what( + sd_bus *bus, + const char *path, + const char *interface, + const char *property, + sd_bus_message *reply, + void *userdata, + sd_bus_error *error) { + + Mount *m = userdata; + const char *d; + + assert(bus); + assert(reply); + assert(m); + + if (m->from_proc_self_mountinfo && m->parameters_proc_self_mountinfo.what) + d = m->parameters_proc_self_mountinfo.what; + else if (m->from_fragment && m->parameters_fragment.what) + d = m->parameters_fragment.what; + else + d = ""; + + return sd_bus_message_append(reply, "s", d); +} + +static int property_get_options( + sd_bus *bus, + const char *path, + const char *interface, + const char *property, + sd_bus_message *reply, + void *userdata, + sd_bus_error *error) { + + Mount *m = userdata; + const char *d; + + assert(bus); + assert(reply); + assert(m); + + if (m->from_proc_self_mountinfo && m->parameters_proc_self_mountinfo.options) + d = m->parameters_proc_self_mountinfo.options; + else if (m->from_fragment && m->parameters_fragment.options) + d = m->parameters_fragment.options; + else + d = ""; + + return sd_bus_message_append(reply, "s", d); +} + +static int property_get_type( + sd_bus *bus, + const char *path, + const char *interface, + const char *property, + sd_bus_message *reply, + void *userdata, + sd_bus_error *error) { + + Mount *m = userdata; + const char *d; + + assert(bus); + assert(reply); + assert(m); + + if (m->from_proc_self_mountinfo && m->parameters_proc_self_mountinfo.fstype) + d = m->parameters_proc_self_mountinfo.fstype; + else if (m->from_fragment && m->parameters_fragment.fstype) + d = m->parameters_fragment.fstype; + else + d = ""; + + return sd_bus_message_append(reply, "s", d); +} + +static BUS_DEFINE_PROPERTY_GET_ENUM(property_get_result, mount_result, MountResult); + +const sd_bus_vtable bus_mount_vtable[] = { + SD_BUS_VTABLE_START(0), + SD_BUS_PROPERTY("Where", "s", NULL, offsetof(Mount, where), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("What", "s", property_get_what, 0, SD_BUS_VTABLE_PROPERTY_EMITS_CHANGE), + SD_BUS_PROPERTY("Options","s", property_get_options, 0, SD_BUS_VTABLE_PROPERTY_EMITS_CHANGE), + SD_BUS_PROPERTY("Type", "s", property_get_type, 0, SD_BUS_VTABLE_PROPERTY_EMITS_CHANGE), + SD_BUS_PROPERTY("TimeoutUSec", "t", bus_property_get_usec, offsetof(Mount, timeout_usec), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("ControlPID", "u", bus_property_get_pid, offsetof(Mount, control_pid), SD_BUS_VTABLE_PROPERTY_EMITS_CHANGE), + SD_BUS_PROPERTY("DirectoryMode", "u", bus_property_get_mode, offsetof(Mount, directory_mode), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("SloppyOptions", "b", bus_property_get_bool, offsetof(Mount, sloppy_options), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("Result", "s", property_get_result, offsetof(Mount, result), SD_BUS_VTABLE_PROPERTY_EMITS_CHANGE), + BUS_EXEC_COMMAND_VTABLE("ExecMount", offsetof(Mount, exec_command[MOUNT_EXEC_MOUNT]), SD_BUS_VTABLE_PROPERTY_EMITS_INVALIDATION), + BUS_EXEC_COMMAND_VTABLE("ExecUnmount", offsetof(Mount, exec_command[MOUNT_EXEC_UNMOUNT]), SD_BUS_VTABLE_PROPERTY_EMITS_INVALIDATION), + BUS_EXEC_COMMAND_VTABLE("ExecRemount", offsetof(Mount, exec_command[MOUNT_EXEC_REMOUNT]), SD_BUS_VTABLE_PROPERTY_EMITS_INVALIDATION), + SD_BUS_VTABLE_END +}; + +static int bus_mount_set_transient_property( + Mount *m, + const char *name, + sd_bus_message *message, + UnitSetPropertiesMode mode, + sd_bus_error *error) { + + const char *new_property; + char **property; + char *p; + int r; + + assert(m); + assert(name); + assert(message); + + if (streq(name, "What")) + property = &m->parameters_fragment.what; + else if (streq(name, "Options")) + property = &m->parameters_fragment.options; + else if (streq(name, "Type")) + property = &m->parameters_fragment.fstype; + else + return 0; + + r = sd_bus_message_read(message, "s", &new_property); + if (r < 0) + return r; + + if (mode != UNIT_CHECK) { + p = strdup(new_property); + if (!p) + return -ENOMEM; + + free(*property); + *property = p; + } + + return 1; +} + +int bus_mount_set_property( + Unit *u, + const char *name, + sd_bus_message *message, + UnitSetPropertiesMode mode, + sd_bus_error *error) { + + Mount *m = MOUNT(u); + int r; + + assert(m); + assert(name); + assert(message); + + r = bus_cgroup_set_property(u, &m->cgroup_context, name, message, mode, error); + if (r != 0) + return r; + + if (u->transient && u->load_state == UNIT_STUB) { + /* This is a transient unit, let's load a little more */ + + r = bus_mount_set_transient_property(m, name, message, mode, error); + if (r != 0) + return r; + + r = bus_exec_context_set_transient_property(u, &m->exec_context, name, message, mode, error); + if (r != 0) + return r; + + r = bus_kill_context_set_transient_property(u, &m->kill_context, name, message, mode, error); + if (r != 0) + return r; + } + + return 0; +} + +int bus_mount_commit_properties(Unit *u) { + assert(u); + + unit_update_cgroup_members_masks(u); + unit_realize_cgroup(u); + + return 0; +} diff --git a/src/grp-system/libcore/src/dbus-mount.h b/src/grp-system/libcore/src/dbus-mount.h new file mode 100644 index 0000000000..81ef769a44 --- /dev/null +++ b/src/grp-system/libcore/src/dbus-mount.h @@ -0,0 +1,29 @@ +#pragma once + +/*** + This file is part of systemd. + + Copyright 2010 Lennart Poettering + + systemd is free software; you can redistribute it and/or modify it + under the terms of the GNU Lesser General Public License as published by + the Free Software Foundation; either version 2.1 of the License, or + (at your option) any later version. + + systemd is distributed in the hope that it will be useful, but + WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + Lesser General Public License for more details. + + You should have received a copy of the GNU Lesser General Public License + along with systemd; If not, see . +***/ + +#include + +#include "core/unit.h" + +extern const sd_bus_vtable bus_mount_vtable[]; + +int bus_mount_set_property(Unit *u, const char *name, sd_bus_message *message, UnitSetPropertiesMode mode, sd_bus_error *error); +int bus_mount_commit_properties(Unit *u); diff --git a/src/grp-system/libcore/src/dbus-path.c b/src/grp-system/libcore/src/dbus-path.c new file mode 100644 index 0000000000..ed57c2133a --- /dev/null +++ b/src/grp-system/libcore/src/dbus-path.c @@ -0,0 +1,87 @@ +/*** + This file is part of systemd. + + Copyright 2010 Lennart Poettering + + systemd is free software; you can redistribute it and/or modify it + under the terms of the GNU Lesser General Public License as published by + the Free Software Foundation; either version 2.1 of the License, or + (at your option) any later version. + + systemd is distributed in the hope that it will be useful, but + WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + Lesser General Public License for more details. + + You should have received a copy of the GNU Lesser General Public License + along with systemd; If not, see . +***/ + +#include "core/path.h" +#include "core/unit.h" +#include "sd-bus/bus-util.h" +#include "systemd-basic/string-util.h" + +#include "dbus-path.h" + +static BUS_DEFINE_PROPERTY_GET_ENUM(property_get_result, path_result, PathResult); + +static int property_get_paths( + sd_bus *bus, + const char *path, + const char *interface, + const char *property, + sd_bus_message *reply, + void *userdata, + sd_bus_error *error) { + + Path *p = userdata; + PathSpec *k; + int r; + + assert(bus); + assert(reply); + assert(p); + + r = sd_bus_message_open_container(reply, 'a', "(ss)"); + if (r < 0) + return r; + + LIST_FOREACH(spec, k, p->specs) { + r = sd_bus_message_append(reply, "(ss)", path_type_to_string(k->type), k->path); + if (r < 0) + return r; + } + + return sd_bus_message_close_container(reply); +} + +static int property_get_unit( + sd_bus *bus, + const char *path, + const char *interface, + const char *property, + sd_bus_message *reply, + void *userdata, + sd_bus_error *error) { + + Unit *p = userdata, *trigger; + + assert(bus); + assert(reply); + assert(p); + + trigger = UNIT_TRIGGER(p); + + return sd_bus_message_append(reply, "s", trigger ? trigger->id : ""); +} + +const sd_bus_vtable bus_path_vtable[] = { + SD_BUS_VTABLE_START(0), + SD_BUS_PROPERTY("Unit", "s", property_get_unit, 0, SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("Paths", "a(ss)", property_get_paths, 0, SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("MakeDirectory", "b", bus_property_get_bool, offsetof(Path, make_directory), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("DirectoryMode", "u", bus_property_get_mode, offsetof(Path, directory_mode), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("Result", "s", property_get_result, offsetof(Path, result), SD_BUS_VTABLE_PROPERTY_EMITS_CHANGE), + SD_BUS_VTABLE_END +}; diff --git a/src/grp-system/libcore/src/dbus-path.h b/src/grp-system/libcore/src/dbus-path.h new file mode 100644 index 0000000000..d3c19e0c2b --- /dev/null +++ b/src/grp-system/libcore/src/dbus-path.h @@ -0,0 +1,24 @@ +#pragma once + +/*** + This file is part of systemd. + + Copyright 2010 Lennart Poettering + + systemd is free software; you can redistribute it and/or modify it + under the terms of the GNU Lesser General Public License as published by + the Free Software Foundation; either version 2.1 of the License, or + (at your option) any later version. + + systemd is distributed in the hope that it will be useful, but + WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + Lesser General Public License for more details. + + You should have received a copy of the GNU Lesser General Public License + along with systemd; If not, see . +***/ + + + +extern const sd_bus_vtable bus_path_vtable[]; diff --git a/src/grp-system/libcore/src/dbus-scope.c b/src/grp-system/libcore/src/dbus-scope.c new file mode 100644 index 0000000000..da930418d5 --- /dev/null +++ b/src/grp-system/libcore/src/dbus-scope.c @@ -0,0 +1,230 @@ +/*** + This file is part of systemd. + + Copyright 2013 Lennart Poettering + + systemd is free software; you can redistribute it and/or modify it + under the terms of the GNU Lesser General Public License as published by + the Free Software Foundation; either version 2.1 of the License, or + (at your option) any later version. + + systemd is distributed in the hope that it will be useful, but + WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + Lesser General Public License for more details. + + You should have received a copy of the GNU Lesser General Public License + along with systemd; If not, see . +***/ + +#include "core/scope.h" +#include "core/unit.h" +#include "sd-bus/bus-common-errors.h" +#include "sd-bus/bus-internal.h" +#include "sd-bus/bus-util.h" +#include "systemd-basic/alloc-util.h" + +#include "dbus-cgroup.h" +#include "dbus-kill.h" +#include "dbus-scope.h" +#include "dbus-unit.h" +#include "dbus.h" +#include "selinux-access.h" + +static int bus_scope_abandon(sd_bus_message *message, void *userdata, sd_bus_error *error) { + Scope *s = userdata; + int r; + + assert(message); + assert(s); + + r = mac_selinux_unit_access_check(UNIT(s), message, "stop", error); + if (r < 0) + return r; + + r = bus_verify_manage_units_async(UNIT(s)->manager, message, error); + if (r < 0) + return r; + if (r == 0) + return 1; /* No authorization for now, but the async polkit stuff will call us again when it has it */ + + r = scope_abandon(s); + if (r == -ESTALE) + return sd_bus_error_setf(error, BUS_ERROR_SCOPE_NOT_RUNNING, "Scope %s is not running, cannot abandon.", UNIT(s)->id); + if (r < 0) + return r; + + return sd_bus_reply_method_return(message, NULL); +} + +static BUS_DEFINE_PROPERTY_GET_ENUM(property_get_result, scope_result, ScopeResult); + +const sd_bus_vtable bus_scope_vtable[] = { + SD_BUS_VTABLE_START(0), + SD_BUS_PROPERTY("Controller", "s", NULL, offsetof(Scope, controller), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("TimeoutStopUSec", "t", bus_property_get_usec, offsetof(Scope, timeout_stop_usec), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("Result", "s", property_get_result, offsetof(Scope, result), SD_BUS_VTABLE_PROPERTY_EMITS_CHANGE), + SD_BUS_SIGNAL("RequestStop", NULL, 0), + SD_BUS_METHOD("Abandon", NULL, NULL, bus_scope_abandon, SD_BUS_VTABLE_UNPRIVILEGED), + SD_BUS_VTABLE_END +}; + +static int bus_scope_set_transient_property( + Scope *s, + const char *name, + sd_bus_message *message, + UnitSetPropertiesMode mode, + sd_bus_error *error) { + + int r; + + assert(s); + assert(name); + assert(message); + + if (streq(name, "PIDs")) { + unsigned n = 0; + uint32_t pid; + + r = sd_bus_message_enter_container(message, 'a', "u"); + if (r < 0) + return r; + + while ((r = sd_bus_message_read(message, "u", &pid)) > 0) { + + if (pid <= 1) + return -EINVAL; + + if (mode != UNIT_CHECK) { + r = unit_watch_pid(UNIT(s), pid); + if (r < 0 && r != -EEXIST) + return r; + } + + n++; + } + if (r < 0) + return r; + + r = sd_bus_message_exit_container(message); + if (r < 0) + return r; + + if (n <= 0) + return -EINVAL; + + return 1; + + } else if (streq(name, "Controller")) { + const char *controller; + char *c; + + r = sd_bus_message_read(message, "s", &controller); + if (r < 0) + return r; + + if (!isempty(controller) && !service_name_is_valid(controller)) + return sd_bus_error_setf(error, SD_BUS_ERROR_INVALID_ARGS, "Controller '%s' is not a valid bus name.", controller); + + if (mode != UNIT_CHECK) { + if (isempty(controller)) + c = NULL; + else { + c = strdup(controller); + if (!c) + return -ENOMEM; + } + + free(s->controller); + s->controller = c; + } + + return 1; + + } else if (streq(name, "TimeoutStopUSec")) { + + if (mode != UNIT_CHECK) { + r = sd_bus_message_read(message, "t", &s->timeout_stop_usec); + if (r < 0) + return r; + + unit_write_drop_in_private_format(UNIT(s), mode, name, "TimeoutStopSec="USEC_FMT"us", s->timeout_stop_usec); + } else { + r = sd_bus_message_skip(message, "t"); + if (r < 0) + return r; + } + + return 1; + } + + return 0; +} + +int bus_scope_set_property( + Unit *u, + const char *name, + sd_bus_message *message, + UnitSetPropertiesMode mode, + sd_bus_error *error) { + + Scope *s = SCOPE(u); + int r; + + assert(s); + assert(name); + assert(message); + + r = bus_cgroup_set_property(u, &s->cgroup_context, name, message, mode, error); + if (r != 0) + return r; + + if (u->load_state == UNIT_STUB) { + /* While we are created we still accept PIDs */ + + r = bus_scope_set_transient_property(s, name, message, mode, error); + if (r != 0) + return r; + + r = bus_kill_context_set_transient_property(u, &s->kill_context, name, message, mode, error); + if (r != 0) + return r; + } + + return 0; +} + +int bus_scope_commit_properties(Unit *u) { + assert(u); + + unit_update_cgroup_members_masks(u); + unit_realize_cgroup(u); + + return 0; +} + +int bus_scope_send_request_stop(Scope *s) { + _cleanup_(sd_bus_message_unrefp) sd_bus_message *m = NULL; + _cleanup_free_ char *p = NULL; + int r; + + assert(s); + + if (!s->controller) + return 0; + + p = unit_dbus_path(UNIT(s)); + if (!p) + return -ENOMEM; + + r = sd_bus_message_new_signal( + UNIT(s)->manager->api_bus, + &m, + p, + "org.freedesktop.systemd1.Scope", + "RequestStop"); + if (r < 0) + return r; + + return sd_bus_send_to(UNIT(s)->manager->api_bus, m, s->controller, NULL); +} diff --git a/src/grp-system/libcore/src/dbus-scope.h b/src/grp-system/libcore/src/dbus-scope.h new file mode 100644 index 0000000000..ad1cb52adc --- /dev/null +++ b/src/grp-system/libcore/src/dbus-scope.h @@ -0,0 +1,31 @@ +#pragma once + +/*** + This file is part of systemd. + + Copyright 2013 Lennart Poettering + + systemd is free software; you can redistribute it and/or modify it + under the terms of the GNU Lesser General Public License as published by + the Free Software Foundation; either version 2.1 of the License, or + (at your option) any later version. + + systemd is distributed in the hope that it will be useful, but + WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + Lesser General Public License for more details. + + You should have received a copy of the GNU Lesser General Public License + along with systemd; If not, see . +***/ + +#include + +#include "core/unit.h" + +extern const sd_bus_vtable bus_scope_vtable[]; + +int bus_scope_set_property(Unit *u, const char *name, sd_bus_message *i, UnitSetPropertiesMode mode, sd_bus_error *error); +int bus_scope_commit_properties(Unit *u); + +int bus_scope_send_request_stop(Scope *s); diff --git a/src/grp-system/libcore/src/dbus-service.c b/src/grp-system/libcore/src/dbus-service.c new file mode 100644 index 0000000000..42cd1c52bd --- /dev/null +++ b/src/grp-system/libcore/src/dbus-service.c @@ -0,0 +1,323 @@ +/*** + This file is part of systemd. + + Copyright 2010 Lennart Poettering + + systemd is free software; you can redistribute it and/or modify it + under the terms of the GNU Lesser General Public License as published by + the Free Software Foundation; either version 2.1 of the License, or + (at your option) any later version. + + systemd is distributed in the hope that it will be useful, but + WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + Lesser General Public License for more details. + + You should have received a copy of the GNU Lesser General Public License + along with systemd; If not, see . +***/ + +#include "core/service.h" +#include "core/unit.h" +#include "sd-bus/bus-util.h" +#include "systemd-basic/alloc-util.h" +#include "systemd-basic/async.h" +#include "systemd-basic/fd-util.h" +#include "systemd-basic/fileio.h" +#include "systemd-basic/path-util.h" +#include "systemd-basic/string-util.h" +#include "systemd-basic/strv.h" + +#include "dbus-cgroup.h" +#include "dbus-execute.h" +#include "dbus-kill.h" +#include "dbus-service.h" + +static BUS_DEFINE_PROPERTY_GET_ENUM(property_get_type, service_type, ServiceType); +static BUS_DEFINE_PROPERTY_GET_ENUM(property_get_result, service_result, ServiceResult); +static BUS_DEFINE_PROPERTY_GET_ENUM(property_get_restart, service_restart, ServiceRestart); +static BUS_DEFINE_PROPERTY_GET_ENUM(property_get_notify_access, notify_access, NotifyAccess); +static BUS_DEFINE_PROPERTY_GET_ENUM(property_get_failure_action, failure_action, FailureAction); + +const sd_bus_vtable bus_service_vtable[] = { + SD_BUS_VTABLE_START(0), + SD_BUS_PROPERTY("Type", "s", property_get_type, offsetof(Service, type), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("Restart", "s", property_get_restart, offsetof(Service, restart), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("PIDFile", "s", NULL, offsetof(Service, pid_file), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("NotifyAccess", "s", property_get_notify_access, offsetof(Service, notify_access), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("RestartUSec", "t", bus_property_get_usec, offsetof(Service, restart_usec), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("TimeoutStartUSec", "t", bus_property_get_usec, offsetof(Service, timeout_start_usec), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("TimeoutStopUSec", "t", bus_property_get_usec, offsetof(Service, timeout_stop_usec), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("RuntimeMaxUSec", "t", bus_property_get_usec, offsetof(Service, runtime_max_usec), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("WatchdogUSec", "t", bus_property_get_usec, offsetof(Service, watchdog_usec), SD_BUS_VTABLE_PROPERTY_CONST), + BUS_PROPERTY_DUAL_TIMESTAMP("WatchdogTimestamp", offsetof(Service, watchdog_timestamp), 0), + /* The following four are obsolete, and thus marked hidden here. They moved into the Unit interface */ + SD_BUS_PROPERTY("StartLimitInterval", "t", bus_property_get_usec, offsetof(Unit, start_limit.interval), SD_BUS_VTABLE_PROPERTY_CONST|SD_BUS_VTABLE_HIDDEN), + SD_BUS_PROPERTY("StartLimitBurst", "u", bus_property_get_unsigned, offsetof(Unit, start_limit.burst), SD_BUS_VTABLE_PROPERTY_CONST|SD_BUS_VTABLE_HIDDEN), + SD_BUS_PROPERTY("StartLimitAction", "s", property_get_failure_action, offsetof(Unit, start_limit_action), SD_BUS_VTABLE_PROPERTY_CONST|SD_BUS_VTABLE_HIDDEN), + SD_BUS_PROPERTY("RebootArgument", "s", NULL, offsetof(Unit, reboot_arg), SD_BUS_VTABLE_PROPERTY_CONST|SD_BUS_VTABLE_HIDDEN), + SD_BUS_PROPERTY("FailureAction", "s", property_get_failure_action, offsetof(Service, failure_action), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("PermissionsStartOnly", "b", bus_property_get_bool, offsetof(Service, permissions_start_only), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("RootDirectoryStartOnly", "b", bus_property_get_bool, offsetof(Service, root_directory_start_only), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("RemainAfterExit", "b", bus_property_get_bool, offsetof(Service, remain_after_exit), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("GuessMainPID", "b", bus_property_get_bool, offsetof(Service, guess_main_pid), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("MainPID", "u", bus_property_get_pid, offsetof(Service, main_pid), SD_BUS_VTABLE_PROPERTY_EMITS_CHANGE), + SD_BUS_PROPERTY("ControlPID", "u", bus_property_get_pid, offsetof(Service, control_pid), SD_BUS_VTABLE_PROPERTY_EMITS_CHANGE), + SD_BUS_PROPERTY("BusName", "s", NULL, offsetof(Service, bus_name), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("FileDescriptorStoreMax", "u", bus_property_get_unsigned, offsetof(Service, n_fd_store_max), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("NFileDescriptorStore", "u", bus_property_get_unsigned, offsetof(Service, n_fd_store), 0), + SD_BUS_PROPERTY("StatusText", "s", NULL, offsetof(Service, status_text), SD_BUS_VTABLE_PROPERTY_EMITS_CHANGE), + SD_BUS_PROPERTY("StatusErrno", "i", NULL, offsetof(Service, status_errno), SD_BUS_VTABLE_PROPERTY_EMITS_CHANGE), + SD_BUS_PROPERTY("Result", "s", property_get_result, offsetof(Service, result), SD_BUS_VTABLE_PROPERTY_EMITS_CHANGE), + SD_BUS_PROPERTY("USBFunctionDescriptors", "s", NULL, offsetof(Service, usb_function_descriptors), SD_BUS_VTABLE_PROPERTY_EMITS_CHANGE), + SD_BUS_PROPERTY("USBFunctionStrings", "s", NULL, offsetof(Service, usb_function_strings), SD_BUS_VTABLE_PROPERTY_EMITS_CHANGE), + BUS_EXEC_STATUS_VTABLE("ExecMain", offsetof(Service, main_exec_status), SD_BUS_VTABLE_PROPERTY_EMITS_CHANGE), + BUS_EXEC_COMMAND_LIST_VTABLE("ExecStartPre", offsetof(Service, exec_command[SERVICE_EXEC_START_PRE]), SD_BUS_VTABLE_PROPERTY_EMITS_INVALIDATION), + BUS_EXEC_COMMAND_LIST_VTABLE("ExecStart", offsetof(Service, exec_command[SERVICE_EXEC_START]), SD_BUS_VTABLE_PROPERTY_EMITS_INVALIDATION), + BUS_EXEC_COMMAND_LIST_VTABLE("ExecStartPost", offsetof(Service, exec_command[SERVICE_EXEC_START_POST]), SD_BUS_VTABLE_PROPERTY_EMITS_INVALIDATION), + BUS_EXEC_COMMAND_LIST_VTABLE("ExecReload", offsetof(Service, exec_command[SERVICE_EXEC_RELOAD]), SD_BUS_VTABLE_PROPERTY_EMITS_INVALIDATION), + BUS_EXEC_COMMAND_LIST_VTABLE("ExecStop", offsetof(Service, exec_command[SERVICE_EXEC_STOP]), SD_BUS_VTABLE_PROPERTY_EMITS_INVALIDATION), + BUS_EXEC_COMMAND_LIST_VTABLE("ExecStopPost", offsetof(Service, exec_command[SERVICE_EXEC_STOP_POST]), SD_BUS_VTABLE_PROPERTY_EMITS_INVALIDATION), + SD_BUS_VTABLE_END +}; + +static int bus_service_set_transient_property( + Service *s, + const char *name, + sd_bus_message *message, + UnitSetPropertiesMode mode, + sd_bus_error *error) { + + int r; + + assert(s); + assert(name); + assert(message); + + if (streq(name, "RemainAfterExit")) { + int b; + + r = sd_bus_message_read(message, "b", &b); + if (r < 0) + return r; + + if (mode != UNIT_CHECK) { + s->remain_after_exit = b; + unit_write_drop_in_private_format(UNIT(s), mode, name, "RemainAfterExit=%s", yes_no(b)); + } + + return 1; + + } else if (streq(name, "Type")) { + const char *t; + ServiceType k; + + r = sd_bus_message_read(message, "s", &t); + if (r < 0) + return r; + + k = service_type_from_string(t); + if (k < 0) + return sd_bus_error_setf(error, SD_BUS_ERROR_INVALID_ARGS, "Invalid service type %s", t); + + if (mode != UNIT_CHECK) { + s->type = k; + unit_write_drop_in_private_format(UNIT(s), mode, name, "Type=%s", service_type_to_string(s->type)); + } + + return 1; + } else if (streq(name, "RuntimeMaxUSec")) { + usec_t u; + + r = sd_bus_message_read(message, "t", &u); + if (r < 0) + return r; + + if (mode != UNIT_CHECK) { + s->runtime_max_usec = u; + unit_write_drop_in_private_format(UNIT(s), mode, name, "RuntimeMaxSec=" USEC_FMT "us", u); + } + + return 1; + + } else if (STR_IN_SET(name, + "StandardInputFileDescriptor", + "StandardOutputFileDescriptor", + "StandardErrorFileDescriptor")) { + int fd; + + r = sd_bus_message_read(message, "h", &fd); + if (r < 0) + return r; + + if (mode != UNIT_CHECK) { + int copy; + + copy = fcntl(fd, F_DUPFD_CLOEXEC, 3); + if (copy < 0) + return -errno; + + if (streq(name, "StandardInputFileDescriptor")) { + asynchronous_close(s->stdin_fd); + s->stdin_fd = copy; + } else if (streq(name, "StandardOutputFileDescriptor")) { + asynchronous_close(s->stdout_fd); + s->stdout_fd = copy; + } else { + asynchronous_close(s->stderr_fd); + s->stderr_fd = copy; + } + + s->exec_context.stdio_as_fds = true; + } + + return 1; + + } else if (streq(name, "ExecStart")) { + unsigned n = 0; + + r = sd_bus_message_enter_container(message, 'a', "(sasb)"); + if (r < 0) + return r; + + while ((r = sd_bus_message_enter_container(message, 'r', "sasb")) > 0) { + _cleanup_strv_free_ char **argv = NULL; + const char *path; + int b; + + r = sd_bus_message_read(message, "s", &path); + if (r < 0) + return r; + + if (!path_is_absolute(path)) + return sd_bus_error_set_errnof(error, EINVAL, "Path %s is not absolute.", path); + + r = sd_bus_message_read_strv(message, &argv); + if (r < 0) + return r; + + r = sd_bus_message_read(message, "b", &b); + if (r < 0) + return r; + + r = sd_bus_message_exit_container(message); + if (r < 0) + return r; + + if (mode != UNIT_CHECK) { + ExecCommand *c; + + c = new0(ExecCommand, 1); + if (!c) + return -ENOMEM; + + c->path = strdup(path); + if (!c->path) { + free(c); + return -ENOMEM; + } + + c->argv = argv; + argv = NULL; + + c->ignore = b; + + path_kill_slashes(c->path); + exec_command_append_list(&s->exec_command[SERVICE_EXEC_START], c); + } + + n++; + } + + if (r < 0) + return r; + + r = sd_bus_message_exit_container(message); + if (r < 0) + return r; + + if (mode != UNIT_CHECK) { + _cleanup_free_ char *buf = NULL; + _cleanup_fclose_ FILE *f = NULL; + ExecCommand *c; + size_t size = 0; + + if (n == 0) + s->exec_command[SERVICE_EXEC_START] = exec_command_free_list(s->exec_command[SERVICE_EXEC_START]); + + f = open_memstream(&buf, &size); + if (!f) + return -ENOMEM; + + fputs("ExecStart=\n", f); + + LIST_FOREACH(command, c, s->exec_command[SERVICE_EXEC_START]) { + _cleanup_free_ char *a; + + a = strv_join_quoted(c->argv); + if (!a) + return -ENOMEM; + + fprintf(f, "ExecStart=%s@%s %s\n", + c->ignore ? "-" : "", + c->path, + a); + } + + r = fflush_and_check(f); + if (r < 0) + return r; + unit_write_drop_in_private(UNIT(s), mode, name, buf); + } + + return 1; + } + + return 0; +} + +int bus_service_set_property( + Unit *u, + const char *name, + sd_bus_message *message, + UnitSetPropertiesMode mode, + sd_bus_error *error) { + + Service *s = SERVICE(u); + int r; + + assert(s); + assert(name); + assert(message); + + r = bus_cgroup_set_property(u, &s->cgroup_context, name, message, mode, error); + if (r != 0) + return r; + + if (u->transient && u->load_state == UNIT_STUB) { + /* This is a transient unit, let's load a little more */ + + r = bus_service_set_transient_property(s, name, message, mode, error); + if (r != 0) + return r; + + r = bus_exec_context_set_transient_property(u, &s->exec_context, name, message, mode, error); + if (r != 0) + return r; + + r = bus_kill_context_set_transient_property(u, &s->kill_context, name, message, mode, error); + if (r != 0) + return r; + } + + return 0; +} + +int bus_service_commit_properties(Unit *u) { + assert(u); + + unit_update_cgroup_members_masks(u); + unit_realize_cgroup(u); + + return 0; +} diff --git a/src/grp-system/libcore/src/dbus-service.h b/src/grp-system/libcore/src/dbus-service.h new file mode 100644 index 0000000000..1d3df67bf0 --- /dev/null +++ b/src/grp-system/libcore/src/dbus-service.h @@ -0,0 +1,29 @@ +#pragma once + +/*** + This file is part of systemd. + + Copyright 2010 Lennart Poettering + + systemd is free software; you can redistribute it and/or modify it + under the terms of the GNU Lesser General Public License as published by + the Free Software Foundation; either version 2.1 of the License, or + (at your option) any later version. + + systemd is distributed in the hope that it will be useful, but + WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + Lesser General Public License for more details. + + You should have received a copy of the GNU Lesser General Public License + along with systemd; If not, see . +***/ + +#include + +#include "core/unit.h" + +extern const sd_bus_vtable bus_service_vtable[]; + +int bus_service_set_property(Unit *u, const char *name, sd_bus_message *i, UnitSetPropertiesMode mode, sd_bus_error *error); +int bus_service_commit_properties(Unit *u); diff --git a/src/grp-system/libcore/src/dbus-slice.c b/src/grp-system/libcore/src/dbus-slice.c new file mode 100644 index 0000000000..ce357cadae --- /dev/null +++ b/src/grp-system/libcore/src/dbus-slice.c @@ -0,0 +1,53 @@ +/*** + This file is part of systemd. + + Copyright 2013 Lennart Poettering + + systemd is free software; you can redistribute it and/or modify it + under the terms of the GNU Lesser General Public License as published by + the Free Software Foundation; either version 2.1 of the License, or + (at your option) any later version. + + systemd is distributed in the hope that it will be useful, but + WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + Lesser General Public License for more details. + + You should have received a copy of the GNU Lesser General Public License + along with systemd; If not, see . +***/ + +#include "core/slice.h" +#include "core/unit.h" + +#include "dbus-cgroup.h" +#include "dbus-slice.h" + +const sd_bus_vtable bus_slice_vtable[] = { + SD_BUS_VTABLE_START(0), + SD_BUS_VTABLE_END +}; + +int bus_slice_set_property( + Unit *u, + const char *name, + sd_bus_message *message, + UnitSetPropertiesMode mode, + sd_bus_error *error) { + + Slice *s = SLICE(u); + + assert(name); + assert(u); + + return bus_cgroup_set_property(u, &s->cgroup_context, name, message, mode, error); +} + +int bus_slice_commit_properties(Unit *u) { + assert(u); + + unit_update_cgroup_members_masks(u); + unit_realize_cgroup(u); + + return 0; +} diff --git a/src/grp-system/libcore/src/dbus-slice.h b/src/grp-system/libcore/src/dbus-slice.h new file mode 100644 index 0000000000..fa039b2c0b --- /dev/null +++ b/src/grp-system/libcore/src/dbus-slice.h @@ -0,0 +1,29 @@ +#pragma once + +/*** + This file is part of systemd. + + Copyright 2013 Lennart Poettering + + systemd is free software; you can redistribute it and/or modify it + under the terms of the GNU Lesser General Public License as published by + the Free Software Foundation; either version 2.1 of the License, or + (at your option) any later version. + + systemd is distributed in the hope that it will be useful, but + WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + Lesser General Public License for more details. + + You should have received a copy of the GNU Lesser General Public License + along with systemd; If not, see . +***/ + +#include + +#include "core/unit.h" + +extern const sd_bus_vtable bus_slice_vtable[]; + +int bus_slice_set_property(Unit *u, const char *name, sd_bus_message *message, UnitSetPropertiesMode mode, sd_bus_error *error); +int bus_slice_commit_properties(Unit *u); diff --git a/src/grp-system/libcore/src/dbus-socket.c b/src/grp-system/libcore/src/dbus-socket.c new file mode 100644 index 0000000000..2bdfd0a47f --- /dev/null +++ b/src/grp-system/libcore/src/dbus-socket.c @@ -0,0 +1,185 @@ +/*** + This file is part of systemd. + + Copyright 2010 Lennart Poettering + + systemd is free software; you can redistribute it and/or modify it + under the terms of the GNU Lesser General Public License as published by + the Free Software Foundation; either version 2.1 of the License, or + (at your option) any later version. + + systemd is distributed in the hope that it will be useful, but + WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + Lesser General Public License for more details. + + You should have received a copy of the GNU Lesser General Public License + along with systemd; If not, see . +***/ + +#include "core/socket.h" +#include "core/unit.h" +#include "sd-bus/bus-util.h" +#include "systemd-basic/alloc-util.h" +#include "systemd-basic/string-util.h" + +#include "dbus-cgroup.h" +#include "dbus-execute.h" +#include "dbus-socket.h" + +static BUS_DEFINE_PROPERTY_GET_ENUM(property_get_result, socket_result, SocketResult); +static BUS_DEFINE_PROPERTY_GET_ENUM(property_get_bind_ipv6_only, socket_address_bind_ipv6_only, SocketAddressBindIPv6Only); + +static int property_get_listen( + sd_bus *bus, + const char *path, + const char *interface, + const char *property, + sd_bus_message *reply, + void *userdata, + sd_bus_error *error) { + + + Socket *s = SOCKET(userdata); + SocketPort *p; + int r; + + assert(bus); + assert(reply); + assert(s); + + r = sd_bus_message_open_container(reply, 'a', "(ss)"); + if (r < 0) + return r; + + LIST_FOREACH(port, p, s->ports) { + _cleanup_free_ char *address = NULL; + const char *a; + + switch (p->type) { + case SOCKET_SOCKET: { + r = socket_address_print(&p->address, &address); + if (r) + return r; + + a = address; + break; + } + + case SOCKET_SPECIAL: + case SOCKET_MQUEUE: + case SOCKET_FIFO: + case SOCKET_USB_FUNCTION: + a = p->path; + break; + + default: + assert_not_reached("Unknown socket type"); + } + + r = sd_bus_message_append(reply, "(ss)", socket_port_type_to_string(p), a); + if (r < 0) + return r; + } + + return sd_bus_message_close_container(reply); +} + + +static int property_get_fdname( + sd_bus *bus, + const char *path, + const char *interface, + const char *property, + sd_bus_message *reply, + void *userdata, + sd_bus_error *error) { + + Socket *s = SOCKET(userdata); + + assert(bus); + assert(reply); + assert(s); + + return sd_bus_message_append(reply, "s", socket_fdname(s)); +} + +const sd_bus_vtable bus_socket_vtable[] = { + SD_BUS_VTABLE_START(0), + SD_BUS_PROPERTY("BindIPv6Only", "s", property_get_bind_ipv6_only, offsetof(Socket, bind_ipv6_only), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("Backlog", "u", bus_property_get_unsigned, offsetof(Socket, backlog), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("TimeoutUSec", "t", bus_property_get_usec, offsetof(Socket, timeout_usec), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("BindToDevice", "s", NULL, offsetof(Socket, bind_to_device), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("SocketUser", "s", NULL, offsetof(Socket, user), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("SocketGroup", "s", NULL, offsetof(Socket, group), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("SocketMode", "u", bus_property_get_mode, offsetof(Socket, socket_mode), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("DirectoryMode", "u", bus_property_get_mode, offsetof(Socket, directory_mode), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("Accept", "b", bus_property_get_bool, offsetof(Socket, accept), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("Writable", "b", bus_property_get_bool, offsetof(Socket, writable), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("KeepAlive", "b", bus_property_get_bool, offsetof(Socket, keep_alive), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("KeepAliveTimeUSec", "t", bus_property_get_usec, offsetof(Socket, keep_alive_time), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("KeepAliveIntervalUSec", "t", bus_property_get_usec, offsetof(Socket, keep_alive_interval), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("KeepAliveProbes", "u", bus_property_get_unsigned, offsetof(Socket, keep_alive_cnt), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("DeferAcceptUSec" , "t", bus_property_get_usec, offsetof(Socket, defer_accept), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("NoDelay", "b", bus_property_get_bool, offsetof(Socket, no_delay), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("Priority", "i", bus_property_get_int, offsetof(Socket, priority), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("ReceiveBuffer", "t", bus_property_get_size, offsetof(Socket, receive_buffer), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("SendBuffer", "t", bus_property_get_size, offsetof(Socket, send_buffer), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("IPTOS", "i", bus_property_get_int, offsetof(Socket, ip_tos), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("IPTTL", "i", bus_property_get_int, offsetof(Socket, ip_ttl), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("PipeSize", "t", bus_property_get_size, offsetof(Socket, pipe_size), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("FreeBind", "b", bus_property_get_bool, offsetof(Socket, free_bind), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("Transparent", "b", bus_property_get_bool, offsetof(Socket, transparent), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("Broadcast", "b", bus_property_get_bool, offsetof(Socket, broadcast), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("PassCredentials", "b", bus_property_get_bool, offsetof(Socket, pass_cred), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("PassSecurity", "b", bus_property_get_bool, offsetof(Socket, pass_sec), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("RemoveOnStop", "b", bus_property_get_bool, offsetof(Socket, remove_on_stop), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("Listen", "a(ss)", property_get_listen, 0, SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("Symlinks", "as", NULL, offsetof(Socket, symlinks), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("Mark", "i", bus_property_get_int, offsetof(Socket, mark), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("MaxConnections", "u", bus_property_get_unsigned, offsetof(Socket, max_connections), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("MessageQueueMaxMessages", "x", bus_property_get_long, offsetof(Socket, mq_maxmsg), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("MessageQueueMessageSize", "x", bus_property_get_long, offsetof(Socket, mq_msgsize), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("ReusePort", "b", bus_property_get_bool, offsetof(Socket, reuse_port), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("SmackLabel", "s", NULL, offsetof(Socket, smack), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("SmackLabelIPIn", "s", NULL, offsetof(Socket, smack_ip_in), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("SmackLabelIPOut", "s", NULL, offsetof(Socket, smack_ip_out), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("ControlPID", "u", bus_property_get_pid, offsetof(Socket, control_pid), SD_BUS_VTABLE_PROPERTY_EMITS_CHANGE), + SD_BUS_PROPERTY("Result", "s", property_get_result, offsetof(Socket, result), SD_BUS_VTABLE_PROPERTY_EMITS_CHANGE), + SD_BUS_PROPERTY("NConnections", "u", bus_property_get_unsigned, offsetof(Socket, n_connections), 0), + SD_BUS_PROPERTY("NAccepted", "u", bus_property_get_unsigned, offsetof(Socket, n_accepted), 0), + SD_BUS_PROPERTY("FileDescriptorName", "s", property_get_fdname, 0, 0), + SD_BUS_PROPERTY("SocketProtocol", "i", bus_property_get_int, offsetof(Socket, socket_protocol), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("TriggerLimitIntervalUSec", "t", bus_property_get_usec, offsetof(Socket, trigger_limit.interval), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("TriggerLimitBurst", "u", bus_property_get_unsigned, offsetof(Socket, trigger_limit.burst), SD_BUS_VTABLE_PROPERTY_CONST), + BUS_EXEC_COMMAND_LIST_VTABLE("ExecStartPre", offsetof(Socket, exec_command[SOCKET_EXEC_START_PRE]), SD_BUS_VTABLE_PROPERTY_EMITS_INVALIDATION), + BUS_EXEC_COMMAND_LIST_VTABLE("ExecStartPost", offsetof(Socket, exec_command[SOCKET_EXEC_START_POST]), SD_BUS_VTABLE_PROPERTY_EMITS_INVALIDATION), + BUS_EXEC_COMMAND_LIST_VTABLE("ExecStopPre", offsetof(Socket, exec_command[SOCKET_EXEC_STOP_PRE]), SD_BUS_VTABLE_PROPERTY_EMITS_INVALIDATION), + BUS_EXEC_COMMAND_LIST_VTABLE("ExecStopPost", offsetof(Socket, exec_command[SOCKET_EXEC_STOP_POST]), SD_BUS_VTABLE_PROPERTY_EMITS_INVALIDATION), + SD_BUS_VTABLE_END +}; + +int bus_socket_set_property( + Unit *u, + const char *name, + sd_bus_message *message, + UnitSetPropertiesMode mode, + sd_bus_error *error) { + + Socket *s = SOCKET(u); + + assert(s); + assert(name); + assert(message); + + return bus_cgroup_set_property(u, &s->cgroup_context, name, message, mode, error); +} + +int bus_socket_commit_properties(Unit *u) { + assert(u); + + unit_update_cgroup_members_masks(u); + unit_realize_cgroup(u); + + return 0; +} diff --git a/src/grp-system/libcore/src/dbus-socket.h b/src/grp-system/libcore/src/dbus-socket.h new file mode 100644 index 0000000000..e68d33ace7 --- /dev/null +++ b/src/grp-system/libcore/src/dbus-socket.h @@ -0,0 +1,29 @@ +#pragma once + +/*** + This file is part of systemd. + + Copyright 2010 Lennart Poettering + + systemd is free software; you can redistribute it and/or modify it + under the terms of the GNU Lesser General Public License as published by + the Free Software Foundation; either version 2.1 of the License, or + (at your option) any later version. + + systemd is distributed in the hope that it will be useful, but + WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + Lesser General Public License for more details. + + You should have received a copy of the GNU Lesser General Public License + along with systemd; If not, see . +***/ + +#include + +#include "core/unit.h" + +extern const sd_bus_vtable bus_socket_vtable[]; + +int bus_socket_set_property(Unit *u, const char *name, sd_bus_message *message, UnitSetPropertiesMode mode, sd_bus_error *error); +int bus_socket_commit_properties(Unit *u); diff --git a/src/grp-system/libcore/src/dbus-swap.c b/src/grp-system/libcore/src/dbus-swap.c new file mode 100644 index 0000000000..3b0f274ba7 --- /dev/null +++ b/src/grp-system/libcore/src/dbus-swap.c @@ -0,0 +1,116 @@ +/*** + This file is part of systemd. + + Copyright 2010 Lennart Poettering + Copyright 2010 Maarten Lankhorst + + systemd is free software; you can redistribute it and/or modify it + under the terms of the GNU Lesser General Public License as published by + the Free Software Foundation; either version 2.1 of the License, or + (at your option) any later version. + + systemd is distributed in the hope that it will be useful, but + WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + Lesser General Public License for more details. + + You should have received a copy of the GNU Lesser General Public License + along with systemd; If not, see . +***/ + +#include "core/swap.h" +#include "core/unit.h" +#include "sd-bus/bus-util.h" +#include "systemd-basic/string-util.h" + +#include "dbus-cgroup.h" +#include "dbus-execute.h" +#include "dbus-swap.h" + +static int property_get_priority( + sd_bus *bus, + const char *path, + const char *interface, + const char *property, + sd_bus_message *reply, + void *userdata, + sd_bus_error *error) { + + Swap *s = SWAP(userdata); + int p; + + assert(bus); + assert(reply); + assert(s); + + if (s->from_proc_swaps) + p = s->parameters_proc_swaps.priority; + else if (s->from_fragment) + p = s->parameters_fragment.priority; + else + p = -1; + + return sd_bus_message_append(reply, "i", p); +} + +static int property_get_options( + sd_bus *bus, + const char *path, + const char *interface, + const char *property, + sd_bus_message *reply, + void *userdata, + sd_bus_error *error) { + + Swap *s = SWAP(userdata); + const char *options = NULL; + + assert(bus); + assert(reply); + assert(s); + + if (s->from_fragment) + options = s->parameters_fragment.options; + + return sd_bus_message_append(reply, "s", options); +} + +static BUS_DEFINE_PROPERTY_GET_ENUM(property_get_result, swap_result, SwapResult); + +const sd_bus_vtable bus_swap_vtable[] = { + SD_BUS_VTABLE_START(0), + SD_BUS_PROPERTY("What", "s", NULL, offsetof(Swap, what), SD_BUS_VTABLE_PROPERTY_EMITS_CHANGE), + SD_BUS_PROPERTY("Priority", "i", property_get_priority, 0, SD_BUS_VTABLE_PROPERTY_EMITS_CHANGE), + SD_BUS_PROPERTY("Options", "s", property_get_options, 0, SD_BUS_VTABLE_PROPERTY_EMITS_CHANGE), + SD_BUS_PROPERTY("TimeoutUSec", "t", bus_property_get_usec, offsetof(Swap, timeout_usec), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("ControlPID", "u", bus_property_get_pid, offsetof(Swap, control_pid), SD_BUS_VTABLE_PROPERTY_EMITS_CHANGE), + SD_BUS_PROPERTY("Result", "s", property_get_result, offsetof(Swap, result), SD_BUS_VTABLE_PROPERTY_EMITS_CHANGE), + BUS_EXEC_COMMAND_VTABLE("ExecActivate", offsetof(Swap, exec_command[SWAP_EXEC_ACTIVATE]), SD_BUS_VTABLE_PROPERTY_EMITS_INVALIDATION), + BUS_EXEC_COMMAND_VTABLE("ExecDeactivate", offsetof(Swap, exec_command[SWAP_EXEC_DEACTIVATE]), SD_BUS_VTABLE_PROPERTY_EMITS_INVALIDATION), + SD_BUS_VTABLE_END +}; + +int bus_swap_set_property( + Unit *u, + const char *name, + sd_bus_message *message, + UnitSetPropertiesMode mode, + sd_bus_error *error) { + + Swap *s = SWAP(u); + + assert(s); + assert(name); + assert(message); + + return bus_cgroup_set_property(u, &s->cgroup_context, name, message, mode, error); +} + +int bus_swap_commit_properties(Unit *u) { + assert(u); + + unit_update_cgroup_members_masks(u); + unit_realize_cgroup(u); + + return 0; +} diff --git a/src/grp-system/libcore/src/dbus-swap.h b/src/grp-system/libcore/src/dbus-swap.h new file mode 100644 index 0000000000..6c1b862665 --- /dev/null +++ b/src/grp-system/libcore/src/dbus-swap.h @@ -0,0 +1,30 @@ +#pragma once + +/*** + This file is part of systemd. + + Copyright 2010 Lennart Poettering + Copyright 2010 Maarten Lankhorst + + systemd is free software; you can redistribute it and/or modify it + under the terms of the GNU Lesser General Public License as published by + the Free Software Foundation; either version 2.1 of the License, or + (at your option) any later version. + + systemd is distributed in the hope that it will be useful, but + WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + Lesser General Public License for more details. + + You should have received a copy of the GNU Lesser General Public License + along with systemd; If not, see . +***/ + +#include + +#include "core/unit.h" + +extern const sd_bus_vtable bus_swap_vtable[]; + +int bus_swap_set_property(Unit *u, const char *name, sd_bus_message *message, UnitSetPropertiesMode mode, sd_bus_error *error); +int bus_swap_commit_properties(Unit *u); diff --git a/src/grp-system/libcore/src/dbus-target.c b/src/grp-system/libcore/src/dbus-target.c new file mode 100644 index 0000000000..5a846432d0 --- /dev/null +++ b/src/grp-system/libcore/src/dbus-target.c @@ -0,0 +1,27 @@ +/*** + This file is part of systemd. + + Copyright 2010 Lennart Poettering + + systemd is free software; you can redistribute it and/or modify it + under the terms of the GNU Lesser General Public License as published by + the Free Software Foundation; either version 2.1 of the License, or + (at your option) any later version. + + systemd is distributed in the hope that it will be useful, but + WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + Lesser General Public License for more details. + + You should have received a copy of the GNU Lesser General Public License + along with systemd; If not, see . +***/ + +#include "core/unit.h" + +#include "dbus-target.h" + +const sd_bus_vtable bus_target_vtable[] = { + SD_BUS_VTABLE_START(0), + SD_BUS_VTABLE_END +}; diff --git a/src/grp-system/libcore/src/dbus-target.h b/src/grp-system/libcore/src/dbus-target.h new file mode 100644 index 0000000000..c97a9d626e --- /dev/null +++ b/src/grp-system/libcore/src/dbus-target.h @@ -0,0 +1,24 @@ +#pragma once + +/*** + This file is part of systemd. + + Copyright 2010 Lennart Poettering + + systemd is free software; you can redistribute it and/or modify it + under the terms of the GNU Lesser General Public License as published by + the Free Software Foundation; either version 2.1 of the License, or + (at your option) any later version. + + systemd is distributed in the hope that it will be useful, but + WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + Lesser General Public License for more details. + + You should have received a copy of the GNU Lesser General Public License + along with systemd; If not, see . +***/ + +#include + +extern const sd_bus_vtable bus_target_vtable[]; diff --git a/src/grp-system/libcore/src/dbus-timer.c b/src/grp-system/libcore/src/dbus-timer.c new file mode 100644 index 0000000000..6b00168361 --- /dev/null +++ b/src/grp-system/libcore/src/dbus-timer.c @@ -0,0 +1,353 @@ +/*** + This file is part of systemd. + + Copyright 2010 Lennart Poettering + + systemd is free software; you can redistribute it and/or modify it + under the terms of the GNU Lesser General Public License as published by + the Free Software Foundation; either version 2.1 of the License, or + (at your option) any later version. + + systemd is distributed in the hope that it will be useful, but + WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + Lesser General Public License for more details. + + You should have received a copy of the GNU Lesser General Public License + along with systemd; If not, see . +***/ + +#include "core/timer.h" +#include "core/unit.h" +#include "sd-bus/bus-util.h" +#include "systemd-basic/alloc-util.h" +#include "systemd-basic/strv.h" + +#include "dbus-timer.h" + +static BUS_DEFINE_PROPERTY_GET_ENUM(property_get_result, timer_result, TimerResult); + +static int property_get_monotonic_timers( + sd_bus *bus, + const char *path, + const char *interface, + const char *property, + sd_bus_message *reply, + void *userdata, + sd_bus_error *error) { + + Timer *t = userdata; + TimerValue *v; + int r; + + assert(bus); + assert(reply); + assert(t); + + r = sd_bus_message_open_container(reply, 'a', "(stt)"); + if (r < 0) + return r; + + LIST_FOREACH(value, v, t->values) { + _cleanup_free_ char *buf = NULL; + const char *s; + size_t l; + + if (v->base == TIMER_CALENDAR) + continue; + + s = timer_base_to_string(v->base); + assert(endswith(s, "Sec")); + + /* s/Sec/USec/ */ + l = strlen(s); + buf = new(char, l+2); + if (!buf) + return -ENOMEM; + + memcpy(buf, s, l-3); + memcpy(buf+l-3, "USec", 5); + + r = sd_bus_message_append(reply, "(stt)", buf, v->value, v->next_elapse); + if (r < 0) + return r; + } + + return sd_bus_message_close_container(reply); +} + +static int property_get_calendar_timers( + sd_bus *bus, + const char *path, + const char *interface, + const char *property, + sd_bus_message *reply, + void *userdata, + sd_bus_error *error) { + + Timer *t = userdata; + TimerValue *v; + int r; + + assert(bus); + assert(reply); + assert(t); + + r = sd_bus_message_open_container(reply, 'a', "(sst)"); + if (r < 0) + return r; + + LIST_FOREACH(value, v, t->values) { + _cleanup_free_ char *buf = NULL; + + if (v->base != TIMER_CALENDAR) + continue; + + r = calendar_spec_to_string(v->calendar_spec, &buf); + if (r < 0) + return r; + + r = sd_bus_message_append(reply, "(sst)", timer_base_to_string(v->base), buf, v->next_elapse); + if (r < 0) + return r; + } + + return sd_bus_message_close_container(reply); +} + +static int property_get_unit( + sd_bus *bus, + const char *path, + const char *interface, + const char *property, + sd_bus_message *reply, + void *userdata, + sd_bus_error *error) { + + Unit *u = userdata, *trigger; + + assert(bus); + assert(reply); + assert(u); + + trigger = UNIT_TRIGGER(u); + + return sd_bus_message_append(reply, "s", trigger ? trigger->id : ""); +} + +static int property_get_next_elapse_monotonic( + sd_bus *bus, + const char *path, + const char *interface, + const char *property, + sd_bus_message *reply, + void *userdata, + sd_bus_error *error) { + + Timer *t = userdata; + usec_t x; + + assert(bus); + assert(reply); + assert(t); + + if (t->next_elapse_monotonic_or_boottime <= 0) + x = 0; + else if (t->wake_system) { + usec_t a, b; + + a = now(CLOCK_MONOTONIC); + b = now(clock_boottime_or_monotonic()); + + if (t->next_elapse_monotonic_or_boottime + a > b) + x = t->next_elapse_monotonic_or_boottime + a - b; + else + x = 0; + } else + x = t->next_elapse_monotonic_or_boottime; + + return sd_bus_message_append(reply, "t", x); +} + +const sd_bus_vtable bus_timer_vtable[] = { + SD_BUS_VTABLE_START(0), + SD_BUS_PROPERTY("Unit", "s", property_get_unit, 0, SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("TimersMonotonic", "a(stt)", property_get_monotonic_timers, 0, SD_BUS_VTABLE_PROPERTY_EMITS_INVALIDATION), + SD_BUS_PROPERTY("TimersCalendar", "a(sst)", property_get_calendar_timers, 0, SD_BUS_VTABLE_PROPERTY_EMITS_INVALIDATION), + SD_BUS_PROPERTY("NextElapseUSecRealtime", "t", bus_property_get_usec, offsetof(Timer, next_elapse_realtime), SD_BUS_VTABLE_PROPERTY_EMITS_CHANGE), + SD_BUS_PROPERTY("NextElapseUSecMonotonic", "t", property_get_next_elapse_monotonic, 0, SD_BUS_VTABLE_PROPERTY_EMITS_CHANGE), + BUS_PROPERTY_DUAL_TIMESTAMP("LastTriggerUSec", offsetof(Timer, last_trigger), SD_BUS_VTABLE_PROPERTY_EMITS_CHANGE), + SD_BUS_PROPERTY("Result", "s", property_get_result, offsetof(Timer, result), SD_BUS_VTABLE_PROPERTY_EMITS_CHANGE), + SD_BUS_PROPERTY("AccuracyUSec", "t", bus_property_get_usec, offsetof(Timer, accuracy_usec), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("RandomizedDelayUSec", "t", bus_property_get_usec, offsetof(Timer, random_usec), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("Persistent", "b", bus_property_get_bool, offsetof(Timer, persistent), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("WakeSystem", "b", bus_property_get_bool, offsetof(Timer, wake_system), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("RemainAfterElapse", "b", bus_property_get_bool, offsetof(Timer, remain_after_elapse), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_VTABLE_END +}; + +static int bus_timer_set_transient_property( + Timer *t, + const char *name, + sd_bus_message *message, + UnitSetPropertiesMode mode, + sd_bus_error *error) { + + int r; + + assert(t); + assert(name); + assert(message); + + if (STR_IN_SET(name, + "OnActiveSec", + "OnBootSec", + "OnStartupSec", + "OnUnitActiveSec", + "OnUnitInactiveSec")) { + + TimerValue *v; + TimerBase b = _TIMER_BASE_INVALID; + usec_t u = 0; + + b = timer_base_from_string(name); + if (b < 0) + return -EINVAL; + + r = sd_bus_message_read(message, "t", &u); + if (r < 0) + return r; + + if (mode != UNIT_CHECK) { + char time[FORMAT_TIMESPAN_MAX]; + + unit_write_drop_in_private_format(UNIT(t), mode, name, "%s=%s", name, format_timespan(time, sizeof(time), u, USEC_PER_MSEC)); + + v = new0(TimerValue, 1); + if (!v) + return -ENOMEM; + + v->base = b; + v->value = u; + + LIST_PREPEND(value, t->values, v); + } + + return 1; + + } else if (streq(name, "OnCalendar")) { + + TimerValue *v; + CalendarSpec *c = NULL; + const char *str; + + r = sd_bus_message_read(message, "s", &str); + if (r < 0) + return r; + + if (mode != UNIT_CHECK) { + r = calendar_spec_from_string(str, &c); + if (r < 0) + return r; + + unit_write_drop_in_private_format(UNIT(t), mode, name, "%s=%s", name, str); + + v = new0(TimerValue, 1); + if (!v) { + calendar_spec_free(c); + return -ENOMEM; + } + + v->base = TIMER_CALENDAR; + v->calendar_spec = c; + + LIST_PREPEND(value, t->values, v); + } + + return 1; + + } else if (STR_IN_SET(name, "AccuracyUSec", "AccuracySec")) { + usec_t u = 0; + + if (streq(name, "AccuracySec")) + log_notice("Client is using obsolete AccuracySec= transient property, please use AccuracyUSec= instead."); + + r = sd_bus_message_read(message, "t", &u); + if (r < 0) + return r; + + if (mode != UNIT_CHECK) { + t->accuracy_usec = u; + unit_write_drop_in_private_format(UNIT(t), mode, name, "AccuracySec=" USEC_FMT "us", u); + } + + return 1; + + } else if (streq(name, "RandomizedDelayUSec")) { + usec_t u = 0; + + r = sd_bus_message_read(message, "t", &u); + if (r < 0) + return r; + + if (mode != UNIT_CHECK) { + t->random_usec = u; + unit_write_drop_in_private_format(UNIT(t), mode, name, "RandomizedDelaySec=" USEC_FMT "us", u); + } + + return 1; + + } else if (streq(name, "WakeSystem")) { + int b; + + r = sd_bus_message_read(message, "b", &b); + if (r < 0) + return r; + + if (mode != UNIT_CHECK) { + t->wake_system = b; + unit_write_drop_in_private_format(UNIT(t), mode, name, "%s=%s", name, yes_no(b)); + } + + return 1; + + } else if (streq(name, "RemainAfterElapse")) { + int b; + + r = sd_bus_message_read(message, "b", &b); + if (r < 0) + return r; + + if (mode != UNIT_CHECK) { + t->remain_after_elapse = b; + unit_write_drop_in_private_format(UNIT(t), mode, name, "%s=%s", name, yes_no(b)); + } + + return 1; + } + + return 0; +} + +int bus_timer_set_property( + Unit *u, + const char *name, + sd_bus_message *message, + UnitSetPropertiesMode mode, + sd_bus_error *error) { + + Timer *t = TIMER(u); + int r; + + assert(t); + assert(name); + assert(message); + + if (u->transient && u->load_state == UNIT_STUB) { + r = bus_timer_set_transient_property(t, name, message, mode, error); + if (r != 0) + return r; + } + + return 0; +} diff --git a/src/grp-system/libcore/src/dbus-timer.h b/src/grp-system/libcore/src/dbus-timer.h new file mode 100644 index 0000000000..5e366918f8 --- /dev/null +++ b/src/grp-system/libcore/src/dbus-timer.h @@ -0,0 +1,28 @@ +#pragma once + +/*** + This file is part of systemd. + + Copyright 2010 Lennart Poettering + + systemd is free software; you can redistribute it and/or modify it + under the terms of the GNU Lesser General Public License as published by + the Free Software Foundation; either version 2.1 of the License, or + (at your option) any later version. + + systemd is distributed in the hope that it will be useful, but + WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + Lesser General Public License for more details. + + You should have received a copy of the GNU Lesser General Public License + along with systemd; If not, see . +***/ + +#include + +#include "core/unit.h" + +extern const sd_bus_vtable bus_timer_vtable[]; + +int bus_timer_set_property(Unit *u, const char *name, sd_bus_message *i, UnitSetPropertiesMode mode, sd_bus_error *error); diff --git a/src/grp-system/libcore/src/dbus-unit.c b/src/grp-system/libcore/src/dbus-unit.c new file mode 100644 index 0000000000..ef12ad66a3 --- /dev/null +++ b/src/grp-system/libcore/src/dbus-unit.c @@ -0,0 +1,1424 @@ +/*** + This file is part of systemd. + + Copyright 2010 Lennart Poettering + + systemd is free software; you can redistribute it and/or modify it + under the terms of the GNU Lesser General Public License as published by + the Free Software Foundation; either version 2.1 of the License, or + (at your option) any later version. + + systemd is distributed in the hope that it will be useful, but + WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + Lesser General Public License for more details. + + You should have received a copy of the GNU Lesser General Public License + along with systemd; If not, see . +***/ + +#include + +#include "sd-bus/bus-common-errors.h" +#include "systemd-basic/alloc-util.h" +#include "systemd-basic/cgroup-util.h" +#include "systemd-basic/fd-util.h" +#include "systemd-basic/locale-util.h" +#include "systemd-basic/log.h" +#include "systemd-basic/process-util.h" +#include "systemd-basic/signal-util.h" +#include "systemd-basic/special.h" +#include "systemd-basic/string-util.h" +#include "systemd-basic/strv.h" +#include "systemd-basic/user-util.h" + +#include "dbus-unit.h" +#include "dbus.h" +#include "selinux-access.h" + +static BUS_DEFINE_PROPERTY_GET_ENUM(property_get_load_state, unit_load_state, UnitLoadState); +static BUS_DEFINE_PROPERTY_GET_ENUM(property_get_job_mode, job_mode, JobMode); +static BUS_DEFINE_PROPERTY_GET_ENUM(property_get_failure_action, failure_action, FailureAction); + +static int property_get_names( + sd_bus *bus, + const char *path, + const char *interface, + const char *property, + sd_bus_message *reply, + void *userdata, + sd_bus_error *error) { + + Unit *u = userdata; + Iterator i; + const char *t; + int r; + + assert(bus); + assert(reply); + assert(u); + + r = sd_bus_message_open_container(reply, 'a', "s"); + if (r < 0) + return r; + + SET_FOREACH(t, u->names, i) { + r = sd_bus_message_append(reply, "s", t); + if (r < 0) + return r; + } + + return sd_bus_message_close_container(reply); +} + +static int property_get_following( + sd_bus *bus, + const char *path, + const char *interface, + const char *property, + sd_bus_message *reply, + void *userdata, + sd_bus_error *error) { + + Unit *u = userdata, *f; + + assert(bus); + assert(reply); + assert(u); + + f = unit_following(u); + return sd_bus_message_append(reply, "s", f ? f->id : ""); +} + +static int property_get_dependencies( + sd_bus *bus, + const char *path, + const char *interface, + const char *property, + sd_bus_message *reply, + void *userdata, + sd_bus_error *error) { + + Set *s = *(Set**) userdata; + Iterator j; + Unit *u; + int r; + + assert(bus); + assert(reply); + + r = sd_bus_message_open_container(reply, 'a', "s"); + if (r < 0) + return r; + + SET_FOREACH(u, s, j) { + r = sd_bus_message_append(reply, "s", u->id); + if (r < 0) + return r; + } + + return sd_bus_message_close_container(reply); +} + +static int property_get_obsolete_dependencies( + sd_bus *bus, + const char *path, + const char *interface, + const char *property, + sd_bus_message *reply, + void *userdata, + sd_bus_error *error) { + + assert(bus); + assert(reply); + + /* For dependency types we don't support anymore always return an empty array */ + return sd_bus_message_append(reply, "as", 0); +} + +static int property_get_description( + sd_bus *bus, + const char *path, + const char *interface, + const char *property, + sd_bus_message *reply, + void *userdata, + sd_bus_error *error) { + + Unit *u = userdata; + + assert(bus); + assert(reply); + assert(u); + + return sd_bus_message_append(reply, "s", unit_description(u)); +} + +static int property_get_active_state( + sd_bus *bus, + const char *path, + const char *interface, + const char *property, + sd_bus_message *reply, + void *userdata, + sd_bus_error *error) { + + Unit *u = userdata; + + assert(bus); + assert(reply); + assert(u); + + return sd_bus_message_append(reply, "s", unit_active_state_to_string(unit_active_state(u))); +} + +static int property_get_sub_state( + sd_bus *bus, + const char *path, + const char *interface, + const char *property, + sd_bus_message *reply, + void *userdata, + sd_bus_error *error) { + + Unit *u = userdata; + + assert(bus); + assert(reply); + assert(u); + + return sd_bus_message_append(reply, "s", unit_sub_state_to_string(u)); +} + +static int property_get_unit_file_preset( + sd_bus *bus, + const char *path, + const char *interface, + const char *property, + sd_bus_message *reply, + void *userdata, + sd_bus_error *error) { + + Unit *u = userdata; + int r; + + assert(bus); + assert(reply); + assert(u); + + r = unit_get_unit_file_preset(u); + + return sd_bus_message_append(reply, "s", + r < 0 ? "": + r > 0 ? "enabled" : "disabled"); +} + +static int property_get_unit_file_state( + sd_bus *bus, + const char *path, + const char *interface, + const char *property, + sd_bus_message *reply, + void *userdata, + sd_bus_error *error) { + + Unit *u = userdata; + + assert(bus); + assert(reply); + assert(u); + + return sd_bus_message_append(reply, "s", unit_file_state_to_string(unit_get_unit_file_state(u))); +} + +static int property_get_can_start( + sd_bus *bus, + const char *path, + const char *interface, + const char *property, + sd_bus_message *reply, + void *userdata, + sd_bus_error *error) { + + Unit *u = userdata; + + assert(bus); + assert(reply); + assert(u); + + return sd_bus_message_append(reply, "b", unit_can_start(u) && !u->refuse_manual_start); +} + +static int property_get_can_stop( + sd_bus *bus, + const char *path, + const char *interface, + const char *property, + sd_bus_message *reply, + void *userdata, + sd_bus_error *error) { + + Unit *u = userdata; + + assert(bus); + assert(reply); + assert(u); + + /* On the lower levels we assume that every unit we can start + * we can also stop */ + + return sd_bus_message_append(reply, "b", unit_can_start(u) && !u->refuse_manual_stop); +} + +static int property_get_can_reload( + sd_bus *bus, + const char *path, + const char *interface, + const char *property, + sd_bus_message *reply, + void *userdata, + sd_bus_error *error) { + + Unit *u = userdata; + + assert(bus); + assert(reply); + assert(u); + + return sd_bus_message_append(reply, "b", unit_can_reload(u)); +} + +static int property_get_can_isolate( + sd_bus *bus, + const char *path, + const char *interface, + const char *property, + sd_bus_message *reply, + void *userdata, + sd_bus_error *error) { + + Unit *u = userdata; + + assert(bus); + assert(reply); + assert(u); + + return sd_bus_message_append(reply, "b", unit_can_isolate(u) && !u->refuse_manual_start); +} + +static int property_get_job( + sd_bus *bus, + const char *path, + const char *interface, + const char *property, + sd_bus_message *reply, + void *userdata, + sd_bus_error *error) { + + _cleanup_free_ char *p = NULL; + Unit *u = userdata; + + assert(bus); + assert(reply); + assert(u); + + if (!u->job) + return sd_bus_message_append(reply, "(uo)", 0, "/"); + + p = job_dbus_path(u->job); + if (!p) + return -ENOMEM; + + return sd_bus_message_append(reply, "(uo)", u->job->id, p); +} + +static int property_get_need_daemon_reload( + sd_bus *bus, + const char *path, + const char *interface, + const char *property, + sd_bus_message *reply, + void *userdata, + sd_bus_error *error) { + + Unit *u = userdata; + + assert(bus); + assert(reply); + assert(u); + + return sd_bus_message_append(reply, "b", unit_need_daemon_reload(u)); +} + +static int property_get_conditions( + sd_bus *bus, + const char *path, + const char *interface, + const char *property, + sd_bus_message *reply, + void *userdata, + sd_bus_error *error) { + + const char *(*to_string)(ConditionType type) = NULL; + Condition **list = userdata, *c; + int r; + + assert(bus); + assert(reply); + assert(list); + + to_string = streq(property, "Asserts") ? assert_type_to_string : condition_type_to_string; + + r = sd_bus_message_open_container(reply, 'a', "(sbbsi)"); + if (r < 0) + return r; + + LIST_FOREACH(conditions, c, *list) { + int tristate; + + tristate = + c->result == CONDITION_UNTESTED ? 0 : + c->result == CONDITION_SUCCEEDED ? 1 : -1; + + r = sd_bus_message_append(reply, "(sbbsi)", + to_string(c->type), + c->trigger, c->negate, + c->parameter, tristate); + if (r < 0) + return r; + + } + + return sd_bus_message_close_container(reply); +} + +static int property_get_load_error( + sd_bus *bus, + const char *path, + const char *interface, + const char *property, + sd_bus_message *reply, + void *userdata, + sd_bus_error *error) { + + _cleanup_(sd_bus_error_free) sd_bus_error e = SD_BUS_ERROR_NULL; + Unit *u = userdata; + + assert(bus); + assert(reply); + assert(u); + + if (u->load_error != 0) + sd_bus_error_set_errno(&e, u->load_error); + + return sd_bus_message_append(reply, "(ss)", e.name, e.message); +} + +static int bus_verify_manage_units_async_full( + Unit *u, + const char *verb, + int capability, + const char *polkit_message, + sd_bus_message *call, + sd_bus_error *error) { + + const char *details[9] = { + "unit", u->id, + "verb", verb, + }; + + if (polkit_message) { + details[4] = "polkit.message"; + details[5] = polkit_message; + details[6] = "polkit.gettext_domain"; + details[7] = GETTEXT_PACKAGE; + } + + return bus_verify_polkit_async(call, capability, "org.freedesktop.systemd1.manage-units", details, false, UID_INVALID, &u->manager->polkit_registry, error); +} + +int bus_unit_method_start_generic( + sd_bus_message *message, + Unit *u, + JobType job_type, + bool reload_if_possible, + sd_bus_error *error) { + + const char *smode; + JobMode mode; + _cleanup_free_ char *verb = NULL; + static const char *const polkit_message_for_job[_JOB_TYPE_MAX] = { + [JOB_START] = N_("Authentication is required to start '$(unit)'."), + [JOB_STOP] = N_("Authentication is required to stop '$(unit)'."), + [JOB_RELOAD] = N_("Authentication is required to reload '$(unit)'."), + [JOB_RESTART] = N_("Authentication is required to restart '$(unit)'."), + [JOB_TRY_RESTART] = N_("Authentication is required to restart '$(unit)'."), + }; + int r; + + assert(message); + assert(u); + assert(job_type >= 0 && job_type < _JOB_TYPE_MAX); + + r = mac_selinux_unit_access_check( + u, message, + job_type_to_access_method(job_type), + error); + if (r < 0) + return r; + + r = sd_bus_message_read(message, "s", &smode); + if (r < 0) + return r; + + mode = job_mode_from_string(smode); + if (mode < 0) + return sd_bus_error_setf(error, SD_BUS_ERROR_INVALID_ARGS, "Job mode %s invalid", smode); + + if (reload_if_possible) + verb = strjoin("reload-or-", job_type_to_string(job_type), NULL); + else + verb = strdup(job_type_to_string(job_type)); + if (!verb) + return -ENOMEM; + + r = bus_verify_manage_units_async_full( + u, + verb, + CAP_SYS_ADMIN, + job_type < _JOB_TYPE_MAX ? polkit_message_for_job[job_type] : NULL, + message, + error); + if (r < 0) + return r; + if (r == 0) + return 1; /* No authorization for now, but the async polkit stuff will call us again when it has it */ + + return bus_unit_queue_job(message, u, job_type, mode, reload_if_possible, error); +} + +static int method_start(sd_bus_message *message, void *userdata, sd_bus_error *error) { + return bus_unit_method_start_generic(message, userdata, JOB_START, false, error); +} + +static int method_stop(sd_bus_message *message, void *userdata, sd_bus_error *error) { + return bus_unit_method_start_generic(message, userdata, JOB_STOP, false, error); +} + +static int method_reload(sd_bus_message *message, void *userdata, sd_bus_error *error) { + return bus_unit_method_start_generic(message, userdata, JOB_RELOAD, false, error); +} + +static int method_restart(sd_bus_message *message, void *userdata, sd_bus_error *error) { + return bus_unit_method_start_generic(message, userdata, JOB_RESTART, false, error); +} + +static int method_try_restart(sd_bus_message *message, void *userdata, sd_bus_error *error) { + return bus_unit_method_start_generic(message, userdata, JOB_TRY_RESTART, false, error); +} + +static int method_reload_or_restart(sd_bus_message *message, void *userdata, sd_bus_error *error) { + return bus_unit_method_start_generic(message, userdata, JOB_RESTART, true, error); +} + +static int method_reload_or_try_restart(sd_bus_message *message, void *userdata, sd_bus_error *error) { + return bus_unit_method_start_generic(message, userdata, JOB_TRY_RESTART, true, error); +} + +int bus_unit_method_kill(sd_bus_message *message, void *userdata, sd_bus_error *error) { + Unit *u = userdata; + const char *swho; + int32_t signo; + KillWho who; + int r; + + assert(message); + assert(u); + + r = mac_selinux_unit_access_check(u, message, "stop", error); + if (r < 0) + return r; + + r = sd_bus_message_read(message, "si", &swho, &signo); + if (r < 0) + return r; + + if (isempty(swho)) + who = KILL_ALL; + else { + who = kill_who_from_string(swho); + if (who < 0) + return sd_bus_error_setf(error, SD_BUS_ERROR_INVALID_ARGS, "Invalid who argument %s", swho); + } + + if (!SIGNAL_VALID(signo)) + return sd_bus_error_setf(error, SD_BUS_ERROR_INVALID_ARGS, "Signal number out of range."); + + r = bus_verify_manage_units_async_full( + u, + "kill", + CAP_KILL, + N_("Authentication is required to kill '$(unit)'."), + message, + error); + if (r < 0) + return r; + if (r == 0) + return 1; /* No authorization for now, but the async polkit stuff will call us again when it has it */ + + r = unit_kill(u, who, signo, error); + if (r < 0) + return r; + + return sd_bus_reply_method_return(message, NULL); +} + +int bus_unit_method_reset_failed(sd_bus_message *message, void *userdata, sd_bus_error *error) { + Unit *u = userdata; + int r; + + assert(message); + assert(u); + + r = mac_selinux_unit_access_check(u, message, "reload", error); + if (r < 0) + return r; + + r = bus_verify_manage_units_async_full( + u, + "reset-failed", + CAP_SYS_ADMIN, + N_("Authentication is required to reset the \"failed\" state of '$(unit)'."), + message, + error); + if (r < 0) + return r; + if (r == 0) + return 1; /* No authorization for now, but the async polkit stuff will call us again when it has it */ + + unit_reset_failed(u); + + return sd_bus_reply_method_return(message, NULL); +} + +int bus_unit_method_set_properties(sd_bus_message *message, void *userdata, sd_bus_error *error) { + Unit *u = userdata; + int runtime, r; + + assert(message); + assert(u); + + r = mac_selinux_unit_access_check(u, message, "start", error); + if (r < 0) + return r; + + r = sd_bus_message_read(message, "b", &runtime); + if (r < 0) + return r; + + r = bus_verify_manage_units_async_full( + u, + "set-property", + CAP_SYS_ADMIN, + N_("Authentication is required to set properties on '$(unit)'."), + message, + error); + if (r < 0) + return r; + if (r == 0) + return 1; /* No authorization for now, but the async polkit stuff will call us again when it has it */ + + r = bus_unit_set_properties(u, message, runtime ? UNIT_RUNTIME : UNIT_PERSISTENT, true, error); + if (r < 0) + return r; + + return sd_bus_reply_method_return(message, NULL); +} + +const sd_bus_vtable bus_unit_vtable[] = { + SD_BUS_VTABLE_START(0), + + SD_BUS_PROPERTY("Id", "s", NULL, offsetof(Unit, id), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("Names", "as", property_get_names, 0, SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("Following", "s", property_get_following, 0, 0), + SD_BUS_PROPERTY("Requires", "as", property_get_dependencies, offsetof(Unit, dependencies[UNIT_REQUIRES]), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("Requisite", "as", property_get_dependencies, offsetof(Unit, dependencies[UNIT_REQUISITE]), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("Wants", "as", property_get_dependencies, offsetof(Unit, dependencies[UNIT_WANTS]), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("BindsTo", "as", property_get_dependencies, offsetof(Unit, dependencies[UNIT_BINDS_TO]), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("PartOf", "as", property_get_dependencies, offsetof(Unit, dependencies[UNIT_PART_OF]), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("RequiredBy", "as", property_get_dependencies, offsetof(Unit, dependencies[UNIT_REQUIRED_BY]), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("RequisiteOf", "as", property_get_dependencies, offsetof(Unit, dependencies[UNIT_REQUISITE_OF]), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("WantedBy", "as", property_get_dependencies, offsetof(Unit, dependencies[UNIT_WANTED_BY]), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("BoundBy", "as", property_get_dependencies, offsetof(Unit, dependencies[UNIT_BOUND_BY]), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("ConsistsOf", "as", property_get_dependencies, offsetof(Unit, dependencies[UNIT_CONSISTS_OF]), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("Conflicts", "as", property_get_dependencies, offsetof(Unit, dependencies[UNIT_CONFLICTS]), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("ConflictedBy", "as", property_get_dependencies, offsetof(Unit, dependencies[UNIT_CONFLICTED_BY]), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("Before", "as", property_get_dependencies, offsetof(Unit, dependencies[UNIT_BEFORE]), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("After", "as", property_get_dependencies, offsetof(Unit, dependencies[UNIT_AFTER]), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("OnFailure", "as", property_get_dependencies, offsetof(Unit, dependencies[UNIT_ON_FAILURE]), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("Triggers", "as", property_get_dependencies, offsetof(Unit, dependencies[UNIT_TRIGGERS]), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("TriggeredBy", "as", property_get_dependencies, offsetof(Unit, dependencies[UNIT_TRIGGERED_BY]), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("PropagatesReloadTo", "as", property_get_dependencies, offsetof(Unit, dependencies[UNIT_PROPAGATES_RELOAD_TO]), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("ReloadPropagatedFrom", "as", property_get_dependencies, offsetof(Unit, dependencies[UNIT_RELOAD_PROPAGATED_FROM]), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("JoinsNamespaceOf", "as", property_get_dependencies, offsetof(Unit, dependencies[UNIT_JOINS_NAMESPACE_OF]), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("RequiresOverridable", "as", property_get_obsolete_dependencies, 0, SD_BUS_VTABLE_HIDDEN), + SD_BUS_PROPERTY("RequisiteOverridable", "as", property_get_obsolete_dependencies, 0, SD_BUS_VTABLE_HIDDEN), + SD_BUS_PROPERTY("RequiredByOverridable", "as", property_get_obsolete_dependencies, 0, SD_BUS_VTABLE_HIDDEN), + SD_BUS_PROPERTY("RequisiteOfOverridable", "as", property_get_obsolete_dependencies, 0, SD_BUS_VTABLE_HIDDEN), + SD_BUS_PROPERTY("RequiresMountsFor", "as", NULL, offsetof(Unit, requires_mounts_for), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("Documentation", "as", NULL, offsetof(Unit, documentation), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("Description", "s", property_get_description, 0, SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("LoadState", "s", property_get_load_state, offsetof(Unit, load_state), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("ActiveState", "s", property_get_active_state, 0, SD_BUS_VTABLE_PROPERTY_EMITS_CHANGE), + SD_BUS_PROPERTY("SubState", "s", property_get_sub_state, 0, SD_BUS_VTABLE_PROPERTY_EMITS_CHANGE), + SD_BUS_PROPERTY("FragmentPath", "s", NULL, offsetof(Unit, fragment_path), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("SourcePath", "s", NULL, offsetof(Unit, source_path), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("DropInPaths", "as", NULL, offsetof(Unit, dropin_paths), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("UnitFileState", "s", property_get_unit_file_state, 0, 0), + SD_BUS_PROPERTY("UnitFilePreset", "s", property_get_unit_file_preset, 0, 0), + BUS_PROPERTY_DUAL_TIMESTAMP("StateChangeTimestamp", offsetof(Unit, state_change_timestamp), SD_BUS_VTABLE_PROPERTY_EMITS_CHANGE), + BUS_PROPERTY_DUAL_TIMESTAMP("InactiveExitTimestamp", offsetof(Unit, inactive_exit_timestamp), SD_BUS_VTABLE_PROPERTY_EMITS_CHANGE), + BUS_PROPERTY_DUAL_TIMESTAMP("ActiveEnterTimestamp", offsetof(Unit, active_enter_timestamp), SD_BUS_VTABLE_PROPERTY_EMITS_CHANGE), + BUS_PROPERTY_DUAL_TIMESTAMP("ActiveExitTimestamp", offsetof(Unit, active_exit_timestamp), SD_BUS_VTABLE_PROPERTY_EMITS_CHANGE), + BUS_PROPERTY_DUAL_TIMESTAMP("InactiveEnterTimestamp", offsetof(Unit, inactive_enter_timestamp), SD_BUS_VTABLE_PROPERTY_EMITS_CHANGE), + SD_BUS_PROPERTY("CanStart", "b", property_get_can_start, 0, SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("CanStop", "b", property_get_can_stop, 0, SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("CanReload", "b", property_get_can_reload, 0, SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("CanIsolate", "b", property_get_can_isolate, 0, SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("Job", "(uo)", property_get_job, 0, SD_BUS_VTABLE_PROPERTY_EMITS_CHANGE), + SD_BUS_PROPERTY("StopWhenUnneeded", "b", bus_property_get_bool, offsetof(Unit, stop_when_unneeded), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("RefuseManualStart", "b", bus_property_get_bool, offsetof(Unit, refuse_manual_start), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("RefuseManualStop", "b", bus_property_get_bool, offsetof(Unit, refuse_manual_stop), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("AllowIsolate", "b", bus_property_get_bool, offsetof(Unit, allow_isolate), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("DefaultDependencies", "b", bus_property_get_bool, offsetof(Unit, default_dependencies), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("OnFailureJobMode", "s", property_get_job_mode, offsetof(Unit, on_failure_job_mode), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("IgnoreOnIsolate", "b", bus_property_get_bool, offsetof(Unit, ignore_on_isolate), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("NeedDaemonReload", "b", property_get_need_daemon_reload, 0, SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("JobTimeoutUSec", "t", bus_property_get_usec, offsetof(Unit, job_timeout), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("JobTimeoutAction", "s", property_get_failure_action, offsetof(Unit, job_timeout_action), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("JobTimeoutRebootArgument", "s", NULL, offsetof(Unit, job_timeout_reboot_arg), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("ConditionResult", "b", bus_property_get_bool, offsetof(Unit, condition_result), SD_BUS_VTABLE_PROPERTY_EMITS_CHANGE), + SD_BUS_PROPERTY("AssertResult", "b", bus_property_get_bool, offsetof(Unit, assert_result), SD_BUS_VTABLE_PROPERTY_EMITS_CHANGE), + BUS_PROPERTY_DUAL_TIMESTAMP("ConditionTimestamp", offsetof(Unit, condition_timestamp), SD_BUS_VTABLE_PROPERTY_EMITS_CHANGE), + BUS_PROPERTY_DUAL_TIMESTAMP("AssertTimestamp", offsetof(Unit, assert_timestamp), SD_BUS_VTABLE_PROPERTY_EMITS_CHANGE), + SD_BUS_PROPERTY("Conditions", "a(sbbsi)", property_get_conditions, offsetof(Unit, conditions), 0), + SD_BUS_PROPERTY("Asserts", "a(sbbsi)", property_get_conditions, offsetof(Unit, asserts), 0), + SD_BUS_PROPERTY("LoadError", "(ss)", property_get_load_error, 0, SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("Transient", "b", bus_property_get_bool, offsetof(Unit, transient), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("StartLimitIntervalSec", "t", bus_property_get_usec, offsetof(Unit, start_limit.interval), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("StartLimitInterval", "t", bus_property_get_usec, offsetof(Unit, start_limit.interval), SD_BUS_VTABLE_PROPERTY_CONST|SD_BUS_VTABLE_HIDDEN), /* obsolete alias name */ + SD_BUS_PROPERTY("StartLimitBurst", "u", bus_property_get_unsigned, offsetof(Unit, start_limit.burst), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("StartLimitAction", "s", property_get_failure_action, offsetof(Unit, start_limit_action), SD_BUS_VTABLE_PROPERTY_CONST), + SD_BUS_PROPERTY("RebootArgument", "s", NULL, offsetof(Unit, reboot_arg), SD_BUS_VTABLE_PROPERTY_CONST), + + SD_BUS_METHOD("Start", "s", "o", method_start, SD_BUS_VTABLE_UNPRIVILEGED), + SD_BUS_METHOD("Stop", "s", "o", method_stop, SD_BUS_VTABLE_UNPRIVILEGED), + SD_BUS_METHOD("Reload", "s", "o", method_reload, SD_BUS_VTABLE_UNPRIVILEGED), + SD_BUS_METHOD("Restart", "s", "o", method_restart, SD_BUS_VTABLE_UNPRIVILEGED), + SD_BUS_METHOD("TryRestart", "s", "o", method_try_restart, SD_BUS_VTABLE_UNPRIVILEGED), + SD_BUS_METHOD("ReloadOrRestart", "s", "o", method_reload_or_restart, SD_BUS_VTABLE_UNPRIVILEGED), + SD_BUS_METHOD("ReloadOrTryRestart", "s", "o", method_reload_or_try_restart, SD_BUS_VTABLE_UNPRIVILEGED), + SD_BUS_METHOD("Kill", "si", NULL, bus_unit_method_kill, SD_BUS_VTABLE_UNPRIVILEGED), + SD_BUS_METHOD("ResetFailed", NULL, NULL, bus_unit_method_reset_failed, SD_BUS_VTABLE_UNPRIVILEGED), + SD_BUS_METHOD("SetProperties", "ba(sv)", NULL, bus_unit_method_set_properties, SD_BUS_VTABLE_UNPRIVILEGED), + + SD_BUS_VTABLE_END +}; + +static int property_get_slice( + sd_bus *bus, + const char *path, + const char *interface, + const char *property, + sd_bus_message *reply, + void *userdata, + sd_bus_error *error) { + + Unit *u = userdata; + + assert(bus); + assert(reply); + assert(u); + + return sd_bus_message_append(reply, "s", unit_slice_name(u)); +} + +static int property_get_current_memory( + sd_bus *bus, + const char *path, + const char *interface, + const char *property, + sd_bus_message *reply, + void *userdata, + sd_bus_error *error) { + + uint64_t sz = (uint64_t) -1; + Unit *u = userdata; + int r; + + assert(bus); + assert(reply); + assert(u); + + r = unit_get_memory_current(u, &sz); + if (r < 0 && r != -ENODATA) + log_unit_warning_errno(u, r, "Failed to get memory.usage_in_bytes attribute: %m"); + + return sd_bus_message_append(reply, "t", sz); +} + +static int property_get_current_tasks( + sd_bus *bus, + const char *path, + const char *interface, + const char *property, + sd_bus_message *reply, + void *userdata, + sd_bus_error *error) { + + uint64_t cn = (uint64_t) -1; + Unit *u = userdata; + int r; + + assert(bus); + assert(reply); + assert(u); + + r = unit_get_tasks_current(u, &cn); + if (r < 0 && r != -ENODATA) + log_unit_warning_errno(u, r, "Failed to get pids.current attribute: %m"); + + return sd_bus_message_append(reply, "t", cn); +} + +static int property_get_cpu_usage( + sd_bus *bus, + const char *path, + const char *interface, + const char *property, + sd_bus_message *reply, + void *userdata, + sd_bus_error *error) { + + nsec_t ns = (nsec_t) -1; + Unit *u = userdata; + int r; + + assert(bus); + assert(reply); + assert(u); + + r = unit_get_cpu_usage(u, &ns); + if (r < 0 && r != -ENODATA) + log_unit_warning_errno(u, r, "Failed to get cpuacct.usage attribute: %m"); + + return sd_bus_message_append(reply, "t", ns); +} + +static int property_get_cgroup( + sd_bus *bus, + const char *path, + const char *interface, + const char *property, + sd_bus_message *reply, + void *userdata, + sd_bus_error *error) { + + Unit *u = userdata; + const char *t; + + assert(bus); + assert(reply); + assert(u); + + /* Three cases: a) u->cgroup_path is NULL, in which case the + * unit has no control group, which we report as the empty + * string. b) u->cgroup_path is the empty string, which + * indicates the root cgroup, which we report as "/". c) all + * other cases we report as-is. */ + + if (u->cgroup_path) + t = isempty(u->cgroup_path) ? "/" : u->cgroup_path; + else + t = ""; + + return sd_bus_message_append(reply, "s", t); +} + +static int append_process(sd_bus_message *reply, const char *p, pid_t pid, Set *pids) { + _cleanup_free_ char *buf = NULL, *cmdline = NULL; + int r; + + assert(reply); + assert(pid > 0); + + r = set_put(pids, PID_TO_PTR(pid)); + if (r == -EEXIST || r == 0) + return 0; + if (r < 0) + return r; + + if (!p) { + r = cg_pid_get_path(SYSTEMD_CGROUP_CONTROLLER, pid, &buf); + if (r == -ESRCH) + return 0; + if (r < 0) + return r; + + p = buf; + } + + (void) get_process_cmdline(pid, 0, true, &cmdline); + + return sd_bus_message_append(reply, + "(sus)", + p, + (uint32_t) pid, + cmdline); +} + +static int append_cgroup(sd_bus_message *reply, const char *p, Set *pids) { + _cleanup_closedir_ DIR *d = NULL; + _cleanup_fclose_ FILE *f = NULL; + int r; + + assert(reply); + assert(p); + + r = cg_enumerate_processes(SYSTEMD_CGROUP_CONTROLLER, p, &f); + if (r == ENOENT) + return 0; + if (r < 0) + return r; + + for (;;) { + pid_t pid; + + r = cg_read_pid(f, &pid); + if (r < 0) + return r; + if (r == 0) + break; + + if (is_kernel_thread(pid) > 0) + continue; + + r = append_process(reply, p, pid, pids); + if (r < 0) + return r; + } + + r = cg_enumerate_subgroups(SYSTEMD_CGROUP_CONTROLLER, p, &d); + if (r == -ENOENT) + return 0; + if (r < 0) + return r; + + for (;;) { + _cleanup_free_ char *g = NULL, *j = NULL; + + r = cg_read_subgroup(d, &g); + if (r < 0) + return r; + if (r == 0) + break; + + j = strjoin(p, "/", g, NULL); + if (!j) + return -ENOMEM; + + r = append_cgroup(reply, j, pids); + if (r < 0) + return r; + } + + return 0; +} + +int bus_unit_method_get_processes(sd_bus_message *message, void *userdata, sd_bus_error *error) { + _cleanup_(sd_bus_message_unrefp) sd_bus_message *reply = NULL; + _cleanup_(set_freep) Set *pids = NULL; + Unit *u = userdata; + pid_t pid; + int r; + + assert(message); + + pids = set_new(NULL); + if (!pids) + return -ENOMEM; + + r = sd_bus_message_new_method_return(message, &reply); + if (r < 0) + return r; + + r = sd_bus_message_open_container(reply, 'a', "(sus)"); + if (r < 0) + return r; + + if (u->cgroup_path) { + r = append_cgroup(reply, u->cgroup_path, pids); + if (r < 0) + return r; + } + + /* The main and control pids might live outside of the cgroup, hence fetch them separately */ + pid = unit_main_pid(u); + if (pid > 0) { + r = append_process(reply, NULL, pid, pids); + if (r < 0) + return r; + } + + pid = unit_control_pid(u); + if (pid > 0) { + r = append_process(reply, NULL, pid, pids); + if (r < 0) + return r; + } + + r = sd_bus_message_close_container(reply); + if (r < 0) + return r; + + return sd_bus_send(NULL, reply, NULL); +} + +const sd_bus_vtable bus_unit_cgroup_vtable[] = { + SD_BUS_VTABLE_START(0), + SD_BUS_PROPERTY("Slice", "s", property_get_slice, 0, 0), + SD_BUS_PROPERTY("ControlGroup", "s", property_get_cgroup, 0, 0), + SD_BUS_PROPERTY("MemoryCurrent", "t", property_get_current_memory, 0, 0), + SD_BUS_PROPERTY("CPUUsageNSec", "t", property_get_cpu_usage, 0, 0), + SD_BUS_PROPERTY("TasksCurrent", "t", property_get_current_tasks, 0, 0), + SD_BUS_METHOD("GetProcesses", NULL, "a(sus)", bus_unit_method_get_processes, SD_BUS_VTABLE_UNPRIVILEGED), + SD_BUS_VTABLE_END +}; + +static int send_new_signal(sd_bus *bus, void *userdata) { + _cleanup_(sd_bus_message_unrefp) sd_bus_message *m = NULL; + _cleanup_free_ char *p = NULL; + Unit *u = userdata; + int r; + + assert(bus); + assert(u); + + p = unit_dbus_path(u); + if (!p) + return -ENOMEM; + + r = sd_bus_message_new_signal( + bus, + &m, + "/org/freedesktop/systemd1", + "org.freedesktop.systemd1.Manager", + "UnitNew"); + if (r < 0) + return r; + + r = sd_bus_message_append(m, "so", u->id, p); + if (r < 0) + return r; + + return sd_bus_send(bus, m, NULL); +} + +static int send_changed_signal(sd_bus *bus, void *userdata) { + _cleanup_free_ char *p = NULL; + Unit *u = userdata; + int r; + + assert(bus); + assert(u); + + p = unit_dbus_path(u); + if (!p) + return -ENOMEM; + + /* Send a properties changed signal. First for the specific + * type, then for the generic unit. The clients may rely on + * this order to get atomic behavior if needed. */ + + r = sd_bus_emit_properties_changed_strv( + bus, p, + unit_dbus_interface_from_type(u->type), + NULL); + if (r < 0) + return r; + + return sd_bus_emit_properties_changed_strv( + bus, p, + "org.freedesktop.systemd1.Unit", + NULL); +} + +void bus_unit_send_change_signal(Unit *u) { + int r; + assert(u); + + if (u->in_dbus_queue) { + LIST_REMOVE(dbus_queue, u->manager->dbus_unit_queue, u); + u->in_dbus_queue = false; + } + + if (!u->id) + return; + + r = bus_foreach_bus(u->manager, NULL, u->sent_dbus_new_signal ? send_changed_signal : send_new_signal, u); + if (r < 0) + log_unit_debug_errno(u, r, "Failed to send unit change signal for %s: %m", u->id); + + u->sent_dbus_new_signal = true; +} + +static int send_removed_signal(sd_bus *bus, void *userdata) { + _cleanup_(sd_bus_message_unrefp) sd_bus_message *m = NULL; + _cleanup_free_ char *p = NULL; + Unit *u = userdata; + int r; + + assert(bus); + assert(u); + + p = unit_dbus_path(u); + if (!p) + return -ENOMEM; + + r = sd_bus_message_new_signal( + bus, + &m, + "/org/freedesktop/systemd1", + "org.freedesktop.systemd1.Manager", + "UnitRemoved"); + if (r < 0) + return r; + + r = sd_bus_message_append(m, "so", u->id, p); + if (r < 0) + return r; + + return sd_bus_send(bus, m, NULL); +} + +void bus_unit_send_removed_signal(Unit *u) { + int r; + assert(u); + + if (!u->sent_dbus_new_signal) + bus_unit_send_change_signal(u); + + if (!u->id) + return; + + r = bus_foreach_bus(u->manager, NULL, send_removed_signal, u); + if (r < 0) + log_unit_debug_errno(u, r, "Failed to send unit remove signal for %s: %m", u->id); +} + +int bus_unit_queue_job( + sd_bus_message *message, + Unit *u, + JobType type, + JobMode mode, + bool reload_if_possible, + sd_bus_error *error) { + + _cleanup_free_ char *path = NULL; + Job *j; + int r; + + assert(message); + assert(u); + assert(type >= 0 && type < _JOB_TYPE_MAX); + assert(mode >= 0 && mode < _JOB_MODE_MAX); + + r = mac_selinux_unit_access_check( + u, message, + job_type_to_access_method(type), + error); + if (r < 0) + return r; + + if (reload_if_possible && unit_can_reload(u)) { + if (type == JOB_RESTART) + type = JOB_RELOAD_OR_START; + else if (type == JOB_TRY_RESTART) + type = JOB_TRY_RELOAD; + } + + if (type == JOB_STOP && + (u->load_state == UNIT_NOT_FOUND || u->load_state == UNIT_ERROR) && + unit_active_state(u) == UNIT_INACTIVE) + return sd_bus_error_setf(error, BUS_ERROR_NO_SUCH_UNIT, "Unit %s not loaded.", u->id); + + if ((type == JOB_START && u->refuse_manual_start) || + (type == JOB_STOP && u->refuse_manual_stop) || + ((type == JOB_RESTART || type == JOB_TRY_RESTART) && (u->refuse_manual_start || u->refuse_manual_stop)) || + (type == JOB_RELOAD_OR_START && job_type_collapse(type, u) == JOB_START && u->refuse_manual_start)) + return sd_bus_error_setf(error, BUS_ERROR_ONLY_BY_DEPENDENCY, "Operation refused, unit %s may be requested by dependency only.", u->id); + + r = manager_add_job(u->manager, type, u, mode, error, &j); + if (r < 0) + return r; + + if (sd_bus_message_get_bus(message) == u->manager->api_bus) { + if (!j->clients) { + r = sd_bus_track_new(sd_bus_message_get_bus(message), &j->clients, NULL, NULL); + if (r < 0) + return r; + } + + r = sd_bus_track_add_sender(j->clients, message); + if (r < 0) + return r; + } + + path = job_dbus_path(j); + if (!path) + return -ENOMEM; + + return sd_bus_reply_method_return(message, "o", path); +} + +static int bus_unit_set_transient_property( + Unit *u, + const char *name, + sd_bus_message *message, + UnitSetPropertiesMode mode, + sd_bus_error *error) { + + int r; + + assert(u); + assert(name); + assert(message); + + if (streq(name, "Description")) { + const char *d; + + r = sd_bus_message_read(message, "s", &d); + if (r < 0) + return r; + + if (mode != UNIT_CHECK) { + r = unit_set_description(u, d); + if (r < 0) + return r; + + unit_write_drop_in_format(u, mode, name, "[Unit]\nDescription=%s", d); + } + + return 1; + + } else if (streq(name, "DefaultDependencies")) { + int b; + + r = sd_bus_message_read(message, "b", &b); + if (r < 0) + return r; + + if (mode != UNIT_CHECK) { + u->default_dependencies = b; + unit_write_drop_in_format(u, mode, name, "[Unit]\nDefaultDependencies=%s", yes_no(b)); + } + + return 1; + + } else if (streq(name, "Slice")) { + Unit *slice; + const char *s; + + if (!UNIT_HAS_CGROUP_CONTEXT(u)) + return sd_bus_error_setf(error, SD_BUS_ERROR_INVALID_ARGS, "The slice property is only available for units with control groups."); + if (u->type == UNIT_SLICE) + return sd_bus_error_setf(error, SD_BUS_ERROR_INVALID_ARGS, "Slice may not be set for slice units."); + if (unit_has_name(u, SPECIAL_INIT_SCOPE)) + return sd_bus_error_setf(error, SD_BUS_ERROR_INVALID_ARGS, "Cannot set slice for init.scope"); + + r = sd_bus_message_read(message, "s", &s); + if (r < 0) + return r; + + if (!unit_name_is_valid(s, UNIT_NAME_PLAIN)) + return sd_bus_error_setf(error, SD_BUS_ERROR_INVALID_ARGS, "Invalid unit name '%s'", s); + + /* Note that we do not dispatch the load queue here yet, as we don't want our own transient unit to be + * loaded while we are still setting it up. Or in other words, we use manager_load_unit_prepare() + * instead of manager_load_unit() on purpose, here. */ + r = manager_load_unit_prepare(u->manager, s, NULL, error, &slice); + if (r < 0) + return r; + + if (slice->type != UNIT_SLICE) + return sd_bus_error_setf(error, SD_BUS_ERROR_INVALID_ARGS, "Unit name '%s' is not a slice", s); + + if (mode != UNIT_CHECK) { + r = unit_set_slice(u, slice); + if (r < 0) + return r; + + unit_write_drop_in_private_format(u, mode, name, "Slice=%s", s); + } + + return 1; + + } else if (STR_IN_SET(name, + "Requires", "RequiresOverridable", + "Requisite", "RequisiteOverridable", + "Wants", + "BindsTo", + "Conflicts", + "Before", "After", + "OnFailure", + "PropagatesReloadTo", "ReloadPropagatedFrom", + "PartOf")) { + + UnitDependency d; + const char *other; + + if (streq(name, "RequiresOverridable")) + d = UNIT_REQUIRES; /* redirect for obsolete unit dependency type */ + else if (streq(name, "RequisiteOverridable")) + d = UNIT_REQUISITE; /* same here */ + else { + d = unit_dependency_from_string(name); + if (d < 0) + return sd_bus_error_setf(error, SD_BUS_ERROR_INVALID_ARGS, "Invalid unit dependency: %s", name); + } + + r = sd_bus_message_enter_container(message, 'a', "s"); + if (r < 0) + return r; + + while ((r = sd_bus_message_read(message, "s", &other)) > 0) { + if (!unit_name_is_valid(other, UNIT_NAME_PLAIN|UNIT_NAME_INSTANCE)) + return sd_bus_error_setf(error, SD_BUS_ERROR_INVALID_ARGS, "Invalid unit name %s", other); + + if (mode != UNIT_CHECK) { + _cleanup_free_ char *label = NULL; + + r = unit_add_dependency_by_name(u, d, other, NULL, true); + if (r < 0) + return r; + + label = strjoin(name, "-", other, NULL); + if (!label) + return -ENOMEM; + + unit_write_drop_in_format(u, mode, label, "[Unit]\n%s=%s", name, other); + } + + } + if (r < 0) + return r; + + r = sd_bus_message_exit_container(message); + if (r < 0) + return r; + + return 1; + } + + return 0; +} + +int bus_unit_set_properties( + Unit *u, + sd_bus_message *message, + UnitSetPropertiesMode mode, + bool commit, + sd_bus_error *error) { + + bool for_real = false; + unsigned n = 0; + int r; + + assert(u); + assert(message); + + /* We iterate through the array twice. First run we just check + * if all passed data is valid, second run actually applies + * it. This is to implement transaction-like behaviour without + * actually providing full transactions. */ + + r = sd_bus_message_enter_container(message, 'a', "(sv)"); + if (r < 0) + return r; + + for (;;) { + const char *name; + + r = sd_bus_message_enter_container(message, 'r', "sv"); + if (r < 0) + return r; + if (r == 0) { + if (for_real || mode == UNIT_CHECK) + break; + + /* Reached EOF. Let's try again, and this time for realz... */ + r = sd_bus_message_rewind(message, false); + if (r < 0) + return r; + + for_real = true; + continue; + } + + r = sd_bus_message_read(message, "s", &name); + if (r < 0) + return r; + + if (!UNIT_VTABLE(u)->bus_set_property) + return sd_bus_error_setf(error, SD_BUS_ERROR_PROPERTY_READ_ONLY, "Objects of this type do not support setting properties."); + + r = sd_bus_message_enter_container(message, 'v', NULL); + if (r < 0) + return r; + + r = UNIT_VTABLE(u)->bus_set_property(u, name, message, for_real ? mode : UNIT_CHECK, error); + if (r == 0 && u->transient && u->load_state == UNIT_STUB) + r = bus_unit_set_transient_property(u, name, message, for_real ? mode : UNIT_CHECK, error); + if (r < 0) + return r; + if (r == 0) + return sd_bus_error_setf(error, SD_BUS_ERROR_PROPERTY_READ_ONLY, "Cannot set property %s, or unknown property.", name); + + r = sd_bus_message_exit_container(message); + if (r < 0) + return r; + + r = sd_bus_message_exit_container(message); + if (r < 0) + return r; + + n += for_real; + } + + r = sd_bus_message_exit_container(message); + if (r < 0) + return r; + + if (commit && n > 0 && UNIT_VTABLE(u)->bus_commit_properties) + UNIT_VTABLE(u)->bus_commit_properties(u); + + return n; +} + +int bus_unit_check_load_state(Unit *u, sd_bus_error *error) { + assert(u); + + if (u->load_state == UNIT_LOADED) + return 0; + + /* Give a better description of the unit error when + * possible. Note that in the case of UNIT_MASKED, load_error + * is not set. */ + if (u->load_state == UNIT_MASKED) + return sd_bus_error_setf(error, BUS_ERROR_UNIT_MASKED, "Unit %s is masked.", u->id); + + if (u->load_state == UNIT_NOT_FOUND) + return sd_bus_error_setf(error, BUS_ERROR_NO_SUCH_UNIT, "Unit %s not found.", u->id); + + return sd_bus_error_set_errnof(error, u->load_error, "Unit %s is not loaded properly: %m.", u->id); +} diff --git a/src/grp-system/libcore/src/dbus-unit.h b/src/grp-system/libcore/src/dbus-unit.h new file mode 100644 index 0000000000..d6351f9552 --- /dev/null +++ b/src/grp-system/libcore/src/dbus-unit.h @@ -0,0 +1,41 @@ +#pragma once + +/*** + This file is part of systemd. + + Copyright 2010 Lennart Poettering + + systemd is free software; you can redistribute it and/or modify it + under the terms of the GNU Lesser General Public License as published by + the Free Software Foundation; either version 2.1 of the License, or + (at your option) any later version. + + systemd is distributed in the hope that it will be useful, but + WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + Lesser General Public License for more details. + + You should have received a copy of the GNU Lesser General Public License + along with systemd; If not, see . +***/ + +#include + +#include "core/unit.h" + +extern const sd_bus_vtable bus_unit_vtable[]; +extern const sd_bus_vtable bus_unit_cgroup_vtable[]; + +void bus_unit_send_change_signal(Unit *u); +void bus_unit_send_removed_signal(Unit *u); + +int bus_unit_method_start_generic(sd_bus_message *message, Unit *u, JobType job_type, bool reload_if_possible, sd_bus_error *error); +int bus_unit_method_kill(sd_bus_message *message, void *userdata, sd_bus_error *error); +int bus_unit_method_reset_failed(sd_bus_message *message, void *userdata, sd_bus_error *error); + +int bus_unit_queue_job(sd_bus_message *message, Unit *u, JobType type, JobMode mode, bool reload_if_possible, sd_bus_error *error); +int bus_unit_set_properties(Unit *u, sd_bus_message *message, UnitSetPropertiesMode mode, bool commit, sd_bus_error *error); +int bus_unit_method_set_properties(sd_bus_message *message, void *userdata, sd_bus_error *error); +int bus_unit_method_get_processes(sd_bus_message *message, void *userdata, sd_bus_error *error); + +int bus_unit_check_load_state(Unit *u, sd_bus_error *error); diff --git a/src/grp-system/libcore/src/dbus.c b/src/grp-system/libcore/src/dbus.c new file mode 100644 index 0000000000..4485b38e2e --- /dev/null +++ b/src/grp-system/libcore/src/dbus.c @@ -0,0 +1,1244 @@ +/*** + This file is part of systemd. + + Copyright 2010 Lennart Poettering + + systemd is free software; you can redistribute it and/or modify it + under the terms of the GNU Lesser General Public License as published by + the Free Software Foundation; either version 2.1 of the License, or + (at your option) any later version. + + systemd is distributed in the hope that it will be useful, but + WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + Lesser General Public License for more details. + + You should have received a copy of the GNU Lesser General Public License + along with systemd; If not, see . +***/ + +#include +#include +#include + +#include + +#include "core/dbus-manager.h" +#include "sd-bus/bus-common-errors.h" +#include "sd-bus/bus-error.h" +#include "sd-bus/bus-internal.h" +#include "sd-bus/bus-util.h" +#include "systemd-basic/alloc-util.h" +#include "systemd-basic/fd-util.h" +#include "systemd-basic/log.h" +#include "systemd-basic/missing.h" +#include "systemd-basic/mkdir.h" +#include "systemd-basic/special.h" +#include "systemd-basic/string-util.h" +#include "systemd-basic/strv.h" +#include "systemd-basic/strxcpyx.h" +#include "systemd-basic/user-util.h" + +#include "dbus-cgroup.h" +#include "dbus-execute.h" +#include "dbus-job.h" +#include "dbus-kill.h" +#include "dbus-unit.h" +#include "dbus.h" +#include "selinux-access.h" + +#define CONNECTIONS_MAX 4096 + +static void destroy_bus(Manager *m, sd_bus **bus); + +int bus_send_queued_message(Manager *m) { + int r; + + assert(m); + + if (!m->queued_message) + return 0; + + /* If we cannot get rid of this message we won't dispatch any + * D-Bus messages, so that we won't end up wanting to queue + * another message. */ + + r = sd_bus_send(NULL, m->queued_message, NULL); + if (r < 0) + log_warning_errno(r, "Failed to send queued message: %m"); + + m->queued_message = sd_bus_message_unref(m->queued_message); + + return 0; +} + +int bus_forward_agent_released(Manager *m, const char *path) { + int r; + + assert(m); + assert(path); + + if (!MANAGER_IS_SYSTEM(m)) + return 0; + + if (!m->system_bus) + return 0; + + /* If we are running a system instance we forward the agent message on the system bus, so that the user + * instances get notified about this, too */ + + r = sd_bus_emit_signal(m->system_bus, + "/org/freedesktop/systemd1/agent", + "org.freedesktop.systemd1.Agent", + "Released", + "s", path); + if (r < 0) + return log_warning_errno(r, "Failed to propagate agent release message: %m"); + + return 1; +} + +static int signal_agent_released(sd_bus_message *message, void *userdata, sd_bus_error *error) { + _cleanup_(sd_bus_creds_unrefp) sd_bus_creds *creds = NULL; + Manager *m = userdata; + const char *cgroup; + uid_t sender_uid; + int r; + + assert(message); + assert(m); + + /* only accept org.freedesktop.systemd1.Agent from UID=0 */ + r = sd_bus_query_sender_creds(message, SD_BUS_CREDS_EUID, &creds); + if (r < 0) + return r; + + r = sd_bus_creds_get_euid(creds, &sender_uid); + if (r < 0 || sender_uid != 0) + return 0; + + /* parse 'cgroup-empty' notification */ + r = sd_bus_message_read(message, "s", &cgroup); + if (r < 0) { + bus_log_parse_error(r); + return 0; + } + + manager_notify_cgroup_empty(m, cgroup); + return 0; +} + +static int signal_disconnected(sd_bus_message *message, void *userdata, sd_bus_error *error) { + Manager *m = userdata; + sd_bus *bus; + + assert(message); + assert(m); + assert_se(bus = sd_bus_message_get_bus(message)); + + if (bus == m->api_bus) + destroy_bus(m, &m->api_bus); + if (bus == m->system_bus) + destroy_bus(m, &m->system_bus); + if (set_remove(m->private_buses, bus)) { + log_debug("Got disconnect on private connection."); + destroy_bus(m, &bus); + } + + return 0; +} + +static int signal_activation_request(sd_bus_message *message, void *userdata, sd_bus_error *ret_error) { + _cleanup_(sd_bus_error_free) sd_bus_error error = SD_BUS_ERROR_NULL; + _cleanup_(sd_bus_message_unrefp) sd_bus_message *reply = NULL; + Manager *m = userdata; + const char *name; + Unit *u; + int r; + + assert(message); + assert(m); + + r = sd_bus_message_read(message, "s", &name); + if (r < 0) { + bus_log_parse_error(r); + return 0; + } + + if (manager_unit_inactive_or_pending(m, SPECIAL_DBUS_SERVICE) || + manager_unit_inactive_or_pending(m, SPECIAL_DBUS_SOCKET)) { + r = sd_bus_error_setf(&error, BUS_ERROR_SHUTTING_DOWN, "Refusing activation, D-Bus is shutting down."); + goto failed; + } + + r = manager_load_unit(m, name, NULL, &error, &u); + if (r < 0) + goto failed; + + if (u->refuse_manual_start) { + r = sd_bus_error_setf(&error, BUS_ERROR_ONLY_BY_DEPENDENCY, "Operation refused, %s may be requested by dependency only.", u->id); + goto failed; + } + + r = manager_add_job(m, JOB_START, u, JOB_REPLACE, &error, NULL); + if (r < 0) + goto failed; + + /* Successfully queued, that's it for us */ + return 0; + +failed: + if (!sd_bus_error_is_set(&error)) + sd_bus_error_set_errno(&error, r); + + log_debug("D-Bus activation failed for %s: %s", name, bus_error_message(&error, r)); + + r = sd_bus_message_new_signal(sd_bus_message_get_bus(message), &reply, "/org/freedesktop/systemd1", "org.freedesktop.systemd1.Activator", "ActivationFailure"); + if (r < 0) { + bus_log_create_error(r); + return 0; + } + + r = sd_bus_message_append(reply, "sss", name, error.name, error.message); + if (r < 0) { + bus_log_create_error(r); + return 0; + } + + r = sd_bus_send_to(NULL, reply, "org.freedesktop.DBus", NULL); + if (r < 0) + return log_error_errno(r, "Failed to respond with to bus activation request: %m"); + + return 0; +} + +#ifdef HAVE_SELINUX +static int mac_selinux_filter(sd_bus_message *message, void *userdata, sd_bus_error *error) { + Manager *m = userdata; + const char *verb, *path; + Unit *u = NULL; + Job *j; + int r; + + assert(message); + + /* Our own method calls are all protected individually with + * selinux checks, but the built-in interfaces need to be + * protected too. */ + + if (sd_bus_message_is_method_call(message, "org.freedesktop.DBus.Properties", "Set")) + verb = "reload"; + else if (sd_bus_message_is_method_call(message, "org.freedesktop.DBus.Introspectable", NULL) || + sd_bus_message_is_method_call(message, "org.freedesktop.DBus.Properties", NULL) || + sd_bus_message_is_method_call(message, "org.freedesktop.DBus.ObjectManager", NULL) || + sd_bus_message_is_method_call(message, "org.freedesktop.DBus.Peer", NULL)) + verb = "status"; + else + return 0; + + path = sd_bus_message_get_path(message); + + if (object_path_startswith("/org/freedesktop/systemd1", path)) { + + r = mac_selinux_access_check(message, verb, error); + if (r < 0) + return r; + + return 0; + } + + if (streq_ptr(path, "/org/freedesktop/systemd1/unit/self")) { + _cleanup_(sd_bus_creds_unrefp) sd_bus_creds *creds = NULL; + pid_t pid; + + r = sd_bus_query_sender_creds(message, SD_BUS_CREDS_PID, &creds); + if (r < 0) + return 0; + + r = sd_bus_creds_get_pid(creds, &pid); + if (r < 0) + return 0; + + u = manager_get_unit_by_pid(m, pid); + } else { + r = manager_get_job_from_dbus_path(m, path, &j); + if (r >= 0) + u = j->unit; + else + manager_load_unit_from_dbus_path(m, path, NULL, &u); + } + + if (!u) + return 0; + + r = mac_selinux_unit_access_check(u, message, verb, error); + if (r < 0) + return r; + + return 0; +} +#endif + +static int bus_job_find(sd_bus *bus, const char *path, const char *interface, void *userdata, void **found, sd_bus_error *error) { + Manager *m = userdata; + Job *j; + int r; + + assert(bus); + assert(path); + assert(interface); + assert(found); + assert(m); + + r = manager_get_job_from_dbus_path(m, path, &j); + if (r < 0) + return 0; + + *found = j; + return 1; +} + +static int find_unit(Manager *m, sd_bus *bus, const char *path, Unit **unit, sd_bus_error *error) { + Unit *u; + int r; + + assert(m); + assert(bus); + assert(path); + + if (streq_ptr(path, "/org/freedesktop/systemd1/unit/self")) { + _cleanup_(sd_bus_creds_unrefp) sd_bus_creds *creds = NULL; + sd_bus_message *message; + pid_t pid; + + message = sd_bus_get_current_message(bus); + if (!message) + return 0; + + r = sd_bus_query_sender_creds(message, SD_BUS_CREDS_PID, &creds); + if (r < 0) + return r; + + r = sd_bus_creds_get_pid(creds, &pid); + if (r < 0) + return r; + + u = manager_get_unit_by_pid(m, pid); + } else { + r = manager_load_unit_from_dbus_path(m, path, error, &u); + if (r < 0) + return 0; + } + + if (!u) + return 0; + + *unit = u; + return 1; +} + +static int bus_unit_find(sd_bus *bus, const char *path, const char *interface, void *userdata, void **found, sd_bus_error *error) { + Manager *m = userdata; + + assert(bus); + assert(path); + assert(interface); + assert(found); + assert(m); + + return find_unit(m, bus, path, (Unit**) found, error); +} + +static int bus_unit_interface_find(sd_bus *bus, const char *path, const char *interface, void *userdata, void **found, sd_bus_error *error) { + Manager *m = userdata; + Unit *u; + int r; + + assert(bus); + assert(path); + assert(interface); + assert(found); + assert(m); + + r = find_unit(m, bus, path, &u, error); + if (r <= 0) + return r; + + if (!streq_ptr(interface, unit_dbus_interface_from_type(u->type))) + return 0; + + *found = u; + return 1; +} + +static int bus_unit_cgroup_find(sd_bus *bus, const char *path, const char *interface, void *userdata, void **found, sd_bus_error *error) { + Manager *m = userdata; + Unit *u; + int r; + + assert(bus); + assert(path); + assert(interface); + assert(found); + assert(m); + + r = find_unit(m, bus, path, &u, error); + if (r <= 0) + return r; + + if (!streq_ptr(interface, unit_dbus_interface_from_type(u->type))) + return 0; + + if (!UNIT_HAS_CGROUP_CONTEXT(u)) + return 0; + + *found = u; + return 1; +} + +static int bus_cgroup_context_find(sd_bus *bus, const char *path, const char *interface, void *userdata, void **found, sd_bus_error *error) { + Manager *m = userdata; + CGroupContext *c; + Unit *u; + int r; + + assert(bus); + assert(path); + assert(interface); + assert(found); + assert(m); + + r = find_unit(m, bus, path, &u, error); + if (r <= 0) + return r; + + if (!streq_ptr(interface, unit_dbus_interface_from_type(u->type))) + return 0; + + c = unit_get_cgroup_context(u); + if (!c) + return 0; + + *found = c; + return 1; +} + +static int bus_exec_context_find(sd_bus *bus, const char *path, const char *interface, void *userdata, void **found, sd_bus_error *error) { + Manager *m = userdata; + ExecContext *c; + Unit *u; + int r; + + assert(bus); + assert(path); + assert(interface); + assert(found); + assert(m); + + r = find_unit(m, bus, path, &u, error); + if (r <= 0) + return r; + + if (!streq_ptr(interface, unit_dbus_interface_from_type(u->type))) + return 0; + + c = unit_get_exec_context(u); + if (!c) + return 0; + + *found = c; + return 1; +} + +static int bus_kill_context_find(sd_bus *bus, const char *path, const char *interface, void *userdata, void **found, sd_bus_error *error) { + Manager *m = userdata; + KillContext *c; + Unit *u; + int r; + + assert(bus); + assert(path); + assert(interface); + assert(found); + assert(m); + + r = find_unit(m, bus, path, &u, error); + if (r <= 0) + return r; + + if (!streq_ptr(interface, unit_dbus_interface_from_type(u->type))) + return 0; + + c = unit_get_kill_context(u); + if (!c) + return 0; + + *found = c; + return 1; +} + +static int bus_job_enumerate(sd_bus *bus, const char *path, void *userdata, char ***nodes, sd_bus_error *error) { + _cleanup_free_ char **l = NULL; + Manager *m = userdata; + unsigned k = 0; + Iterator i; + Job *j; + + l = new0(char*, hashmap_size(m->jobs)+1); + if (!l) + return -ENOMEM; + + HASHMAP_FOREACH(j, m->jobs, i) { + l[k] = job_dbus_path(j); + if (!l[k]) + return -ENOMEM; + + k++; + } + + assert(hashmap_size(m->jobs) == k); + + *nodes = l; + l = NULL; + + return k; +} + +static int bus_unit_enumerate(sd_bus *bus, const char *path, void *userdata, char ***nodes, sd_bus_error *error) { + _cleanup_free_ char **l = NULL; + Manager *m = userdata; + unsigned k = 0; + Iterator i; + Unit *u; + + l = new0(char*, hashmap_size(m->units)+1); + if (!l) + return -ENOMEM; + + HASHMAP_FOREACH(u, m->units, i) { + l[k] = unit_dbus_path(u); + if (!l[k]) + return -ENOMEM; + + k++; + } + + *nodes = l; + l = NULL; + + return k; +} + +static int bus_setup_api_vtables(Manager *m, sd_bus *bus) { + UnitType t; + int r; + + assert(m); + assert(bus); + +#ifdef HAVE_SELINUX + r = sd_bus_add_filter(bus, NULL, mac_selinux_filter, m); + if (r < 0) + return log_error_errno(r, "Failed to add SELinux access filter: %m"); +#endif + + r = sd_bus_add_object_vtable(bus, NULL, "/org/freedesktop/systemd1", "org.freedesktop.systemd1.Manager", bus_manager_vtable, m); + if (r < 0) + return log_error_errno(r, "Failed to register Manager vtable: %m"); + + r = sd_bus_add_fallback_vtable(bus, NULL, "/org/freedesktop/systemd1/job", "org.freedesktop.systemd1.Job", bus_job_vtable, bus_job_find, m); + if (r < 0) + return log_error_errno(r, "Failed to register Job vtable: %m"); + + r = sd_bus_add_node_enumerator(bus, NULL, "/org/freedesktop/systemd1/job", bus_job_enumerate, m); + if (r < 0) + return log_error_errno(r, "Failed to add job enumerator: %m"); + + r = sd_bus_add_fallback_vtable(bus, NULL, "/org/freedesktop/systemd1/unit", "org.freedesktop.systemd1.Unit", bus_unit_vtable, bus_unit_find, m); + if (r < 0) + return log_error_errno(r, "Failed to register Unit vtable: %m"); + + r = sd_bus_add_node_enumerator(bus, NULL, "/org/freedesktop/systemd1/unit", bus_unit_enumerate, m); + if (r < 0) + return log_error_errno(r, "Failed to add job enumerator: %m"); + + for (t = 0; t < _UNIT_TYPE_MAX; t++) { + const char *interface; + + assert_se(interface = unit_dbus_interface_from_type(t)); + + r = sd_bus_add_fallback_vtable(bus, NULL, "/org/freedesktop/systemd1/unit", interface, unit_vtable[t]->bus_vtable, bus_unit_interface_find, m); + if (r < 0) + return log_error_errno(r, "Failed to register type specific vtable for %s: %m", interface); + + if (unit_vtable[t]->cgroup_context_offset > 0) { + r = sd_bus_add_fallback_vtable(bus, NULL, "/org/freedesktop/systemd1/unit", interface, bus_unit_cgroup_vtable, bus_unit_cgroup_find, m); + if (r < 0) + return log_error_errno(r, "Failed to register control group unit vtable for %s: %m", interface); + + r = sd_bus_add_fallback_vtable(bus, NULL, "/org/freedesktop/systemd1/unit", interface, bus_cgroup_vtable, bus_cgroup_context_find, m); + if (r < 0) + return log_error_errno(r, "Failed to register control group vtable for %s: %m", interface); + } + + if (unit_vtable[t]->exec_context_offset > 0) { + r = sd_bus_add_fallback_vtable(bus, NULL, "/org/freedesktop/systemd1/unit", interface, bus_exec_vtable, bus_exec_context_find, m); + if (r < 0) + return log_error_errno(r, "Failed to register execute vtable for %s: %m", interface); + } + + if (unit_vtable[t]->kill_context_offset > 0) { + r = sd_bus_add_fallback_vtable(bus, NULL, "/org/freedesktop/systemd1/unit", interface, bus_kill_vtable, bus_kill_context_find, m); + if (r < 0) + return log_error_errno(r, "Failed to register kill vtable for %s: %m", interface); + } + } + + return 0; +} + +static int bus_setup_disconnected_match(Manager *m, sd_bus *bus) { + int r; + + assert(m); + assert(bus); + + r = sd_bus_add_match( + bus, + NULL, + "sender='org.freedesktop.DBus.Local'," + "type='signal'," + "path='/org/freedesktop/DBus/Local'," + "interface='org.freedesktop.DBus.Local'," + "member='Disconnected'", + signal_disconnected, m); + + if (r < 0) + return log_error_errno(r, "Failed to register match for Disconnected message: %m"); + + return 0; +} + +static int bus_on_connection(sd_event_source *s, int fd, uint32_t revents, void *userdata) { + _cleanup_(sd_bus_unrefp) sd_bus *bus = NULL; + _cleanup_close_ int nfd = -1; + Manager *m = userdata; + sd_id128_t id; + int r; + + assert(s); + assert(m); + + nfd = accept4(fd, NULL, NULL, SOCK_NONBLOCK|SOCK_CLOEXEC); + if (nfd < 0) { + log_warning_errno(errno, "Failed to accept private connection, ignoring: %m"); + return 0; + } + + if (set_size(m->private_buses) >= CONNECTIONS_MAX) { + log_warning("Too many concurrent connections, refusing"); + return 0; + } + + r = set_ensure_allocated(&m->private_buses, NULL); + if (r < 0) { + log_oom(); + return 0; + } + + r = sd_bus_new(&bus); + if (r < 0) { + log_warning_errno(r, "Failed to allocate new private connection bus: %m"); + return 0; + } + + r = sd_bus_set_fd(bus, nfd, nfd); + if (r < 0) { + log_warning_errno(r, "Failed to set fd on new connection bus: %m"); + return 0; + } + + nfd = -1; + + r = bus_check_peercred(bus); + if (r < 0) { + log_warning_errno(r, "Incoming private connection from unprivileged client, refusing: %m"); + return 0; + } + + assert_se(sd_id128_randomize(&id) >= 0); + + r = sd_bus_set_server(bus, 1, id); + if (r < 0) { + log_warning_errno(r, "Failed to enable server support for new connection bus: %m"); + return 0; + } + + r = sd_bus_negotiate_creds(bus, 1, + SD_BUS_CREDS_PID|SD_BUS_CREDS_UID| + SD_BUS_CREDS_EUID|SD_BUS_CREDS_EFFECTIVE_CAPS| + SD_BUS_CREDS_SELINUX_CONTEXT); + if (r < 0) { + log_warning_errno(r, "Failed to enable credentials for new connection: %m"); + return 0; + } + + r = sd_bus_start(bus); + if (r < 0) { + log_warning_errno(r, "Failed to start new connection bus: %m"); + return 0; + } + + r = sd_bus_attach_event(bus, m->event, SD_EVENT_PRIORITY_NORMAL); + if (r < 0) { + log_warning_errno(r, "Failed to attach new connection bus to event loop: %m"); + return 0; + } + + r = bus_setup_disconnected_match(m, bus); + if (r < 0) + return 0; + + r = bus_setup_api_vtables(m, bus); + if (r < 0) { + log_warning_errno(r, "Failed to set up API vtables on new connection bus: %m"); + return 0; + } + + r = set_put(m->private_buses, bus); + if (r < 0) { + log_warning_errno(r, "Failed to add new connection bus to set: %m"); + return 0; + } + + bus = NULL; + + log_debug("Accepted new private connection."); + + return 0; +} + +int manager_sync_bus_names(Manager *m, sd_bus *bus) { + _cleanup_strv_free_ char **names = NULL; + const char *name; + Iterator i; + Unit *u; + int r; + + assert(m); + assert(bus); + + r = sd_bus_list_names(bus, &names, NULL); + if (r < 0) + return log_error_errno(r, "Failed to get initial list of names: %m"); + + /* We have to synchronize the current bus names with the + * list of active services. To do this, walk the list of + * all units with bus names. */ + HASHMAP_FOREACH_KEY(u, name, m->watch_bus, i) { + Service *s = SERVICE(u); + + assert(s); + + if (!streq_ptr(s->bus_name, name)) { + log_unit_warning(u, "Bus name has changed from %s → %s, ignoring.", s->bus_name, name); + continue; + } + + /* Check if a service's bus name is in the list of currently + * active names */ + if (strv_contains(names, name)) { + _cleanup_(sd_bus_creds_unrefp) sd_bus_creds *creds = NULL; + const char *unique; + + /* If it is, determine its current owner */ + r = sd_bus_get_name_creds(bus, name, SD_BUS_CREDS_UNIQUE_NAME, &creds); + if (r < 0) { + log_error_errno(r, "Failed to get bus name owner %s: %m", name); + continue; + } + + r = sd_bus_creds_get_unique_name(creds, &unique); + if (r < 0) { + log_error_errno(r, "Failed to get unique name for %s: %m", name); + continue; + } + + /* Now, let's compare that to the previous bus owner, and + * if it's still the same, all is fine, so just don't + * bother the service. Otherwise, the name has apparently + * changed, so synthesize a name owner changed signal. */ + + if (!streq_ptr(unique, s->bus_name_owner)) + UNIT_VTABLE(u)->bus_name_owner_change(u, name, s->bus_name_owner, unique); + } else { + /* So, the name we're watching is not on the bus. + * This either means it simply hasn't appeared yet, + * or it was lost during the daemon reload. + * Check if the service has a stored name owner, + * and synthesize a name loss signal in this case. */ + + if (s->bus_name_owner) + UNIT_VTABLE(u)->bus_name_owner_change(u, name, s->bus_name_owner, NULL); + } + } + + return 0; +} + +static int bus_setup_api(Manager *m, sd_bus *bus) { + Iterator i; + char *name; + Unit *u; + int r; + + assert(m); + assert(bus); + + /* Let's make sure we have enough credential bits so that we can make security and selinux decisions */ + r = sd_bus_negotiate_creds(bus, 1, + SD_BUS_CREDS_PID|SD_BUS_CREDS_UID| + SD_BUS_CREDS_EUID|SD_BUS_CREDS_EFFECTIVE_CAPS| + SD_BUS_CREDS_SELINUX_CONTEXT); + if (r < 0) + log_warning_errno(r, "Failed to enable credential passing, ignoring: %m"); + + r = bus_setup_api_vtables(m, bus); + if (r < 0) + return r; + + HASHMAP_FOREACH_KEY(u, name, m->watch_bus, i) { + r = unit_install_bus_match(u, bus, name); + if (r < 0) + log_error_errno(r, "Failed to subscribe to NameOwnerChanged signal for '%s': %m", name); + } + + r = sd_bus_add_match( + bus, + NULL, + "type='signal'," + "sender='org.freedesktop.DBus'," + "path='/org/freedesktop/DBus'," + "interface='org.freedesktop.systemd1.Activator'," + "member='ActivationRequest'", + signal_activation_request, m); + if (r < 0) + log_warning_errno(r, "Failed to subscribe to activation signal: %m"); + + /* Allow replacing of our name, to ease implementation of + * reexecution, where we keep the old connection open until + * after the new connection is set up and the name installed + * to allow clients to synchronously wait for reexecution to + * finish */ + r = sd_bus_request_name(bus,"org.freedesktop.systemd1", SD_BUS_NAME_REPLACE_EXISTING|SD_BUS_NAME_ALLOW_REPLACEMENT); + if (r < 0) + return log_error_errno(r, "Failed to register name: %m"); + + r = manager_sync_bus_names(m, bus); + if (r < 0) + return r; + + log_debug("Successfully connected to API bus."); + return 0; +} + +static int bus_init_api(Manager *m) { + _cleanup_(sd_bus_unrefp) sd_bus *bus = NULL; + int r; + + if (m->api_bus) + return 0; + + /* The API and system bus is the same if we are running in system mode */ + if (MANAGER_IS_SYSTEM(m) && m->system_bus) + bus = sd_bus_ref(m->system_bus); + else { + if (MANAGER_IS_SYSTEM(m)) + r = sd_bus_open_system(&bus); + else + r = sd_bus_open_user(&bus); + + if (r < 0) { + log_debug("Failed to connect to API bus, retrying later..."); + return 0; + } + + r = sd_bus_attach_event(bus, m->event, SD_EVENT_PRIORITY_NORMAL); + if (r < 0) { + log_error_errno(r, "Failed to attach API bus to event loop: %m"); + return 0; + } + + r = bus_setup_disconnected_match(m, bus); + if (r < 0) + return 0; + } + + r = bus_setup_api(m, bus); + if (r < 0) { + log_error_errno(r, "Failed to set up API bus: %m"); + return 0; + } + + m->api_bus = bus; + bus = NULL; + + return 0; +} + +static int bus_setup_system(Manager *m, sd_bus *bus) { + int r; + + assert(m); + assert(bus); + + /* if we are a user instance we get the Released message via the system bus */ + if (MANAGER_IS_USER(m)) { + r = sd_bus_add_match( + bus, + NULL, + "type='signal'," + "interface='org.freedesktop.systemd1.Agent'," + "member='Released'," + "path='/org/freedesktop/systemd1/agent'", + signal_agent_released, m); + if (r < 0) + log_warning_errno(r, "Failed to register Released match on system bus: %m"); + } + + log_debug("Successfully connected to system bus."); + return 0; +} + +static int bus_init_system(Manager *m) { + _cleanup_(sd_bus_unrefp) sd_bus *bus = NULL; + int r; + + if (m->system_bus) + return 0; + + /* The API and system bus is the same if we are running in system mode */ + if (MANAGER_IS_SYSTEM(m) && m->api_bus) { + m->system_bus = sd_bus_ref(m->api_bus); + return 0; + } + + r = sd_bus_open_system(&bus); + if (r < 0) { + log_debug("Failed to connect to system bus, retrying later..."); + return 0; + } + + r = bus_setup_disconnected_match(m, bus); + if (r < 0) + return 0; + + r = sd_bus_attach_event(bus, m->event, SD_EVENT_PRIORITY_NORMAL); + if (r < 0) { + log_error_errno(r, "Failed to attach system bus to event loop: %m"); + return 0; + } + + r = bus_setup_system(m, bus); + if (r < 0) { + log_error_errno(r, "Failed to set up system bus: %m"); + return 0; + } + + m->system_bus = bus; + bus = NULL; + + return 0; +} + +static int bus_init_private(Manager *m) { + _cleanup_close_ int fd = -1; + union sockaddr_union sa = { + .un.sun_family = AF_UNIX + }; + sd_event_source *s; + socklen_t salen; + int r; + + assert(m); + + if (m->private_listen_fd >= 0) + return 0; + + /* We don't need the private socket if we have kdbus */ + if (m->kdbus_fd >= 0) + return 0; + + if (MANAGER_IS_SYSTEM(m)) { + + /* We want the private bus only when running as init */ + if (getpid() != 1) + return 0; + + strcpy(sa.un.sun_path, "/run/systemd/private"); + salen = SOCKADDR_UN_LEN(sa.un); + } else { + size_t left = sizeof(sa.un.sun_path); + char *p = sa.un.sun_path; + const char *e; + + e = secure_getenv("XDG_RUNTIME_DIR"); + if (!e) { + log_error("Failed to determine XDG_RUNTIME_DIR"); + return -EHOSTDOWN; + } + + left = strpcpy(&p, left, e); + left = strpcpy(&p, left, "/systemd/private"); + + salen = sizeof(sa.un) - left; + } + + (void) mkdir_parents_label(sa.un.sun_path, 0755); + (void) unlink(sa.un.sun_path); + + fd = socket(AF_UNIX, SOCK_STREAM|SOCK_CLOEXEC|SOCK_NONBLOCK, 0); + if (fd < 0) + return log_error_errno(errno, "Failed to allocate private socket: %m"); + + r = bind(fd, &sa.sa, salen); + if (r < 0) + return log_error_errno(errno, "Failed to bind private socket: %m"); + + r = listen(fd, SOMAXCONN); + if (r < 0) + return log_error_errno(errno, "Failed to make private socket listening: %m"); + + r = sd_event_add_io(m->event, &s, fd, EPOLLIN, bus_on_connection, m); + if (r < 0) + return log_error_errno(r, "Failed to allocate event source: %m"); + + (void) sd_event_source_set_description(s, "bus-connection"); + + m->private_listen_fd = fd; + m->private_listen_event_source = s; + fd = -1; + + log_debug("Successfully created private D-Bus server."); + + return 0; +} + +int bus_init(Manager *m, bool try_bus_connect) { + int r; + + if (try_bus_connect) { + r = bus_init_system(m); + if (r < 0) + return r; + + r = bus_init_api(m); + if (r < 0) + return r; + } + + r = bus_init_private(m); + if (r < 0) + return r; + + return 0; +} + +static void destroy_bus(Manager *m, sd_bus **bus) { + Iterator i; + Job *j; + + assert(m); + assert(bus); + + if (!*bus) + return; + + /* Get rid of tracked clients on this bus */ + if (m->subscribed && sd_bus_track_get_bus(m->subscribed) == *bus) + m->subscribed = sd_bus_track_unref(m->subscribed); + + HASHMAP_FOREACH(j, m->jobs, i) + if (j->clients && sd_bus_track_get_bus(j->clients) == *bus) + j->clients = sd_bus_track_unref(j->clients); + + /* Get rid of queued message on this bus */ + if (m->queued_message && sd_bus_message_get_bus(m->queued_message) == *bus) + m->queued_message = sd_bus_message_unref(m->queued_message); + + /* Possibly flush unwritten data, but only if we are + * unprivileged, since we don't want to sync here */ + if (!MANAGER_IS_SYSTEM(m)) + sd_bus_flush(*bus); + + /* And destroy the object */ + sd_bus_close(*bus); + *bus = sd_bus_unref(*bus); +} + +void bus_done(Manager *m) { + sd_bus *b; + + assert(m); + + if (m->api_bus) + destroy_bus(m, &m->api_bus); + if (m->system_bus) + destroy_bus(m, &m->system_bus); + while ((b = set_steal_first(m->private_buses))) + destroy_bus(m, &b); + + m->private_buses = set_free(m->private_buses); + + m->subscribed = sd_bus_track_unref(m->subscribed); + m->deserialized_subscribed = strv_free(m->deserialized_subscribed); + + if (m->private_listen_event_source) + m->private_listen_event_source = sd_event_source_unref(m->private_listen_event_source); + + m->private_listen_fd = safe_close(m->private_listen_fd); + + bus_verify_polkit_async_registry_free(m->polkit_registry); +} + +int bus_fdset_add_all(Manager *m, FDSet *fds) { + Iterator i; + sd_bus *b; + int fd; + + assert(m); + assert(fds); + + /* When we are about to reexecute we add all D-Bus fds to the + * set to pass over to the newly executed systemd. They won't + * be used there however, except thatt they are closed at the + * very end of deserialization, those making it possible for + * clients to synchronously wait for systemd to reexec by + * simply waiting for disconnection */ + + if (m->api_bus) { + fd = sd_bus_get_fd(m->api_bus); + if (fd >= 0) { + fd = fdset_put_dup(fds, fd); + if (fd < 0) + return fd; + } + } + + SET_FOREACH(b, m->private_buses, i) { + fd = sd_bus_get_fd(b); + if (fd >= 0) { + fd = fdset_put_dup(fds, fd); + if (fd < 0) + return fd; + } + } + + /* We don't offer any APIs on the system bus (well, unless it + * is the same as the API bus) hence we don't bother with it + * here */ + + return 0; +} + +int bus_foreach_bus( + Manager *m, + sd_bus_track *subscribed2, + int (*send_message)(sd_bus *bus, void *userdata), + void *userdata) { + + Iterator i; + sd_bus *b; + int r, ret = 0; + + /* Send to all direct buses, unconditionally */ + SET_FOREACH(b, m->private_buses, i) { + r = send_message(b, userdata); + if (r < 0) + ret = r; + } + + /* Send to API bus, but only if somebody is subscribed */ + if (sd_bus_track_count(m->subscribed) > 0 || + sd_bus_track_count(subscribed2) > 0) { + r = send_message(m->api_bus, userdata); + if (r < 0) + ret = r; + } + + return ret; +} + +void bus_track_serialize(sd_bus_track *t, FILE *f) { + const char *n; + + assert(f); + + for (n = sd_bus_track_first(t); n; n = sd_bus_track_next(t)) + fprintf(f, "subscribed=%s\n", n); +} + +int bus_track_deserialize_item(char ***l, const char *line) { + const char *e; + int r; + + assert(l); + assert(line); + + e = startswith(line, "subscribed="); + if (!e) + return 0; + + r = strv_extend(l, e); + if (r < 0) + return r; + + return 1; +} + +int bus_track_coldplug(Manager *m, sd_bus_track **t, char ***l) { + int r = 0; + + assert(m); + assert(t); + assert(l); + + if (!strv_isempty(*l) && m->api_bus) { + char **i; + + if (!*t) { + r = sd_bus_track_new(m->api_bus, t, NULL, NULL); + if (r < 0) + return r; + } + + r = 0; + STRV_FOREACH(i, *l) { + int k; + + k = sd_bus_track_add_name(*t, *i); + if (k < 0) + r = k; + } + } + + *l = strv_free(*l); + + return r; +} + +int bus_verify_manage_units_async(Manager *m, sd_bus_message *call, sd_bus_error *error) { + return bus_verify_polkit_async(call, CAP_SYS_ADMIN, "org.freedesktop.systemd1.manage-units", NULL, false, UID_INVALID, &m->polkit_registry, error); +} + +int bus_verify_manage_unit_files_async(Manager *m, sd_bus_message *call, sd_bus_error *error) { + return bus_verify_polkit_async(call, CAP_SYS_ADMIN, "org.freedesktop.systemd1.manage-unit-files", NULL, false, UID_INVALID, &m->polkit_registry, error); +} + +int bus_verify_reload_daemon_async(Manager *m, sd_bus_message *call, sd_bus_error *error) { + return bus_verify_polkit_async(call, CAP_SYS_ADMIN, "org.freedesktop.systemd1.reload-daemon", NULL, false, UID_INVALID, &m->polkit_registry, error); +} + +int bus_verify_set_environment_async(Manager *m, sd_bus_message *call, sd_bus_error *error) { + return bus_verify_polkit_async(call, CAP_SYS_ADMIN, "org.freedesktop.systemd1.set-environment", NULL, false, UID_INVALID, &m->polkit_registry, error); +} diff --git a/src/grp-system/libcore/src/dbus.h b/src/grp-system/libcore/src/dbus.h new file mode 100644 index 0000000000..36f8d34a1d --- /dev/null +++ b/src/grp-system/libcore/src/dbus.h @@ -0,0 +1,44 @@ +#pragma once + +/*** + This file is part of systemd. + + Copyright 2010 Lennart Poettering + + systemd is free software; you can redistribute it and/or modify it + under the terms of the GNU Lesser General Public License as published by + the Free Software Foundation; either version 2.1 of the License, or + (at your option) any later version. + + systemd is distributed in the hope that it will be useful, but + WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + Lesser General Public License for more details. + + You should have received a copy of the GNU Lesser General Public License + along with systemd; If not, see . +***/ + +#include "core/manager.h" + +int bus_send_queued_message(Manager *m); + +int bus_init(Manager *m, bool try_bus_connect); +void bus_done(Manager *m); + +int bus_fdset_add_all(Manager *m, FDSet *fds); + +void bus_track_serialize(sd_bus_track *t, FILE *f); +int bus_track_deserialize_item(char ***l, const char *line); +int bus_track_coldplug(Manager *m, sd_bus_track **t, char ***l); + +int manager_sync_bus_names(Manager *m, sd_bus *bus); + +int bus_foreach_bus(Manager *m, sd_bus_track *subscribed2, int (*send_message)(sd_bus *bus, void *userdata), void *userdata); + +int bus_verify_manage_units_async(Manager *m, sd_bus_message *call, sd_bus_error *error); +int bus_verify_manage_unit_files_async(Manager *m, sd_bus_message *call, sd_bus_error *error); +int bus_verify_reload_daemon_async(Manager *m, sd_bus_message *call, sd_bus_error *error); +int bus_verify_set_environment_async(Manager *m, sd_bus_message *call, sd_bus_error *error); + +int bus_forward_agent_released(Manager *m, const char *path); diff --git a/src/grp-system/libcore/src/device.c b/src/grp-system/libcore/src/device.c new file mode 100644 index 0000000000..a147de3280 --- /dev/null +++ b/src/grp-system/libcore/src/device.c @@ -0,0 +1,877 @@ +/*** + This file is part of systemd. + + Copyright 2010 Lennart Poettering + + systemd is free software; you can redistribute it and/or modify it + under the terms of the GNU Lesser General Public License as published by + the Free Software Foundation; either version 2.1 of the License, or + (at your option) any later version. + + systemd is distributed in the hope that it will be useful, but + WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + Lesser General Public License for more details. + + You should have received a copy of the GNU Lesser General Public License + along with systemd; If not, see . +***/ + +#include +#include + +#include + +#include "core/device.h" +#include "core/swap.h" +#include "core/unit.h" +#include "systemd-basic/alloc-util.h" +#include "systemd-basic/log.h" +#include "systemd-basic/parse-util.h" +#include "systemd-basic/path-util.h" +#include "systemd-basic/stat-util.h" +#include "systemd-basic/string-util.h" +#include "systemd-basic/unit-name.h" +#include "systemd-shared/udev-util.h" + +#include "dbus-device.h" + +static const UnitActiveState state_translation_table[_DEVICE_STATE_MAX] = { + [DEVICE_DEAD] = UNIT_INACTIVE, + [DEVICE_TENTATIVE] = UNIT_ACTIVATING, + [DEVICE_PLUGGED] = UNIT_ACTIVE, +}; + +static int device_dispatch_io(sd_event_source *source, int fd, uint32_t revents, void *userdata); + +static void device_unset_sysfs(Device *d) { + Hashmap *devices; + Device *first; + + assert(d); + + if (!d->sysfs) + return; + + /* Remove this unit from the chain of devices which share the + * same sysfs path. */ + devices = UNIT(d)->manager->devices_by_sysfs; + first = hashmap_get(devices, d->sysfs); + LIST_REMOVE(same_sysfs, first, d); + + if (first) + hashmap_remove_and_replace(devices, d->sysfs, first->sysfs, first); + else + hashmap_remove(devices, d->sysfs); + + d->sysfs = mfree(d->sysfs); +} + +static int device_set_sysfs(Device *d, const char *sysfs) { + Device *first; + char *copy; + int r; + + assert(d); + + if (streq_ptr(d->sysfs, sysfs)) + return 0; + + r = hashmap_ensure_allocated(&UNIT(d)->manager->devices_by_sysfs, &string_hash_ops); + if (r < 0) + return r; + + copy = strdup(sysfs); + if (!copy) + return -ENOMEM; + + device_unset_sysfs(d); + + first = hashmap_get(UNIT(d)->manager->devices_by_sysfs, sysfs); + LIST_PREPEND(same_sysfs, first, d); + + r = hashmap_replace(UNIT(d)->manager->devices_by_sysfs, copy, first); + if (r < 0) { + LIST_REMOVE(same_sysfs, first, d); + free(copy); + return r; + } + + d->sysfs = copy; + + return 0; +} + +static void device_init(Unit *u) { + Device *d = DEVICE(u); + + assert(d); + assert(UNIT(d)->load_state == UNIT_STUB); + + /* In contrast to all other unit types we timeout jobs waiting + * for devices by default. This is because they otherwise wait + * indefinitely for plugged in devices, something which cannot + * happen for the other units since their operations time out + * anyway. */ + u->job_timeout = u->manager->default_timeout_start_usec; + + u->ignore_on_isolate = true; +} + +static void device_done(Unit *u) { + Device *d = DEVICE(u); + + assert(d); + + device_unset_sysfs(d); +} + +static void device_set_state(Device *d, DeviceState state) { + DeviceState old_state; + assert(d); + + old_state = d->state; + d->state = state; + + if (state != old_state) + log_unit_debug(UNIT(d), "Changed %s -> %s", device_state_to_string(old_state), device_state_to_string(state)); + + unit_notify(UNIT(d), state_translation_table[old_state], state_translation_table[state], true); +} + +static int device_coldplug(Unit *u) { + Device *d = DEVICE(u); + + assert(d); + assert(d->state == DEVICE_DEAD); + + if (d->found & DEVICE_FOUND_UDEV) + /* If udev says the device is around, it's around */ + device_set_state(d, DEVICE_PLUGGED); + else if (d->found != DEVICE_NOT_FOUND && d->deserialized_state != DEVICE_PLUGGED) + /* If a device is found in /proc/self/mountinfo or + * /proc/swaps, and was not yet announced via udev, + * it's "tentatively" around. */ + device_set_state(d, DEVICE_TENTATIVE); + + return 0; +} + +static int device_serialize(Unit *u, FILE *f, FDSet *fds) { + Device *d = DEVICE(u); + + assert(u); + assert(f); + assert(fds); + + unit_serialize_item(u, f, "state", device_state_to_string(d->state)); + + return 0; +} + +static int device_deserialize_item(Unit *u, const char *key, const char *value, FDSet *fds) { + Device *d = DEVICE(u); + + assert(u); + assert(key); + assert(value); + assert(fds); + + if (streq(key, "state")) { + DeviceState state; + + state = device_state_from_string(value); + if (state < 0) + log_unit_debug(u, "Failed to parse state value: %s", value); + else + d->deserialized_state = state; + } else + log_unit_debug(u, "Unknown serialization key: %s", key); + + return 0; +} + +static void device_dump(Unit *u, FILE *f, const char *prefix) { + Device *d = DEVICE(u); + + assert(d); + + fprintf(f, + "%sDevice State: %s\n" + "%sSysfs Path: %s\n", + prefix, device_state_to_string(d->state), + prefix, strna(d->sysfs)); +} + +_pure_ static UnitActiveState device_active_state(Unit *u) { + assert(u); + + return state_translation_table[DEVICE(u)->state]; +} + +_pure_ static const char *device_sub_state_to_string(Unit *u) { + assert(u); + + return device_state_to_string(DEVICE(u)->state); +} + +static int device_update_description(Unit *u, struct udev_device *dev, const char *path) { + const char *model; + int r; + + assert(u); + assert(dev); + assert(path); + + model = udev_device_get_property_value(dev, "ID_MODEL_FROM_DATABASE"); + if (!model) + model = udev_device_get_property_value(dev, "ID_MODEL"); + + if (model) { + const char *label; + + /* Try to concatenate the device model string with a label, if there is one */ + label = udev_device_get_property_value(dev, "ID_FS_LABEL"); + if (!label) + label = udev_device_get_property_value(dev, "ID_PART_ENTRY_NAME"); + if (!label) + label = udev_device_get_property_value(dev, "ID_PART_ENTRY_NUMBER"); + + if (label) { + _cleanup_free_ char *j; + + j = strjoin(model, " ", label, NULL); + if (j) + r = unit_set_description(u, j); + else + r = -ENOMEM; + } else + r = unit_set_description(u, model); + } else + r = unit_set_description(u, path); + + if (r < 0) + log_unit_error_errno(u, r, "Failed to set device description: %m"); + + return r; +} + +static int device_add_udev_wants(Unit *u, struct udev_device *dev) { + const char *wants; + const char *word, *state; + size_t l; + int r; + const char *property; + + assert(u); + assert(dev); + + property = MANAGER_IS_USER(u->manager) ? "SYSTEMD_USER_WANTS" : "SYSTEMD_WANTS"; + wants = udev_device_get_property_value(dev, property); + if (!wants) + return 0; + + FOREACH_WORD_QUOTED(word, l, wants, state) { + _cleanup_free_ char *n = NULL; + char e[l+1]; + + memcpy(e, word, l); + e[l] = 0; + + r = unit_name_mangle(e, UNIT_NAME_NOGLOB, &n); + if (r < 0) + return log_unit_error_errno(u, r, "Failed to mangle unit name: %m"); + + r = unit_add_dependency_by_name(u, UNIT_WANTS, n, NULL, true); + if (r < 0) + return log_unit_error_errno(u, r, "Failed to add wants dependency: %m"); + } + if (!isempty(state)) + log_unit_warning(u, "Property %s on %s has trailing garbage, ignoring.", property, strna(udev_device_get_syspath(dev))); + + return 0; +} + +static int device_setup_unit(Manager *m, struct udev_device *dev, const char *path, bool main) { + _cleanup_free_ char *e = NULL; + const char *sysfs = NULL; + Unit *u = NULL; + bool delete; + int r; + + assert(m); + assert(path); + + if (dev) { + sysfs = udev_device_get_syspath(dev); + if (!sysfs) + return 0; + } + + r = unit_name_from_path(path, ".device", &e); + if (r < 0) + return log_error_errno(r, "Failed to generate unit name from device path: %m"); + + u = manager_get_unit(m, e); + + /* The device unit can still be present even if the device was + * unplugged: a mount unit can reference it hence preventing + * the GC to have garbaged it. That's desired since the device + * unit may have a dependency on the mount unit which was + * added during the loading of the later. */ + if (dev && u && DEVICE(u)->state == DEVICE_PLUGGED) { + /* This unit is in plugged state: we're sure it's + * attached to a device. */ + if (!path_equal(DEVICE(u)->sysfs, sysfs)) { + log_unit_debug(u, "Dev %s appeared twice with different sysfs paths %s and %s", + e, DEVICE(u)->sysfs, sysfs); + return -EEXIST; + } + } + + if (!u) { + delete = true; + + u = unit_new(m, sizeof(Device)); + if (!u) + return log_oom(); + + r = unit_add_name(u, e); + if (r < 0) + goto fail; + + unit_add_to_load_queue(u); + } else + delete = false; + + /* If this was created via some dependency and has not + * actually been seen yet ->sysfs will not be + * initialized. Hence initialize it if necessary. */ + if (sysfs) { + r = device_set_sysfs(DEVICE(u), sysfs); + if (r < 0) + goto fail; + + (void) device_update_description(u, dev, path); + + /* The additional systemd udev properties we only interpret + * for the main object */ + if (main) + (void) device_add_udev_wants(u, dev); + } + + + /* Note that this won't dispatch the load queue, the caller + * has to do that if needed and appropriate */ + + unit_add_to_dbus_queue(u); + return 0; + +fail: + log_unit_warning_errno(u, r, "Failed to set up device unit: %m"); + + if (delete) + unit_free(u); + + return r; +} + +static int device_process_new(Manager *m, struct udev_device *dev) { + const char *sysfs, *dn, *alias; + struct udev_list_entry *item = NULL, *first = NULL; + int r; + + assert(m); + + sysfs = udev_device_get_syspath(dev); + if (!sysfs) + return 0; + + /* Add the main unit named after the sysfs path */ + r = device_setup_unit(m, dev, sysfs, true); + if (r < 0) + return r; + + /* Add an additional unit for the device node */ + dn = udev_device_get_devnode(dev); + if (dn) + (void) device_setup_unit(m, dev, dn, false); + + /* Add additional units for all symlinks */ + first = udev_device_get_devlinks_list_entry(dev); + udev_list_entry_foreach(item, first) { + const char *p; + struct stat st; + + /* Don't bother with the /dev/block links */ + p = udev_list_entry_get_name(item); + + if (path_startswith(p, "/dev/block/") || + path_startswith(p, "/dev/char/")) + continue; + + /* Verify that the symlink in the FS actually belongs + * to this device. This is useful to deal with + * conflicting devices, e.g. when two disks want the + * same /dev/disk/by-label/xxx link because they have + * the same label. We want to make sure that the same + * device that won the symlink wins in systemd, so we + * check the device node major/minor */ + if (stat(p, &st) >= 0) + if ((!S_ISBLK(st.st_mode) && !S_ISCHR(st.st_mode)) || + st.st_rdev != udev_device_get_devnum(dev)) + continue; + + (void) device_setup_unit(m, dev, p, false); + } + + /* Add additional units for all explicitly configured + * aliases */ + alias = udev_device_get_property_value(dev, "SYSTEMD_ALIAS"); + if (alias) { + const char *word, *state; + size_t l; + + FOREACH_WORD_QUOTED(word, l, alias, state) { + char e[l+1]; + + memcpy(e, word, l); + e[l] = 0; + + if (path_is_absolute(e)) + (void) device_setup_unit(m, dev, e, false); + else + log_warning("SYSTEMD_ALIAS for %s is not an absolute path, ignoring: %s", sysfs, e); + } + if (!isempty(state)) + log_warning("SYSTEMD_ALIAS for %s has trailing garbage, ignoring.", sysfs); + } + + return 0; +} + +static void device_update_found_one(Device *d, bool add, DeviceFound found, bool now) { + DeviceFound n, previous; + + assert(d); + + n = add ? (d->found | found) : (d->found & ~found); + if (n == d->found) + return; + + previous = d->found; + d->found = n; + + if (!now) + return; + + if (d->found & DEVICE_FOUND_UDEV) + /* When the device is known to udev we consider it + * plugged. */ + device_set_state(d, DEVICE_PLUGGED); + else if (d->found != DEVICE_NOT_FOUND && (previous & DEVICE_FOUND_UDEV) == 0) + /* If the device has not been seen by udev yet, but is + * now referenced by the kernel, then we assume the + * kernel knows it now, and udev might soon too. */ + device_set_state(d, DEVICE_TENTATIVE); + else + /* If nobody sees the device, or if the device was + * previously seen by udev and now is only referenced + * from the kernel, then we consider the device is + * gone, the kernel just hasn't noticed it yet. */ + device_set_state(d, DEVICE_DEAD); +} + +static int device_update_found_by_sysfs(Manager *m, const char *sysfs, bool add, DeviceFound found, bool now) { + Device *d, *l; + + assert(m); + assert(sysfs); + + if (found == DEVICE_NOT_FOUND) + return 0; + + l = hashmap_get(m->devices_by_sysfs, sysfs); + LIST_FOREACH(same_sysfs, d, l) + device_update_found_one(d, add, found, now); + + return 0; +} + +static int device_update_found_by_name(Manager *m, const char *path, bool add, DeviceFound found, bool now) { + _cleanup_free_ char *e = NULL; + Unit *u; + int r; + + assert(m); + assert(path); + + if (found == DEVICE_NOT_FOUND) + return 0; + + r = unit_name_from_path(path, ".device", &e); + if (r < 0) + return log_error_errno(r, "Failed to generate unit name from device path: %m"); + + u = manager_get_unit(m, e); + if (!u) + return 0; + + device_update_found_one(DEVICE(u), add, found, now); + return 0; +} + +static bool device_is_ready(struct udev_device *dev) { + const char *ready; + + assert(dev); + + ready = udev_device_get_property_value(dev, "SYSTEMD_READY"); + if (!ready) + return true; + + return parse_boolean(ready) != 0; +} + +static Unit *device_following(Unit *u) { + Device *d = DEVICE(u); + Device *other, *first = NULL; + + assert(d); + + if (startswith(u->id, "sys-")) + return NULL; + + /* Make everybody follow the unit that's named after the sysfs path */ + for (other = d->same_sysfs_next; other; other = other->same_sysfs_next) + if (startswith(UNIT(other)->id, "sys-")) + return UNIT(other); + + for (other = d->same_sysfs_prev; other; other = other->same_sysfs_prev) { + if (startswith(UNIT(other)->id, "sys-")) + return UNIT(other); + + first = other; + } + + return UNIT(first); +} + +static int device_following_set(Unit *u, Set **_set) { + Device *d = DEVICE(u), *other; + Set *set; + int r; + + assert(d); + assert(_set); + + if (LIST_JUST_US(same_sysfs, d)) { + *_set = NULL; + return 0; + } + + set = set_new(NULL); + if (!set) + return -ENOMEM; + + LIST_FOREACH_AFTER(same_sysfs, other, d) { + r = set_put(set, other); + if (r < 0) + goto fail; + } + + LIST_FOREACH_BEFORE(same_sysfs, other, d) { + r = set_put(set, other); + if (r < 0) + goto fail; + } + + *_set = set; + return 1; + +fail: + set_free(set); + return r; +} + +static void device_shutdown(Manager *m) { + assert(m); + + m->udev_event_source = sd_event_source_unref(m->udev_event_source); + + if (m->udev_monitor) { + udev_monitor_unref(m->udev_monitor); + m->udev_monitor = NULL; + } + + m->devices_by_sysfs = hashmap_free(m->devices_by_sysfs); +} + +static void device_enumerate(Manager *m) { + _cleanup_udev_enumerate_unref_ struct udev_enumerate *e = NULL; + struct udev_list_entry *item = NULL, *first = NULL; + int r; + + assert(m); + + if (!m->udev_monitor) { + m->udev_monitor = udev_monitor_new_from_netlink(m->udev, "udev"); + if (!m->udev_monitor) { + log_oom(); + goto fail; + } + + /* This will fail if we are unprivileged, but that + * should not matter much, as user instances won't run + * during boot. */ + (void) udev_monitor_set_receive_buffer_size(m->udev_monitor, 128*1024*1024); + + r = udev_monitor_filter_add_match_tag(m->udev_monitor, "systemd"); + if (r < 0) { + log_error_errno(r, "Failed to add udev tag match: %m"); + goto fail; + } + + r = udev_monitor_enable_receiving(m->udev_monitor); + if (r < 0) { + log_error_errno(r, "Failed to enable udev event reception: %m"); + goto fail; + } + + r = sd_event_add_io(m->event, &m->udev_event_source, udev_monitor_get_fd(m->udev_monitor), EPOLLIN, device_dispatch_io, m); + if (r < 0) { + log_error_errno(r, "Failed to watch udev file descriptor: %m"); + goto fail; + } + + (void) sd_event_source_set_description(m->udev_event_source, "device"); + } + + e = udev_enumerate_new(m->udev); + if (!e) { + log_oom(); + goto fail; + } + + r = udev_enumerate_add_match_tag(e, "systemd"); + if (r < 0) { + log_error_errno(r, "Failed to create udev tag enumeration: %m"); + goto fail; + } + + r = udev_enumerate_add_match_is_initialized(e); + if (r < 0) { + log_error_errno(r, "Failed to install initialization match into enumeration: %m"); + goto fail; + } + + r = udev_enumerate_scan_devices(e); + if (r < 0) { + log_error_errno(r, "Failed to enumerate devices: %m"); + goto fail; + } + + first = udev_enumerate_get_list_entry(e); + udev_list_entry_foreach(item, first) { + _cleanup_udev_device_unref_ struct udev_device *dev = NULL; + const char *sysfs; + + sysfs = udev_list_entry_get_name(item); + + dev = udev_device_new_from_syspath(m->udev, sysfs); + if (!dev) { + log_oom(); + continue; + } + + if (!device_is_ready(dev)) + continue; + + (void) device_process_new(m, dev); + + device_update_found_by_sysfs(m, sysfs, true, DEVICE_FOUND_UDEV, false); + } + + return; + +fail: + device_shutdown(m); +} + +static int device_dispatch_io(sd_event_source *source, int fd, uint32_t revents, void *userdata) { + _cleanup_udev_device_unref_ struct udev_device *dev = NULL; + Manager *m = userdata; + const char *action, *sysfs; + int r; + + assert(m); + + if (revents != EPOLLIN) { + static RATELIMIT_DEFINE(limit, 10*USEC_PER_SEC, 5); + + if (!ratelimit_test(&limit)) + log_error_errno(errno, "Failed to get udev event: %m"); + if (!(revents & EPOLLIN)) + return 0; + } + + /* + * libudev might filter-out devices which pass the bloom + * filter, so getting NULL here is not necessarily an error. + */ + dev = udev_monitor_receive_device(m->udev_monitor); + if (!dev) + return 0; + + sysfs = udev_device_get_syspath(dev); + if (!sysfs) { + log_error("Failed to get udev sys path."); + return 0; + } + + action = udev_device_get_action(dev); + if (!action) { + log_error("Failed to get udev action string."); + return 0; + } + + if (streq(action, "remove")) { + r = swap_process_device_remove(m, dev); + if (r < 0) + log_error_errno(r, "Failed to process swap device remove event: %m"); + + /* If we get notified that a device was removed by + * udev, then it's completely gone, hence unset all + * found bits */ + device_update_found_by_sysfs(m, sysfs, false, DEVICE_FOUND_UDEV|DEVICE_FOUND_MOUNT|DEVICE_FOUND_SWAP, true); + + } else if (device_is_ready(dev)) { + + (void) device_process_new(m, dev); + + r = swap_process_device_new(m, dev); + if (r < 0) + log_error_errno(r, "Failed to process swap device new event: %m"); + + manager_dispatch_load_queue(m); + + /* The device is found now, set the udev found bit */ + device_update_found_by_sysfs(m, sysfs, true, DEVICE_FOUND_UDEV, true); + + } else { + /* The device is nominally around, but not ready for + * us. Hence unset the udev bit, but leave the rest + * around. */ + + device_update_found_by_sysfs(m, sysfs, false, DEVICE_FOUND_UDEV, true); + } + + return 0; +} + +static bool device_supported(void) { + static int read_only = -1; + + /* If /sys is read-only we don't support device units, and any + * attempts to start one should fail immediately. */ + + if (read_only < 0) + read_only = path_is_read_only_fs("/sys"); + + return read_only <= 0; +} + +int device_found_node(Manager *m, const char *node, bool add, DeviceFound found, bool now) { + _cleanup_udev_device_unref_ struct udev_device *dev = NULL; + struct stat st; + + assert(m); + assert(node); + + if (!device_supported()) + return 0; + + /* This is called whenever we find a device referenced in + * /proc/swaps or /proc/self/mounts. Such a device might be + * mounted/enabled at a time where udev has not finished + * probing it yet, and we thus haven't learned about it + * yet. In this case we will set the device unit to + * "tentative" state. */ + + if (add) { + if (!path_startswith(node, "/dev")) + return 0; + + /* We make an extra check here, if the device node + * actually exists. If it's missing, then this is an + * indication that device was unplugged but is still + * referenced in /proc/swaps or + * /proc/self/mountinfo. Note that this check doesn't + * really cover all cases where a device might be gone + * away, since drives that can have a medium inserted + * will still have a device node even when the medium + * is not there... */ + + if (stat(node, &st) >= 0) { + if (!S_ISBLK(st.st_mode) && !S_ISCHR(st.st_mode)) + return 0; + + dev = udev_device_new_from_devnum(m->udev, S_ISBLK(st.st_mode) ? 'b' : 'c', st.st_rdev); + if (!dev && errno != ENOENT) + return log_error_errno(errno, "Failed to get udev device from devnum %u:%u: %m", major(st.st_rdev), minor(st.st_rdev)); + + } else if (errno != ENOENT) + return log_error_errno(errno, "Failed to stat device node file %s: %m", node); + + /* If the device is known in the kernel and newly + * appeared, then we'll create a device unit for it, + * under the name referenced in /proc/swaps or + * /proc/self/mountinfo. */ + + (void) device_setup_unit(m, dev, node, false); + } + + /* Update the device unit's state, should it exist */ + return device_update_found_by_name(m, node, add, found, now); +} + +const UnitVTable device_vtable = { + .object_size = sizeof(Device), + .sections = + "Unit\0" + "Device\0" + "Install\0", + + .init = device_init, + .done = device_done, + .load = unit_load_fragment_and_dropin_optional, + + .coldplug = device_coldplug, + + .serialize = device_serialize, + .deserialize_item = device_deserialize_item, + + .dump = device_dump, + + .active_state = device_active_state, + .sub_state_to_string = device_sub_state_to_string, + + .bus_vtable = bus_device_vtable, + + .following = device_following, + .following_set = device_following_set, + + .enumerate = device_enumerate, + .shutdown = device_shutdown, + .supported = device_supported, + + .status_message_formats = { + .starting_stopping = { + [0] = "Expecting device %s...", + }, + .finished_start_job = { + [JOB_DONE] = "Found device %s.", + [JOB_TIMEOUT] = "Timed out waiting for device %s.", + }, + }, +}; diff --git a/src/grp-system/libcore/src/execute.c b/src/grp-system/libcore/src/execute.c new file mode 100644 index 0000000000..83b43c3535 --- /dev/null +++ b/src/grp-system/libcore/src/execute.c @@ -0,0 +1,3284 @@ +/*** + This file is part of systemd. + + Copyright 2010 Lennart Poettering + + systemd is free software; you can redistribute it and/or modify it + under the terms of the GNU Lesser General Public License as published by + the Free Software Foundation; either version 2.1 of the License, or + (at your option) any later version. + + systemd is distributed in the hope that it will be useful, but + WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + Lesser General Public License for more details. + + You should have received a copy of the GNU Lesser General Public License + along with systemd; If not, see . +***/ + +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include + +#ifdef HAVE_PAM +#include +#endif + +#ifdef HAVE_SELINUX +#include +#endif + +#ifdef HAVE_SECCOMP +#include +#endif + +#ifdef HAVE_APPARMOR +#include +#endif + +#include + +#include "systemd-basic/af-list.h" +#include "systemd-basic/alloc-util.h" +#ifdef HAVE_APPARMOR +#include "systemd-shared/apparmor-util.h" +#endif +#include "core/execute.h" +#include "core/namespace.h" +#include "systemd-basic/async.h" +#include "systemd-basic/barrier.h" +#include "systemd-basic/cap-list.h" +#include "systemd-basic/capability-util.h" +#include "systemd-basic/def.h" +#include "systemd-basic/env-util.h" +#include "systemd-basic/errno-list.h" +#include "systemd-basic/exit-status.h" +#include "systemd-basic/fd-util.h" +#include "systemd-basic/fileio.h" +#include "systemd-basic/formats-util.h" +#include "systemd-basic/fs-util.h" +#include "systemd-basic/glob-util.h" +#include "systemd-basic/io-util.h" +#include "systemd-basic/ioprio.h" +#include "systemd-basic/log.h" +#include "systemd-basic/macro.h" +#include "systemd-basic/missing.h" +#include "systemd-basic/mkdir.h" +#include "systemd-basic/parse-util.h" +#include "systemd-basic/path-util.h" +#include "systemd-basic/process-util.h" +#include "systemd-basic/rlimit-util.h" +#include "systemd-basic/rm-rf.h" +#ifdef HAVE_SECCOMP +#include "systemd-shared/seccomp-util.h" +#endif +#include "core/unit.h" +#include "systemd-basic/securebits.h" +#include "systemd-basic/selinux-util.h" +#include "systemd-basic/signal-util.h" +#include "systemd-basic/smack-util.h" +#include "systemd-basic/string-table.h" +#include "systemd-basic/string-util.h" +#include "systemd-basic/strv.h" +#include "systemd-basic/syslog-util.h" +#include "systemd-basic/terminal-util.h" +#include "systemd-basic/user-util.h" +#include "systemd-basic/util.h" +#include "systemd-shared/utmp-wtmp.h" + +#define IDLE_TIMEOUT_USEC (5*USEC_PER_SEC) +#define IDLE_TIMEOUT2_USEC (1*USEC_PER_SEC) + +/* This assumes there is a 'tty' group */ +#define TTY_MODE 0620 + +#define SNDBUF_SIZE (8*1024*1024) + +static int shift_fds(int fds[], unsigned n_fds) { + int start, restart_from; + + if (n_fds <= 0) + return 0; + + /* Modifies the fds array! (sorts it) */ + + assert(fds); + + start = 0; + for (;;) { + int i; + + restart_from = -1; + + for (i = start; i < (int) n_fds; i++) { + int nfd; + + /* Already at right index? */ + if (fds[i] == i+3) + continue; + + nfd = fcntl(fds[i], F_DUPFD, i + 3); + if (nfd < 0) + return -errno; + + safe_close(fds[i]); + fds[i] = nfd; + + /* Hmm, the fd we wanted isn't free? Then + * let's remember that and try again from here */ + if (nfd != i+3 && restart_from < 0) + restart_from = i; + } + + if (restart_from < 0) + break; + + start = restart_from; + } + + return 0; +} + +static int flags_fds(const int fds[], unsigned n_fds, bool nonblock) { + unsigned i; + int r; + + if (n_fds <= 0) + return 0; + + assert(fds); + + /* Drops/Sets O_NONBLOCK and FD_CLOEXEC from the file flags */ + + for (i = 0; i < n_fds; i++) { + + r = fd_nonblock(fds[i], nonblock); + if (r < 0) + return r; + + /* We unconditionally drop FD_CLOEXEC from the fds, + * since after all we want to pass these fds to our + * children */ + + r = fd_cloexec(fds[i], false); + if (r < 0) + return r; + } + + return 0; +} + +static const char *exec_context_tty_path(const ExecContext *context) { + assert(context); + + if (context->stdio_as_fds) + return NULL; + + if (context->tty_path) + return context->tty_path; + + return "/dev/console"; +} + +static void exec_context_tty_reset(const ExecContext *context, const ExecParameters *p) { + const char *path; + + assert(context); + + path = exec_context_tty_path(context); + + if (context->tty_vhangup) { + if (p && p->stdin_fd >= 0) + (void) terminal_vhangup_fd(p->stdin_fd); + else if (path) + (void) terminal_vhangup(path); + } + + if (context->tty_reset) { + if (p && p->stdin_fd >= 0) + (void) reset_terminal_fd(p->stdin_fd, true); + else if (path) + (void) reset_terminal(path); + } + + if (context->tty_vt_disallocate && path) + (void) vt_disallocate(path); +} + +static bool is_terminal_output(ExecOutput o) { + return + o == EXEC_OUTPUT_TTY || + o == EXEC_OUTPUT_SYSLOG_AND_CONSOLE || + o == EXEC_OUTPUT_KMSG_AND_CONSOLE || + o == EXEC_OUTPUT_JOURNAL_AND_CONSOLE; +} + +static int open_null_as(int flags, int nfd) { + int fd, r; + + assert(nfd >= 0); + + fd = open("/dev/null", flags|O_NOCTTY); + if (fd < 0) + return -errno; + + if (fd != nfd) { + r = dup2(fd, nfd) < 0 ? -errno : nfd; + safe_close(fd); + } else + r = nfd; + + return r; +} + +static int connect_journal_socket(int fd, uid_t uid, gid_t gid) { + union sockaddr_union sa = { + .un.sun_family = AF_UNIX, + .un.sun_path = "/run/systemd/journal/stdout", + }; + uid_t olduid = UID_INVALID; + gid_t oldgid = GID_INVALID; + int r; + + if (gid != GID_INVALID) { + oldgid = getgid(); + + r = setegid(gid); + if (r < 0) + return -errno; + } + + if (uid != UID_INVALID) { + olduid = getuid(); + + r = seteuid(uid); + if (r < 0) { + r = -errno; + goto restore_gid; + } + } + + r = connect(fd, &sa.sa, SOCKADDR_UN_LEN(sa.un)); + if (r < 0) + r = -errno; + + /* If we fail to restore the uid or gid, things will likely + fail later on. This should only happen if an LSM interferes. */ + + if (uid != UID_INVALID) + (void) seteuid(olduid); + + restore_gid: + if (gid != GID_INVALID) + (void) setegid(oldgid); + + return r; +} + +static int connect_logger_as( + Unit *unit, + const ExecContext *context, + ExecOutput output, + const char *ident, + int nfd, + uid_t uid, + gid_t gid) { + + int fd, r; + + assert(context); + assert(output < _EXEC_OUTPUT_MAX); + assert(ident); + assert(nfd >= 0); + + fd = socket(AF_UNIX, SOCK_STREAM, 0); + if (fd < 0) + return -errno; + + r = connect_journal_socket(fd, uid, gid); + if (r < 0) + return r; + + if (shutdown(fd, SHUT_RD) < 0) { + safe_close(fd); + return -errno; + } + + (void) fd_inc_sndbuf(fd, SNDBUF_SIZE); + + dprintf(fd, + "%s\n" + "%s\n" + "%i\n" + "%i\n" + "%i\n" + "%i\n" + "%i\n", + context->syslog_identifier ? context->syslog_identifier : ident, + unit->id, + context->syslog_priority, + !!context->syslog_level_prefix, + output == EXEC_OUTPUT_SYSLOG || output == EXEC_OUTPUT_SYSLOG_AND_CONSOLE, + output == EXEC_OUTPUT_KMSG || output == EXEC_OUTPUT_KMSG_AND_CONSOLE, + is_terminal_output(output)); + + if (fd == nfd) + return nfd; + + r = dup2(fd, nfd) < 0 ? -errno : nfd; + safe_close(fd); + + return r; +} +static int open_terminal_as(const char *path, mode_t mode, int nfd) { + int fd, r; + + assert(path); + assert(nfd >= 0); + + fd = open_terminal(path, mode | O_NOCTTY); + if (fd < 0) + return fd; + + if (fd != nfd) { + r = dup2(fd, nfd) < 0 ? -errno : nfd; + safe_close(fd); + } else + r = nfd; + + return r; +} + +static bool is_terminal_input(ExecInput i) { + return + i == EXEC_INPUT_TTY || + i == EXEC_INPUT_TTY_FORCE || + i == EXEC_INPUT_TTY_FAIL; +} + +static int fixup_input(ExecInput std_input, int socket_fd, bool apply_tty_stdin) { + + if (is_terminal_input(std_input) && !apply_tty_stdin) + return EXEC_INPUT_NULL; + + if (std_input == EXEC_INPUT_SOCKET && socket_fd < 0) + return EXEC_INPUT_NULL; + + return std_input; +} + +static int fixup_output(ExecOutput std_output, int socket_fd) { + + if (std_output == EXEC_OUTPUT_SOCKET && socket_fd < 0) + return EXEC_OUTPUT_INHERIT; + + return std_output; +} + +static int setup_input( + const ExecContext *context, + const ExecParameters *params, + int socket_fd) { + + ExecInput i; + + assert(context); + assert(params); + + if (params->stdin_fd >= 0) { + if (dup2(params->stdin_fd, STDIN_FILENO) < 0) + return -errno; + + /* Try to make this the controlling tty, if it is a tty, and reset it */ + (void) ioctl(STDIN_FILENO, TIOCSCTTY, context->std_input == EXEC_INPUT_TTY_FORCE); + (void) reset_terminal_fd(STDIN_FILENO, true); + + return STDIN_FILENO; + } + + i = fixup_input(context->std_input, socket_fd, params->apply_tty_stdin); + + switch (i) { + + case EXEC_INPUT_NULL: + return open_null_as(O_RDONLY, STDIN_FILENO); + + case EXEC_INPUT_TTY: + case EXEC_INPUT_TTY_FORCE: + case EXEC_INPUT_TTY_FAIL: { + int fd, r; + + fd = acquire_terminal(exec_context_tty_path(context), + i == EXEC_INPUT_TTY_FAIL, + i == EXEC_INPUT_TTY_FORCE, + false, + USEC_INFINITY); + if (fd < 0) + return fd; + + if (fd != STDIN_FILENO) { + r = dup2(fd, STDIN_FILENO) < 0 ? -errno : STDIN_FILENO; + safe_close(fd); + } else + r = STDIN_FILENO; + + return r; + } + + case EXEC_INPUT_SOCKET: + return dup2(socket_fd, STDIN_FILENO) < 0 ? -errno : STDIN_FILENO; + + default: + assert_not_reached("Unknown input type"); + } +} + +static int setup_output( + Unit *unit, + const ExecContext *context, + const ExecParameters *params, + int fileno, + int socket_fd, + const char *ident, + uid_t uid, + gid_t gid, + dev_t *journal_stream_dev, + ino_t *journal_stream_ino) { + + ExecOutput o; + ExecInput i; + int r; + + assert(unit); + assert(context); + assert(params); + assert(ident); + assert(journal_stream_dev); + assert(journal_stream_ino); + + if (fileno == STDOUT_FILENO && params->stdout_fd >= 0) { + + if (dup2(params->stdout_fd, STDOUT_FILENO) < 0) + return -errno; + + return STDOUT_FILENO; + } + + if (fileno == STDERR_FILENO && params->stderr_fd >= 0) { + if (dup2(params->stderr_fd, STDERR_FILENO) < 0) + return -errno; + + return STDERR_FILENO; + } + + i = fixup_input(context->std_input, socket_fd, params->apply_tty_stdin); + o = fixup_output(context->std_output, socket_fd); + + if (fileno == STDERR_FILENO) { + ExecOutput e; + e = fixup_output(context->std_error, socket_fd); + + /* This expects the input and output are already set up */ + + /* Don't change the stderr file descriptor if we inherit all + * the way and are not on a tty */ + if (e == EXEC_OUTPUT_INHERIT && + o == EXEC_OUTPUT_INHERIT && + i == EXEC_INPUT_NULL && + !is_terminal_input(context->std_input) && + getppid () != 1) + return fileno; + + /* Duplicate from stdout if possible */ + if (e == o || e == EXEC_OUTPUT_INHERIT) + return dup2(STDOUT_FILENO, fileno) < 0 ? -errno : fileno; + + o = e; + + } else if (o == EXEC_OUTPUT_INHERIT) { + /* If input got downgraded, inherit the original value */ + if (i == EXEC_INPUT_NULL && is_terminal_input(context->std_input)) + return open_terminal_as(exec_context_tty_path(context), O_WRONLY, fileno); + + /* If the input is connected to anything that's not a /dev/null, inherit that... */ + if (i != EXEC_INPUT_NULL) + return dup2(STDIN_FILENO, fileno) < 0 ? -errno : fileno; + + /* If we are not started from PID 1 we just inherit STDOUT from our parent process. */ + if (getppid() != 1) + return fileno; + + /* We need to open /dev/null here anew, to get the right access mode. */ + return open_null_as(O_WRONLY, fileno); + } + + switch (o) { + + case EXEC_OUTPUT_NULL: + return open_null_as(O_WRONLY, fileno); + + case EXEC_OUTPUT_TTY: + if (is_terminal_input(i)) + return dup2(STDIN_FILENO, fileno) < 0 ? -errno : fileno; + + /* We don't reset the terminal if this is just about output */ + return open_terminal_as(exec_context_tty_path(context), O_WRONLY, fileno); + + case EXEC_OUTPUT_SYSLOG: + case EXEC_OUTPUT_SYSLOG_AND_CONSOLE: + case EXEC_OUTPUT_KMSG: + case EXEC_OUTPUT_KMSG_AND_CONSOLE: + case EXEC_OUTPUT_JOURNAL: + case EXEC_OUTPUT_JOURNAL_AND_CONSOLE: + r = connect_logger_as(unit, context, o, ident, fileno, uid, gid); + if (r < 0) { + log_unit_error_errno(unit, r, "Failed to connect %s to the journal socket, ignoring: %m", fileno == STDOUT_FILENO ? "stdout" : "stderr"); + r = open_null_as(O_WRONLY, fileno); + } else { + struct stat st; + + /* If we connected this fd to the journal via a stream, patch the device/inode into the passed + * parameters, but only then. This is useful so that we can set $JOURNAL_STREAM that permits + * services to detect whether they are connected to the journal or not. */ + + if (fstat(fileno, &st) >= 0) { + *journal_stream_dev = st.st_dev; + *journal_stream_ino = st.st_ino; + } + } + return r; + + case EXEC_OUTPUT_SOCKET: + assert(socket_fd >= 0); + return dup2(socket_fd, fileno) < 0 ? -errno : fileno; + + default: + assert_not_reached("Unknown error type"); + } +} + +static int chown_terminal(int fd, uid_t uid) { + struct stat st; + + assert(fd >= 0); + + /* Before we chown/chmod the TTY, let's ensure this is actually a tty */ + if (isatty(fd) < 1) + return 0; + + /* This might fail. What matters are the results. */ + (void) fchown(fd, uid, -1); + (void) fchmod(fd, TTY_MODE); + + if (fstat(fd, &st) < 0) + return -errno; + + if (st.st_uid != uid || (st.st_mode & 0777) != TTY_MODE) + return -EPERM; + + return 0; +} + +static int setup_confirm_stdio(int *_saved_stdin, int *_saved_stdout) { + _cleanup_close_ int fd = -1, saved_stdin = -1, saved_stdout = -1; + int r; + + assert(_saved_stdin); + assert(_saved_stdout); + + saved_stdin = fcntl(STDIN_FILENO, F_DUPFD, 3); + if (saved_stdin < 0) + return -errno; + + saved_stdout = fcntl(STDOUT_FILENO, F_DUPFD, 3); + if (saved_stdout < 0) + return -errno; + + fd = acquire_terminal( + "/dev/console", + false, + false, + false, + DEFAULT_CONFIRM_USEC); + if (fd < 0) + return fd; + + r = chown_terminal(fd, getuid()); + if (r < 0) + return r; + + r = reset_terminal_fd(fd, true); + if (r < 0) + return r; + + if (dup2(fd, STDIN_FILENO) < 0) + return -errno; + + if (dup2(fd, STDOUT_FILENO) < 0) + return -errno; + + if (fd >= 2) + safe_close(fd); + fd = -1; + + *_saved_stdin = saved_stdin; + *_saved_stdout = saved_stdout; + + saved_stdin = saved_stdout = -1; + + return 0; +} + +_printf_(1, 2) static int write_confirm_message(const char *format, ...) { + _cleanup_close_ int fd = -1; + va_list ap; + + assert(format); + + fd = open_terminal("/dev/console", O_WRONLY|O_NOCTTY|O_CLOEXEC); + if (fd < 0) + return fd; + + va_start(ap, format); + vdprintf(fd, format, ap); + va_end(ap); + + return 0; +} + +static int restore_confirm_stdio(int *saved_stdin, int *saved_stdout) { + int r = 0; + + assert(saved_stdin); + assert(saved_stdout); + + release_terminal(); + + if (*saved_stdin >= 0) + if (dup2(*saved_stdin, STDIN_FILENO) < 0) + r = -errno; + + if (*saved_stdout >= 0) + if (dup2(*saved_stdout, STDOUT_FILENO) < 0) + r = -errno; + + *saved_stdin = safe_close(*saved_stdin); + *saved_stdout = safe_close(*saved_stdout); + + return r; +} + +static int ask_for_confirmation(char *response, char **argv) { + int saved_stdout = -1, saved_stdin = -1, r; + _cleanup_free_ char *line = NULL; + + r = setup_confirm_stdio(&saved_stdin, &saved_stdout); + if (r < 0) + return r; + + line = exec_command_line(argv); + if (!line) + return -ENOMEM; + + r = ask_char(response, "yns", "Execute %s? [Yes, No, Skip] ", line); + + restore_confirm_stdio(&saved_stdin, &saved_stdout); + + return r; +} + +static int enforce_groups(const ExecContext *context, const char *username, gid_t gid) { + bool keep_groups = false; + int r; + + assert(context); + + /* Lookup and set GID and supplementary group list. Here too + * we avoid NSS lookups for gid=0. */ + + if (context->group || username) { + /* First step, initialize groups from /etc/groups */ + if (username && gid != 0) { + if (initgroups(username, gid) < 0) + return -errno; + + keep_groups = true; + } + + /* Second step, set our gids */ + if (setresgid(gid, gid, gid) < 0) + return -errno; + } + + if (context->supplementary_groups) { + int ngroups_max, k; + gid_t *gids; + char **i; + + /* Final step, initialize any manually set supplementary groups */ + assert_se((ngroups_max = (int) sysconf(_SC_NGROUPS_MAX)) > 0); + + if (!(gids = new(gid_t, ngroups_max))) + return -ENOMEM; + + if (keep_groups) { + k = getgroups(ngroups_max, gids); + if (k < 0) { + free(gids); + return -errno; + } + } else + k = 0; + + STRV_FOREACH(i, context->supplementary_groups) { + const char *g; + + if (k >= ngroups_max) { + free(gids); + return -E2BIG; + } + + g = *i; + r = get_group_creds(&g, gids+k); + if (r < 0) { + free(gids); + return r; + } + + k++; + } + + if (setgroups(k, gids) < 0) { + free(gids); + return -errno; + } + + free(gids); + } + + return 0; +} + +static int enforce_user(const ExecContext *context, uid_t uid) { + assert(context); + + /* Sets (but doesn't look up) the uid and make sure we keep the + * capabilities while doing so. */ + + if (context->capability_ambient_set != 0) { + + /* First step: If we need to keep capabilities but + * drop privileges we need to make sure we keep our + * caps, while we drop privileges. */ + if (uid != 0) { + int sb = context->secure_bits | 1<= 0); + + parent_pid = getpid(); + + pam_pid = fork(); + if (pam_pid < 0) { + r = -errno; + goto fail; + } + + if (pam_pid == 0) { + int sig, ret = EXIT_PAM; + + /* The child's job is to reset the PAM session on + * termination */ + barrier_set_role(&barrier, BARRIER_CHILD); + + /* This string must fit in 10 chars (i.e. the length + * of "/sbin/init"), to look pretty in /bin/ps */ + rename_process("(sd-pam)"); + + /* Make sure we don't keep open the passed fds in this + child. We assume that otherwise only those fds are + open here that have been opened by PAM. */ + close_many(fds, n_fds); + + /* Drop privileges - we don't need any to pam_close_session + * and this will make PR_SET_PDEATHSIG work in most cases. + * If this fails, ignore the error - but expect sd-pam threads + * to fail to exit normally */ + if (setresuid(uid, uid, uid) < 0) + log_error_errno(r, "Error: Failed to setresuid() in sd-pam: %m"); + + (void) ignore_signals(SIGPIPE, -1); + + /* Wait until our parent died. This will only work if + * the above setresuid() succeeds, otherwise the kernel + * will not allow unprivileged parents kill their privileged + * children this way. We rely on the control groups kill logic + * to do the rest for us. */ + if (prctl(PR_SET_PDEATHSIG, SIGTERM) < 0) + goto child_finish; + + /* Tell the parent that our setup is done. This is especially + * important regarding dropping privileges. Otherwise, unit + * setup might race against our setresuid(2) call. */ + barrier_place(&barrier); + + /* Check if our parent process might already have + * died? */ + if (getppid() == parent_pid) { + sigset_t ss; + + assert_se(sigemptyset(&ss) >= 0); + assert_se(sigaddset(&ss, SIGTERM) >= 0); + + for (;;) { + if (sigwait(&ss, &sig) < 0) { + if (errno == EINTR) + continue; + + goto child_finish; + } + + assert(sig == SIGTERM); + break; + } + } + + /* If our parent died we'll end the session */ + if (getppid() != parent_pid) { + pam_code = pam_close_session(handle, flags); + if (pam_code != PAM_SUCCESS) + goto child_finish; + } + + ret = 0; + + child_finish: + pam_end(handle, pam_code | flags); + _exit(ret); + } + + barrier_set_role(&barrier, BARRIER_PARENT); + + /* If the child was forked off successfully it will do all the + * cleanups, so forget about the handle here. */ + handle = NULL; + + /* Unblock SIGTERM again in the parent */ + assert_se(sigprocmask(SIG_SETMASK, &old_ss, NULL) >= 0); + + /* We close the log explicitly here, since the PAM modules + * might have opened it, but we don't want this fd around. */ + closelog(); + + /* Synchronously wait for the child to initialize. We don't care for + * errors as we cannot recover. However, warn loudly if it happens. */ + if (!barrier_place_and_sync(&barrier)) + log_error("PAM initialization failed"); + + strv_free(*env); + *env = e; + + return 0; + +fail: + if (pam_code != PAM_SUCCESS) { + log_error("PAM failed: %s", pam_strerror(handle, pam_code)); + r = -EPERM; /* PAM errors do not map to errno */ + } else + log_error_errno(r, "PAM failed: %m"); + + if (handle) { + if (close_session) + pam_code = pam_close_session(handle, flags); + + pam_end(handle, pam_code | flags); + } + + strv_free(e); + closelog(); + + return r; +} +#endif + +static void rename_process_from_path(const char *path) { + char process_name[11]; + const char *p; + size_t l; + + /* This resulting string must fit in 10 chars (i.e. the length + * of "/sbin/init") to look pretty in /bin/ps */ + + p = basename(path); + if (isempty(p)) { + rename_process("(...)"); + return; + } + + l = strlen(p); + if (l > 8) { + /* The end of the process name is usually more + * interesting, since the first bit might just be + * "systemd-" */ + p = p + l - 8; + l = 8; + } + + process_name[0] = '('; + memcpy(process_name+1, p, l); + process_name[1+l] = ')'; + process_name[1+l+1] = 0; + + rename_process(process_name); +} + +#ifdef HAVE_SECCOMP + +static int apply_seccomp(const ExecContext *c) { + uint32_t negative_action, action; + scmp_filter_ctx *seccomp; + Iterator i; + void *id; + int r; + + assert(c); + + negative_action = c->syscall_errno == 0 ? SCMP_ACT_KILL : SCMP_ACT_ERRNO(c->syscall_errno); + + seccomp = seccomp_init(c->syscall_whitelist ? negative_action : SCMP_ACT_ALLOW); + if (!seccomp) + return -ENOMEM; + + if (c->syscall_archs) { + + SET_FOREACH(id, c->syscall_archs, i) { + r = seccomp_arch_add(seccomp, PTR_TO_UINT32(id) - 1); + if (r == -EEXIST) + continue; + if (r < 0) + goto finish; + } + + } else { + r = seccomp_add_secondary_archs(seccomp); + if (r < 0) + goto finish; + } + + action = c->syscall_whitelist ? SCMP_ACT_ALLOW : negative_action; + SET_FOREACH(id, c->syscall_filter, i) { + r = seccomp_rule_add(seccomp, action, PTR_TO_INT(id) - 1, 0); + if (r < 0) + goto finish; + } + + r = seccomp_attr_set(seccomp, SCMP_FLTATR_CTL_NNP, 0); + if (r < 0) + goto finish; + + r = seccomp_load(seccomp); + +finish: + seccomp_release(seccomp); + return r; +} + +static int apply_address_families(const ExecContext *c) { + scmp_filter_ctx *seccomp; + Iterator i; + int r; + + assert(c); + + seccomp = seccomp_init(SCMP_ACT_ALLOW); + if (!seccomp) + return -ENOMEM; + + r = seccomp_add_secondary_archs(seccomp); + if (r < 0) + goto finish; + + if (c->address_families_whitelist) { + int af, first = 0, last = 0; + void *afp; + + /* If this is a whitelist, we first block the address + * families that are out of range and then everything + * that is not in the set. First, we find the lowest + * and highest address family in the set. */ + + SET_FOREACH(afp, c->address_families, i) { + af = PTR_TO_INT(afp); + + if (af <= 0 || af >= af_max()) + continue; + + if (first == 0 || af < first) + first = af; + + if (last == 0 || af > last) + last = af; + } + + assert((first == 0) == (last == 0)); + + if (first == 0) { + + /* No entries in the valid range, block everything */ + r = seccomp_rule_add( + seccomp, + SCMP_ACT_ERRNO(EPROTONOSUPPORT), + SCMP_SYS(socket), + 0); + if (r < 0) + goto finish; + + } else { + + /* Block everything below the first entry */ + r = seccomp_rule_add( + seccomp, + SCMP_ACT_ERRNO(EPROTONOSUPPORT), + SCMP_SYS(socket), + 1, + SCMP_A0(SCMP_CMP_LT, first)); + if (r < 0) + goto finish; + + /* Block everything above the last entry */ + r = seccomp_rule_add( + seccomp, + SCMP_ACT_ERRNO(EPROTONOSUPPORT), + SCMP_SYS(socket), + 1, + SCMP_A0(SCMP_CMP_GT, last)); + if (r < 0) + goto finish; + + /* Block everything between the first and last + * entry */ + for (af = 1; af < af_max(); af++) { + + if (set_contains(c->address_families, INT_TO_PTR(af))) + continue; + + r = seccomp_rule_add( + seccomp, + SCMP_ACT_ERRNO(EPROTONOSUPPORT), + SCMP_SYS(socket), + 1, + SCMP_A0(SCMP_CMP_EQ, af)); + if (r < 0) + goto finish; + } + } + + } else { + void *af; + + /* If this is a blacklist, then generate one rule for + * each address family that are then combined in OR + * checks. */ + + SET_FOREACH(af, c->address_families, i) { + + r = seccomp_rule_add( + seccomp, + SCMP_ACT_ERRNO(EPROTONOSUPPORT), + SCMP_SYS(socket), + 1, + SCMP_A0(SCMP_CMP_EQ, PTR_TO_INT(af))); + if (r < 0) + goto finish; + } + } + + r = seccomp_attr_set(seccomp, SCMP_FLTATR_CTL_NNP, 0); + if (r < 0) + goto finish; + + r = seccomp_load(seccomp); + +finish: + seccomp_release(seccomp); + return r; +} + +static int apply_memory_deny_write_execute(const ExecContext *c) { + scmp_filter_ctx *seccomp; + int r; + + assert(c); + + seccomp = seccomp_init(SCMP_ACT_ALLOW); + if (!seccomp) + return -ENOMEM; + + r = seccomp_rule_add( + seccomp, + SCMP_ACT_ERRNO(EPERM), + SCMP_SYS(mmap), + 1, + SCMP_A2(SCMP_CMP_MASKED_EQ, PROT_EXEC|PROT_WRITE, PROT_EXEC|PROT_WRITE)); + if (r < 0) + goto finish; + + r = seccomp_rule_add( + seccomp, + SCMP_ACT_ERRNO(EPERM), + SCMP_SYS(mprotect), + 1, + SCMP_A2(SCMP_CMP_MASKED_EQ, PROT_EXEC, PROT_EXEC)); + if (r < 0) + goto finish; + + r = seccomp_attr_set(seccomp, SCMP_FLTATR_CTL_NNP, 0); + if (r < 0) + goto finish; + + r = seccomp_load(seccomp); + +finish: + seccomp_release(seccomp); + return r; +} + +static int apply_restrict_realtime(const ExecContext *c) { + static const int permitted_policies[] = { + SCHED_OTHER, + SCHED_BATCH, + SCHED_IDLE, + }; + + scmp_filter_ctx *seccomp; + unsigned i; + int r, p, max_policy = 0; + + assert(c); + + seccomp = seccomp_init(SCMP_ACT_ALLOW); + if (!seccomp) + return -ENOMEM; + + /* Determine the highest policy constant we want to allow */ + for (i = 0; i < ELEMENTSOF(permitted_policies); i++) + if (permitted_policies[i] > max_policy) + max_policy = permitted_policies[i]; + + /* Go through all policies with lower values than that, and block them -- unless they appear in the + * whitelist. */ + for (p = 0; p < max_policy; p++) { + bool good = false; + + /* Check if this is in the whitelist. */ + for (i = 0; i < ELEMENTSOF(permitted_policies); i++) + if (permitted_policies[i] == p) { + good = true; + break; + } + + if (good) + continue; + + /* Deny this policy */ + r = seccomp_rule_add( + seccomp, + SCMP_ACT_ERRNO(EPERM), + SCMP_SYS(sched_setscheduler), + 1, + SCMP_A1(SCMP_CMP_EQ, p)); + if (r < 0) + goto finish; + } + + /* Blacklist all other policies, i.e. the ones with higher values. Note that all comparisons are unsigned here, + * hence no need no check for < 0 values. */ + r = seccomp_rule_add( + seccomp, + SCMP_ACT_ERRNO(EPERM), + SCMP_SYS(sched_setscheduler), + 1, + SCMP_A1(SCMP_CMP_GT, max_policy)); + if (r < 0) + goto finish; + + r = seccomp_attr_set(seccomp, SCMP_FLTATR_CTL_NNP, 0); + if (r < 0) + goto finish; + + r = seccomp_load(seccomp); + +finish: + seccomp_release(seccomp); + return r; +} + +#endif + +static void do_idle_pipe_dance(int idle_pipe[4]) { + assert(idle_pipe); + + + idle_pipe[1] = safe_close(idle_pipe[1]); + idle_pipe[2] = safe_close(idle_pipe[2]); + + if (idle_pipe[0] >= 0) { + int r; + + r = fd_wait_for_event(idle_pipe[0], POLLHUP, IDLE_TIMEOUT_USEC); + + if (idle_pipe[3] >= 0 && r == 0 /* timeout */) { + ssize_t n; + + /* Signal systemd that we are bored and want to continue. */ + n = write(idle_pipe[3], "x", 1); + if (n > 0) + /* Wait for systemd to react to the signal above. */ + fd_wait_for_event(idle_pipe[0], POLLHUP, IDLE_TIMEOUT2_USEC); + } + + idle_pipe[0] = safe_close(idle_pipe[0]); + + } + + idle_pipe[3] = safe_close(idle_pipe[3]); +} + +static int build_environment( + const ExecContext *c, + const ExecParameters *p, + unsigned n_fds, + const char *home, + const char *username, + const char *shell, + dev_t journal_stream_dev, + ino_t journal_stream_ino, + char ***ret) { + + _cleanup_strv_free_ char **our_env = NULL; + unsigned n_env = 0; + char *x; + + assert(c); + assert(ret); + + our_env = new0(char*, 12); + if (!our_env) + return -ENOMEM; + + if (n_fds > 0) { + _cleanup_free_ char *joined = NULL; + + if (asprintf(&x, "LISTEN_PID="PID_FMT, getpid()) < 0) + return -ENOMEM; + our_env[n_env++] = x; + + if (asprintf(&x, "LISTEN_FDS=%u", n_fds) < 0) + return -ENOMEM; + our_env[n_env++] = x; + + joined = strv_join(p->fd_names, ":"); + if (!joined) + return -ENOMEM; + + x = strjoin("LISTEN_FDNAMES=", joined, NULL); + if (!x) + return -ENOMEM; + our_env[n_env++] = x; + } + + if (p->watchdog_usec > 0) { + if (asprintf(&x, "WATCHDOG_PID="PID_FMT, getpid()) < 0) + return -ENOMEM; + our_env[n_env++] = x; + + if (asprintf(&x, "WATCHDOG_USEC="USEC_FMT, p->watchdog_usec) < 0) + return -ENOMEM; + our_env[n_env++] = x; + } + + if (home) { + x = strappend("HOME=", home); + if (!x) + return -ENOMEM; + our_env[n_env++] = x; + } + + if (username) { + x = strappend("LOGNAME=", username); + if (!x) + return -ENOMEM; + our_env[n_env++] = x; + + x = strappend("USER=", username); + if (!x) + return -ENOMEM; + our_env[n_env++] = x; + } + + if (shell) { + x = strappend("SHELL=", shell); + if (!x) + return -ENOMEM; + our_env[n_env++] = x; + } + + if (is_terminal_input(c->std_input) || + c->std_output == EXEC_OUTPUT_TTY || + c->std_error == EXEC_OUTPUT_TTY || + c->tty_path) { + + x = strdup(default_term_for_tty(exec_context_tty_path(c))); + if (!x) + return -ENOMEM; + our_env[n_env++] = x; + } + + if (journal_stream_dev != 0 && journal_stream_ino != 0) { + if (asprintf(&x, "JOURNAL_STREAM=" DEV_FMT ":" INO_FMT, journal_stream_dev, journal_stream_ino) < 0) + return -ENOMEM; + + our_env[n_env++] = x; + } + + our_env[n_env++] = NULL; + assert(n_env <= 12); + + *ret = our_env; + our_env = NULL; + + return 0; +} + +static int build_pass_environment(const ExecContext *c, char ***ret) { + _cleanup_strv_free_ char **pass_env = NULL; + size_t n_env = 0, n_bufsize = 0; + char **i; + + STRV_FOREACH(i, c->pass_environment) { + _cleanup_free_ char *x = NULL; + char *v; + + v = getenv(*i); + if (!v) + continue; + x = strjoin(*i, "=", v, NULL); + if (!x) + return -ENOMEM; + if (!GREEDY_REALLOC(pass_env, n_bufsize, n_env + 2)) + return -ENOMEM; + pass_env[n_env++] = x; + pass_env[n_env] = NULL; + x = NULL; + } + + *ret = pass_env; + pass_env = NULL; + + return 0; +} + +static bool exec_needs_mount_namespace( + const ExecContext *context, + const ExecParameters *params, + ExecRuntime *runtime) { + + assert(context); + assert(params); + + if (!strv_isempty(context->read_write_paths) || + !strv_isempty(context->read_only_paths) || + !strv_isempty(context->inaccessible_paths)) + return true; + + if (context->mount_flags != 0) + return true; + + if (context->private_tmp && runtime && (runtime->tmp_dir || runtime->var_tmp_dir)) + return true; + + if (context->private_devices || + context->protect_system != PROTECT_SYSTEM_NO || + context->protect_home != PROTECT_HOME_NO) + return true; + + return false; +} + +static int close_remaining_fds( + const ExecParameters *params, + ExecRuntime *runtime, + int socket_fd, + int *fds, unsigned n_fds) { + + unsigned n_dont_close = 0; + int dont_close[n_fds + 7]; + + assert(params); + + if (params->stdin_fd >= 0) + dont_close[n_dont_close++] = params->stdin_fd; + if (params->stdout_fd >= 0) + dont_close[n_dont_close++] = params->stdout_fd; + if (params->stderr_fd >= 0) + dont_close[n_dont_close++] = params->stderr_fd; + + if (socket_fd >= 0) + dont_close[n_dont_close++] = socket_fd; + if (n_fds > 0) { + memcpy(dont_close + n_dont_close, fds, sizeof(int) * n_fds); + n_dont_close += n_fds; + } + + if (runtime) { + if (runtime->netns_storage_socket[0] >= 0) + dont_close[n_dont_close++] = runtime->netns_storage_socket[0]; + if (runtime->netns_storage_socket[1] >= 0) + dont_close[n_dont_close++] = runtime->netns_storage_socket[1]; + } + + return close_all_fds(dont_close, n_dont_close); +} + +static int exec_child( + Unit *unit, + ExecCommand *command, + const ExecContext *context, + const ExecParameters *params, + ExecRuntime *runtime, + char **argv, + int socket_fd, + int *fds, unsigned n_fds, + char **files_env, + int *exit_status) { + + _cleanup_strv_free_ char **our_env = NULL, **pass_env = NULL, **accum_env = NULL, **final_argv = NULL; + _cleanup_free_ char *mac_selinux_context_net = NULL; + const char *username = NULL, *home = NULL, *shell = NULL, *wd; + dev_t journal_stream_dev = 0; + ino_t journal_stream_ino = 0; + bool needs_mount_namespace; + uid_t uid = UID_INVALID; + gid_t gid = GID_INVALID; + int i, r; + + assert(unit); + assert(command); + assert(context); + assert(params); + assert(exit_status); + + rename_process_from_path(command->path); + + /* We reset exactly these signals, since they are the + * only ones we set to SIG_IGN in the main daemon. All + * others we leave untouched because we set them to + * SIG_DFL or a valid handler initially, both of which + * will be demoted to SIG_DFL. */ + (void) default_signals(SIGNALS_CRASH_HANDLER, + SIGNALS_IGNORE, -1); + + if (context->ignore_sigpipe) + (void) ignore_signals(SIGPIPE, -1); + + r = reset_signal_mask(); + if (r < 0) { + *exit_status = EXIT_SIGNAL_MASK; + return r; + } + + if (params->idle_pipe) + do_idle_pipe_dance(params->idle_pipe); + + /* Close sockets very early to make sure we don't + * block init reexecution because it cannot bind its + * sockets */ + + log_forget_fds(); + + r = close_remaining_fds(params, runtime, socket_fd, fds, n_fds); + if (r < 0) { + *exit_status = EXIT_FDS; + return r; + } + + if (!context->same_pgrp) + if (setsid() < 0) { + *exit_status = EXIT_SETSID; + return -errno; + } + + exec_context_tty_reset(context, params); + + if (params->confirm_spawn) { + char response; + + r = ask_for_confirmation(&response, argv); + if (r == -ETIMEDOUT) + write_confirm_message("Confirmation question timed out, assuming positive response.\n"); + else if (r < 0) + write_confirm_message("Couldn't ask confirmation question, assuming positive response: %s\n", strerror(-r)); + else if (response == 's') { + write_confirm_message("Skipping execution.\n"); + *exit_status = EXIT_CONFIRM; + return -ECANCELED; + } else if (response == 'n') { + write_confirm_message("Failing execution.\n"); + *exit_status = 0; + return 0; + } + } + + if (context->user) { + username = context->user; + r = get_user_creds(&username, &uid, &gid, &home, &shell); + if (r < 0) { + *exit_status = EXIT_USER; + return r; + } + } + + if (context->group) { + const char *g = context->group; + + r = get_group_creds(&g, &gid); + if (r < 0) { + *exit_status = EXIT_GROUP; + return r; + } + } + + + /* If a socket is connected to STDIN/STDOUT/STDERR, we + * must sure to drop O_NONBLOCK */ + if (socket_fd >= 0) + (void) fd_nonblock(socket_fd, false); + + r = setup_input(context, params, socket_fd); + if (r < 0) { + *exit_status = EXIT_STDIN; + return r; + } + + r = setup_output(unit, context, params, STDOUT_FILENO, socket_fd, basename(command->path), uid, gid, &journal_stream_dev, &journal_stream_ino); + if (r < 0) { + *exit_status = EXIT_STDOUT; + return r; + } + + r = setup_output(unit, context, params, STDERR_FILENO, socket_fd, basename(command->path), uid, gid, &journal_stream_dev, &journal_stream_ino); + if (r < 0) { + *exit_status = EXIT_STDERR; + return r; + } + + if (params->cgroup_path) { + r = cg_attach_everywhere(params->cgroup_supported, params->cgroup_path, 0, NULL, NULL); + if (r < 0) { + *exit_status = EXIT_CGROUP; + return r; + } + } + + if (context->oom_score_adjust_set) { + char t[DECIMAL_STR_MAX(context->oom_score_adjust)]; + + /* When we can't make this change due to EPERM, then + * let's silently skip over it. User namespaces + * prohibit write access to this file, and we + * shouldn't trip up over that. */ + + sprintf(t, "%i", context->oom_score_adjust); + r = write_string_file("/proc/self/oom_score_adj", t, 0); + if (r == -EPERM || r == -EACCES) { + log_open(); + log_unit_debug_errno(unit, r, "Failed to adjust OOM setting, assuming containerized execution, ignoring: %m"); + log_close(); + } else if (r < 0) { + *exit_status = EXIT_OOM_ADJUST; + return -errno; + } + } + + if (context->nice_set) + if (setpriority(PRIO_PROCESS, 0, context->nice) < 0) { + *exit_status = EXIT_NICE; + return -errno; + } + + if (context->cpu_sched_set) { + struct sched_param param = { + .sched_priority = context->cpu_sched_priority, + }; + + r = sched_setscheduler(0, + context->cpu_sched_policy | + (context->cpu_sched_reset_on_fork ? + SCHED_RESET_ON_FORK : 0), + ¶m); + if (r < 0) { + *exit_status = EXIT_SETSCHEDULER; + return -errno; + } + } + + if (context->cpuset) + if (sched_setaffinity(0, CPU_ALLOC_SIZE(context->cpuset_ncpus), context->cpuset) < 0) { + *exit_status = EXIT_CPUAFFINITY; + return -errno; + } + + if (context->ioprio_set) + if (ioprio_set(IOPRIO_WHO_PROCESS, 0, context->ioprio) < 0) { + *exit_status = EXIT_IOPRIO; + return -errno; + } + + if (context->timer_slack_nsec != NSEC_INFINITY) + if (prctl(PR_SET_TIMERSLACK, context->timer_slack_nsec) < 0) { + *exit_status = EXIT_TIMERSLACK; + return -errno; + } + + if (context->personality != PERSONALITY_INVALID) + if (personality(context->personality) < 0) { + *exit_status = EXIT_PERSONALITY; + return -errno; + } + + if (context->utmp_id) + utmp_put_init_process(context->utmp_id, getpid(), getsid(0), context->tty_path, + context->utmp_mode == EXEC_UTMP_INIT ? INIT_PROCESS : + context->utmp_mode == EXEC_UTMP_LOGIN ? LOGIN_PROCESS : + USER_PROCESS, + username ? "root" : context->user); + + if (context->user && is_terminal_input(context->std_input)) { + r = chown_terminal(STDIN_FILENO, uid); + if (r < 0) { + *exit_status = EXIT_STDIN; + return r; + } + } + + /* If delegation is enabled we'll pass ownership of the cgroup + * (but only in systemd's own controller hierarchy!) to the + * user of the new process. */ + if (params->cgroup_path && context->user && params->cgroup_delegate) { + r = cg_set_task_access(SYSTEMD_CGROUP_CONTROLLER, params->cgroup_path, 0644, uid, gid); + if (r < 0) { + *exit_status = EXIT_CGROUP; + return r; + } + + + r = cg_set_group_access(SYSTEMD_CGROUP_CONTROLLER, params->cgroup_path, 0755, uid, gid); + if (r < 0) { + *exit_status = EXIT_CGROUP; + return r; + } + } + + if (!strv_isempty(context->runtime_directory) && params->runtime_prefix) { + char **rt; + + STRV_FOREACH(rt, context->runtime_directory) { + _cleanup_free_ char *p; + + p = strjoin(params->runtime_prefix, "/", *rt, NULL); + if (!p) { + *exit_status = EXIT_RUNTIME_DIRECTORY; + return -ENOMEM; + } + + r = mkdir_p_label(p, context->runtime_directory_mode); + if (r < 0) { + *exit_status = EXIT_RUNTIME_DIRECTORY; + return r; + } + + r = chmod_and_chown(p, context->runtime_directory_mode, uid, gid); + if (r < 0) { + *exit_status = EXIT_RUNTIME_DIRECTORY; + return r; + } + } + } + + r = build_environment( + context, + params, + n_fds, + home, + username, + shell, + journal_stream_dev, + journal_stream_ino, + &our_env); + if (r < 0) { + *exit_status = EXIT_MEMORY; + return r; + } + + r = build_pass_environment(context, &pass_env); + if (r < 0) { + *exit_status = EXIT_MEMORY; + return r; + } + + accum_env = strv_env_merge(5, + params->environment, + our_env, + pass_env, + context->environment, + files_env, + NULL); + if (!accum_env) { + *exit_status = EXIT_MEMORY; + return -ENOMEM; + } + accum_env = strv_env_clean(accum_env); + + umask(context->umask); + + if (params->apply_permissions && !command->privileged) { + r = enforce_groups(context, username, gid); + if (r < 0) { + *exit_status = EXIT_GROUP; + return r; + } +#ifdef HAVE_SMACK + if (context->smack_process_label) { + r = mac_smack_apply_pid(0, context->smack_process_label); + if (r < 0) { + *exit_status = EXIT_SMACK_PROCESS_LABEL; + return r; + } + } +#ifdef SMACK_DEFAULT_PROCESS_LABEL + else { + _cleanup_free_ char *exec_label = NULL; + + r = mac_smack_read(command->path, SMACK_ATTR_EXEC, &exec_label); + if (r < 0 && r != -ENODATA && r != -EOPNOTSUPP) { + *exit_status = EXIT_SMACK_PROCESS_LABEL; + return r; + } + + r = mac_smack_apply_pid(0, exec_label ? : SMACK_DEFAULT_PROCESS_LABEL); + if (r < 0) { + *exit_status = EXIT_SMACK_PROCESS_LABEL; + return r; + } + } +#endif +#endif +#ifdef HAVE_PAM + if (context->pam_name && username) { + r = setup_pam(context->pam_name, username, uid, context->tty_path, &accum_env, fds, n_fds); + if (r < 0) { + *exit_status = EXIT_PAM; + return r; + } + } +#endif + } + + if (context->private_network && runtime && runtime->netns_storage_socket[0] >= 0) { + r = setup_netns(runtime->netns_storage_socket); + if (r < 0) { + *exit_status = EXIT_NETWORK; + return r; + } + } + + needs_mount_namespace = exec_needs_mount_namespace(context, params, runtime); + + if (needs_mount_namespace) { + char *tmp = NULL, *var = NULL; + + /* The runtime struct only contains the parent + * of the private /tmp, which is + * non-accessible to world users. Inside of it + * there's a /tmp that is sticky, and that's + * the one we want to use here. */ + + if (context->private_tmp && runtime) { + if (runtime->tmp_dir) + tmp = strjoina(runtime->tmp_dir, "/tmp"); + if (runtime->var_tmp_dir) + var = strjoina(runtime->var_tmp_dir, "/tmp"); + } + + r = setup_namespace( + params->apply_chroot ? context->root_directory : NULL, + context->read_write_paths, + context->read_only_paths, + context->inaccessible_paths, + tmp, + var, + context->private_devices, + context->protect_home, + context->protect_system, + context->mount_flags); + + /* If we couldn't set up the namespace this is + * probably due to a missing capability. In this case, + * silently proceeed. */ + if (r == -EPERM || r == -EACCES) { + log_open(); + log_unit_debug_errno(unit, r, "Failed to set up namespace, assuming containerized execution, ignoring: %m"); + log_close(); + } else if (r < 0) { + *exit_status = EXIT_NAMESPACE; + return r; + } + } + + if (context->working_directory_home) + wd = home; + else if (context->working_directory) + wd = context->working_directory; + else + wd = "/"; + + if (params->apply_chroot) { + if (!needs_mount_namespace && context->root_directory) + if (chroot(context->root_directory) < 0) { + *exit_status = EXIT_CHROOT; + return -errno; + } + + if (chdir(wd) < 0 && + !context->working_directory_missing_ok) { + *exit_status = EXIT_CHDIR; + return -errno; + } + } else { + const char *d; + + d = strjoina(strempty(context->root_directory), "/", strempty(wd)); + if (chdir(d) < 0 && + !context->working_directory_missing_ok) { + *exit_status = EXIT_CHDIR; + return -errno; + } + } + +#ifdef HAVE_SELINUX + if (params->apply_permissions && mac_selinux_use() && params->selinux_context_net && socket_fd >= 0 && !command->privileged) { + r = mac_selinux_get_child_mls_label(socket_fd, command->path, context->selinux_context, &mac_selinux_context_net); + if (r < 0) { + *exit_status = EXIT_SELINUX_CONTEXT; + return r; + } + } +#endif + + /* We repeat the fd closing here, to make sure that + * nothing is leaked from the PAM modules. Note that + * we are more aggressive this time since socket_fd + * and the netns fds we don't need anymore. The custom + * endpoint fd was needed to upload the policy and can + * now be closed as well. */ + r = close_all_fds(fds, n_fds); + if (r >= 0) + r = shift_fds(fds, n_fds); + if (r >= 0) + r = flags_fds(fds, n_fds, context->non_blocking); + if (r < 0) { + *exit_status = EXIT_FDS; + return r; + } + + if (params->apply_permissions && !command->privileged) { + + bool use_address_families = context->address_families_whitelist || + !set_isempty(context->address_families); + bool use_syscall_filter = context->syscall_whitelist || + !set_isempty(context->syscall_filter) || + !set_isempty(context->syscall_archs); + int secure_bits = context->secure_bits; + + for (i = 0; i < _RLIMIT_MAX; i++) { + + if (!context->rlimit[i]) + continue; + + r = setrlimit_closest(i, context->rlimit[i]); + if (r < 0) { + *exit_status = EXIT_LIMITS; + return r; + } + } + + /* Set the RTPRIO resource limit to 0, but only if nothing else was explicitly requested. */ + if (context->restrict_realtime && !context->rlimit[RLIMIT_RTPRIO]) { + if (setrlimit(RLIMIT_RTPRIO, &RLIMIT_MAKE_CONST(0)) < 0) { + *exit_status = EXIT_LIMITS; + return -errno; + } + } + + if (!cap_test_all(context->capability_bounding_set)) { + r = capability_bounding_set_drop(context->capability_bounding_set, false); + if (r < 0) { + *exit_status = EXIT_CAPABILITIES; + return r; + } + } + + /* This is done before enforce_user, but ambient set + * does not survive over setresuid() if keep_caps is not set. */ + if (context->capability_ambient_set != 0) { + r = capability_ambient_set_apply(context->capability_ambient_set, true); + if (r < 0) { + *exit_status = EXIT_CAPABILITIES; + return r; + } + } + + if (context->user) { + r = enforce_user(context, uid); + if (r < 0) { + *exit_status = EXIT_USER; + return r; + } + if (context->capability_ambient_set != 0) { + + /* Fix the ambient capabilities after user change. */ + r = capability_ambient_set_apply(context->capability_ambient_set, false); + if (r < 0) { + *exit_status = EXIT_CAPABILITIES; + return r; + } + + /* If we were asked to change user and ambient capabilities + * were requested, we had to add keep-caps to the securebits + * so that we would maintain the inherited capability set + * through the setresuid(). Make sure that the bit is added + * also to the context secure_bits so that we don't try to + * drop the bit away next. */ + + secure_bits |= 1<no_new_privileges || + (!have_effective_cap(CAP_SYS_ADMIN) && (use_address_families || context->memory_deny_write_execute || context->restrict_realtime || use_syscall_filter))) + if (prctl(PR_SET_NO_NEW_PRIVS, 1, 0, 0, 0) < 0) { + *exit_status = EXIT_NO_NEW_PRIVILEGES; + return -errno; + } + +#ifdef HAVE_SECCOMP + if (use_address_families) { + r = apply_address_families(context); + if (r < 0) { + *exit_status = EXIT_ADDRESS_FAMILIES; + return r; + } + } + + if (context->memory_deny_write_execute) { + r = apply_memory_deny_write_execute(context); + if (r < 0) { + *exit_status = EXIT_SECCOMP; + return r; + } + } + + if (context->restrict_realtime) { + r = apply_restrict_realtime(context); + if (r < 0) { + *exit_status = EXIT_SECCOMP; + return r; + } + } + + if (use_syscall_filter) { + r = apply_seccomp(context); + if (r < 0) { + *exit_status = EXIT_SECCOMP; + return r; + } + } +#endif + +#ifdef HAVE_SELINUX + if (mac_selinux_use()) { + char *exec_context = mac_selinux_context_net ?: context->selinux_context; + + if (exec_context) { + r = setexeccon(exec_context); + if (r < 0) { + *exit_status = EXIT_SELINUX_CONTEXT; + return r; + } + } + } +#endif + +#ifdef HAVE_APPARMOR + if (context->apparmor_profile && mac_apparmor_use()) { + r = aa_change_onexec(context->apparmor_profile); + if (r < 0 && !context->apparmor_profile_ignore) { + *exit_status = EXIT_APPARMOR_PROFILE; + return -errno; + } + } +#endif + } + + final_argv = replace_env_argv(argv, accum_env); + if (!final_argv) { + *exit_status = EXIT_MEMORY; + return -ENOMEM; + } + + if (_unlikely_(log_get_max_level() >= LOG_DEBUG)) { + _cleanup_free_ char *line; + + line = exec_command_line(final_argv); + if (line) { + log_open(); + log_struct(LOG_DEBUG, + LOG_UNIT_ID(unit), + "EXECUTABLE=%s", command->path, + LOG_UNIT_MESSAGE(unit, "Executing: %s", line), + NULL); + log_close(); + } + } + + execve(command->path, final_argv, accum_env); + *exit_status = EXIT_EXEC; + return -errno; +} + +int exec_spawn(Unit *unit, + ExecCommand *command, + const ExecContext *context, + const ExecParameters *params, + ExecRuntime *runtime, + pid_t *ret) { + + _cleanup_strv_free_ char **files_env = NULL; + int *fds = NULL; unsigned n_fds = 0; + _cleanup_free_ char *line = NULL; + int socket_fd, r; + char **argv; + pid_t pid; + + assert(unit); + assert(command); + assert(context); + assert(ret); + assert(params); + assert(params->fds || params->n_fds <= 0); + + if (context->std_input == EXEC_INPUT_SOCKET || + context->std_output == EXEC_OUTPUT_SOCKET || + context->std_error == EXEC_OUTPUT_SOCKET) { + + if (params->n_fds != 1) { + log_unit_error(unit, "Got more than one socket."); + return -EINVAL; + } + + socket_fd = params->fds[0]; + } else { + socket_fd = -1; + fds = params->fds; + n_fds = params->n_fds; + } + + r = exec_context_load_environment(unit, context, &files_env); + if (r < 0) + return log_unit_error_errno(unit, r, "Failed to load environment files: %m"); + + argv = params->argv ?: command->argv; + line = exec_command_line(argv); + if (!line) + return log_oom(); + + log_struct(LOG_DEBUG, + LOG_UNIT_ID(unit), + LOG_UNIT_MESSAGE(unit, "About to execute: %s", line), + "EXECUTABLE=%s", command->path, + NULL); + pid = fork(); + if (pid < 0) + return log_unit_error_errno(unit, errno, "Failed to fork: %m"); + + if (pid == 0) { + int exit_status; + + r = exec_child(unit, + command, + context, + params, + runtime, + argv, + socket_fd, + fds, n_fds, + files_env, + &exit_status); + if (r < 0) { + log_open(); + log_struct_errno(LOG_ERR, r, + LOG_MESSAGE_ID(SD_MESSAGE_SPAWN_FAILED), + LOG_UNIT_ID(unit), + LOG_UNIT_MESSAGE(unit, "Failed at step %s spawning %s: %m", + exit_status_to_string(exit_status, EXIT_STATUS_SYSTEMD), + command->path), + "EXECUTABLE=%s", command->path, + NULL); + } + + _exit(exit_status); + } + + log_unit_debug(unit, "Forked %s as "PID_FMT, command->path, pid); + + /* We add the new process to the cgroup both in the child (so + * that we can be sure that no user code is ever executed + * outside of the cgroup) and in the parent (so that we can be + * sure that when we kill the cgroup the process will be + * killed too). */ + if (params->cgroup_path) + (void) cg_attach(SYSTEMD_CGROUP_CONTROLLER, params->cgroup_path, pid); + + exec_status_start(&command->exec_status, pid); + + *ret = pid; + return 0; +} + +void exec_context_init(ExecContext *c) { + assert(c); + + c->umask = 0022; + c->ioprio = IOPRIO_PRIO_VALUE(IOPRIO_CLASS_BE, 0); + c->cpu_sched_policy = SCHED_OTHER; + c->syslog_priority = LOG_DAEMON|LOG_INFO; + c->syslog_level_prefix = true; + c->ignore_sigpipe = true; + c->timer_slack_nsec = NSEC_INFINITY; + c->personality = PERSONALITY_INVALID; + c->runtime_directory_mode = 0755; + c->capability_bounding_set = CAP_ALL; +} + +void exec_context_done(ExecContext *c) { + unsigned l; + + assert(c); + + c->environment = strv_free(c->environment); + c->environment_files = strv_free(c->environment_files); + c->pass_environment = strv_free(c->pass_environment); + + for (l = 0; l < ELEMENTSOF(c->rlimit); l++) + c->rlimit[l] = mfree(c->rlimit[l]); + + c->working_directory = mfree(c->working_directory); + c->root_directory = mfree(c->root_directory); + c->tty_path = mfree(c->tty_path); + c->syslog_identifier = mfree(c->syslog_identifier); + c->user = mfree(c->user); + c->group = mfree(c->group); + + c->supplementary_groups = strv_free(c->supplementary_groups); + + c->pam_name = mfree(c->pam_name); + + c->read_only_paths = strv_free(c->read_only_paths); + c->read_write_paths = strv_free(c->read_write_paths); + c->inaccessible_paths = strv_free(c->inaccessible_paths); + + if (c->cpuset) + CPU_FREE(c->cpuset); + + c->utmp_id = mfree(c->utmp_id); + c->selinux_context = mfree(c->selinux_context); + c->apparmor_profile = mfree(c->apparmor_profile); + + c->syscall_filter = set_free(c->syscall_filter); + c->syscall_archs = set_free(c->syscall_archs); + c->address_families = set_free(c->address_families); + + c->runtime_directory = strv_free(c->runtime_directory); +} + +int exec_context_destroy_runtime_directory(ExecContext *c, const char *runtime_prefix) { + char **i; + + assert(c); + + if (!runtime_prefix) + return 0; + + STRV_FOREACH(i, c->runtime_directory) { + _cleanup_free_ char *p; + + p = strjoin(runtime_prefix, "/", *i, NULL); + if (!p) + return -ENOMEM; + + /* We execute this synchronously, since we need to be + * sure this is gone when we start the service + * next. */ + (void) rm_rf(p, REMOVE_ROOT); + } + + return 0; +} + +void exec_command_done(ExecCommand *c) { + assert(c); + + c->path = mfree(c->path); + + c->argv = strv_free(c->argv); +} + +void exec_command_done_array(ExecCommand *c, unsigned n) { + unsigned i; + + for (i = 0; i < n; i++) + exec_command_done(c+i); +} + +ExecCommand* exec_command_free_list(ExecCommand *c) { + ExecCommand *i; + + while ((i = c)) { + LIST_REMOVE(command, c, i); + exec_command_done(i); + free(i); + } + + return NULL; +} + +void exec_command_free_array(ExecCommand **c, unsigned n) { + unsigned i; + + for (i = 0; i < n; i++) + c[i] = exec_command_free_list(c[i]); +} + +typedef struct InvalidEnvInfo { + Unit *unit; + const char *path; +} InvalidEnvInfo; + +static void invalid_env(const char *p, void *userdata) { + InvalidEnvInfo *info = userdata; + + log_unit_error(info->unit, "Ignoring invalid environment assignment '%s': %s", p, info->path); +} + +int exec_context_load_environment(Unit *unit, const ExecContext *c, char ***l) { + char **i, **r = NULL; + + assert(c); + assert(l); + + STRV_FOREACH(i, c->environment_files) { + char *fn; + int k; + bool ignore = false; + char **p; + _cleanup_globfree_ glob_t pglob = {}; + int count, n; + + fn = *i; + + if (fn[0] == '-') { + ignore = true; + fn++; + } + + if (!path_is_absolute(fn)) { + if (ignore) + continue; + + strv_free(r); + return -EINVAL; + } + + /* Filename supports globbing, take all matching files */ + errno = 0; + if (glob(fn, 0, NULL, &pglob) != 0) { + if (ignore) + continue; + + strv_free(r); + return errno > 0 ? -errno : -EINVAL; + } + count = pglob.gl_pathc; + if (count == 0) { + if (ignore) + continue; + + strv_free(r); + return -EINVAL; + } + for (n = 0; n < count; n++) { + k = load_env_file(NULL, pglob.gl_pathv[n], NULL, &p); + if (k < 0) { + if (ignore) + continue; + + strv_free(r); + return k; + } + /* Log invalid environment variables with filename */ + if (p) { + InvalidEnvInfo info = { + .unit = unit, + .path = pglob.gl_pathv[n] + }; + + p = strv_env_clean_with_callback(p, invalid_env, &info); + } + + if (r == NULL) + r = p; + else { + char **m; + + m = strv_env_merge(2, r, p); + strv_free(r); + strv_free(p); + if (!m) + return -ENOMEM; + + r = m; + } + } + } + + *l = r; + + return 0; +} + +static bool tty_may_match_dev_console(const char *tty) { + _cleanup_free_ char *active = NULL; + char *console; + + if (!tty) + return true; + + if (startswith(tty, "/dev/")) + tty += 5; + + /* trivial identity? */ + if (streq(tty, "console")) + return true; + + console = resolve_dev_console(&active); + /* if we could not resolve, assume it may */ + if (!console) + return true; + + /* "tty0" means the active VC, so it may be the same sometimes */ + return streq(console, tty) || (streq(console, "tty0") && tty_is_vc(tty)); +} + +bool exec_context_may_touch_console(ExecContext *ec) { + + return (ec->tty_reset || + ec->tty_vhangup || + ec->tty_vt_disallocate || + is_terminal_input(ec->std_input) || + is_terminal_output(ec->std_output) || + is_terminal_output(ec->std_error)) && + tty_may_match_dev_console(exec_context_tty_path(ec)); +} + +static void strv_fprintf(FILE *f, char **l) { + char **g; + + assert(f); + + STRV_FOREACH(g, l) + fprintf(f, " %s", *g); +} + +void exec_context_dump(ExecContext *c, FILE* f, const char *prefix) { + char **e, **d; + unsigned i; + + assert(c); + assert(f); + + prefix = strempty(prefix); + + fprintf(f, + "%sUMask: %04o\n" + "%sWorkingDirectory: %s\n" + "%sRootDirectory: %s\n" + "%sNonBlocking: %s\n" + "%sPrivateTmp: %s\n" + "%sPrivateNetwork: %s\n" + "%sPrivateDevices: %s\n" + "%sProtectHome: %s\n" + "%sProtectSystem: %s\n" + "%sIgnoreSIGPIPE: %s\n" + "%sMemoryDenyWriteExecute: %s\n" + "%sRestrictRealtime: %s\n", + prefix, c->umask, + prefix, c->working_directory ? c->working_directory : "/", + prefix, c->root_directory ? c->root_directory : "/", + prefix, yes_no(c->non_blocking), + prefix, yes_no(c->private_tmp), + prefix, yes_no(c->private_network), + prefix, yes_no(c->private_devices), + prefix, protect_home_to_string(c->protect_home), + prefix, protect_system_to_string(c->protect_system), + prefix, yes_no(c->ignore_sigpipe), + prefix, yes_no(c->memory_deny_write_execute), + prefix, yes_no(c->restrict_realtime)); + + STRV_FOREACH(e, c->environment) + fprintf(f, "%sEnvironment: %s\n", prefix, *e); + + STRV_FOREACH(e, c->environment_files) + fprintf(f, "%sEnvironmentFile: %s\n", prefix, *e); + + STRV_FOREACH(e, c->pass_environment) + fprintf(f, "%sPassEnvironment: %s\n", prefix, *e); + + fprintf(f, "%sRuntimeDirectoryMode: %04o\n", prefix, c->runtime_directory_mode); + + STRV_FOREACH(d, c->runtime_directory) + fprintf(f, "%sRuntimeDirectory: %s\n", prefix, *d); + + if (c->nice_set) + fprintf(f, + "%sNice: %i\n", + prefix, c->nice); + + if (c->oom_score_adjust_set) + fprintf(f, + "%sOOMScoreAdjust: %i\n", + prefix, c->oom_score_adjust); + + for (i = 0; i < RLIM_NLIMITS; i++) + if (c->rlimit[i]) { + fprintf(f, "%s%s: " RLIM_FMT "\n", + prefix, rlimit_to_string(i), c->rlimit[i]->rlim_max); + fprintf(f, "%s%sSoft: " RLIM_FMT "\n", + prefix, rlimit_to_string(i), c->rlimit[i]->rlim_cur); + } + + if (c->ioprio_set) { + _cleanup_free_ char *class_str = NULL; + + ioprio_class_to_string_alloc(IOPRIO_PRIO_CLASS(c->ioprio), &class_str); + fprintf(f, + "%sIOSchedulingClass: %s\n" + "%sIOPriority: %i\n", + prefix, strna(class_str), + prefix, (int) IOPRIO_PRIO_DATA(c->ioprio)); + } + + if (c->cpu_sched_set) { + _cleanup_free_ char *policy_str = NULL; + + sched_policy_to_string_alloc(c->cpu_sched_policy, &policy_str); + fprintf(f, + "%sCPUSchedulingPolicy: %s\n" + "%sCPUSchedulingPriority: %i\n" + "%sCPUSchedulingResetOnFork: %s\n", + prefix, strna(policy_str), + prefix, c->cpu_sched_priority, + prefix, yes_no(c->cpu_sched_reset_on_fork)); + } + + if (c->cpuset) { + fprintf(f, "%sCPUAffinity:", prefix); + for (i = 0; i < c->cpuset_ncpus; i++) + if (CPU_ISSET_S(i, CPU_ALLOC_SIZE(c->cpuset_ncpus), c->cpuset)) + fprintf(f, " %u", i); + fputs("\n", f); + } + + if (c->timer_slack_nsec != NSEC_INFINITY) + fprintf(f, "%sTimerSlackNSec: "NSEC_FMT "\n", prefix, c->timer_slack_nsec); + + fprintf(f, + "%sStandardInput: %s\n" + "%sStandardOutput: %s\n" + "%sStandardError: %s\n", + prefix, exec_input_to_string(c->std_input), + prefix, exec_output_to_string(c->std_output), + prefix, exec_output_to_string(c->std_error)); + + if (c->tty_path) + fprintf(f, + "%sTTYPath: %s\n" + "%sTTYReset: %s\n" + "%sTTYVHangup: %s\n" + "%sTTYVTDisallocate: %s\n", + prefix, c->tty_path, + prefix, yes_no(c->tty_reset), + prefix, yes_no(c->tty_vhangup), + prefix, yes_no(c->tty_vt_disallocate)); + + if (c->std_output == EXEC_OUTPUT_SYSLOG || + c->std_output == EXEC_OUTPUT_KMSG || + c->std_output == EXEC_OUTPUT_JOURNAL || + c->std_output == EXEC_OUTPUT_SYSLOG_AND_CONSOLE || + c->std_output == EXEC_OUTPUT_KMSG_AND_CONSOLE || + c->std_output == EXEC_OUTPUT_JOURNAL_AND_CONSOLE || + c->std_error == EXEC_OUTPUT_SYSLOG || + c->std_error == EXEC_OUTPUT_KMSG || + c->std_error == EXEC_OUTPUT_JOURNAL || + c->std_error == EXEC_OUTPUT_SYSLOG_AND_CONSOLE || + c->std_error == EXEC_OUTPUT_KMSG_AND_CONSOLE || + c->std_error == EXEC_OUTPUT_JOURNAL_AND_CONSOLE) { + + _cleanup_free_ char *fac_str = NULL, *lvl_str = NULL; + + log_facility_unshifted_to_string_alloc(c->syslog_priority >> 3, &fac_str); + log_level_to_string_alloc(LOG_PRI(c->syslog_priority), &lvl_str); + + fprintf(f, + "%sSyslogFacility: %s\n" + "%sSyslogLevel: %s\n", + prefix, strna(fac_str), + prefix, strna(lvl_str)); + } + + if (c->secure_bits) + fprintf(f, "%sSecure Bits:%s%s%s%s%s%s\n", + prefix, + (c->secure_bits & 1<secure_bits & 1<secure_bits & 1<secure_bits & 1<secure_bits & 1<secure_bits & 1<capability_bounding_set != CAP_ALL) { + unsigned long l; + fprintf(f, "%sCapabilityBoundingSet:", prefix); + + for (l = 0; l <= cap_last_cap(); l++) + if (c->capability_bounding_set & (UINT64_C(1) << l)) + fprintf(f, " %s", strna(capability_to_name(l))); + + fputs("\n", f); + } + + if (c->capability_ambient_set != 0) { + unsigned long l; + fprintf(f, "%sAmbientCapabilities:", prefix); + + for (l = 0; l <= cap_last_cap(); l++) + if (c->capability_ambient_set & (UINT64_C(1) << l)) + fprintf(f, " %s", strna(capability_to_name(l))); + + fputs("\n", f); + } + + if (c->user) + fprintf(f, "%sUser: %s\n", prefix, c->user); + if (c->group) + fprintf(f, "%sGroup: %s\n", prefix, c->group); + + if (strv_length(c->supplementary_groups) > 0) { + fprintf(f, "%sSupplementaryGroups:", prefix); + strv_fprintf(f, c->supplementary_groups); + fputs("\n", f); + } + + if (c->pam_name) + fprintf(f, "%sPAMName: %s\n", prefix, c->pam_name); + + if (strv_length(c->read_write_paths) > 0) { + fprintf(f, "%sReadWritePaths:", prefix); + strv_fprintf(f, c->read_write_paths); + fputs("\n", f); + } + + if (strv_length(c->read_only_paths) > 0) { + fprintf(f, "%sReadOnlyPaths:", prefix); + strv_fprintf(f, c->read_only_paths); + fputs("\n", f); + } + + if (strv_length(c->inaccessible_paths) > 0) { + fprintf(f, "%sInaccessiblePaths:", prefix); + strv_fprintf(f, c->inaccessible_paths); + fputs("\n", f); + } + + if (c->utmp_id) + fprintf(f, + "%sUtmpIdentifier: %s\n", + prefix, c->utmp_id); + + if (c->selinux_context) + fprintf(f, + "%sSELinuxContext: %s%s\n", + prefix, c->selinux_context_ignore ? "-" : "", c->selinux_context); + + if (c->personality != PERSONALITY_INVALID) + fprintf(f, + "%sPersonality: %s\n", + prefix, strna(personality_to_string(c->personality))); + + if (c->syscall_filter) { +#ifdef HAVE_SECCOMP + Iterator j; + void *id; + bool first = true; +#endif + + fprintf(f, + "%sSystemCallFilter: ", + prefix); + + if (!c->syscall_whitelist) + fputc('~', f); + +#ifdef HAVE_SECCOMP + SET_FOREACH(id, c->syscall_filter, j) { + _cleanup_free_ char *name = NULL; + + if (first) + first = false; + else + fputc(' ', f); + + name = seccomp_syscall_resolve_num_arch(SCMP_ARCH_NATIVE, PTR_TO_INT(id) - 1); + fputs(strna(name), f); + } +#endif + + fputc('\n', f); + } + + if (c->syscall_archs) { +#ifdef HAVE_SECCOMP + Iterator j; + void *id; +#endif + + fprintf(f, + "%sSystemCallArchitectures:", + prefix); + +#ifdef HAVE_SECCOMP + SET_FOREACH(id, c->syscall_archs, j) + fprintf(f, " %s", strna(seccomp_arch_to_string(PTR_TO_UINT32(id) - 1))); +#endif + fputc('\n', f); + } + + if (c->syscall_errno > 0) + fprintf(f, + "%sSystemCallErrorNumber: %s\n", + prefix, strna(errno_to_name(c->syscall_errno))); + + if (c->apparmor_profile) + fprintf(f, + "%sAppArmorProfile: %s%s\n", + prefix, c->apparmor_profile_ignore ? "-" : "", c->apparmor_profile); +} + +bool exec_context_maintains_privileges(ExecContext *c) { + assert(c); + + /* Returns true if the process forked off would run under + * an unchanged UID or as root. */ + + if (!c->user) + return true; + + if (streq(c->user, "root") || streq(c->user, "0")) + return true; + + return false; +} + +void exec_status_start(ExecStatus *s, pid_t pid) { + assert(s); + + zero(*s); + s->pid = pid; + dual_timestamp_get(&s->start_timestamp); +} + +void exec_status_exit(ExecStatus *s, ExecContext *context, pid_t pid, int code, int status) { + assert(s); + + if (s->pid && s->pid != pid) + zero(*s); + + s->pid = pid; + dual_timestamp_get(&s->exit_timestamp); + + s->code = code; + s->status = status; + + if (context) { + if (context->utmp_id) + utmp_put_dead_process(context->utmp_id, pid, code, status); + + exec_context_tty_reset(context, NULL); + } +} + +void exec_status_dump(ExecStatus *s, FILE *f, const char *prefix) { + char buf[FORMAT_TIMESTAMP_MAX]; + + assert(s); + assert(f); + + if (s->pid <= 0) + return; + + prefix = strempty(prefix); + + fprintf(f, + "%sPID: "PID_FMT"\n", + prefix, s->pid); + + if (s->start_timestamp.realtime > 0) + fprintf(f, + "%sStart Timestamp: %s\n", + prefix, format_timestamp(buf, sizeof(buf), s->start_timestamp.realtime)); + + if (s->exit_timestamp.realtime > 0) + fprintf(f, + "%sExit Timestamp: %s\n" + "%sExit Code: %s\n" + "%sExit Status: %i\n", + prefix, format_timestamp(buf, sizeof(buf), s->exit_timestamp.realtime), + prefix, sigchld_code_to_string(s->code), + prefix, s->status); +} + +char *exec_command_line(char **argv) { + size_t k; + char *n, *p, **a; + bool first = true; + + assert(argv); + + k = 1; + STRV_FOREACH(a, argv) + k += strlen(*a)+3; + + if (!(n = new(char, k))) + return NULL; + + p = n; + STRV_FOREACH(a, argv) { + + if (!first) + *(p++) = ' '; + else + first = false; + + if (strpbrk(*a, WHITESPACE)) { + *(p++) = '\''; + p = stpcpy(p, *a); + *(p++) = '\''; + } else + p = stpcpy(p, *a); + + } + + *p = 0; + + /* FIXME: this doesn't really handle arguments that have + * spaces and ticks in them */ + + return n; +} + +void exec_command_dump(ExecCommand *c, FILE *f, const char *prefix) { + _cleanup_free_ char *cmd = NULL; + const char *prefix2; + + assert(c); + assert(f); + + prefix = strempty(prefix); + prefix2 = strjoina(prefix, "\t"); + + cmd = exec_command_line(c->argv); + fprintf(f, + "%sCommand Line: %s\n", + prefix, cmd ? cmd : strerror(ENOMEM)); + + exec_status_dump(&c->exec_status, f, prefix2); +} + +void exec_command_dump_list(ExecCommand *c, FILE *f, const char *prefix) { + assert(f); + + prefix = strempty(prefix); + + LIST_FOREACH(command, c, c) + exec_command_dump(c, f, prefix); +} + +void exec_command_append_list(ExecCommand **l, ExecCommand *e) { + ExecCommand *end; + + assert(l); + assert(e); + + if (*l) { + /* It's kind of important, that we keep the order here */ + LIST_FIND_TAIL(command, *l, end); + LIST_INSERT_AFTER(command, *l, end, e); + } else + *l = e; +} + +int exec_command_set(ExecCommand *c, const char *path, ...) { + va_list ap; + char **l, *p; + + assert(c); + assert(path); + + va_start(ap, path); + l = strv_new_ap(path, ap); + va_end(ap); + + if (!l) + return -ENOMEM; + + p = strdup(path); + if (!p) { + strv_free(l); + return -ENOMEM; + } + + free(c->path); + c->path = p; + + strv_free(c->argv); + c->argv = l; + + return 0; +} + +int exec_command_append(ExecCommand *c, const char *path, ...) { + _cleanup_strv_free_ char **l = NULL; + va_list ap; + int r; + + assert(c); + assert(path); + + va_start(ap, path); + l = strv_new_ap(path, ap); + va_end(ap); + + if (!l) + return -ENOMEM; + + r = strv_extend_strv(&c->argv, l, false); + if (r < 0) + return r; + + return 0; +} + + +static int exec_runtime_allocate(ExecRuntime **rt) { + + if (*rt) + return 0; + + *rt = new0(ExecRuntime, 1); + if (!*rt) + return -ENOMEM; + + (*rt)->n_ref = 1; + (*rt)->netns_storage_socket[0] = (*rt)->netns_storage_socket[1] = -1; + + return 0; +} + +int exec_runtime_make(ExecRuntime **rt, ExecContext *c, const char *id) { + int r; + + assert(rt); + assert(c); + assert(id); + + if (*rt) + return 1; + + if (!c->private_network && !c->private_tmp) + return 0; + + r = exec_runtime_allocate(rt); + if (r < 0) + return r; + + if (c->private_network && (*rt)->netns_storage_socket[0] < 0) { + if (socketpair(AF_UNIX, SOCK_DGRAM|SOCK_CLOEXEC, 0, (*rt)->netns_storage_socket) < 0) + return -errno; + } + + if (c->private_tmp && !(*rt)->tmp_dir) { + r = setup_tmp_dirs(id, &(*rt)->tmp_dir, &(*rt)->var_tmp_dir); + if (r < 0) + return r; + } + + return 1; +} + +ExecRuntime *exec_runtime_ref(ExecRuntime *r) { + assert(r); + assert(r->n_ref > 0); + + r->n_ref++; + return r; +} + +ExecRuntime *exec_runtime_unref(ExecRuntime *r) { + + if (!r) + return NULL; + + assert(r->n_ref > 0); + + r->n_ref--; + if (r->n_ref > 0) + return NULL; + + free(r->tmp_dir); + free(r->var_tmp_dir); + safe_close_pair(r->netns_storage_socket); + free(r); + + return NULL; +} + +int exec_runtime_serialize(Unit *u, ExecRuntime *rt, FILE *f, FDSet *fds) { + assert(u); + assert(f); + assert(fds); + + if (!rt) + return 0; + + if (rt->tmp_dir) + unit_serialize_item(u, f, "tmp-dir", rt->tmp_dir); + + if (rt->var_tmp_dir) + unit_serialize_item(u, f, "var-tmp-dir", rt->var_tmp_dir); + + if (rt->netns_storage_socket[0] >= 0) { + int copy; + + copy = fdset_put_dup(fds, rt->netns_storage_socket[0]); + if (copy < 0) + return copy; + + unit_serialize_item_format(u, f, "netns-socket-0", "%i", copy); + } + + if (rt->netns_storage_socket[1] >= 0) { + int copy; + + copy = fdset_put_dup(fds, rt->netns_storage_socket[1]); + if (copy < 0) + return copy; + + unit_serialize_item_format(u, f, "netns-socket-1", "%i", copy); + } + + return 0; +} + +int exec_runtime_deserialize_item(Unit *u, ExecRuntime **rt, const char *key, const char *value, FDSet *fds) { + int r; + + assert(rt); + assert(key); + assert(value); + + if (streq(key, "tmp-dir")) { + char *copy; + + r = exec_runtime_allocate(rt); + if (r < 0) + return log_oom(); + + copy = strdup(value); + if (!copy) + return log_oom(); + + free((*rt)->tmp_dir); + (*rt)->tmp_dir = copy; + + } else if (streq(key, "var-tmp-dir")) { + char *copy; + + r = exec_runtime_allocate(rt); + if (r < 0) + return log_oom(); + + copy = strdup(value); + if (!copy) + return log_oom(); + + free((*rt)->var_tmp_dir); + (*rt)->var_tmp_dir = copy; + + } else if (streq(key, "netns-socket-0")) { + int fd; + + r = exec_runtime_allocate(rt); + if (r < 0) + return log_oom(); + + if (safe_atoi(value, &fd) < 0 || !fdset_contains(fds, fd)) + log_unit_debug(u, "Failed to parse netns socket value: %s", value); + else { + safe_close((*rt)->netns_storage_socket[0]); + (*rt)->netns_storage_socket[0] = fdset_remove(fds, fd); + } + } else if (streq(key, "netns-socket-1")) { + int fd; + + r = exec_runtime_allocate(rt); + if (r < 0) + return log_oom(); + + if (safe_atoi(value, &fd) < 0 || !fdset_contains(fds, fd)) + log_unit_debug(u, "Failed to parse netns socket value: %s", value); + else { + safe_close((*rt)->netns_storage_socket[1]); + (*rt)->netns_storage_socket[1] = fdset_remove(fds, fd); + } + } else + return 0; + + return 1; +} + +static void *remove_tmpdir_thread(void *p) { + _cleanup_free_ char *path = p; + + (void) rm_rf(path, REMOVE_ROOT|REMOVE_PHYSICAL); + return NULL; +} + +void exec_runtime_destroy(ExecRuntime *rt) { + int r; + + if (!rt) + return; + + /* If there are multiple users of this, let's leave the stuff around */ + if (rt->n_ref > 1) + return; + + if (rt->tmp_dir) { + log_debug("Spawning thread to nuke %s", rt->tmp_dir); + + r = asynchronous_job(remove_tmpdir_thread, rt->tmp_dir); + if (r < 0) { + log_warning_errno(r, "Failed to nuke %s: %m", rt->tmp_dir); + free(rt->tmp_dir); + } + + rt->tmp_dir = NULL; + } + + if (rt->var_tmp_dir) { + log_debug("Spawning thread to nuke %s", rt->var_tmp_dir); + + r = asynchronous_job(remove_tmpdir_thread, rt->var_tmp_dir); + if (r < 0) { + log_warning_errno(r, "Failed to nuke %s: %m", rt->var_tmp_dir); + free(rt->var_tmp_dir); + } + + rt->var_tmp_dir = NULL; + } + + safe_close_pair(rt->netns_storage_socket); +} + +static const char* const exec_input_table[_EXEC_INPUT_MAX] = { + [EXEC_INPUT_NULL] = "null", + [EXEC_INPUT_TTY] = "tty", + [EXEC_INPUT_TTY_FORCE] = "tty-force", + [EXEC_INPUT_TTY_FAIL] = "tty-fail", + [EXEC_INPUT_SOCKET] = "socket" +}; + +DEFINE_STRING_TABLE_LOOKUP(exec_input, ExecInput); + +static const char* const exec_output_table[_EXEC_OUTPUT_MAX] = { + [EXEC_OUTPUT_INHERIT] = "inherit", + [EXEC_OUTPUT_NULL] = "null", + [EXEC_OUTPUT_TTY] = "tty", + [EXEC_OUTPUT_SYSLOG] = "syslog", + [EXEC_OUTPUT_SYSLOG_AND_CONSOLE] = "syslog+console", + [EXEC_OUTPUT_KMSG] = "kmsg", + [EXEC_OUTPUT_KMSG_AND_CONSOLE] = "kmsg+console", + [EXEC_OUTPUT_JOURNAL] = "journal", + [EXEC_OUTPUT_JOURNAL_AND_CONSOLE] = "journal+console", + [EXEC_OUTPUT_SOCKET] = "socket" +}; + +DEFINE_STRING_TABLE_LOOKUP(exec_output, ExecOutput); + +static const char* const exec_utmp_mode_table[_EXEC_UTMP_MODE_MAX] = { + [EXEC_UTMP_INIT] = "init", + [EXEC_UTMP_LOGIN] = "login", + [EXEC_UTMP_USER] = "user", +}; + +DEFINE_STRING_TABLE_LOOKUP(exec_utmp_mode, ExecUtmpMode); diff --git a/src/grp-system/libcore/src/failure-action.c b/src/grp-system/libcore/src/failure-action.c new file mode 100644 index 0000000000..e17ed7219b --- /dev/null +++ b/src/grp-system/libcore/src/failure-action.c @@ -0,0 +1,128 @@ +/*** + This file is part of systemd. + + Copyright 2014 Lennart Poettering + Copyright 2012 Michael Olbrich + + systemd is free software; you can redistribute it and/or modify it + under the terms of the GNU Lesser General Public License as published by + the Free Software Foundation; either version 2.1 of the License, or + (at your option) any later version. + + systemd is distributed in the hope that it will be useful, but + WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + Lesser General Public License for more details. + + You should have received a copy of the GNU Lesser General Public License + along with systemd; If not, see . +***/ + +#include + +#include + +#include "core/failure-action.h" +#include "sd-bus/bus-error.h" +#include "sd-bus/bus-util.h" +#include "systemd-basic/special.h" +#include "systemd-basic/string-table.h" +#include "systemd-basic/terminal-util.h" + +static void log_and_status(Manager *m, const char *message) { + log_warning("%s", message); + manager_status_printf(m, STATUS_TYPE_EMERGENCY, + ANSI_HIGHLIGHT_RED " !! " ANSI_NORMAL, + "%s", message); +} + +int failure_action( + Manager *m, + FailureAction action, + const char *reboot_arg) { + + assert(m); + assert(action >= 0); + assert(action < _FAILURE_ACTION_MAX); + + if (action == FAILURE_ACTION_NONE) + return -ECANCELED; + + if (!MANAGER_IS_SYSTEM(m)) { + /* Downgrade all options to simply exiting if we run + * in user mode */ + + log_warning("Exiting as result of failure."); + m->exit_code = MANAGER_EXIT; + return -ECANCELED; + } + + switch (action) { + + case FAILURE_ACTION_REBOOT: + log_and_status(m, "Rebooting as result of failure."); + + (void) update_reboot_parameter_and_warn(reboot_arg); + (void) manager_add_job_by_name_and_warn(m, JOB_START, SPECIAL_REBOOT_TARGET, JOB_REPLACE_IRREVERSIBLY, NULL); + + break; + + case FAILURE_ACTION_REBOOT_FORCE: + log_and_status(m, "Forcibly rebooting as result of failure."); + + (void) update_reboot_parameter_and_warn(reboot_arg); + m->exit_code = MANAGER_REBOOT; + + break; + + case FAILURE_ACTION_REBOOT_IMMEDIATE: + log_and_status(m, "Rebooting immediately as result of failure."); + + sync(); + + if (!isempty(reboot_arg)) { + log_info("Rebooting with argument '%s'.", reboot_arg); + syscall(SYS_reboot, LINUX_REBOOT_MAGIC1, LINUX_REBOOT_MAGIC2, LINUX_REBOOT_CMD_RESTART2, reboot_arg); + log_warning_errno(errno, "Failed to reboot with parameter, retrying without: %m"); + } + + log_info("Rebooting."); + reboot(RB_AUTOBOOT); + break; + + case FAILURE_ACTION_POWEROFF: + log_and_status(m, "Powering off as result of failure."); + (void) manager_add_job_by_name_and_warn(m, JOB_START, SPECIAL_POWEROFF_TARGET, JOB_REPLACE_IRREVERSIBLY, NULL); + break; + + case FAILURE_ACTION_POWEROFF_FORCE: + log_and_status(m, "Forcibly powering off as result of failure."); + m->exit_code = MANAGER_POWEROFF; + break; + + case FAILURE_ACTION_POWEROFF_IMMEDIATE: + log_and_status(m, "Powering off immediately as result of failure."); + + sync(); + + log_info("Powering off."); + reboot(RB_POWER_OFF); + break; + + default: + assert_not_reached("Unknown failure action"); + } + + return -ECANCELED; +} + +static const char* const failure_action_table[_FAILURE_ACTION_MAX] = { + [FAILURE_ACTION_NONE] = "none", + [FAILURE_ACTION_REBOOT] = "reboot", + [FAILURE_ACTION_REBOOT_FORCE] = "reboot-force", + [FAILURE_ACTION_REBOOT_IMMEDIATE] = "reboot-immediate", + [FAILURE_ACTION_POWEROFF] = "poweroff", + [FAILURE_ACTION_POWEROFF_FORCE] = "poweroff-force", + [FAILURE_ACTION_POWEROFF_IMMEDIATE] = "poweroff-immediate" +}; +DEFINE_STRING_TABLE_LOOKUP(failure_action, FailureAction); diff --git a/src/grp-system/libcore/src/hostname-setup.c b/src/grp-system/libcore/src/hostname-setup.c new file mode 100644 index 0000000000..6170d5e030 --- /dev/null +++ b/src/grp-system/libcore/src/hostname-setup.c @@ -0,0 +1,68 @@ +/*** + This file is part of systemd. + + Copyright 2010 Lennart Poettering + + systemd is free software; you can redistribute it and/or modify it + under the terms of the GNU Lesser General Public License as published by + the Free Software Foundation; either version 2.1 of the License, or + (at your option) any later version. + + systemd is distributed in the hope that it will be useful, but + WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + Lesser General Public License for more details. + + You should have received a copy of the GNU Lesser General Public License + along with systemd; If not, see . +***/ + +#include +#include +#include + +#include "core/hostname-setup.h" +#include "systemd-basic/alloc-util.h" +#include "systemd-basic/fileio.h" +#include "systemd-basic/hostname-util.h" +#include "systemd-basic/log.h" +#include "systemd-basic/macro.h" +#include "systemd-basic/string-util.h" +#include "systemd-basic/util.h" + +int hostname_setup(void) { + int r; + _cleanup_free_ char *b = NULL; + const char *hn; + bool enoent = false; + + r = read_hostname_config("/etc/hostname", &b); + if (r < 0) { + if (r == -ENOENT) + enoent = true; + else + log_warning_errno(r, "Failed to read configured hostname: %m"); + + hn = NULL; + } else + hn = b; + + if (isempty(hn)) { + /* Don't override the hostname if it is already set + * and not explicitly configured */ + if (hostname_is_set()) + return 0; + + if (enoent) + log_info("No hostname configured."); + + hn = "localhost"; + } + + r = sethostname_idempotent(hn); + if (r < 0) + return log_warning_errno(r, "Failed to set hostname to <%s>: %m", hn); + + log_info("Set hostname to <%s>.", hn); + return 0; +} diff --git a/src/grp-system/libcore/src/ima-setup.c b/src/grp-system/libcore/src/ima-setup.c new file mode 100644 index 0000000000..d1ec4852c9 --- /dev/null +++ b/src/grp-system/libcore/src/ima-setup.c @@ -0,0 +1,80 @@ +/*** + This file is part of systemd. + + Copyright 2010 Lennart Poettering + Copyright (C) 2012 Roberto Sassu - Politecnico di Torino, Italy + TORSEC group — http://security.polito.it + + systemd is free software; you can redistribute it and/or modify it + under the terms of the GNU Lesser General Public License as published by + the Free Software Foundation; either version 2.1 of the License, or + (at your option) any later version. + + systemd is distributed in the hope that it will be useful, but + WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + Lesser General Public License for more details. + + You should have received a copy of the GNU Lesser General Public License + along with systemd; If not, see . +***/ + +#include +#include + +#include "core/ima-setup.h" +#include "systemd-basic/fd-util.h" +#include "systemd-basic/fileio.h" +#include "systemd-basic/log.h" +#include "systemd-basic/util.h" + +#define IMA_SECFS_DIR "/sys/kernel/security/ima" +#define IMA_SECFS_POLICY IMA_SECFS_DIR "/policy" +#define IMA_POLICY_PATH "/etc/ima/ima-policy" + +int ima_setup(void) { +#ifdef HAVE_IMA + _cleanup_fclose_ FILE *input = NULL; + _cleanup_close_ int imafd = -1; + unsigned lineno = 0; + char line[page_size()]; + + if (access(IMA_SECFS_DIR, F_OK) < 0) { + log_debug("IMA support is disabled in the kernel, ignoring."); + return 0; + } + + input = fopen(IMA_POLICY_PATH, "re"); + if (!input) { + log_full_errno(errno == ENOENT ? LOG_DEBUG : LOG_WARNING, errno, + "Failed to open the IMA custom policy file "IMA_POLICY_PATH", ignoring: %m"); + return 0; + } + + if (access(IMA_SECFS_POLICY, F_OK) < 0) { + log_warning("Another IMA custom policy has already been loaded, ignoring."); + return 0; + } + + imafd = open(IMA_SECFS_POLICY, O_WRONLY|O_CLOEXEC); + if (imafd < 0) { + log_error_errno(errno, "Failed to open the IMA kernel interface "IMA_SECFS_POLICY", ignoring: %m"); + return 0; + } + + FOREACH_LINE(line, input, + return log_error_errno(errno, "Failed to read the IMA custom policy file "IMA_POLICY_PATH": %m")) { + size_t len; + + len = strlen(line); + lineno++; + + if (len > 0 && write(imafd, line, len) < 0) + return log_error_errno(errno, "Failed to load the IMA custom policy file "IMA_POLICY_PATH"%u: %m", + lineno); + } + + log_info("Successfully loaded the IMA custom policy "IMA_POLICY_PATH"."); +#endif /* HAVE_IMA */ + return 0; +} diff --git a/src/grp-system/libcore/src/job.c b/src/grp-system/libcore/src/job.c new file mode 100644 index 0000000000..5c28ea1bd6 --- /dev/null +++ b/src/grp-system/libcore/src/job.c @@ -0,0 +1,1260 @@ +/*** + This file is part of systemd. + + Copyright 2010 Lennart Poettering + + systemd is free software; you can redistribute it and/or modify it + under the terms of the GNU Lesser General Public License as published by + the Free Software Foundation; either version 2.1 of the License, or + (at your option) any later version. + + systemd is distributed in the hope that it will be useful, but + WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + Lesser General Public License for more details. + + You should have received a copy of the GNU Lesser General Public License + along with systemd; If not, see . +***/ + +#include + +#include +#include + +#include "core/job.h" +#include "core/unit.h" +#include "systemd-basic/alloc-util.h" +#include "systemd-basic/async.h" +#include "systemd-basic/escape.h" +#include "systemd-basic/log.h" +#include "systemd-basic/macro.h" +#include "systemd-basic/parse-util.h" +#include "systemd-basic/set.h" +#include "systemd-basic/special.h" +#include "systemd-basic/stdio-util.h" +#include "systemd-basic/string-table.h" +#include "systemd-basic/string-util.h" +#include "systemd-basic/strv.h" +#include "systemd-basic/terminal-util.h" +#include "systemd-basic/virt.h" + +#include "dbus-job.h" +#include "dbus.h" + +Job* job_new_raw(Unit *unit) { + Job *j; + + /* used for deserialization */ + + assert(unit); + + j = new0(Job, 1); + if (!j) + return NULL; + + j->manager = unit->manager; + j->unit = unit; + j->type = _JOB_TYPE_INVALID; + + return j; +} + +Job* job_new(Unit *unit, JobType type) { + Job *j; + + assert(type < _JOB_TYPE_MAX); + + j = job_new_raw(unit); + if (!j) + return NULL; + + j->id = j->manager->current_job_id++; + j->type = type; + + /* We don't link it here, that's what job_dependency() is for */ + + return j; +} + +void job_free(Job *j) { + assert(j); + assert(!j->installed); + assert(!j->transaction_prev); + assert(!j->transaction_next); + assert(!j->subject_list); + assert(!j->object_list); + + if (j->in_run_queue) + LIST_REMOVE(run_queue, j->manager->run_queue, j); + + if (j->in_dbus_queue) + LIST_REMOVE(dbus_queue, j->manager->dbus_job_queue, j); + + sd_event_source_unref(j->timer_event_source); + + sd_bus_track_unref(j->clients); + strv_free(j->deserialized_clients); + + free(j); +} + +static void job_set_state(Job *j, JobState state) { + assert(j); + assert(state >= 0); + assert(state < _JOB_STATE_MAX); + + if (j->state == state) + return; + + j->state = state; + + if (!j->installed) + return; + + if (j->state == JOB_RUNNING) + j->unit->manager->n_running_jobs++; + else { + assert(j->state == JOB_WAITING); + assert(j->unit->manager->n_running_jobs > 0); + + j->unit->manager->n_running_jobs--; + + if (j->unit->manager->n_running_jobs <= 0) + j->unit->manager->jobs_in_progress_event_source = sd_event_source_unref(j->unit->manager->jobs_in_progress_event_source); + } +} + +void job_uninstall(Job *j) { + Job **pj; + + assert(j->installed); + + job_set_state(j, JOB_WAITING); + + pj = (j->type == JOB_NOP) ? &j->unit->nop_job : &j->unit->job; + assert(*pj == j); + + /* Detach from next 'bigger' objects */ + + /* daemon-reload should be transparent to job observers */ + if (!MANAGER_IS_RELOADING(j->manager)) + bus_job_send_removed_signal(j); + + *pj = NULL; + + unit_add_to_gc_queue(j->unit); + + hashmap_remove(j->manager->jobs, UINT32_TO_PTR(j->id)); + j->installed = false; +} + +static bool job_type_allows_late_merge(JobType t) { + /* Tells whether it is OK to merge a job of type 't' with an already + * running job. + * Reloads cannot be merged this way. Think of the sequence: + * 1. Reload of a daemon is in progress; the daemon has already loaded + * its config file, but hasn't completed the reload operation yet. + * 2. Edit foo's config file. + * 3. Trigger another reload to have the daemon use the new config. + * Should the second reload job be merged into the first one, the daemon + * would not know about the new config. + * JOB_RESTART jobs on the other hand can be merged, because they get + * patched into JOB_START after stopping the unit. So if we see a + * JOB_RESTART running, it means the unit hasn't stopped yet and at + * this time the merge is still allowed. */ + return t != JOB_RELOAD; +} + +static void job_merge_into_installed(Job *j, Job *other) { + assert(j->installed); + assert(j->unit == other->unit); + + if (j->type != JOB_NOP) + job_type_merge_and_collapse(&j->type, other->type, j->unit); + else + assert(other->type == JOB_NOP); + + j->irreversible = j->irreversible || other->irreversible; + j->ignore_order = j->ignore_order || other->ignore_order; +} + +Job* job_install(Job *j) { + Job **pj; + Job *uj; + + assert(!j->installed); + assert(j->type < _JOB_TYPE_MAX_IN_TRANSACTION); + assert(j->state == JOB_WAITING); + + pj = (j->type == JOB_NOP) ? &j->unit->nop_job : &j->unit->job; + uj = *pj; + + if (uj) { + if (job_type_is_conflicting(uj->type, j->type)) + job_finish_and_invalidate(uj, JOB_CANCELED, false, false); + else { + /* not conflicting, i.e. mergeable */ + + if (uj->state == JOB_WAITING || + (job_type_allows_late_merge(j->type) && job_type_is_superset(uj->type, j->type))) { + job_merge_into_installed(uj, j); + log_unit_debug(uj->unit, + "Merged into installed job %s/%s as %u", + uj->unit->id, job_type_to_string(uj->type), (unsigned) uj->id); + return uj; + } else { + /* already running and not safe to merge into */ + /* Patch uj to become a merged job and re-run it. */ + /* XXX It should be safer to queue j to run after uj finishes, but it is + * not currently possible to have more than one installed job per unit. */ + job_merge_into_installed(uj, j); + log_unit_debug(uj->unit, + "Merged into running job, re-running: %s/%s as %u", + uj->unit->id, job_type_to_string(uj->type), (unsigned) uj->id); + + job_set_state(uj, JOB_WAITING); + return uj; + } + } + } + + /* Install the job */ + *pj = j; + j->installed = true; + + j->manager->n_installed_jobs++; + log_unit_debug(j->unit, + "Installed new job %s/%s as %u", + j->unit->id, job_type_to_string(j->type), (unsigned) j->id); + return j; +} + +int job_install_deserialized(Job *j) { + Job **pj; + + assert(!j->installed); + + if (j->type < 0 || j->type >= _JOB_TYPE_MAX_IN_TRANSACTION) { + log_debug("Invalid job type %s in deserialization.", strna(job_type_to_string(j->type))); + return -EINVAL; + } + + pj = (j->type == JOB_NOP) ? &j->unit->nop_job : &j->unit->job; + if (*pj) { + log_unit_debug(j->unit, "Unit already has a job installed. Not installing deserialized job."); + return -EEXIST; + } + + *pj = j; + j->installed = true; + + if (j->state == JOB_RUNNING) + j->unit->manager->n_running_jobs++; + + log_unit_debug(j->unit, + "Reinstalled deserialized job %s/%s as %u", + j->unit->id, job_type_to_string(j->type), (unsigned) j->id); + return 0; +} + +JobDependency* job_dependency_new(Job *subject, Job *object, bool matters, bool conflicts) { + JobDependency *l; + + assert(object); + + /* Adds a new job link, which encodes that the 'subject' job + * needs the 'object' job in some way. If 'subject' is NULL + * this means the 'anchor' job (i.e. the one the user + * explicitly asked for) is the requester. */ + + if (!(l = new0(JobDependency, 1))) + return NULL; + + l->subject = subject; + l->object = object; + l->matters = matters; + l->conflicts = conflicts; + + if (subject) + LIST_PREPEND(subject, subject->subject_list, l); + + LIST_PREPEND(object, object->object_list, l); + + return l; +} + +void job_dependency_free(JobDependency *l) { + assert(l); + + if (l->subject) + LIST_REMOVE(subject, l->subject->subject_list, l); + + LIST_REMOVE(object, l->object->object_list, l); + + free(l); +} + +void job_dump(Job *j, FILE*f, const char *prefix) { + assert(j); + assert(f); + + if (!prefix) + prefix = ""; + + fprintf(f, + "%s-> Job %u:\n" + "%s\tAction: %s -> %s\n" + "%s\tState: %s\n" + "%s\tIrreversible: %s\n", + prefix, j->id, + prefix, j->unit->id, job_type_to_string(j->type), + prefix, job_state_to_string(j->state), + prefix, yes_no(j->irreversible)); +} + +/* + * Merging is commutative, so imagine the matrix as symmetric. We store only + * its lower triangle to avoid duplication. We don't store the main diagonal, + * because A merged with A is simply A. + * + * If the resulting type is collapsed immediately afterwards (to get rid of + * the JOB_RELOAD_OR_START, which lies outside the lookup function's domain), + * the following properties hold: + * + * Merging is associative! A merged with B, and then merged with C is the same + * as A merged with the result of B merged with C. + * + * Mergeability is transitive! If A can be merged with B and B with C then + * A also with C. + * + * Also, if A merged with B cannot be merged with C, then either A or B cannot + * be merged with C either. + */ +static const JobType job_merging_table[] = { +/* What \ With * JOB_START JOB_VERIFY_ACTIVE JOB_STOP JOB_RELOAD */ +/*********************************************************************************/ +/*JOB_START */ +/*JOB_VERIFY_ACTIVE */ JOB_START, +/*JOB_STOP */ -1, -1, +/*JOB_RELOAD */ JOB_RELOAD_OR_START, JOB_RELOAD, -1, +/*JOB_RESTART */ JOB_RESTART, JOB_RESTART, -1, JOB_RESTART, +}; + +JobType job_type_lookup_merge(JobType a, JobType b) { + assert_cc(ELEMENTSOF(job_merging_table) == _JOB_TYPE_MAX_MERGING * (_JOB_TYPE_MAX_MERGING - 1) / 2); + assert(a >= 0 && a < _JOB_TYPE_MAX_MERGING); + assert(b >= 0 && b < _JOB_TYPE_MAX_MERGING); + + if (a == b) + return a; + + if (a < b) { + JobType tmp = a; + a = b; + b = tmp; + } + + return job_merging_table[(a - 1) * a / 2 + b]; +} + +bool job_type_is_redundant(JobType a, UnitActiveState b) { + switch (a) { + + case JOB_START: + return + b == UNIT_ACTIVE || + b == UNIT_RELOADING; + + case JOB_STOP: + return + b == UNIT_INACTIVE || + b == UNIT_FAILED; + + case JOB_VERIFY_ACTIVE: + return + b == UNIT_ACTIVE || + b == UNIT_RELOADING; + + case JOB_RELOAD: + return + b == UNIT_RELOADING; + + case JOB_RESTART: + return + b == UNIT_ACTIVATING; + + case JOB_NOP: + return true; + + default: + assert_not_reached("Invalid job type"); + } +} + +JobType job_type_collapse(JobType t, Unit *u) { + UnitActiveState s; + + switch (t) { + + case JOB_TRY_RESTART: + s = unit_active_state(u); + if (UNIT_IS_INACTIVE_OR_DEACTIVATING(s)) + return JOB_NOP; + + return JOB_RESTART; + + case JOB_TRY_RELOAD: + s = unit_active_state(u); + if (UNIT_IS_INACTIVE_OR_DEACTIVATING(s)) + return JOB_NOP; + + return JOB_RELOAD; + + case JOB_RELOAD_OR_START: + s = unit_active_state(u); + if (UNIT_IS_INACTIVE_OR_DEACTIVATING(s)) + return JOB_START; + + return JOB_RELOAD; + + default: + return t; + } +} + +int job_type_merge_and_collapse(JobType *a, JobType b, Unit *u) { + JobType t; + + t = job_type_lookup_merge(*a, b); + if (t < 0) + return -EEXIST; + + *a = job_type_collapse(t, u); + return 0; +} + +static bool job_is_runnable(Job *j) { + Iterator i; + Unit *other; + + assert(j); + assert(j->installed); + + /* Checks whether there is any job running for the units this + * job needs to be running after (in the case of a 'positive' + * job type) or before (in the case of a 'negative' job + * type. */ + + /* Note that unit types have a say in what is runnable, + * too. For example, if they return -EAGAIN from + * unit_start() they can indicate they are not + * runnable yet. */ + + /* First check if there is an override */ + if (j->ignore_order) + return true; + + if (j->type == JOB_NOP) + return true; + + if (j->type == JOB_START || + j->type == JOB_VERIFY_ACTIVE || + j->type == JOB_RELOAD) { + + /* Immediate result is that the job is or might be + * started. In this case let's wait for the + * dependencies, regardless whether they are + * starting or stopping something. */ + + SET_FOREACH(other, j->unit->dependencies[UNIT_AFTER], i) + if (other->job) + return false; + } + + /* Also, if something else is being stopped and we should + * change state after it, then let's wait. */ + + SET_FOREACH(other, j->unit->dependencies[UNIT_BEFORE], i) + if (other->job && + (other->job->type == JOB_STOP || + other->job->type == JOB_RESTART)) + return false; + + /* This means that for a service a and a service b where b + * shall be started after a: + * + * start a + start b → 1st step start a, 2nd step start b + * start a + stop b → 1st step stop b, 2nd step start a + * stop a + start b → 1st step stop a, 2nd step start b + * stop a + stop b → 1st step stop b, 2nd step stop a + * + * This has the side effect that restarts are properly + * synchronized too. */ + + return true; +} + +static void job_change_type(Job *j, JobType newtype) { + assert(j); + + log_unit_debug(j->unit, + "Converting job %s/%s -> %s/%s", + j->unit->id, job_type_to_string(j->type), + j->unit->id, job_type_to_string(newtype)); + + j->type = newtype; +} + +static int job_perform_on_unit(Job **j) { + uint32_t id; + Manager *m; + JobType t; + Unit *u; + int r; + + /* While we execute this operation the job might go away (for + * example: because it finishes immediately or is replaced by + * a new, conflicting job.) To make sure we don't access a + * freed job later on we store the id here, so that we can + * verify the job is still valid. */ + + assert(j); + assert(*j); + + m = (*j)->manager; + u = (*j)->unit; + t = (*j)->type; + id = (*j)->id; + + switch (t) { + case JOB_START: + r = unit_start(u); + break; + + case JOB_RESTART: + t = JOB_STOP; + /* fall through */ + case JOB_STOP: + r = unit_stop(u); + break; + + case JOB_RELOAD: + r = unit_reload(u); + break; + + default: + assert_not_reached("Invalid job type"); + } + + /* Log if the job still exists and the start/stop/reload function + * actually did something. */ + *j = manager_get_job(m, id); + if (*j && r > 0) + unit_status_emit_starting_stopping_reloading(u, t); + + return r; +} + +int job_run_and_invalidate(Job *j) { + int r; + + assert(j); + assert(j->installed); + assert(j->type < _JOB_TYPE_MAX_IN_TRANSACTION); + assert(j->in_run_queue); + + LIST_REMOVE(run_queue, j->manager->run_queue, j); + j->in_run_queue = false; + + if (j->state != JOB_WAITING) + return 0; + + if (!job_is_runnable(j)) + return -EAGAIN; + + job_set_state(j, JOB_RUNNING); + job_add_to_dbus_queue(j); + + + switch (j->type) { + + case JOB_VERIFY_ACTIVE: { + UnitActiveState t = unit_active_state(j->unit); + if (UNIT_IS_ACTIVE_OR_RELOADING(t)) + r = -EALREADY; + else if (t == UNIT_ACTIVATING) + r = -EAGAIN; + else + r = -EBADR; + break; + } + + case JOB_START: + case JOB_STOP: + case JOB_RESTART: + r = job_perform_on_unit(&j); + + /* If the unit type does not support starting/stopping, + * then simply wait. */ + if (r == -EBADR) + r = 0; + break; + + case JOB_RELOAD: + r = job_perform_on_unit(&j); + break; + + case JOB_NOP: + r = -EALREADY; + break; + + default: + assert_not_reached("Unknown job type"); + } + + if (j) { + if (r == -EALREADY) + r = job_finish_and_invalidate(j, JOB_DONE, true, true); + else if (r == -EBADR) + r = job_finish_and_invalidate(j, JOB_SKIPPED, true, false); + else if (r == -ENOEXEC) + r = job_finish_and_invalidate(j, JOB_INVALID, true, false); + else if (r == -EPROTO) + r = job_finish_and_invalidate(j, JOB_ASSERT, true, false); + else if (r == -EOPNOTSUPP) + r = job_finish_and_invalidate(j, JOB_UNSUPPORTED, true, false); + else if (r == -EAGAIN) + job_set_state(j, JOB_WAITING); + else if (r < 0) + r = job_finish_and_invalidate(j, JOB_FAILED, true, false); + } + + return r; +} + +_pure_ static const char *job_get_status_message_format(Unit *u, JobType t, JobResult result) { + + static const char *const generic_finished_start_job[_JOB_RESULT_MAX] = { + [JOB_DONE] = "Started %s.", + [JOB_TIMEOUT] = "Timed out starting %s.", + [JOB_FAILED] = "Failed to start %s.", + [JOB_DEPENDENCY] = "Dependency failed for %s.", + [JOB_ASSERT] = "Assertion failed for %s.", + [JOB_UNSUPPORTED] = "Starting of %s not supported.", + }; + static const char *const generic_finished_stop_job[_JOB_RESULT_MAX] = { + [JOB_DONE] = "Stopped %s.", + [JOB_FAILED] = "Stopped (with error) %s.", + [JOB_TIMEOUT] = "Timed out stopping %s.", + }; + static const char *const generic_finished_reload_job[_JOB_RESULT_MAX] = { + [JOB_DONE] = "Reloaded %s.", + [JOB_FAILED] = "Reload failed for %s.", + [JOB_TIMEOUT] = "Timed out reloading %s.", + }; + /* When verify-active detects the unit is inactive, report it. + * Most likely a DEPEND warning from a requisiting unit will + * occur next and it's nice to see what was requisited. */ + static const char *const generic_finished_verify_active_job[_JOB_RESULT_MAX] = { + [JOB_SKIPPED] = "%s is not active.", + }; + + const UnitStatusMessageFormats *format_table; + const char *format; + + assert(u); + assert(t >= 0); + assert(t < _JOB_TYPE_MAX); + + if (IN_SET(t, JOB_START, JOB_STOP, JOB_RESTART)) { + format_table = &UNIT_VTABLE(u)->status_message_formats; + if (format_table) { + format = t == JOB_START ? format_table->finished_start_job[result] : + format_table->finished_stop_job[result]; + if (format) + return format; + } + } + + /* Return generic strings */ + if (t == JOB_START) + return generic_finished_start_job[result]; + else if (t == JOB_STOP || t == JOB_RESTART) + return generic_finished_stop_job[result]; + else if (t == JOB_RELOAD) + return generic_finished_reload_job[result]; + else if (t == JOB_VERIFY_ACTIVE) + return generic_finished_verify_active_job[result]; + + return NULL; +} + +static void job_print_status_message(Unit *u, JobType t, JobResult result) { + static struct { + const char *color, *word; + } const statuses[_JOB_RESULT_MAX] = { + [JOB_DONE] = {ANSI_GREEN, " OK "}, + [JOB_TIMEOUT] = {ANSI_HIGHLIGHT_RED, " TIME "}, + [JOB_FAILED] = {ANSI_HIGHLIGHT_RED, "FAILED"}, + [JOB_DEPENDENCY] = {ANSI_HIGHLIGHT_YELLOW, "DEPEND"}, + [JOB_SKIPPED] = {ANSI_HIGHLIGHT, " INFO "}, + [JOB_ASSERT] = {ANSI_HIGHLIGHT_YELLOW, "ASSERT"}, + [JOB_UNSUPPORTED] = {ANSI_HIGHLIGHT_YELLOW, "UNSUPP"}, + }; + + const char *format; + const char *status; + + assert(u); + assert(t >= 0); + assert(t < _JOB_TYPE_MAX); + + /* Reload status messages have traditionally not been printed to console. */ + if (t == JOB_RELOAD) + return; + + format = job_get_status_message_format(u, t, result); + if (!format) + return; + + if (log_get_show_color()) + status = strjoina(statuses[result].color, statuses[result].word, ANSI_NORMAL); + else + status = statuses[result].word; + + if (result != JOB_DONE) + manager_flip_auto_status(u->manager, true); + + DISABLE_WARNING_FORMAT_NONLITERAL; + unit_status_printf(u, status, format); + REENABLE_WARNING; + + if (t == JOB_START && result == JOB_FAILED) { + _cleanup_free_ char *quoted; + + quoted = shell_maybe_quote(u->id); + manager_status_printf(u->manager, STATUS_TYPE_NORMAL, NULL, "See 'systemctl status %s' for details.", strna(quoted)); + } +} + +static void job_log_status_message(Unit *u, JobType t, JobResult result) { + const char *format; + char buf[LINE_MAX]; + sd_id128_t mid; + static const int job_result_log_level[_JOB_RESULT_MAX] = { + [JOB_DONE] = LOG_INFO, + [JOB_CANCELED] = LOG_INFO, + [JOB_TIMEOUT] = LOG_ERR, + [JOB_FAILED] = LOG_ERR, + [JOB_DEPENDENCY] = LOG_WARNING, + [JOB_SKIPPED] = LOG_NOTICE, + [JOB_INVALID] = LOG_INFO, + [JOB_ASSERT] = LOG_WARNING, + [JOB_UNSUPPORTED] = LOG_WARNING, + }; + + assert(u); + assert(t >= 0); + assert(t < _JOB_TYPE_MAX); + + /* Skip this if it goes to the console. since we already print + * to the console anyway... */ + + if (log_on_console()) + return; + + format = job_get_status_message_format(u, t, result); + if (!format) + return; + + DISABLE_WARNING_FORMAT_NONLITERAL; + xsprintf(buf, format, unit_description(u)); + REENABLE_WARNING; + + switch (t) { + + case JOB_START: + mid = result == JOB_DONE ? SD_MESSAGE_UNIT_STARTED : SD_MESSAGE_UNIT_FAILED; + break; + + case JOB_RELOAD: + mid = SD_MESSAGE_UNIT_RELOADED; + break; + + case JOB_STOP: + case JOB_RESTART: + mid = SD_MESSAGE_UNIT_STOPPED; + break; + + default: + log_struct(job_result_log_level[result], + LOG_UNIT_ID(u), + LOG_MESSAGE("%s", buf), + "RESULT=%s", job_result_to_string(result), + NULL); + return; + } + + log_struct(job_result_log_level[result], + LOG_MESSAGE_ID(mid), + LOG_UNIT_ID(u), + LOG_MESSAGE("%s", buf), + "RESULT=%s", job_result_to_string(result), + NULL); +} + +static void job_emit_status_message(Unit *u, JobType t, JobResult result) { + + /* No message if the job did not actually do anything due to failed condition. */ + if (t == JOB_START && result == JOB_DONE && !u->condition_result) + return; + + job_log_status_message(u, t, result); + job_print_status_message(u, t, result); +} + +static void job_fail_dependencies(Unit *u, UnitDependency d) { + Unit *other; + Iterator i; + + assert(u); + + SET_FOREACH(other, u->dependencies[d], i) { + Job *j = other->job; + + if (!j) + continue; + if (!IN_SET(j->type, JOB_START, JOB_VERIFY_ACTIVE)) + continue; + + job_finish_and_invalidate(j, JOB_DEPENDENCY, true, false); + } +} + +int job_finish_and_invalidate(Job *j, JobResult result, bool recursive, bool already) { + Unit *u; + Unit *other; + JobType t; + Iterator i; + + assert(j); + assert(j->installed); + assert(j->type < _JOB_TYPE_MAX_IN_TRANSACTION); + + u = j->unit; + t = j->type; + + j->result = result; + + log_unit_debug(u, "Job %s/%s finished, result=%s", u->id, job_type_to_string(t), job_result_to_string(result)); + + /* If this job did nothing to respective unit we don't log the status message */ + if (!already) + job_emit_status_message(u, t, result); + + job_add_to_dbus_queue(j); + + /* Patch restart jobs so that they become normal start jobs */ + if (result == JOB_DONE && t == JOB_RESTART) { + + job_change_type(j, JOB_START); + job_set_state(j, JOB_WAITING); + + job_add_to_run_queue(j); + + goto finish; + } + + if (result == JOB_FAILED || result == JOB_INVALID) + j->manager->n_failed_jobs++; + + job_uninstall(j); + job_free(j); + + /* Fail depending jobs on failure */ + if (result != JOB_DONE && recursive) { + if (IN_SET(t, JOB_START, JOB_VERIFY_ACTIVE)) { + job_fail_dependencies(u, UNIT_REQUIRED_BY); + job_fail_dependencies(u, UNIT_REQUISITE_OF); + job_fail_dependencies(u, UNIT_BOUND_BY); + } else if (t == JOB_STOP) + job_fail_dependencies(u, UNIT_CONFLICTED_BY); + } + + /* Trigger OnFailure dependencies that are not generated by + * the unit itself. We don't treat JOB_CANCELED as failure in + * this context. And JOB_FAILURE is already handled by the + * unit itself. */ + if (result == JOB_TIMEOUT || result == JOB_DEPENDENCY) { + log_struct(LOG_NOTICE, + "JOB_TYPE=%s", job_type_to_string(t), + "JOB_RESULT=%s", job_result_to_string(result), + LOG_UNIT_ID(u), + LOG_UNIT_MESSAGE(u, "Job %s/%s failed with result '%s'.", + u->id, + job_type_to_string(t), + job_result_to_string(result)), + NULL); + + unit_start_on_failure(u); + } + + unit_trigger_notify(u); + +finish: + /* Try to start the next jobs that can be started */ + SET_FOREACH(other, u->dependencies[UNIT_AFTER], i) + if (other->job) + job_add_to_run_queue(other->job); + SET_FOREACH(other, u->dependencies[UNIT_BEFORE], i) + if (other->job) + job_add_to_run_queue(other->job); + + manager_check_finished(u->manager); + + return 0; +} + +static int job_dispatch_timer(sd_event_source *s, uint64_t monotonic, void *userdata) { + Job *j = userdata; + Unit *u; + + assert(j); + assert(s == j->timer_event_source); + + log_unit_warning(j->unit, "Job %s/%s timed out.", j->unit->id, job_type_to_string(j->type)); + + u = j->unit; + job_finish_and_invalidate(j, JOB_TIMEOUT, true, false); + + failure_action(u->manager, u->job_timeout_action, u->job_timeout_reboot_arg); + + return 0; +} + +int job_start_timer(Job *j) { + int r; + + if (j->timer_event_source) + return 0; + + j->begin_usec = now(CLOCK_MONOTONIC); + + if (j->unit->job_timeout == USEC_INFINITY) + return 0; + + r = sd_event_add_time( + j->manager->event, + &j->timer_event_source, + CLOCK_MONOTONIC, + usec_add(j->begin_usec, j->unit->job_timeout), 0, + job_dispatch_timer, j); + if (r < 0) + return r; + + (void) sd_event_source_set_description(j->timer_event_source, "job-start"); + + return 0; +} + +void job_add_to_run_queue(Job *j) { + assert(j); + assert(j->installed); + + if (j->in_run_queue) + return; + + if (!j->manager->run_queue) + sd_event_source_set_enabled(j->manager->run_queue_event_source, SD_EVENT_ONESHOT); + + LIST_PREPEND(run_queue, j->manager->run_queue, j); + j->in_run_queue = true; +} + +void job_add_to_dbus_queue(Job *j) { + assert(j); + assert(j->installed); + + if (j->in_dbus_queue) + return; + + /* We don't check if anybody is subscribed here, since this + * job might just have been created and not yet assigned to a + * connection/client. */ + + LIST_PREPEND(dbus_queue, j->manager->dbus_job_queue, j); + j->in_dbus_queue = true; +} + +char *job_dbus_path(Job *j) { + char *p; + + assert(j); + + if (asprintf(&p, "/org/freedesktop/systemd1/job/%"PRIu32, j->id) < 0) + return NULL; + + return p; +} + +int job_serialize(Job *j, FILE *f, FDSet *fds) { + fprintf(f, "job-id=%u\n", j->id); + fprintf(f, "job-type=%s\n", job_type_to_string(j->type)); + fprintf(f, "job-state=%s\n", job_state_to_string(j->state)); + fprintf(f, "job-irreversible=%s\n", yes_no(j->irreversible)); + fprintf(f, "job-sent-dbus-new-signal=%s\n", yes_no(j->sent_dbus_new_signal)); + fprintf(f, "job-ignore-order=%s\n", yes_no(j->ignore_order)); + + if (j->begin_usec > 0) + fprintf(f, "job-begin="USEC_FMT"\n", j->begin_usec); + + bus_track_serialize(j->clients, f); + + /* End marker */ + fputc('\n', f); + return 0; +} + +int job_deserialize(Job *j, FILE *f, FDSet *fds) { + assert(j); + + for (;;) { + char line[LINE_MAX], *l, *v; + size_t k; + + if (!fgets(line, sizeof(line), f)) { + if (feof(f)) + return 0; + return -errno; + } + + char_array_0(line); + l = strstrip(line); + + /* End marker */ + if (l[0] == 0) + return 0; + + k = strcspn(l, "="); + + if (l[k] == '=') { + l[k] = 0; + v = l+k+1; + } else + v = l+k; + + if (streq(l, "job-id")) { + + if (safe_atou32(v, &j->id) < 0) + log_debug("Failed to parse job id value %s", v); + + } else if (streq(l, "job-type")) { + JobType t; + + t = job_type_from_string(v); + if (t < 0) + log_debug("Failed to parse job type %s", v); + else if (t >= _JOB_TYPE_MAX_IN_TRANSACTION) + log_debug("Cannot deserialize job of type %s", v); + else + j->type = t; + + } else if (streq(l, "job-state")) { + JobState s; + + s = job_state_from_string(v); + if (s < 0) + log_debug("Failed to parse job state %s", v); + else + job_set_state(j, s); + + } else if (streq(l, "job-irreversible")) { + int b; + + b = parse_boolean(v); + if (b < 0) + log_debug("Failed to parse job irreversible flag %s", v); + else + j->irreversible = j->irreversible || b; + + } else if (streq(l, "job-sent-dbus-new-signal")) { + int b; + + b = parse_boolean(v); + if (b < 0) + log_debug("Failed to parse job sent_dbus_new_signal flag %s", v); + else + j->sent_dbus_new_signal = j->sent_dbus_new_signal || b; + + } else if (streq(l, "job-ignore-order")) { + int b; + + b = parse_boolean(v); + if (b < 0) + log_debug("Failed to parse job ignore_order flag %s", v); + else + j->ignore_order = j->ignore_order || b; + + } else if (streq(l, "job-begin")) { + unsigned long long ull; + + if (sscanf(v, "%llu", &ull) != 1) + log_debug("Failed to parse job-begin value %s", v); + else + j->begin_usec = ull; + + } else if (streq(l, "subscribed")) { + + if (strv_extend(&j->deserialized_clients, v) < 0) + return log_oom(); + } + } +} + +int job_coldplug(Job *j) { + int r; + + assert(j); + + /* After deserialization is complete and the bus connection + * set up again, let's start watching our subscribers again */ + r = bus_track_coldplug(j->manager, &j->clients, &j->deserialized_clients); + if (r < 0) + return r; + + if (j->state == JOB_WAITING) + job_add_to_run_queue(j); + + if (j->begin_usec == 0 || j->unit->job_timeout == USEC_INFINITY) + return 0; + + j->timer_event_source = sd_event_source_unref(j->timer_event_source); + + r = sd_event_add_time( + j->manager->event, + &j->timer_event_source, + CLOCK_MONOTONIC, + usec_add(j->begin_usec, j->unit->job_timeout), 0, + job_dispatch_timer, j); + if (r < 0) + log_debug_errno(r, "Failed to restart timeout for job: %m"); + + (void) sd_event_source_set_description(j->timer_event_source, "job-timeout"); + + return r; +} + +void job_shutdown_magic(Job *j) { + assert(j); + + /* The shutdown target gets some special treatment here: we + * tell the kernel to begin with flushing its disk caches, to + * optimize shutdown time a bit. Ideally we wouldn't hardcode + * this magic into PID 1. However all other processes aren't + * options either since they'd exit much sooner than PID 1 and + * asynchronous sync() would cause their exit to be + * delayed. */ + + if (j->type != JOB_START) + return; + + if (!MANAGER_IS_SYSTEM(j->unit->manager)) + return; + + if (!unit_has_name(j->unit, SPECIAL_SHUTDOWN_TARGET)) + return; + + /* In case messages on console has been disabled on boot */ + j->unit->manager->no_console_output = false; + + if (detect_container() > 0) + return; + + asynchronous_sync(); +} + +int job_get_timeout(Job *j, usec_t *timeout) { + usec_t x = USEC_INFINITY, y = USEC_INFINITY; + Unit *u = j->unit; + int r; + + assert(u); + + if (j->timer_event_source) { + r = sd_event_source_get_time(j->timer_event_source, &x); + if (r < 0) + return r; + } + + if (UNIT_VTABLE(u)->get_timeout) { + r = UNIT_VTABLE(u)->get_timeout(u, &y); + if (r < 0) + return r; + } + + if (x == USEC_INFINITY && y == USEC_INFINITY) + return 0; + + *timeout = MIN(x, y); + return 1; +} + +static const char* const job_state_table[_JOB_STATE_MAX] = { + [JOB_WAITING] = "waiting", + [JOB_RUNNING] = "running" +}; + +DEFINE_STRING_TABLE_LOOKUP(job_state, JobState); + +static const char* const job_type_table[_JOB_TYPE_MAX] = { + [JOB_START] = "start", + [JOB_VERIFY_ACTIVE] = "verify-active", + [JOB_STOP] = "stop", + [JOB_RELOAD] = "reload", + [JOB_RELOAD_OR_START] = "reload-or-start", + [JOB_RESTART] = "restart", + [JOB_TRY_RESTART] = "try-restart", + [JOB_TRY_RELOAD] = "try-reload", + [JOB_NOP] = "nop", +}; + +DEFINE_STRING_TABLE_LOOKUP(job_type, JobType); + +static const char* const job_mode_table[_JOB_MODE_MAX] = { + [JOB_FAIL] = "fail", + [JOB_REPLACE] = "replace", + [JOB_REPLACE_IRREVERSIBLY] = "replace-irreversibly", + [JOB_ISOLATE] = "isolate", + [JOB_FLUSH] = "flush", + [JOB_IGNORE_DEPENDENCIES] = "ignore-dependencies", + [JOB_IGNORE_REQUIREMENTS] = "ignore-requirements", +}; + +DEFINE_STRING_TABLE_LOOKUP(job_mode, JobMode); + +static const char* const job_result_table[_JOB_RESULT_MAX] = { + [JOB_DONE] = "done", + [JOB_CANCELED] = "canceled", + [JOB_TIMEOUT] = "timeout", + [JOB_FAILED] = "failed", + [JOB_DEPENDENCY] = "dependency", + [JOB_SKIPPED] = "skipped", + [JOB_INVALID] = "invalid", + [JOB_ASSERT] = "assert", + [JOB_UNSUPPORTED] = "unsupported", +}; + +DEFINE_STRING_TABLE_LOOKUP(job_result, JobResult); + +const char* job_type_to_access_method(JobType t) { + assert(t >= 0); + assert(t < _JOB_TYPE_MAX); + + if (IN_SET(t, JOB_START, JOB_RESTART, JOB_TRY_RESTART)) + return "start"; + else if (t == JOB_STOP) + return "stop"; + else + return "reload"; +} diff --git a/src/grp-system/libcore/src/kill.c b/src/grp-system/libcore/src/kill.c new file mode 100644 index 0000000000..2f6d81f4eb --- /dev/null +++ b/src/grp-system/libcore/src/kill.c @@ -0,0 +1,68 @@ +/*** + This file is part of systemd. + + Copyright 2012 Lennart Poettering + + systemd is free software; you can redistribute it and/or modify it + under the terms of the GNU Lesser General Public License as published by + the Free Software Foundation; either version 2.1 of the License, or + (at your option) any later version. + + systemd is distributed in the hope that it will be useful, but + WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + Lesser General Public License for more details. + + You should have received a copy of the GNU Lesser General Public License + along with systemd; If not, see . +***/ + +#include "core/kill.h" +#include "systemd-basic/signal-util.h" +#include "systemd-basic/string-table.h" +#include "systemd-basic/util.h" + +void kill_context_init(KillContext *c) { + assert(c); + + c->kill_signal = SIGTERM; + c->send_sigkill = true; + c->send_sighup = false; +} + +void kill_context_dump(KillContext *c, FILE *f, const char *prefix) { + assert(c); + + if (!prefix) + prefix = ""; + + fprintf(f, + "%sKillMode: %s\n" + "%sKillSignal: SIG%s\n" + "%sSendSIGKILL: %s\n" + "%sSendSIGHUP: %s\n", + prefix, kill_mode_to_string(c->kill_mode), + prefix, signal_to_string(c->kill_signal), + prefix, yes_no(c->send_sigkill), + prefix, yes_no(c->send_sighup)); +} + +static const char* const kill_mode_table[_KILL_MODE_MAX] = { + [KILL_CONTROL_GROUP] = "control-group", + [KILL_PROCESS] = "process", + [KILL_MIXED] = "mixed", + [KILL_NONE] = "none" +}; + +DEFINE_STRING_TABLE_LOOKUP(kill_mode, KillMode); + +static const char* const kill_who_table[_KILL_WHO_MAX] = { + [KILL_MAIN] = "main", + [KILL_CONTROL] = "control", + [KILL_ALL] = "all", + [KILL_MAIN_FAIL] = "main-fail", + [KILL_CONTROL_FAIL] = "control-fail", + [KILL_ALL_FAIL] = "all-fail" +}; + +DEFINE_STRING_TABLE_LOOKUP(kill_who, KillWho); diff --git a/src/grp-system/libcore/src/killall.c b/src/grp-system/libcore/src/killall.c new file mode 100644 index 0000000000..3e4b5e5186 --- /dev/null +++ b/src/grp-system/libcore/src/killall.c @@ -0,0 +1,248 @@ +/*** + This file is part of systemd. + + Copyright 2010 ProFUSION embedded systems + + systemd is free software; you can redistribute it and/or modify it + under the terms of the GNU Lesser General Public License as published by + the Free Software Foundation; either version 2.1 of the License, or + (at your option) any later version. + + systemd is distributed in the hope that it will be useful, but + WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + Lesser General Public License for more details. + + You should have received a copy of the GNU Lesser General Public License + along with systemd; If not, see . +***/ + +#include +#include +#include +#include + +#include "core/killall.h" +#include "systemd-basic/alloc-util.h" +#include "systemd-basic/def.h" +#include "systemd-basic/fd-util.h" +#include "systemd-basic/formats-util.h" +#include "systemd-basic/parse-util.h" +#include "systemd-basic/process-util.h" +#include "systemd-basic/set.h" +#include "systemd-basic/string-util.h" +#include "systemd-basic/terminal-util.h" +#include "systemd-basic/util.h" + +static bool ignore_proc(pid_t pid, bool warn_rootfs) { + _cleanup_fclose_ FILE *f = NULL; + char c; + const char *p; + size_t count; + uid_t uid; + int r; + + /* We are PID 1, let's not commit suicide */ + if (pid == 1) + return true; + + r = get_process_uid(pid, &uid); + if (r < 0) + return true; /* not really, but better safe than sorry */ + + /* Non-root processes otherwise are always subject to be killed */ + if (uid != 0) + return false; + + p = procfs_file_alloca(pid, "cmdline"); + f = fopen(p, "re"); + if (!f) + return true; /* not really, but has the desired effect */ + + count = fread(&c, 1, 1, f); + + /* Kernel threads have an empty cmdline */ + if (count <= 0) + return true; + + /* Processes with argv[0][0] = '@' we ignore from the killing + * spree. + * + * http://www.freedesktop.org/wiki/Software/systemd/RootStorageDaemons */ + if (c == '@' && warn_rootfs) { + _cleanup_free_ char *comm = NULL; + + r = pid_from_same_root_fs(pid); + if (r < 0) + return true; + + get_process_comm(pid, &comm); + + if (r) + log_notice("Process " PID_FMT " (%s) has been marked to be excluded from killing. It is " + "running from the root file system, and thus likely to block re-mounting of the " + "root file system to read-only. Please consider moving it into an initrd file " + "system instead.", pid, strna(comm)); + return true; + } else if (c == '@') + return true; + + return false; +} + +static void wait_for_children(Set *pids, sigset_t *mask) { + usec_t until; + + assert(mask); + + if (set_isempty(pids)) + return; + + until = now(CLOCK_MONOTONIC) + DEFAULT_TIMEOUT_USEC; + for (;;) { + struct timespec ts; + int k; + usec_t n; + void *p; + Iterator i; + + /* First, let the kernel inform us about killed + * children. Most processes will probably be our + * children, but some are not (might be our + * grandchildren instead...). */ + for (;;) { + pid_t pid; + + pid = waitpid(-1, NULL, WNOHANG); + if (pid == 0) + break; + if (pid < 0) { + if (errno == ECHILD) + break; + + log_error_errno(errno, "waitpid() failed: %m"); + return; + } + + (void) set_remove(pids, PID_TO_PTR(pid)); + } + + /* Now explicitly check who might be remaining, who + * might not be our child. */ + SET_FOREACH(p, pids, i) { + + /* We misuse getpgid as a check whether a + * process still exists. */ + if (getpgid(PTR_TO_PID(p)) >= 0) + continue; + + if (errno != ESRCH) + continue; + + set_remove(pids, p); + } + + if (set_isempty(pids)) + return; + + n = now(CLOCK_MONOTONIC); + if (n >= until) + return; + + timespec_store(&ts, until - n); + k = sigtimedwait(mask, NULL, &ts); + if (k != SIGCHLD) { + + if (k < 0 && errno != EAGAIN) { + log_error_errno(errno, "sigtimedwait() failed: %m"); + return; + } + + if (k >= 0) + log_warning("sigtimedwait() returned unexpected signal."); + } + } +} + +static int killall(int sig, Set *pids, bool send_sighup) { + _cleanup_closedir_ DIR *dir = NULL; + struct dirent *d; + + dir = opendir("/proc"); + if (!dir) + return -errno; + + while ((d = readdir(dir))) { + pid_t pid; + int r; + + if (d->d_type != DT_DIR && + d->d_type != DT_UNKNOWN) + continue; + + if (parse_pid(d->d_name, &pid) < 0) + continue; + + if (ignore_proc(pid, sig == SIGKILL && !in_initrd())) + continue; + + if (sig == SIGKILL) { + _cleanup_free_ char *s = NULL; + + get_process_comm(pid, &s); + log_notice("Sending SIGKILL to PID "PID_FMT" (%s).", pid, strna(s)); + } + + if (kill(pid, sig) >= 0) { + if (pids) { + r = set_put(pids, PID_TO_PTR(pid)); + if (r < 0) + log_oom(); + } + } else if (errno != ENOENT) + log_warning_errno(errno, "Could not kill %d: %m", pid); + + if (send_sighup) { + /* Optionally, also send a SIGHUP signal, but + only if the process has a controlling + tty. This is useful to allow handling of + shells which ignore SIGTERM but react to + SIGHUP. We do not send this to processes that + have no controlling TTY since we don't want to + trigger reloads of daemon processes. Also we + make sure to only send this after SIGTERM so + that SIGTERM is always first in the queue. */ + + + if (get_ctty_devnr(pid, NULL) >= 0) + kill(pid, SIGHUP); + } + } + + return set_size(pids); +} + +void broadcast_signal(int sig, bool wait_for_exit, bool send_sighup) { + sigset_t mask, oldmask; + _cleanup_set_free_ Set *pids = NULL; + + if (wait_for_exit) + pids = set_new(NULL); + + assert_se(sigemptyset(&mask) == 0); + assert_se(sigaddset(&mask, SIGCHLD) == 0); + assert_se(sigprocmask(SIG_BLOCK, &mask, &oldmask) == 0); + + if (kill(-1, SIGSTOP) < 0 && errno != ESRCH) + log_warning_errno(errno, "kill(-1, SIGSTOP) failed: %m"); + + killall(sig, pids, send_sighup); + + if (kill(-1, SIGCONT) < 0 && errno != ESRCH) + log_warning_errno(errno, "kill(-1, SIGCONT) failed: %m"); + + if (wait_for_exit) + wait_for_children(pids, &mask); + + assert_se(sigprocmask(SIG_SETMASK, &oldmask, NULL) == 0); +} diff --git a/src/grp-system/libcore/src/kmod-setup.c b/src/grp-system/libcore/src/kmod-setup.c new file mode 100644 index 0000000000..0f935380b2 --- /dev/null +++ b/src/grp-system/libcore/src/kmod-setup.c @@ -0,0 +1,128 @@ +/*** + This file is part of systemd. + + Copyright 2010 Lennart Poettering + + systemd is free software; you can redistribute it and/or modify it + under the terms of the GNU Lesser General Public License as published by + the Free Software Foundation; either version 2.1 of the License, or + (at your option) any later version. + + systemd is distributed in the hope that it will be useful, but + WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + Lesser General Public License for more details. + + You should have received a copy of the GNU Lesser General Public License + along with systemd; If not, see . +***/ + +#include +#include + +#ifdef HAVE_KMOD +#include +#endif + +#include "core/kmod-setup.h" +#include "sd-bus/bus-util.h" +#include "systemd-basic/capability-util.h" +#include "systemd-basic/macro.h" + +#ifdef HAVE_KMOD +static void systemd_kmod_log( + void *data, + int priority, + const char *file, int line, + const char *fn, + const char *format, + va_list args) { + + /* library logging is enabled at debug only */ + DISABLE_WARNING_FORMAT_NONLITERAL; + log_internalv(LOG_DEBUG, 0, file, line, fn, format, args); + REENABLE_WARNING; +} +#endif + +int kmod_setup(void) { +#ifdef HAVE_KMOD + + static const struct { + const char *module; + const char *path; + bool warn_if_unavailable:1; + bool warn_if_module:1; + bool (*condition_fn)(void); + } kmod_table[] = { + /* auto-loading on use doesn't work before udev is up */ + { "autofs4", "/sys/class/misc/autofs", true, false, NULL }, + + /* early configure of ::1 on the loopback device */ + { "ipv6", "/sys/module/ipv6", false, true, NULL }, + + /* this should never be a module */ + { "unix", "/proc/net/unix", true, true, NULL }, + +#ifdef HAVE_LIBIPTC + /* netfilter is needed by networkd, nspawn among others, and cannot be autoloaded */ + { "ip_tables", "/proc/net/ip_tables_names", false, false, NULL }, +#endif + }; + struct kmod_ctx *ctx = NULL; + unsigned int i; + int r; + + if (have_effective_cap(CAP_SYS_MODULE) == 0) + return 0; + + for (i = 0; i < ELEMENTSOF(kmod_table); i++) { + struct kmod_module *mod; + + if (kmod_table[i].path && access(kmod_table[i].path, F_OK) >= 0) + continue; + + if (kmod_table[i].condition_fn && !kmod_table[i].condition_fn()) + continue; + + if (kmod_table[i].warn_if_module) + log_debug("Your kernel apparently lacks built-in %s support. Might be " + "a good idea to compile it in. We'll now try to work around " + "this by loading the module...", kmod_table[i].module); + + if (!ctx) { + ctx = kmod_new(NULL, NULL); + if (!ctx) + return log_oom(); + + kmod_set_log_fn(ctx, systemd_kmod_log, NULL); + kmod_load_resources(ctx); + } + + r = kmod_module_new_from_name(ctx, kmod_table[i].module, &mod); + if (r < 0) { + log_error("Failed to lookup module '%s'", kmod_table[i].module); + continue; + } + + r = kmod_module_probe_insert_module(mod, KMOD_PROBE_APPLY_BLACKLIST, NULL, NULL, NULL, NULL); + if (r == 0) + log_debug("Inserted module '%s'", kmod_module_get_name(mod)); + else if (r == KMOD_PROBE_APPLY_BLACKLIST) + log_info("Module '%s' is blacklisted", kmod_module_get_name(mod)); + else { + bool print_warning = kmod_table[i].warn_if_unavailable || (r < 0 && r != -ENOENT); + + log_full_errno(print_warning ? LOG_WARNING : LOG_DEBUG, r, + "Failed to insert module '%s': %m", kmod_module_get_name(mod)); + } + + kmod_module_unref(mod); + } + + if (ctx) + kmod_unref(ctx); + +#endif + return 0; +} diff --git a/src/grp-system/libcore/src/linux/auto_dev-ioctl.h b/src/grp-system/libcore/src/linux/auto_dev-ioctl.h new file mode 100644 index 0000000000..aeaeb3ea7a --- /dev/null +++ b/src/grp-system/libcore/src/linux/auto_dev-ioctl.h @@ -0,0 +1,228 @@ +/* + * Copyright 2008 Red Hat, Inc. All rights reserved. + * Copyright 2008 Ian Kent + * + * This file is part of the Linux kernel and is made available under + * the terms of the GNU General Public License, version 2, or at your + * option, any later version, incorporated herein by reference. + */ + +#ifndef _LINUX_AUTO_DEV_IOCTL_H +#define _LINUX_AUTO_DEV_IOCTL_H + +#include + +#ifdef __KERNEL__ +#include +#else +#include +#endif /* __KERNEL__ */ + +#define AUTOFS_DEVICE_NAME "autofs" + +#define AUTOFS_DEV_IOCTL_VERSION_MAJOR 1 +#define AUTOFS_DEV_IOCTL_VERSION_MINOR 0 + +#define AUTOFS_DEVID_LEN 16 + +#define AUTOFS_DEV_IOCTL_SIZE sizeof(struct autofs_dev_ioctl) + +/* + * An ioctl interface for autofs mount point control. + */ + +struct args_protover { + __u32 version; +}; + +struct args_protosubver { + __u32 sub_version; +}; + +struct args_openmount { + __u32 devid; +}; + +struct args_ready { + __u32 token; +}; + +struct args_fail { + __u32 token; + __s32 status; +}; + +struct args_setpipefd { + __s32 pipefd; +}; + +struct args_timeout { + __u64 timeout; +}; + +struct args_requester { + __u32 uid; + __u32 gid; +}; + +struct args_expire { + __u32 how; +}; + +struct args_askumount { + __u32 may_umount; +}; + +struct args_ismountpoint { + union { + struct args_in { + __u32 type; + } in; + struct args_out { + __u32 devid; + __u32 magic; + } out; + }; +}; + +/* + * All the ioctls use this structure. + * When sending a path size must account for the total length + * of the chunk of memory otherwise is is the size of the + * structure. + */ + +struct autofs_dev_ioctl { + __u32 ver_major; + __u32 ver_minor; + __u32 size; /* total size of data passed in + * including this struct */ + __s32 ioctlfd; /* automount command fd */ + + /* Command parameters */ + + union { + struct args_protover protover; + struct args_protosubver protosubver; + struct args_openmount openmount; + struct args_ready ready; + struct args_fail fail; + struct args_setpipefd setpipefd; + struct args_timeout timeout; + struct args_requester requester; + struct args_expire expire; + struct args_askumount askumount; + struct args_ismountpoint ismountpoint; + }; + + char path[0]; +}; + +static inline void init_autofs_dev_ioctl(struct autofs_dev_ioctl *in) { + memset(in, 0, sizeof(struct autofs_dev_ioctl)); + in->ver_major = AUTOFS_DEV_IOCTL_VERSION_MAJOR; + in->ver_minor = AUTOFS_DEV_IOCTL_VERSION_MINOR; + in->size = sizeof(struct autofs_dev_ioctl); + in->ioctlfd = -1; + return; +} + +/* + * If you change this make sure you make the corresponding change + * to autofs-dev-ioctl.c:lookup_ioctl() + */ +enum { + /* Get various version info */ + AUTOFS_DEV_IOCTL_VERSION_CMD = 0x71, + AUTOFS_DEV_IOCTL_PROTOVER_CMD, + AUTOFS_DEV_IOCTL_PROTOSUBVER_CMD, + + /* Open mount ioctl fd */ + AUTOFS_DEV_IOCTL_OPENMOUNT_CMD, + + /* Close mount ioctl fd */ + AUTOFS_DEV_IOCTL_CLOSEMOUNT_CMD, + + /* Mount/expire status returns */ + AUTOFS_DEV_IOCTL_READY_CMD, + AUTOFS_DEV_IOCTL_FAIL_CMD, + + /* Activate/deactivate autofs mount */ + AUTOFS_DEV_IOCTL_SETPIPEFD_CMD, + AUTOFS_DEV_IOCTL_CATATONIC_CMD, + + /* Expiry timeout */ + AUTOFS_DEV_IOCTL_TIMEOUT_CMD, + + /* Get mount last requesting uid and gid */ + AUTOFS_DEV_IOCTL_REQUESTER_CMD, + + /* Check for eligible expire candidates */ + AUTOFS_DEV_IOCTL_EXPIRE_CMD, + + /* Request busy status */ + AUTOFS_DEV_IOCTL_ASKUMOUNT_CMD, + + /* Check if path is a mountpoint */ + AUTOFS_DEV_IOCTL_ISMOUNTPOINT_CMD, +}; + +#define AUTOFS_IOCTL 0x93 + +#define AUTOFS_DEV_IOCTL_VERSION \ + _IOWR(AUTOFS_IOCTL, \ + AUTOFS_DEV_IOCTL_VERSION_CMD, struct autofs_dev_ioctl) + +#define AUTOFS_DEV_IOCTL_PROTOVER \ + _IOWR(AUTOFS_IOCTL, \ + AUTOFS_DEV_IOCTL_PROTOVER_CMD, struct autofs_dev_ioctl) + +#define AUTOFS_DEV_IOCTL_PROTOSUBVER \ + _IOWR(AUTOFS_IOCTL, \ + AUTOFS_DEV_IOCTL_PROTOSUBVER_CMD, struct autofs_dev_ioctl) + +#define AUTOFS_DEV_IOCTL_OPENMOUNT \ + _IOWR(AUTOFS_IOCTL, \ + AUTOFS_DEV_IOCTL_OPENMOUNT_CMD, struct autofs_dev_ioctl) + +#define AUTOFS_DEV_IOCTL_CLOSEMOUNT \ + _IOWR(AUTOFS_IOCTL, \ + AUTOFS_DEV_IOCTL_CLOSEMOUNT_CMD, struct autofs_dev_ioctl) + +#define AUTOFS_DEV_IOCTL_READY \ + _IOWR(AUTOFS_IOCTL, \ + AUTOFS_DEV_IOCTL_READY_CMD, struct autofs_dev_ioctl) + +#define AUTOFS_DEV_IOCTL_FAIL \ + _IOWR(AUTOFS_IOCTL, \ + AUTOFS_DEV_IOCTL_FAIL_CMD, struct autofs_dev_ioctl) + +#define AUTOFS_DEV_IOCTL_SETPIPEFD \ + _IOWR(AUTOFS_IOCTL, \ + AUTOFS_DEV_IOCTL_SETPIPEFD_CMD, struct autofs_dev_ioctl) + +#define AUTOFS_DEV_IOCTL_CATATONIC \ + _IOWR(AUTOFS_IOCTL, \ + AUTOFS_DEV_IOCTL_CATATONIC_CMD, struct autofs_dev_ioctl) + +#define AUTOFS_DEV_IOCTL_TIMEOUT \ + _IOWR(AUTOFS_IOCTL, \ + AUTOFS_DEV_IOCTL_TIMEOUT_CMD, struct autofs_dev_ioctl) + +#define AUTOFS_DEV_IOCTL_REQUESTER \ + _IOWR(AUTOFS_IOCTL, \ + AUTOFS_DEV_IOCTL_REQUESTER_CMD, struct autofs_dev_ioctl) + +#define AUTOFS_DEV_IOCTL_EXPIRE \ + _IOWR(AUTOFS_IOCTL, \ + AUTOFS_DEV_IOCTL_EXPIRE_CMD, struct autofs_dev_ioctl) + +#define AUTOFS_DEV_IOCTL_ASKUMOUNT \ + _IOWR(AUTOFS_IOCTL, \ + AUTOFS_DEV_IOCTL_ASKUMOUNT_CMD, struct autofs_dev_ioctl) + +#define AUTOFS_DEV_IOCTL_ISMOUNTPOINT \ + _IOWR(AUTOFS_IOCTL, \ + AUTOFS_DEV_IOCTL_ISMOUNTPOINT_CMD, struct autofs_dev_ioctl) + +#endif /* _LINUX_AUTO_DEV_IOCTL_H */ diff --git a/src/grp-system/libcore/src/load-dropin.c b/src/grp-system/libcore/src/load-dropin.c new file mode 100644 index 0000000000..4374a1fc25 --- /dev/null +++ b/src/grp-system/libcore/src/load-dropin.c @@ -0,0 +1,91 @@ +/*** + This file is part of systemd. + + Copyright 2010 Lennart Poettering + + systemd is free software; you can redistribute it and/or modify it + under the terms of the GNU Lesser General Public License as published by + the Free Software Foundation; either version 2.1 of the License, or + (at your option) any later version. + + systemd is distributed in the hope that it will be useful, but + WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + Lesser General Public License for more details. + + You should have received a copy of the GNU Lesser General Public License + along with systemd; If not, see . +***/ + + +#include "core/load-fragment.h" +#include "core/unit.h" +#include "systemd-basic/log.h" +#include "systemd-basic/strv.h" +#include "systemd-basic/unit-name.h" +#include "systemd-shared/conf-parser.h" + +#include "load-dropin.h" + +static int add_dependency_consumer( + UnitDependency dependency, + const char *entry, + const char* filepath, + void *arg) { + Unit *u = arg; + int r; + + assert(u); + + r = unit_add_dependency_by_name(u, dependency, entry, filepath, true); + if (r < 0) + log_error_errno(r, "Cannot add dependency %s to %s, ignoring: %m", entry, u->id); + + return 0; +} + +int unit_load_dropin(Unit *u) { + _cleanup_strv_free_ char **l = NULL; + Iterator i; + char *t, **f; + int r; + + assert(u); + + /* Load dependencies from supplementary drop-in directories */ + + SET_FOREACH(t, u->names, i) { + char **p; + + STRV_FOREACH(p, u->manager->lookup_paths.search_path) { + unit_file_process_dir(u->manager->unit_path_cache, *p, t, ".wants", UNIT_WANTS, + add_dependency_consumer, u, NULL); + unit_file_process_dir(u->manager->unit_path_cache, *p, t, ".requires", UNIT_REQUIRES, + add_dependency_consumer, u, NULL); + } + } + + r = unit_find_dropin_paths(u, &l); + if (r <= 0) + return 0; + + if (!u->dropin_paths) { + u->dropin_paths = l; + l = NULL; + } else { + r = strv_extend_strv(&u->dropin_paths, l, true); + if (r < 0) + return log_oom(); + } + + STRV_FOREACH(f, u->dropin_paths) { + config_parse(u->id, *f, NULL, + UNIT_VTABLE(u)->sections, + config_item_perf_lookup, load_fragment_gperf_lookup, + false, false, false, u); + } + + u->dropin_mtime = now(CLOCK_REALTIME); + + return 0; +} diff --git a/src/grp-system/libcore/src/load-dropin.h b/src/grp-system/libcore/src/load-dropin.h new file mode 100644 index 0000000000..aa0fa024de --- /dev/null +++ b/src/grp-system/libcore/src/load-dropin.h @@ -0,0 +1,34 @@ +#pragma once + +/*** + This file is part of systemd. + + Copyright 2010 Lennart Poettering + + systemd is free software; you can redistribute it and/or modify it + under the terms of the GNU Lesser General Public License as published by + the Free Software Foundation; either version 2.1 of the License, or + (at your option) any later version. + + systemd is distributed in the hope that it will be useful, but + WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + Lesser General Public License for more details. + + You should have received a copy of the GNU Lesser General Public License + along with systemd; If not, see . +***/ + +#include "core/unit.h" +#include "systemd-shared/dropin.h" + +/* Read service data supplementary drop-in directories */ + +static inline int unit_find_dropin_paths(Unit *u, char ***paths) { + return unit_file_find_dropin_paths(u->manager->lookup_paths.search_path, + u->manager->unit_path_cache, + u->names, + paths); +} + +int unit_load_dropin(Unit *u); diff --git a/src/grp-system/libcore/src/load-fragment-gperf.gperf.m4 b/src/grp-system/libcore/src/load-fragment-gperf.gperf.m4 new file mode 100644 index 0000000000..4658fe71b8 --- /dev/null +++ b/src/grp-system/libcore/src/load-fragment-gperf.gperf.m4 @@ -0,0 +1,402 @@ +%{ +#include + +#include "core/load-fragment.h" +#include "systemd-basic/missing.h" +#include "systemd-shared/conf-parser.h" +%} +struct ConfigPerfItem; +%null_strings +%language=ANSI-C +%define slot-name section_and_lvalue +%define hash-function-name load_fragment_gperf_hash +%define lookup-function-name load_fragment_gperf_lookup +%readonly-tables +%omit-struct-type +%struct-type +%includes +%% +m4_dnl Define the context options only once +m4_define(`EXEC_CONTEXT_CONFIG_ITEMS', +`$1.WorkingDirectory, config_parse_working_directory, 0, offsetof($1, exec_context) +$1.RootDirectory, config_parse_unit_path_printf, 0, offsetof($1, exec_context.root_directory) +$1.User, config_parse_unit_string_printf, 0, offsetof($1, exec_context.user) +$1.Group, config_parse_unit_string_printf, 0, offsetof($1, exec_context.group) +$1.SupplementaryGroups, config_parse_strv, 0, offsetof($1, exec_context.supplementary_groups) +$1.Nice, config_parse_exec_nice, 0, offsetof($1, exec_context) +$1.OOMScoreAdjust, config_parse_exec_oom_score_adjust, 0, offsetof($1, exec_context) +$1.IOSchedulingClass, config_parse_exec_io_class, 0, offsetof($1, exec_context) +$1.IOSchedulingPriority, config_parse_exec_io_priority, 0, offsetof($1, exec_context) +$1.CPUSchedulingPolicy, config_parse_exec_cpu_sched_policy, 0, offsetof($1, exec_context) +$1.CPUSchedulingPriority, config_parse_exec_cpu_sched_prio, 0, offsetof($1, exec_context) +$1.CPUSchedulingResetOnFork, config_parse_bool, 0, offsetof($1, exec_context.cpu_sched_reset_on_fork) +$1.CPUAffinity, config_parse_exec_cpu_affinity, 0, offsetof($1, exec_context) +$1.UMask, config_parse_mode, 0, offsetof($1, exec_context.umask) +$1.Environment, config_parse_environ, 0, offsetof($1, exec_context.environment) +$1.EnvironmentFile, config_parse_unit_env_file, 0, offsetof($1, exec_context.environment_files) +$1.PassEnvironment, config_parse_pass_environ, 0, offsetof($1, exec_context.pass_environment) +$1.StandardInput, config_parse_input, 0, offsetof($1, exec_context.std_input) +$1.StandardOutput, config_parse_output, 0, offsetof($1, exec_context.std_output) +$1.StandardError, config_parse_output, 0, offsetof($1, exec_context.std_error) +$1.TTYPath, config_parse_unit_path_printf, 0, offsetof($1, exec_context.tty_path) +$1.TTYReset, config_parse_bool, 0, offsetof($1, exec_context.tty_reset) +$1.TTYVHangup, config_parse_bool, 0, offsetof($1, exec_context.tty_vhangup) +$1.TTYVTDisallocate, config_parse_bool, 0, offsetof($1, exec_context.tty_vt_disallocate) +$1.SyslogIdentifier, config_parse_unit_string_printf, 0, offsetof($1, exec_context.syslog_identifier) +$1.SyslogFacility, config_parse_log_facility, 0, offsetof($1, exec_context.syslog_priority) +$1.SyslogLevel, config_parse_log_level, 0, offsetof($1, exec_context.syslog_priority) +$1.SyslogLevelPrefix, config_parse_bool, 0, offsetof($1, exec_context.syslog_level_prefix) +$1.Capabilities, config_parse_warn_compat, DISABLED_LEGACY, offsetof($1, exec_context) +$1.SecureBits, config_parse_exec_secure_bits, 0, offsetof($1, exec_context) +$1.CapabilityBoundingSet, config_parse_capability_set, 0, offsetof($1, exec_context.capability_bounding_set) +$1.AmbientCapabilities, config_parse_capability_set, 0, offsetof($1, exec_context.capability_ambient_set) +$1.TimerSlackNSec, config_parse_nsec, 0, offsetof($1, exec_context.timer_slack_nsec) +$1.NoNewPrivileges, config_parse_no_new_privileges, 0, offsetof($1, exec_context) +m4_ifdef(`HAVE_SECCOMP', +`$1.SystemCallFilter, config_parse_syscall_filter, 0, offsetof($1, exec_context) +$1.SystemCallArchitectures, config_parse_syscall_archs, 0, offsetof($1, exec_context.syscall_archs) +$1.SystemCallErrorNumber, config_parse_syscall_errno, 0, offsetof($1, exec_context) +$1.MemoryDenyWriteExecute, config_parse_bool, 0, offsetof($1, exec_context.memory_deny_write_execute) +$1.RestrictRealtime, config_parse_bool, 0, offsetof($1, exec_context.restrict_realtime) +$1.RestrictAddressFamilies, config_parse_address_families, 0, offsetof($1, exec_context)', +`$1.SystemCallFilter, config_parse_warn_compat, DISABLED_CONFIGURATION, 0 +$1.SystemCallArchitectures, config_parse_warn_compat, DISABLED_CONFIGURATION, 0 +$1.SystemCallErrorNumber, config_parse_warn_compat, DISABLED_CONFIGURATION, 0 +$1.MemoryDenyWriteExecute, config_parse_warn_compat, DISABLED_CONFIGURATION, 0 +$1.RestrictRealtime, config_parse_warn_compat, DISABLED_CONFIGURATION, 0 +$1.RestrictAddressFamilies, config_parse_warn_compat, DISABLED_CONFIGURATION, 0') +$1.LimitCPU, config_parse_limit, RLIMIT_CPU, offsetof($1, exec_context.rlimit) +$1.LimitFSIZE, config_parse_limit, RLIMIT_FSIZE, offsetof($1, exec_context.rlimit) +$1.LimitDATA, config_parse_limit, RLIMIT_DATA, offsetof($1, exec_context.rlimit) +$1.LimitSTACK, config_parse_limit, RLIMIT_STACK, offsetof($1, exec_context.rlimit) +$1.LimitCORE, config_parse_limit, RLIMIT_CORE, offsetof($1, exec_context.rlimit) +$1.LimitRSS, config_parse_limit, RLIMIT_RSS, offsetof($1, exec_context.rlimit) +$1.LimitNOFILE, config_parse_limit, RLIMIT_NOFILE, offsetof($1, exec_context.rlimit) +$1.LimitAS, config_parse_limit, RLIMIT_AS, offsetof($1, exec_context.rlimit) +$1.LimitNPROC, config_parse_limit, RLIMIT_NPROC, offsetof($1, exec_context.rlimit) +$1.LimitMEMLOCK, config_parse_limit, RLIMIT_MEMLOCK, offsetof($1, exec_context.rlimit) +$1.LimitLOCKS, config_parse_limit, RLIMIT_LOCKS, offsetof($1, exec_context.rlimit) +$1.LimitSIGPENDING, config_parse_limit, RLIMIT_SIGPENDING, offsetof($1, exec_context.rlimit) +$1.LimitMSGQUEUE, config_parse_limit, RLIMIT_MSGQUEUE, offsetof($1, exec_context.rlimit) +$1.LimitNICE, config_parse_limit, RLIMIT_NICE, offsetof($1, exec_context.rlimit) +$1.LimitRTPRIO, config_parse_limit, RLIMIT_RTPRIO, offsetof($1, exec_context.rlimit) +$1.LimitRTTIME, config_parse_limit, RLIMIT_RTTIME, offsetof($1, exec_context.rlimit) +$1.ReadWriteDirectories, config_parse_namespace_path_strv, 0, offsetof($1, exec_context.read_write_paths) +$1.ReadOnlyDirectories, config_parse_namespace_path_strv, 0, offsetof($1, exec_context.read_only_paths) +$1.InaccessibleDirectories, config_parse_namespace_path_strv, 0, offsetof($1, exec_context.inaccessible_paths) +$1.ReadWritePaths, config_parse_namespace_path_strv, 0, offsetof($1, exec_context.read_write_paths) +$1.ReadOnlyPaths, config_parse_namespace_path_strv, 0, offsetof($1, exec_context.read_only_paths) +$1.InaccessiblePaths, config_parse_namespace_path_strv, 0, offsetof($1, exec_context.inaccessible_paths) +$1.PrivateTmp, config_parse_bool, 0, offsetof($1, exec_context.private_tmp) +$1.PrivateNetwork, config_parse_bool, 0, offsetof($1, exec_context.private_network) +$1.PrivateDevices, config_parse_bool, 0, offsetof($1, exec_context.private_devices) +$1.ProtectSystem, config_parse_protect_system, 0, offsetof($1, exec_context) +$1.ProtectHome, config_parse_protect_home, 0, offsetof($1, exec_context) +$1.MountFlags, config_parse_exec_mount_flags, 0, offsetof($1, exec_context) +$1.Personality, config_parse_personality, 0, offsetof($1, exec_context.personality) +$1.RuntimeDirectoryMode, config_parse_mode, 0, offsetof($1, exec_context.runtime_directory_mode) +$1.RuntimeDirectory, config_parse_runtime_directory, 0, offsetof($1, exec_context.runtime_directory) +m4_ifdef(`HAVE_PAM', +`$1.PAMName, config_parse_unit_string_printf, 0, offsetof($1, exec_context.pam_name)', +`$1.PAMName, config_parse_warn_compat, DISABLED_CONFIGURATION, 0') +$1.IgnoreSIGPIPE, config_parse_bool, 0, offsetof($1, exec_context.ignore_sigpipe) +$1.UtmpIdentifier, config_parse_unit_string_printf, 0, offsetof($1, exec_context.utmp_id) +$1.UtmpMode, config_parse_exec_utmp_mode, 0, offsetof($1, exec_context.utmp_mode) +m4_ifdef(`HAVE_SELINUX', +`$1.SELinuxContext, config_parse_exec_selinux_context, 0, offsetof($1, exec_context)', +`$1.SELinuxContext, config_parse_warn_compat, DISABLED_CONFIGURATION, 0') +m4_ifdef(`HAVE_APPARMOR', +`$1.AppArmorProfile, config_parse_exec_apparmor_profile, 0, offsetof($1, exec_context)', +`$1.AppArmorProfile, config_parse_warn_compat, DISABLED_CONFIGURATION, 0') +m4_ifdef(`HAVE_SMACK', +`$1.SmackProcessLabel, config_parse_exec_smack_process_label, 0, offsetof($1, exec_context)', +`$1.SmackProcessLabel, config_parse_warn_compat, DISABLED_CONFIGURATION, 0')' +)m4_dnl +m4_define(`KILL_CONTEXT_CONFIG_ITEMS', +`$1.SendSIGKILL, config_parse_bool, 0, offsetof($1, kill_context.send_sigkill) +$1.SendSIGHUP, config_parse_bool, 0, offsetof($1, kill_context.send_sighup) +$1.KillMode, config_parse_kill_mode, 0, offsetof($1, kill_context.kill_mode) +$1.KillSignal, config_parse_signal, 0, offsetof($1, kill_context.kill_signal)' +)m4_dnl +m4_define(`CGROUP_CONTEXT_CONFIG_ITEMS', +`$1.Slice, config_parse_unit_slice, 0, 0 +$1.CPUAccounting, config_parse_bool, 0, offsetof($1, cgroup_context.cpu_accounting) +$1.CPUShares, config_parse_cpu_shares, 0, offsetof($1, cgroup_context.cpu_shares) +$1.StartupCPUShares, config_parse_cpu_shares, 0, offsetof($1, cgroup_context.startup_cpu_shares) +$1.CPUQuota, config_parse_cpu_quota, 0, offsetof($1, cgroup_context) +$1.MemoryAccounting, config_parse_bool, 0, offsetof($1, cgroup_context.memory_accounting) +$1.MemoryLow, config_parse_memory_limit, 0, offsetof($1, cgroup_context) +$1.MemoryHigh, config_parse_memory_limit, 0, offsetof($1, cgroup_context) +$1.MemoryMax, config_parse_memory_limit, 0, offsetof($1, cgroup_context) +$1.MemoryLimit, config_parse_memory_limit, 0, offsetof($1, cgroup_context) +$1.DeviceAllow, config_parse_device_allow, 0, offsetof($1, cgroup_context) +$1.DevicePolicy, config_parse_device_policy, 0, offsetof($1, cgroup_context.device_policy) +$1.IOAccounting, config_parse_bool, 0, offsetof($1, cgroup_context.io_accounting) +$1.IOWeight, config_parse_io_weight, 0, offsetof($1, cgroup_context.io_weight) +$1.StartupIOWeight, config_parse_io_weight, 0, offsetof($1, cgroup_context.startup_io_weight) +$1.IODeviceWeight, config_parse_io_device_weight, 0, offsetof($1, cgroup_context) +$1.IOReadBandwidthMax, config_parse_io_limit, 0, offsetof($1, cgroup_context) +$1.IOWriteBandwidthMax, config_parse_io_limit, 0, offsetof($1, cgroup_context) +$1.IOReadIOPSMax, config_parse_io_limit, 0, offsetof($1, cgroup_context) +$1.IOWriteIOPSMax, config_parse_io_limit, 0, offsetof($1, cgroup_context) +$1.BlockIOAccounting, config_parse_bool, 0, offsetof($1, cgroup_context.blockio_accounting) +$1.BlockIOWeight, config_parse_blockio_weight, 0, offsetof($1, cgroup_context.blockio_weight) +$1.StartupBlockIOWeight, config_parse_blockio_weight, 0, offsetof($1, cgroup_context.startup_blockio_weight) +$1.BlockIODeviceWeight, config_parse_blockio_device_weight, 0, offsetof($1, cgroup_context) +$1.BlockIOReadBandwidth, config_parse_blockio_bandwidth, 0, offsetof($1, cgroup_context) +$1.BlockIOWriteBandwidth, config_parse_blockio_bandwidth, 0, offsetof($1, cgroup_context) +$1.TasksAccounting, config_parse_bool, 0, offsetof($1, cgroup_context.tasks_accounting) +$1.TasksMax, config_parse_tasks_max, 0, offsetof($1, cgroup_context.tasks_max) +$1.Delegate, config_parse_bool, 0, offsetof($1, cgroup_context.delegate) +$1.NetClass, config_parse_warn_compat, DISABLED_LEGACY, 0' +)m4_dnl +Unit.Description, config_parse_unit_string_printf, 0, offsetof(Unit, description) +Unit.Documentation, config_parse_documentation, 0, offsetof(Unit, documentation) +Unit.SourcePath, config_parse_path, 0, offsetof(Unit, source_path) +Unit.Requires, config_parse_unit_deps, UNIT_REQUIRES, 0 +Unit.Requisite, config_parse_unit_deps, UNIT_REQUISITE, 0 +Unit.Wants, config_parse_unit_deps, UNIT_WANTS, 0 +Unit.BindsTo, config_parse_unit_deps, UNIT_BINDS_TO, 0 +Unit.BindTo, config_parse_unit_deps, UNIT_BINDS_TO, 0 +Unit.Conflicts, config_parse_unit_deps, UNIT_CONFLICTS, 0 +Unit.Before, config_parse_unit_deps, UNIT_BEFORE, 0 +Unit.After, config_parse_unit_deps, UNIT_AFTER, 0 +Unit.OnFailure, config_parse_unit_deps, UNIT_ON_FAILURE, 0 +Unit.PropagatesReloadTo, config_parse_unit_deps, UNIT_PROPAGATES_RELOAD_TO, 0 +Unit.PropagateReloadTo, config_parse_unit_deps, UNIT_PROPAGATES_RELOAD_TO, 0 +Unit.ReloadPropagatedFrom, config_parse_unit_deps, UNIT_RELOAD_PROPAGATED_FROM, 0 +Unit.PropagateReloadFrom, config_parse_unit_deps, UNIT_RELOAD_PROPAGATED_FROM, 0 +Unit.PartOf, config_parse_unit_deps, UNIT_PART_OF, 0 +Unit.JoinsNamespaceOf, config_parse_unit_deps, UNIT_JOINS_NAMESPACE_OF, 0 +Unit.RequiresOverridable, config_parse_obsolete_unit_deps, UNIT_REQUIRES, 0 +Unit.RequisiteOverridable, config_parse_obsolete_unit_deps, UNIT_REQUISITE, 0 +Unit.RequiresMountsFor, config_parse_unit_requires_mounts_for, 0, 0 +Unit.StopWhenUnneeded, config_parse_bool, 0, offsetof(Unit, stop_when_unneeded) +Unit.RefuseManualStart, config_parse_bool, 0, offsetof(Unit, refuse_manual_start) +Unit.RefuseManualStop, config_parse_bool, 0, offsetof(Unit, refuse_manual_stop) +Unit.AllowIsolate, config_parse_bool, 0, offsetof(Unit, allow_isolate) +Unit.DefaultDependencies, config_parse_bool, 0, offsetof(Unit, default_dependencies) +Unit.OnFailureJobMode, config_parse_job_mode, 0, offsetof(Unit, on_failure_job_mode) +Unit.OnFailureIsolate, config_parse_job_mode_isolate, 0, offsetof(Unit, on_failure_job_mode) +Unit.IgnoreOnIsolate, config_parse_bool, 0, offsetof(Unit, ignore_on_isolate) +Unit.IgnoreOnSnapshot, config_parse_warn_compat, DISABLED_LEGACY, 0 +Unit.JobTimeoutSec, config_parse_sec_fix_0, 0, offsetof(Unit, job_timeout) +Unit.JobTimeoutAction, config_parse_failure_action, 0, offsetof(Unit, job_timeout_action) +Unit.JobTimeoutRebootArgument, config_parse_string, 0, offsetof(Unit, job_timeout_reboot_arg) +Unit.StartLimitIntervalSec, config_parse_sec, 0, offsetof(Unit, start_limit.interval) +m4_dnl The following is a legacy alias name for compatibility +Unit.StartLimitInterval, config_parse_sec, 0, offsetof(Unit, start_limit.interval) +Unit.StartLimitBurst, config_parse_unsigned, 0, offsetof(Unit, start_limit.burst) +Unit.StartLimitAction, config_parse_failure_action, 0, offsetof(Unit, start_limit_action) +Unit.RebootArgument, config_parse_string, 0, offsetof(Unit, reboot_arg) +Unit.ConditionPathExists, config_parse_unit_condition_path, CONDITION_PATH_EXISTS, offsetof(Unit, conditions) +Unit.ConditionPathExistsGlob, config_parse_unit_condition_path, CONDITION_PATH_EXISTS_GLOB, offsetof(Unit, conditions) +Unit.ConditionPathIsDirectory, config_parse_unit_condition_path, CONDITION_PATH_IS_DIRECTORY, offsetof(Unit, conditions) +Unit.ConditionPathIsSymbolicLink,config_parse_unit_condition_path, CONDITION_PATH_IS_SYMBOLIC_LINK,offsetof(Unit, conditions) +Unit.ConditionPathIsMountPoint, config_parse_unit_condition_path, CONDITION_PATH_IS_MOUNT_POINT, offsetof(Unit, conditions) +Unit.ConditionPathIsReadWrite, config_parse_unit_condition_path, CONDITION_PATH_IS_READ_WRITE, offsetof(Unit, conditions) +Unit.ConditionDirectoryNotEmpty, config_parse_unit_condition_path, CONDITION_DIRECTORY_NOT_EMPTY, offsetof(Unit, conditions) +Unit.ConditionFileNotEmpty, config_parse_unit_condition_path, CONDITION_FILE_NOT_EMPTY, offsetof(Unit, conditions) +Unit.ConditionFileIsExecutable, config_parse_unit_condition_path, CONDITION_FILE_IS_EXECUTABLE, offsetof(Unit, conditions) +Unit.ConditionNeedsUpdate, config_parse_unit_condition_path, CONDITION_NEEDS_UPDATE, offsetof(Unit, conditions) +Unit.ConditionFirstBoot, config_parse_unit_condition_string, CONDITION_FIRST_BOOT, offsetof(Unit, conditions) +Unit.ConditionKernelCommandLine, config_parse_unit_condition_string, CONDITION_KERNEL_COMMAND_LINE, offsetof(Unit, conditions) +Unit.ConditionArchitecture, config_parse_unit_condition_string, CONDITION_ARCHITECTURE, offsetof(Unit, conditions) +Unit.ConditionVirtualization, config_parse_unit_condition_string, CONDITION_VIRTUALIZATION, offsetof(Unit, conditions) +Unit.ConditionSecurity, config_parse_unit_condition_string, CONDITION_SECURITY, offsetof(Unit, conditions) +Unit.ConditionCapability, config_parse_unit_condition_string, CONDITION_CAPABILITY, offsetof(Unit, conditions) +Unit.ConditionHost, config_parse_unit_condition_string, CONDITION_HOST, offsetof(Unit, conditions) +Unit.ConditionACPower, config_parse_unit_condition_string, CONDITION_AC_POWER, offsetof(Unit, conditions) +Unit.ConditionNull, config_parse_unit_condition_null, 0, offsetof(Unit, conditions) +Unit.AssertPathExists, config_parse_unit_condition_path, CONDITION_PATH_EXISTS, offsetof(Unit, asserts) +Unit.AssertPathExistsGlob, config_parse_unit_condition_path, CONDITION_PATH_EXISTS_GLOB, offsetof(Unit, asserts) +Unit.AssertPathIsDirectory, config_parse_unit_condition_path, CONDITION_PATH_IS_DIRECTORY, offsetof(Unit, asserts) +Unit.AssertPathIsSymbolicLink, config_parse_unit_condition_path, CONDITION_PATH_IS_SYMBOLIC_LINK,offsetof(Unit, asserts) +Unit.AssertPathIsMountPoint, config_parse_unit_condition_path, CONDITION_PATH_IS_MOUNT_POINT, offsetof(Unit, asserts) +Unit.AssertPathIsReadWrite, config_parse_unit_condition_path, CONDITION_PATH_IS_READ_WRITE, offsetof(Unit, asserts) +Unit.AssertDirectoryNotEmpty, config_parse_unit_condition_path, CONDITION_DIRECTORY_NOT_EMPTY, offsetof(Unit, asserts) +Unit.AssertFileNotEmpty, config_parse_unit_condition_path, CONDITION_FILE_NOT_EMPTY, offsetof(Unit, asserts) +Unit.AssertFileIsExecutable, config_parse_unit_condition_path, CONDITION_FILE_IS_EXECUTABLE, offsetof(Unit, asserts) +Unit.AssertNeedsUpdate, config_parse_unit_condition_path, CONDITION_NEEDS_UPDATE, offsetof(Unit, asserts) +Unit.AssertFirstBoot, config_parse_unit_condition_string, CONDITION_FIRST_BOOT, offsetof(Unit, asserts) +Unit.AssertKernelCommandLine, config_parse_unit_condition_string, CONDITION_KERNEL_COMMAND_LINE, offsetof(Unit, asserts) +Unit.AssertArchitecture, config_parse_unit_condition_string, CONDITION_ARCHITECTURE, offsetof(Unit, asserts) +Unit.AssertVirtualization, config_parse_unit_condition_string, CONDITION_VIRTUALIZATION, offsetof(Unit, asserts) +Unit.AssertSecurity, config_parse_unit_condition_string, CONDITION_SECURITY, offsetof(Unit, asserts) +Unit.AssertCapability, config_parse_unit_condition_string, CONDITION_CAPABILITY, offsetof(Unit, asserts) +Unit.AssertHost, config_parse_unit_condition_string, CONDITION_HOST, offsetof(Unit, asserts) +Unit.AssertACPower, config_parse_unit_condition_string, CONDITION_AC_POWER, offsetof(Unit, asserts) +Unit.AssertNull, config_parse_unit_condition_null, 0, offsetof(Unit, asserts) +m4_dnl +Service.PIDFile, config_parse_unit_path_printf, 0, offsetof(Service, pid_file) +Service.ExecStartPre, config_parse_exec, SERVICE_EXEC_START_PRE, offsetof(Service, exec_command) +Service.ExecStart, config_parse_exec, SERVICE_EXEC_START, offsetof(Service, exec_command) +Service.ExecStartPost, config_parse_exec, SERVICE_EXEC_START_POST, offsetof(Service, exec_command) +Service.ExecReload, config_parse_exec, SERVICE_EXEC_RELOAD, offsetof(Service, exec_command) +Service.ExecStop, config_parse_exec, SERVICE_EXEC_STOP, offsetof(Service, exec_command) +Service.ExecStopPost, config_parse_exec, SERVICE_EXEC_STOP_POST, offsetof(Service, exec_command) +Service.RestartSec, config_parse_sec, 0, offsetof(Service, restart_usec) +Service.TimeoutSec, config_parse_service_timeout, 0, 0 +Service.TimeoutStartSec, config_parse_service_timeout, 0, 0 +Service.TimeoutStopSec, config_parse_service_timeout, 0, 0 +Service.RuntimeMaxSec, config_parse_sec, 0, offsetof(Service, runtime_max_usec) +Service.WatchdogSec, config_parse_sec, 0, offsetof(Service, watchdog_usec) +m4_dnl The following three only exist for compatibility, they moved into Unit, see above +Service.StartLimitInterval, config_parse_sec, 0, offsetof(Unit, start_limit.interval) +Service.StartLimitBurst, config_parse_unsigned, 0, offsetof(Unit, start_limit.burst) +Service.StartLimitAction, config_parse_failure_action, 0, offsetof(Unit, start_limit_action) +Service.RebootArgument, config_parse_string, 0, offsetof(Unit, reboot_arg) +Service.FailureAction, config_parse_failure_action, 0, offsetof(Service, failure_action) +Service.Type, config_parse_service_type, 0, offsetof(Service, type) +Service.Restart, config_parse_service_restart, 0, offsetof(Service, restart) +Service.PermissionsStartOnly, config_parse_bool, 0, offsetof(Service, permissions_start_only) +Service.RootDirectoryStartOnly, config_parse_bool, 0, offsetof(Service, root_directory_start_only) +Service.RemainAfterExit, config_parse_bool, 0, offsetof(Service, remain_after_exit) +Service.GuessMainPID, config_parse_bool, 0, offsetof(Service, guess_main_pid) +Service.RestartPreventExitStatus, config_parse_set_status, 0, offsetof(Service, restart_prevent_status) +Service.RestartForceExitStatus, config_parse_set_status, 0, offsetof(Service, restart_force_status) +Service.SuccessExitStatus, config_parse_set_status, 0, offsetof(Service, success_status) +Service.SysVStartPriority, config_parse_warn_compat, DISABLED_LEGACY, 0 +Service.NonBlocking, config_parse_bool, 0, offsetof(Service, exec_context.non_blocking) +Service.BusName, config_parse_bus_name, 0, offsetof(Service, bus_name) +Service.FileDescriptorStoreMax, config_parse_unsigned, 0, offsetof(Service, n_fd_store_max) +Service.NotifyAccess, config_parse_notify_access, 0, offsetof(Service, notify_access) +Service.Sockets, config_parse_service_sockets, 0, 0 +Service.BusPolicy, config_parse_warn_compat, DISABLED_LEGACY, 0 +Service.USBFunctionDescriptors, config_parse_path, 0, offsetof(Service, usb_function_descriptors) +Service.USBFunctionStrings, config_parse_path, 0, offsetof(Service, usb_function_strings) +EXEC_CONTEXT_CONFIG_ITEMS(Service)m4_dnl +CGROUP_CONTEXT_CONFIG_ITEMS(Service)m4_dnl +KILL_CONTEXT_CONFIG_ITEMS(Service)m4_dnl +m4_dnl +Socket.ListenStream, config_parse_socket_listen, SOCKET_SOCKET, 0 +Socket.ListenDatagram, config_parse_socket_listen, SOCKET_SOCKET, 0 +Socket.ListenSequentialPacket, config_parse_socket_listen, SOCKET_SOCKET, 0 +Socket.ListenFIFO, config_parse_socket_listen, SOCKET_FIFO, 0 +Socket.ListenNetlink, config_parse_socket_listen, SOCKET_SOCKET, 0 +Socket.ListenSpecial, config_parse_socket_listen, SOCKET_SPECIAL, 0 +Socket.ListenMessageQueue, config_parse_socket_listen, SOCKET_MQUEUE, 0 +Socket.ListenUSBFunction, config_parse_socket_listen, SOCKET_USB_FUNCTION, 0 +Socket.SocketProtocol, config_parse_socket_protocol, 0, 0 +Socket.BindIPv6Only, config_parse_socket_bind, 0, 0, +Socket.Backlog, config_parse_unsigned, 0, offsetof(Socket, backlog) +Socket.BindToDevice, config_parse_socket_bindtodevice, 0, 0 +Socket.ExecStartPre, config_parse_exec, SOCKET_EXEC_START_PRE, offsetof(Socket, exec_command) +Socket.ExecStartPost, config_parse_exec, SOCKET_EXEC_START_POST, offsetof(Socket, exec_command) +Socket.ExecStopPre, config_parse_exec, SOCKET_EXEC_STOP_PRE, offsetof(Socket, exec_command) +Socket.ExecStopPost, config_parse_exec, SOCKET_EXEC_STOP_POST, offsetof(Socket, exec_command) +Socket.TimeoutSec, config_parse_sec, 0, offsetof(Socket, timeout_usec) +Socket.SocketUser, config_parse_unit_string_printf, 0, offsetof(Socket, user) +Socket.SocketGroup, config_parse_unit_string_printf, 0, offsetof(Socket, group) +Socket.SocketMode, config_parse_mode, 0, offsetof(Socket, socket_mode) +Socket.DirectoryMode, config_parse_mode, 0, offsetof(Socket, directory_mode) +Socket.Accept, config_parse_bool, 0, offsetof(Socket, accept) +Socket.Writable, config_parse_bool, 0, offsetof(Socket, writable) +Socket.MaxConnections, config_parse_unsigned, 0, offsetof(Socket, max_connections) +Socket.KeepAlive, config_parse_bool, 0, offsetof(Socket, keep_alive) +Socket.KeepAliveTimeSec, config_parse_sec, 0, offsetof(Socket, keep_alive_time) +Socket.KeepAliveIntervalSec, config_parse_sec, 0, offsetof(Socket, keep_alive_interval) +Socket.KeepAliveProbes, config_parse_unsigned, 0, offsetof(Socket, keep_alive_cnt) +Socket.DeferAcceptSec, config_parse_sec, 0, offsetof(Socket, defer_accept) +Socket.NoDelay, config_parse_bool, 0, offsetof(Socket, no_delay) +Socket.Priority, config_parse_int, 0, offsetof(Socket, priority) +Socket.ReceiveBuffer, config_parse_iec_size, 0, offsetof(Socket, receive_buffer) +Socket.SendBuffer, config_parse_iec_size, 0, offsetof(Socket, send_buffer) +Socket.IPTOS, config_parse_ip_tos, 0, offsetof(Socket, ip_tos) +Socket.IPTTL, config_parse_int, 0, offsetof(Socket, ip_ttl) +Socket.Mark, config_parse_int, 0, offsetof(Socket, mark) +Socket.PipeSize, config_parse_iec_size, 0, offsetof(Socket, pipe_size) +Socket.FreeBind, config_parse_bool, 0, offsetof(Socket, free_bind) +Socket.Transparent, config_parse_bool, 0, offsetof(Socket, transparent) +Socket.Broadcast, config_parse_bool, 0, offsetof(Socket, broadcast) +Socket.PassCredentials, config_parse_bool, 0, offsetof(Socket, pass_cred) +Socket.PassSecurity, config_parse_bool, 0, offsetof(Socket, pass_sec) +Socket.TCPCongestion, config_parse_string, 0, offsetof(Socket, tcp_congestion) +Socket.ReusePort, config_parse_bool, 0, offsetof(Socket, reuse_port) +Socket.MessageQueueMaxMessages, config_parse_long, 0, offsetof(Socket, mq_maxmsg) +Socket.MessageQueueMessageSize, config_parse_long, 0, offsetof(Socket, mq_msgsize) +Socket.RemoveOnStop, config_parse_bool, 0, offsetof(Socket, remove_on_stop) +Socket.Symlinks, config_parse_unit_path_strv_printf, 0, offsetof(Socket, symlinks) +Socket.FileDescriptorName, config_parse_fdname, 0, 0 +Socket.Service, config_parse_socket_service, 0, 0 +Socket.TriggerLimitIntervalSec, config_parse_sec, 0, offsetof(Socket, trigger_limit.interval) +Socket.TriggerLimitBurst, config_parse_unsigned, 0, offsetof(Socket, trigger_limit.burst) +m4_ifdef(`HAVE_SMACK', +`Socket.SmackLabel, config_parse_string, 0, offsetof(Socket, smack) +Socket.SmackLabelIPIn, config_parse_string, 0, offsetof(Socket, smack_ip_in) +Socket.SmackLabelIPOut, config_parse_string, 0, offsetof(Socket, smack_ip_out)', +`Socket.SmackLabel, config_parse_warn_compat, DISABLED_CONFIGURATION, 0 +Socket.SmackLabelIPIn, config_parse_warn_compat, DISABLED_CONFIGURATION, 0 +Socket.SmackLabelIPOut, config_parse_warn_compat, DISABLED_CONFIGURATION, 0') +m4_ifdef(`HAVE_SELINUX', +`Socket.SELinuxContextFromNet, config_parse_bool, 0, offsetof(Socket, selinux_context_from_net)', +`Socket.SELinuxContextFromNet, config_parse_warn_compat, DISABLED_CONFIGURATION, 0') +EXEC_CONTEXT_CONFIG_ITEMS(Socket)m4_dnl +CGROUP_CONTEXT_CONFIG_ITEMS(Socket)m4_dnl +KILL_CONTEXT_CONFIG_ITEMS(Socket)m4_dnl +m4_dnl +BusName.Name, config_parse_string, 0, offsetof(BusName, name) +BusName.Activating, config_parse_bool, 0, offsetof(BusName, activating) +BusName.Service, config_parse_busname_service, 0, 0 +BusName.AllowUser, config_parse_bus_policy, 0, 0 +BusName.AllowGroup, config_parse_bus_policy, 0, 0 +BusName.AllowWorld, config_parse_bus_policy_world, 0, offsetof(BusName, policy_world) +BusName.SELinuxContext, config_parse_exec_selinux_context, 0, 0 +BusName.AcceptFileDescriptors, config_parse_bool, 0, offsetof(BusName, accept_fd) +m4_dnl +Mount.What, config_parse_string, 0, offsetof(Mount, parameters_fragment.what) +Mount.Where, config_parse_path, 0, offsetof(Mount, where) +Mount.Options, config_parse_string, 0, offsetof(Mount, parameters_fragment.options) +Mount.Type, config_parse_string, 0, offsetof(Mount, parameters_fragment.fstype) +Mount.TimeoutSec, config_parse_sec, 0, offsetof(Mount, timeout_usec) +Mount.DirectoryMode, config_parse_mode, 0, offsetof(Mount, directory_mode) +Mount.SloppyOptions, config_parse_bool, 0, offsetof(Mount, sloppy_options) +EXEC_CONTEXT_CONFIG_ITEMS(Mount)m4_dnl +CGROUP_CONTEXT_CONFIG_ITEMS(Mount)m4_dnl +KILL_CONTEXT_CONFIG_ITEMS(Mount)m4_dnl +m4_dnl +Automount.Where, config_parse_path, 0, offsetof(Automount, where) +Automount.DirectoryMode, config_parse_mode, 0, offsetof(Automount, directory_mode) +Automount.TimeoutIdleSec, config_parse_sec, 0, offsetof(Automount, timeout_idle_usec) +m4_dnl +Swap.What, config_parse_path, 0, offsetof(Swap, parameters_fragment.what) +Swap.Priority, config_parse_int, 0, offsetof(Swap, parameters_fragment.priority) +Swap.Options, config_parse_string, 0, offsetof(Swap, parameters_fragment.options) +Swap.TimeoutSec, config_parse_sec, 0, offsetof(Swap, timeout_usec) +EXEC_CONTEXT_CONFIG_ITEMS(Swap)m4_dnl +CGROUP_CONTEXT_CONFIG_ITEMS(Swap)m4_dnl +KILL_CONTEXT_CONFIG_ITEMS(Swap)m4_dnl +m4_dnl +Timer.OnCalendar, config_parse_timer, 0, 0 +Timer.OnActiveSec, config_parse_timer, 0, 0 +Timer.OnBootSec, config_parse_timer, 0, 0 +Timer.OnStartupSec, config_parse_timer, 0, 0 +Timer.OnUnitActiveSec, config_parse_timer, 0, 0 +Timer.OnUnitInactiveSec, config_parse_timer, 0, 0 +Timer.Persistent, config_parse_bool, 0, offsetof(Timer, persistent) +Timer.WakeSystem, config_parse_bool, 0, offsetof(Timer, wake_system) +Timer.RemainAfterElapse, config_parse_bool, 0, offsetof(Timer, remain_after_elapse) +Timer.AccuracySec, config_parse_sec, 0, offsetof(Timer, accuracy_usec) +Timer.RandomizedDelaySec, config_parse_sec, 0, offsetof(Timer, random_usec) +Timer.Unit, config_parse_trigger_unit, 0, 0 +m4_dnl +Path.PathExists, config_parse_path_spec, 0, 0 +Path.PathExistsGlob, config_parse_path_spec, 0, 0 +Path.PathChanged, config_parse_path_spec, 0, 0 +Path.PathModified, config_parse_path_spec, 0, 0 +Path.DirectoryNotEmpty, config_parse_path_spec, 0, 0 +Path.Unit, config_parse_trigger_unit, 0, 0 +Path.MakeDirectory, config_parse_bool, 0, offsetof(Path, make_directory) +Path.DirectoryMode, config_parse_mode, 0, offsetof(Path, directory_mode) +m4_dnl +CGROUP_CONTEXT_CONFIG_ITEMS(Slice)m4_dnl +m4_dnl +CGROUP_CONTEXT_CONFIG_ITEMS(Scope)m4_dnl +KILL_CONTEXT_CONFIG_ITEMS(Scope)m4_dnl +Scope.TimeoutStopSec, config_parse_sec, 0, offsetof(Scope, timeout_stop_usec) +m4_dnl The [Install] section is ignored here. +Install.Alias, NULL, 0, 0 +Install.WantedBy, NULL, 0, 0 +Install.RequiredBy, NULL, 0, 0 +Install.Also, NULL, 0, 0 +Install.DefaultInstance, NULL, 0, 0 diff --git a/src/grp-system/libcore/src/load-fragment.c b/src/grp-system/libcore/src/load-fragment.c new file mode 100644 index 0000000000..53f3746f4d --- /dev/null +++ b/src/grp-system/libcore/src/load-fragment.c @@ -0,0 +1,4132 @@ +/*** + This file is part of systemd. + + Copyright 2010 Lennart Poettering + Copyright 2012 Holger Hans Peter Freyther + + systemd is free software; you can redistribute it and/or modify it + under the terms of the GNU Lesser General Public License as published by + the Free Software Foundation; either version 2.1 of the License, or + (at your option) any later version. + + systemd is distributed in the hope that it will be useful, but + WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + Lesser General Public License for more details. + + You should have received a copy of the GNU Lesser General Public License + along with systemd; If not, see . +***/ + +#include +#include + +#include +#include +#ifdef HAVE_SECCOMP +#include +#endif +#include +#include +#include +#include + +#include "core/cgroup.h" +#include "core/load-fragment.h" +#include "sd-bus/bus-error.h" +#include "sd-bus/bus-internal.h" +#include "sd-bus/bus-util.h" +#include "systemd-basic/af-list.h" +#include "systemd-basic/alloc-util.h" +#include "systemd-basic/cap-list.h" +#include "systemd-basic/capability-util.h" +#include "systemd-basic/cpu-set-util.h" +#include "systemd-basic/env-util.h" +#include "systemd-basic/errno-list.h" +#include "systemd-basic/escape.h" +#include "systemd-basic/fd-util.h" +#include "systemd-basic/fs-util.h" +#include "systemd-basic/ioprio.h" +#include "systemd-basic/log.h" +#include "systemd-basic/missing.h" +#include "systemd-basic/parse-util.h" +#include "systemd-basic/path-util.h" +#include "systemd-basic/process-util.h" +#include "systemd-basic/rlimit-util.h" +#include "systemd-shared/conf-parser.h" +#ifdef HAVE_SECCOMP +#include "systemd-shared/seccomp-util.h" +#endif +#include "core/unit.h" +#include "systemd-basic/securebits.h" +#include "systemd-basic/signal-util.h" +#include "systemd-basic/stat-util.h" +#include "systemd-basic/string-util.h" +#include "systemd-basic/strv.h" +#include "systemd-basic/unit-name.h" +#include "systemd-basic/utf8.h" +#include "systemd-basic/web-util.h" + +#include "unit-printf.h" + +int config_parse_warn_compat( + const char *unit, + const char *filename, + unsigned line, + const char *section, + unsigned section_line, + const char *lvalue, + int ltype, + const char *rvalue, + void *data, + void *userdata) { + Disabled reason = ltype; + + switch(reason) { + case DISABLED_CONFIGURATION: + log_syntax(unit, LOG_DEBUG, filename, line, 0, + "Support for option %s= has been disabled at compile time and it is ignored", lvalue); + break; + case DISABLED_LEGACY: + log_syntax(unit, LOG_INFO, filename, line, 0, + "Support for option %s= has been removed and it is ignored", lvalue); + break; + case DISABLED_EXPERIMENTAL: + log_syntax(unit, LOG_INFO, filename, line, 0, + "Support for option %s= has not yet been enabled and it is ignored", lvalue); + break; + }; + + return 0; +} + +int config_parse_unit_deps( + const char *unit, + const char *filename, + unsigned line, + const char *section, + unsigned section_line, + const char *lvalue, + int ltype, + const char *rvalue, + void *data, + void *userdata) { + + UnitDependency d = ltype; + Unit *u = userdata; + const char *p; + + assert(filename); + assert(lvalue); + assert(rvalue); + + p = rvalue; + for (;;) { + _cleanup_free_ char *word = NULL, *k = NULL; + int r; + + r = extract_first_word(&p, &word, NULL, EXTRACT_RETAIN_ESCAPE); + if (r == 0) + break; + if (r == -ENOMEM) + return log_oom(); + if (r < 0) { + log_syntax(unit, LOG_ERR, filename, line, r, "Invalid syntax, ignoring: %s", rvalue); + break; + } + + r = unit_name_printf(u, word, &k); + if (r < 0) { + log_syntax(unit, LOG_ERR, filename, line, r, "Failed to resolve specifiers, ignoring: %m"); + continue; + } + + r = unit_add_dependency_by_name(u, d, k, NULL, true); + if (r < 0) + log_syntax(unit, LOG_ERR, filename, line, r, "Failed to add dependency on %s, ignoring: %m", k); + } + + return 0; +} + +int config_parse_obsolete_unit_deps( + const char *unit, + const char *filename, + unsigned line, + const char *section, + unsigned section_line, + const char *lvalue, + int ltype, + const char *rvalue, + void *data, + void *userdata) { + + log_syntax(unit, LOG_WARNING, filename, line, 0, + "Unit dependency type %s= is obsolete, replacing by %s=, please update your unit file", lvalue, unit_dependency_to_string(ltype)); + + return config_parse_unit_deps(unit, filename, line, section, section_line, lvalue, ltype, rvalue, data, userdata); +} + +int config_parse_unit_string_printf( + const char *unit, + const char *filename, + unsigned line, + const char *section, + unsigned section_line, + const char *lvalue, + int ltype, + const char *rvalue, + void *data, + void *userdata) { + + _cleanup_free_ char *k = NULL; + Unit *u = userdata; + int r; + + assert(filename); + assert(lvalue); + assert(rvalue); + assert(u); + + r = unit_full_printf(u, rvalue, &k); + if (r < 0) { + log_syntax(unit, LOG_ERR, filename, line, r, "Failed to resolve unit specifiers on %s, ignoring: %m", rvalue); + return 0; + } + + return config_parse_string(unit, filename, line, section, section_line, lvalue, ltype, k, data, userdata); +} + +int config_parse_unit_strv_printf( + const char *unit, + const char *filename, + unsigned line, + const char *section, + unsigned section_line, + const char *lvalue, + int ltype, + const char *rvalue, + void *data, + void *userdata) { + + Unit *u = userdata; + _cleanup_free_ char *k = NULL; + int r; + + assert(filename); + assert(lvalue); + assert(rvalue); + assert(u); + + r = unit_full_printf(u, rvalue, &k); + if (r < 0) { + log_syntax(unit, LOG_ERR, filename, line, r, "Failed to resolve unit specifiers on %s, ignoring: %m", rvalue); + return 0; + } + + return config_parse_strv(unit, filename, line, section, section_line, lvalue, ltype, k, data, userdata); +} + +int config_parse_unit_path_printf( + const char *unit, + const char *filename, + unsigned line, + const char *section, + unsigned section_line, + const char *lvalue, + int ltype, + const char *rvalue, + void *data, + void *userdata) { + + _cleanup_free_ char *k = NULL; + Unit *u = userdata; + int r; + + assert(filename); + assert(lvalue); + assert(rvalue); + assert(u); + + r = unit_full_printf(u, rvalue, &k); + if (r < 0) { + log_syntax(unit, LOG_ERR, filename, line, r, "Failed to resolve unit specifiers on %s, ignoring: %m", rvalue); + return 0; + } + + return config_parse_path(unit, filename, line, section, section_line, lvalue, ltype, k, data, userdata); +} + +int config_parse_unit_path_strv_printf( + const char *unit, + const char *filename, + unsigned line, + const char *section, + unsigned section_line, + const char *lvalue, + int ltype, + const char *rvalue, + void *data, + void *userdata) { + + char ***x = data; + const char *word, *state; + Unit *u = userdata; + size_t l; + int r; + + assert(filename); + assert(lvalue); + assert(rvalue); + assert(u); + + FOREACH_WORD_QUOTED(word, l, rvalue, state) { + _cleanup_free_ char *k = NULL; + char t[l+1]; + + memcpy(t, word, l); + t[l] = 0; + + r = unit_full_printf(u, t, &k); + if (r < 0) { + log_syntax(unit, LOG_ERR, filename, line, r, "Failed to resolve unit specifiers on %s, ignoring: %m", t); + return 0; + } + + if (!utf8_is_valid(k)) { + log_syntax_invalid_utf8(unit, LOG_ERR, filename, line, rvalue); + return 0; + } + + if (!path_is_absolute(k)) { + log_syntax(unit, LOG_ERR, filename, line, 0, "Symlink path %s is not absolute, ignoring: %m", k); + return 0; + } + + path_kill_slashes(k); + + r = strv_push(x, k); + if (r < 0) + return log_oom(); + + k = NULL; + } + if (!isempty(state)) + log_syntax(unit, LOG_ERR, filename, line, 0, "Invalid syntax, ignoring."); + + return 0; +} + +int config_parse_socket_listen(const char *unit, + const char *filename, + unsigned line, + const char *section, + unsigned section_line, + const char *lvalue, + int ltype, + const char *rvalue, + void *data, + void *userdata) { + + _cleanup_free_ SocketPort *p = NULL; + SocketPort *tail; + Socket *s; + int r; + + assert(filename); + assert(lvalue); + assert(rvalue); + assert(data); + + s = SOCKET(data); + + if (isempty(rvalue)) { + /* An empty assignment removes all ports */ + socket_free_ports(s); + return 0; + } + + p = new0(SocketPort, 1); + if (!p) + return log_oom(); + + if (ltype != SOCKET_SOCKET) { + + p->type = ltype; + r = unit_full_printf(UNIT(s), rvalue, &p->path); + if (r < 0) { + log_syntax(unit, LOG_ERR, filename, line, r, "Failed to resolve unit specifiers on %s, ignoring: %m", rvalue); + return 0; + } + + path_kill_slashes(p->path); + + } else if (streq(lvalue, "ListenNetlink")) { + _cleanup_free_ char *k = NULL; + + p->type = SOCKET_SOCKET; + r = unit_full_printf(UNIT(s), rvalue, &k); + if (r < 0) { + log_syntax(unit, LOG_ERR, filename, line, r, "Failed to resolve unit specifiers on %s, ignoring: %m", rvalue); + return 0; + } + + r = socket_address_parse_netlink(&p->address, k); + if (r < 0) { + log_syntax(unit, LOG_ERR, filename, line, r, "Failed to parse address value, ignoring: %s", rvalue); + return 0; + } + + } else { + _cleanup_free_ char *k = NULL; + + p->type = SOCKET_SOCKET; + r = unit_full_printf(UNIT(s), rvalue, &k); + if (r < 0) { + log_syntax(unit, LOG_ERR, filename, line, r,"Failed to resolve unit specifiers on %s, ignoring: %m", rvalue); + return 0; + } + + r = socket_address_parse_and_warn(&p->address, k); + if (r < 0) { + log_syntax(unit, LOG_ERR, filename, line, r, "Failed to parse address value, ignoring: %s", rvalue); + return 0; + } + + if (streq(lvalue, "ListenStream")) + p->address.type = SOCK_STREAM; + else if (streq(lvalue, "ListenDatagram")) + p->address.type = SOCK_DGRAM; + else { + assert(streq(lvalue, "ListenSequentialPacket")); + p->address.type = SOCK_SEQPACKET; + } + + if (socket_address_family(&p->address) != AF_LOCAL && p->address.type == SOCK_SEQPACKET) { + log_syntax(unit, LOG_ERR, filename, line, 0, "Address family not supported, ignoring: %s", rvalue); + return 0; + } + } + + p->fd = -1; + p->auxiliary_fds = NULL; + p->n_auxiliary_fds = 0; + p->socket = s; + + if (s->ports) { + LIST_FIND_TAIL(port, s->ports, tail); + LIST_INSERT_AFTER(port, s->ports, tail, p); + } else + LIST_PREPEND(port, s->ports, p); + p = NULL; + + return 0; +} + +int config_parse_socket_protocol(const char *unit, + const char *filename, + unsigned line, + const char *section, + unsigned section_line, + const char *lvalue, + int ltype, + const char *rvalue, + void *data, + void *userdata) { + Socket *s; + + assert(filename); + assert(lvalue); + assert(rvalue); + assert(data); + + s = SOCKET(data); + + if (streq(rvalue, "udplite")) + s->socket_protocol = IPPROTO_UDPLITE; + else if (streq(rvalue, "sctp")) + s->socket_protocol = IPPROTO_SCTP; + else { + log_syntax(unit, LOG_ERR, filename, line, 0, "Socket protocol not supported, ignoring: %s", rvalue); + return 0; + } + + return 0; +} + +int config_parse_socket_bind(const char *unit, + const char *filename, + unsigned line, + const char *section, + unsigned section_line, + const char *lvalue, + int ltype, + const char *rvalue, + void *data, + void *userdata) { + + Socket *s; + SocketAddressBindIPv6Only b; + + assert(filename); + assert(lvalue); + assert(rvalue); + assert(data); + + s = SOCKET(data); + + b = socket_address_bind_ipv6_only_from_string(rvalue); + if (b < 0) { + int r; + + r = parse_boolean(rvalue); + if (r < 0) { + log_syntax(unit, LOG_ERR, filename, line, r, "Failed to parse bind IPv6 only value, ignoring: %s", rvalue); + return 0; + } + + s->bind_ipv6_only = r ? SOCKET_ADDRESS_IPV6_ONLY : SOCKET_ADDRESS_BOTH; + } else + s->bind_ipv6_only = b; + + return 0; +} + +int config_parse_exec_nice(const char *unit, + const char *filename, + unsigned line, + const char *section, + unsigned section_line, + const char *lvalue, + int ltype, + const char *rvalue, + void *data, + void *userdata) { + + ExecContext *c = data; + int priority, r; + + assert(filename); + assert(lvalue); + assert(rvalue); + assert(data); + + r = safe_atoi(rvalue, &priority); + if (r < 0) { + log_syntax(unit, LOG_ERR, filename, line, r, "Failed to parse nice priority, ignoring: %s", rvalue); + return 0; + } + + if (priority < PRIO_MIN || priority >= PRIO_MAX) { + log_syntax(unit, LOG_ERR, filename, line, 0, "Nice priority out of range, ignoring: %s", rvalue); + return 0; + } + + c->nice = priority; + c->nice_set = true; + + return 0; +} + +int config_parse_exec_oom_score_adjust(const char* unit, + const char *filename, + unsigned line, + const char *section, + unsigned section_line, + const char *lvalue, + int ltype, + const char *rvalue, + void *data, + void *userdata) { + + ExecContext *c = data; + int oa, r; + + assert(filename); + assert(lvalue); + assert(rvalue); + assert(data); + + r = safe_atoi(rvalue, &oa); + if (r < 0) { + log_syntax(unit, LOG_ERR, filename, line, r, "Failed to parse the OOM score adjust value, ignoring: %s", rvalue); + return 0; + } + + if (oa < OOM_SCORE_ADJ_MIN || oa > OOM_SCORE_ADJ_MAX) { + log_syntax(unit, LOG_ERR, filename, line, 0, "OOM score adjust value out of range, ignoring: %s", rvalue); + return 0; + } + + c->oom_score_adjust = oa; + c->oom_score_adjust_set = true; + + return 0; +} + +int config_parse_exec( + const char *unit, + const char *filename, + unsigned line, + const char *section, + unsigned section_line, + const char *lvalue, + int ltype, + const char *rvalue, + void *data, + void *userdata) { + + ExecCommand **e = data; + const char *p; + bool semicolon; + int r; + + assert(filename); + assert(lvalue); + assert(rvalue); + assert(e); + + e += ltype; + rvalue += strspn(rvalue, WHITESPACE); + + if (isempty(rvalue)) { + /* An empty assignment resets the list */ + *e = exec_command_free_list(*e); + return 0; + } + + p = rvalue; + do { + _cleanup_free_ char *path = NULL, *firstword = NULL; + bool separate_argv0 = false, ignore = false, privileged = false; + _cleanup_free_ ExecCommand *nce = NULL; + _cleanup_strv_free_ char **n = NULL; + size_t nlen = 0, nbufsize = 0; + char *f; + int i; + + semicolon = false; + + r = extract_first_word_and_warn(&p, &firstword, WHITESPACE, EXTRACT_QUOTES|EXTRACT_CUNESCAPE, unit, filename, line, rvalue); + if (r <= 0) + return 0; + + f = firstword; + for (i = 0; i < 3; i++) { + /* We accept an absolute path as first argument. + * If it's prefixed with - and the path doesn't exist, + * we ignore it instead of erroring out; + * if it's prefixed with @, we allow overriding of argv[0]; + * and if it's prefixed with !, it will be run with full privileges */ + if (*f == '-' && !ignore) + ignore = true; + else if (*f == '@' && !separate_argv0) + separate_argv0 = true; + else if (*f == '+' && !privileged) + privileged = true; + else + break; + f++; + } + + if (isempty(f)) { + /* First word is either "-" or "@" with no command. */ + log_syntax(unit, LOG_ERR, filename, line, 0, "Empty path in command line, ignoring: \"%s\"", rvalue); + return 0; + } + if (!string_is_safe(f)) { + log_syntax(unit, LOG_ERR, filename, line, 0, "Executable path contains special characters, ignoring: %s", rvalue); + return 0; + } + if (!path_is_absolute(f)) { + log_syntax(unit, LOG_ERR, filename, line, 0, "Executable path is not absolute, ignoring: %s", rvalue); + return 0; + } + if (endswith(f, "/")) { + log_syntax(unit, LOG_ERR, filename, line, 0, "Executable path specifies a directory, ignoring: %s", rvalue); + return 0; + } + + if (f == firstword) { + path = firstword; + firstword = NULL; + } else { + path = strdup(f); + if (!path) + return log_oom(); + } + + if (!separate_argv0) { + if (!GREEDY_REALLOC(n, nbufsize, nlen + 2)) + return log_oom(); + f = strdup(path); + if (!f) + return log_oom(); + n[nlen++] = f; + n[nlen] = NULL; + } + + path_kill_slashes(path); + + while (!isempty(p)) { + _cleanup_free_ char *word = NULL; + + /* Check explicitly for an unquoted semicolon as + * command separator token. */ + if (p[0] == ';' && (!p[1] || strchr(WHITESPACE, p[1]))) { + p++; + p += strspn(p, WHITESPACE); + semicolon = true; + break; + } + + /* Check for \; explicitly, to not confuse it with \\; + * or "\;" or "\\;" etc. extract_first_word would + * return the same for all of those. */ + if (p[0] == '\\' && p[1] == ';' && (!p[2] || strchr(WHITESPACE, p[2]))) { + p += 2; + p += strspn(p, WHITESPACE); + if (!GREEDY_REALLOC(n, nbufsize, nlen + 2)) + return log_oom(); + f = strdup(";"); + if (!f) + return log_oom(); + n[nlen++] = f; + n[nlen] = NULL; + continue; + } + + r = extract_first_word_and_warn(&p, &word, WHITESPACE, EXTRACT_QUOTES|EXTRACT_CUNESCAPE, unit, filename, line, rvalue); + if (r == 0) + break; + else if (r < 0) + return 0; + + if (!GREEDY_REALLOC(n, nbufsize, nlen + 2)) + return log_oom(); + n[nlen++] = word; + n[nlen] = NULL; + word = NULL; + } + + if (!n || !n[0]) { + log_syntax(unit, LOG_ERR, filename, line, 0, "Empty executable name or zeroeth argument, ignoring: %s", rvalue); + return 0; + } + + nce = new0(ExecCommand, 1); + if (!nce) + return log_oom(); + + nce->argv = n; + nce->path = path; + nce->ignore = ignore; + nce->privileged = privileged; + + exec_command_append_list(e, nce); + + /* Do not _cleanup_free_ these. */ + n = NULL; + path = NULL; + nce = NULL; + + rvalue = p; + } while (semicolon); + + return 0; +} + +DEFINE_CONFIG_PARSE_ENUM(config_parse_service_type, service_type, ServiceType, "Failed to parse service type"); +DEFINE_CONFIG_PARSE_ENUM(config_parse_service_restart, service_restart, ServiceRestart, "Failed to parse service restart specifier"); + +int config_parse_socket_bindtodevice( + const char* unit, + const char *filename, + unsigned line, + const char *section, + unsigned section_line, + const char *lvalue, + int ltype, + const char *rvalue, + void *data, + void *userdata) { + + Socket *s = data; + char *n; + + assert(filename); + assert(lvalue); + assert(rvalue); + assert(data); + + if (rvalue[0] && !streq(rvalue, "*")) { + if (!ifname_valid(rvalue)) { + log_syntax(unit, LOG_ERR, filename, line, 0, "Interface name is invalid, ignoring: %s", rvalue); + return 0; + } + + n = strdup(rvalue); + if (!n) + return log_oom(); + } else + n = NULL; + + free(s->bind_to_device); + s->bind_to_device = n; + + return 0; +} + +DEFINE_CONFIG_PARSE_ENUM(config_parse_output, exec_output, ExecOutput, "Failed to parse output specifier"); +DEFINE_CONFIG_PARSE_ENUM(config_parse_input, exec_input, ExecInput, "Failed to parse input specifier"); + +int config_parse_exec_io_class(const char *unit, + const char *filename, + unsigned line, + const char *section, + unsigned section_line, + const char *lvalue, + int ltype, + const char *rvalue, + void *data, + void *userdata) { + + ExecContext *c = data; + int x; + + assert(filename); + assert(lvalue); + assert(rvalue); + assert(data); + + x = ioprio_class_from_string(rvalue); + if (x < 0) { + log_syntax(unit, LOG_ERR, filename, line, 0, "Failed to parse IO scheduling class, ignoring: %s", rvalue); + return 0; + } + + c->ioprio = IOPRIO_PRIO_VALUE(x, IOPRIO_PRIO_DATA(c->ioprio)); + c->ioprio_set = true; + + return 0; +} + +int config_parse_exec_io_priority(const char *unit, + const char *filename, + unsigned line, + const char *section, + unsigned section_line, + const char *lvalue, + int ltype, + const char *rvalue, + void *data, + void *userdata) { + + ExecContext *c = data; + int i, r; + + assert(filename); + assert(lvalue); + assert(rvalue); + assert(data); + + r = safe_atoi(rvalue, &i); + if (r < 0 || i < 0 || i >= IOPRIO_BE_NR) { + log_syntax(unit, LOG_ERR, filename, line, r, "Failed to parse IO priority, ignoring: %s", rvalue); + return 0; + } + + c->ioprio = IOPRIO_PRIO_VALUE(IOPRIO_PRIO_CLASS(c->ioprio), i); + c->ioprio_set = true; + + return 0; +} + +int config_parse_exec_cpu_sched_policy(const char *unit, + const char *filename, + unsigned line, + const char *section, + unsigned section_line, + const char *lvalue, + int ltype, + const char *rvalue, + void *data, + void *userdata) { + + + ExecContext *c = data; + int x; + + assert(filename); + assert(lvalue); + assert(rvalue); + assert(data); + + x = sched_policy_from_string(rvalue); + if (x < 0) { + log_syntax(unit, LOG_ERR, filename, line, 0, "Failed to parse CPU scheduling policy, ignoring: %s", rvalue); + return 0; + } + + c->cpu_sched_policy = x; + /* Moving to or from real-time policy? We need to adjust the priority */ + c->cpu_sched_priority = CLAMP(c->cpu_sched_priority, sched_get_priority_min(x), sched_get_priority_max(x)); + c->cpu_sched_set = true; + + return 0; +} + +int config_parse_exec_cpu_sched_prio(const char *unit, + const char *filename, + unsigned line, + const char *section, + unsigned section_line, + const char *lvalue, + int ltype, + const char *rvalue, + void *data, + void *userdata) { + + ExecContext *c = data; + int i, min, max, r; + + assert(filename); + assert(lvalue); + assert(rvalue); + assert(data); + + r = safe_atoi(rvalue, &i); + if (r < 0) { + log_syntax(unit, LOG_ERR, filename, line, r, "Failed to parse CPU scheduling policy, ignoring: %s", rvalue); + return 0; + } + + /* On Linux RR/FIFO range from 1 to 99 and OTHER/BATCH may only be 0 */ + min = sched_get_priority_min(c->cpu_sched_policy); + max = sched_get_priority_max(c->cpu_sched_policy); + + if (i < min || i > max) { + log_syntax(unit, LOG_ERR, filename, line, 0, "CPU scheduling priority is out of range, ignoring: %s", rvalue); + return 0; + } + + c->cpu_sched_priority = i; + c->cpu_sched_set = true; + + return 0; +} + +int config_parse_exec_cpu_affinity(const char *unit, + const char *filename, + unsigned line, + const char *section, + unsigned section_line, + const char *lvalue, + int ltype, + const char *rvalue, + void *data, + void *userdata) { + + ExecContext *c = data; + _cleanup_cpu_free_ cpu_set_t *cpuset = NULL; + int ncpus; + + assert(filename); + assert(lvalue); + assert(rvalue); + assert(data); + + ncpus = parse_cpu_set_and_warn(rvalue, &cpuset, unit, filename, line, lvalue); + if (ncpus < 0) + return ncpus; + + if (c->cpuset) + CPU_FREE(c->cpuset); + + if (ncpus == 0) + /* An empty assignment resets the CPU list */ + c->cpuset = NULL; + else { + c->cpuset = cpuset; + cpuset = NULL; + } + c->cpuset_ncpus = ncpus; + + return 0; +} + +int config_parse_exec_secure_bits(const char *unit, + const char *filename, + unsigned line, + const char *section, + unsigned section_line, + const char *lvalue, + int ltype, + const char *rvalue, + void *data, + void *userdata) { + + ExecContext *c = data; + size_t l; + const char *word, *state; + + assert(filename); + assert(lvalue); + assert(rvalue); + assert(data); + + if (isempty(rvalue)) { + /* An empty assignment resets the field */ + c->secure_bits = 0; + return 0; + } + + FOREACH_WORD_QUOTED(word, l, rvalue, state) { + if (first_word(word, "keep-caps")) + c->secure_bits |= 1<secure_bits |= 1<secure_bits |= 1<secure_bits |= 1<secure_bits |= 1<secure_bits |= 1< replace */ + *capability_set = sum; + else + /* previous data -> merge */ + *capability_set |= sum; + + return 0; +} + +int config_parse_limit( + const char *unit, + const char *filename, + unsigned line, + const char *section, + unsigned section_line, + const char *lvalue, + int ltype, + const char *rvalue, + void *data, + void *userdata) { + + struct rlimit **rl = data, d = {}; + int r; + + assert(filename); + assert(lvalue); + assert(rvalue); + assert(data); + + r = rlimit_parse(ltype, rvalue, &d); + if (r == -EILSEQ) { + log_syntax(unit, LOG_WARNING, filename, line, r, "Soft resource limit chosen higher than hard limit, ignoring: %s", rvalue); + return 0; + } + if (r < 0) { + log_syntax(unit, LOG_ERR, filename, line, r, "Failed to parse resource value, ignoring: %s", rvalue); + return 0; + } + + if (rl[ltype]) + *rl[ltype] = d; + else { + rl[ltype] = newdup(struct rlimit, &d, 1); + if (!rl[ltype]) + return log_oom(); + } + + return 0; +} + +#ifdef HAVE_SYSV_COMPAT +int config_parse_sysv_priority(const char *unit, + const char *filename, + unsigned line, + const char *section, + unsigned section_line, + const char *lvalue, + int ltype, + const char *rvalue, + void *data, + void *userdata) { + + int *priority = data; + int i, r; + + assert(filename); + assert(lvalue); + assert(rvalue); + assert(data); + + r = safe_atoi(rvalue, &i); + if (r < 0 || i < 0) { + log_syntax(unit, LOG_ERR, filename, line, r, "Failed to parse SysV start priority, ignoring: %s", rvalue); + return 0; + } + + *priority = (int) i; + return 0; +} +#endif + +DEFINE_CONFIG_PARSE_ENUM(config_parse_exec_utmp_mode, exec_utmp_mode, ExecUtmpMode, "Failed to parse utmp mode"); +DEFINE_CONFIG_PARSE_ENUM(config_parse_kill_mode, kill_mode, KillMode, "Failed to parse kill mode"); + +int config_parse_exec_mount_flags(const char *unit, + const char *filename, + unsigned line, + const char *section, + unsigned section_line, + const char *lvalue, + int ltype, + const char *rvalue, + void *data, + void *userdata) { + + + unsigned long flags = 0; + ExecContext *c = data; + + assert(filename); + assert(lvalue); + assert(rvalue); + assert(data); + + if (streq(rvalue, "shared")) + flags = MS_SHARED; + else if (streq(rvalue, "slave")) + flags = MS_SLAVE; + else if (streq(rvalue, "private")) + flags = MS_PRIVATE; + else { + log_syntax(unit, LOG_ERR, filename, line, 0, "Failed to parse mount flag %s, ignoring.", rvalue); + return 0; + } + + c->mount_flags = flags; + + return 0; +} + +int config_parse_exec_selinux_context( + const char *unit, + const char *filename, + unsigned line, + const char *section, + unsigned section_line, + const char *lvalue, + int ltype, + const char *rvalue, + void *data, + void *userdata) { + + ExecContext *c = data; + Unit *u = userdata; + bool ignore; + char *k; + int r; + + assert(filename); + assert(lvalue); + assert(rvalue); + assert(data); + + if (isempty(rvalue)) { + c->selinux_context = mfree(c->selinux_context); + c->selinux_context_ignore = false; + return 0; + } + + if (rvalue[0] == '-') { + ignore = true; + rvalue++; + } else + ignore = false; + + r = unit_name_printf(u, rvalue, &k); + if (r < 0) { + log_syntax(unit, LOG_ERR, filename, line, r, "Failed to resolve specifiers, ignoring: %m"); + return 0; + } + + free(c->selinux_context); + c->selinux_context = k; + c->selinux_context_ignore = ignore; + + return 0; +} + +int config_parse_exec_apparmor_profile( + const char *unit, + const char *filename, + unsigned line, + const char *section, + unsigned section_line, + const char *lvalue, + int ltype, + const char *rvalue, + void *data, + void *userdata) { + + ExecContext *c = data; + Unit *u = userdata; + bool ignore; + char *k; + int r; + + assert(filename); + assert(lvalue); + assert(rvalue); + assert(data); + + if (isempty(rvalue)) { + c->apparmor_profile = mfree(c->apparmor_profile); + c->apparmor_profile_ignore = false; + return 0; + } + + if (rvalue[0] == '-') { + ignore = true; + rvalue++; + } else + ignore = false; + + r = unit_name_printf(u, rvalue, &k); + if (r < 0) { + log_syntax(unit, LOG_ERR, filename, line, r, "Failed to resolve specifiers, ignoring: %m"); + return 0; + } + + free(c->apparmor_profile); + c->apparmor_profile = k; + c->apparmor_profile_ignore = ignore; + + return 0; +} + +int config_parse_exec_smack_process_label( + const char *unit, + const char *filename, + unsigned line, + const char *section, + unsigned section_line, + const char *lvalue, + int ltype, + const char *rvalue, + void *data, + void *userdata) { + + ExecContext *c = data; + Unit *u = userdata; + bool ignore; + char *k; + int r; + + assert(filename); + assert(lvalue); + assert(rvalue); + assert(data); + + if (isempty(rvalue)) { + c->smack_process_label = mfree(c->smack_process_label); + c->smack_process_label_ignore = false; + return 0; + } + + if (rvalue[0] == '-') { + ignore = true; + rvalue++; + } else + ignore = false; + + r = unit_name_printf(u, rvalue, &k); + if (r < 0) { + log_syntax(unit, LOG_ERR, filename, line, r, "Failed to resolve specifiers, ignoring: %m"); + return 0; + } + + free(c->smack_process_label); + c->smack_process_label = k; + c->smack_process_label_ignore = ignore; + + return 0; +} + +int config_parse_timer(const char *unit, + const char *filename, + unsigned line, + const char *section, + unsigned section_line, + const char *lvalue, + int ltype, + const char *rvalue, + void *data, + void *userdata) { + + Timer *t = data; + usec_t u = 0; + TimerValue *v; + TimerBase b; + CalendarSpec *c = NULL; + + assert(filename); + assert(lvalue); + assert(rvalue); + assert(data); + + if (isempty(rvalue)) { + /* Empty assignment resets list */ + timer_free_values(t); + return 0; + } + + b = timer_base_from_string(lvalue); + if (b < 0) { + log_syntax(unit, LOG_ERR, filename, line, 0, "Failed to parse timer base, ignoring: %s", lvalue); + return 0; + } + + if (b == TIMER_CALENDAR) { + if (calendar_spec_from_string(rvalue, &c) < 0) { + log_syntax(unit, LOG_ERR, filename, line, 0, "Failed to parse calendar specification, ignoring: %s", rvalue); + return 0; + } + } else { + if (parse_sec(rvalue, &u) < 0) { + log_syntax(unit, LOG_ERR, filename, line, 0, "Failed to parse timer value, ignoring: %s", rvalue); + return 0; + } + } + + v = new0(TimerValue, 1); + if (!v) { + calendar_spec_free(c); + return log_oom(); + } + + v->base = b; + v->value = u; + v->calendar_spec = c; + + LIST_PREPEND(value, t->values, v); + + return 0; +} + +int config_parse_trigger_unit( + const char *unit, + const char *filename, + unsigned line, + const char *section, + unsigned section_line, + const char *lvalue, + int ltype, + const char *rvalue, + void *data, + void *userdata) { + + _cleanup_free_ char *p = NULL; + Unit *u = data; + UnitType type; + int r; + + assert(filename); + assert(lvalue); + assert(rvalue); + assert(data); + + if (!set_isempty(u->dependencies[UNIT_TRIGGERS])) { + log_syntax(unit, LOG_ERR, filename, line, 0, "Multiple units to trigger specified, ignoring: %s", rvalue); + return 0; + } + + r = unit_name_printf(u, rvalue, &p); + if (r < 0) { + log_syntax(unit, LOG_ERR, filename, line, r, "Failed to resolve specifiers, ignoring: %m"); + return 0; + } + + type = unit_name_to_type(p); + if (type < 0) { + log_syntax(unit, LOG_ERR, filename, line, 0, "Unit type not valid, ignoring: %s", rvalue); + return 0; + } + + if (type == u->type) { + log_syntax(unit, LOG_ERR, filename, line, 0, "Trigger cannot be of same type, ignoring: %s", rvalue); + return 0; + } + + r = unit_add_two_dependencies_by_name(u, UNIT_BEFORE, UNIT_TRIGGERS, p, NULL, true); + if (r < 0) { + log_syntax(unit, LOG_ERR, filename, line, r, "Failed to add trigger on %s, ignoring: %m", p); + return 0; + } + + return 0; +} + +int config_parse_path_spec(const char *unit, + const char *filename, + unsigned line, + const char *section, + unsigned section_line, + const char *lvalue, + int ltype, + const char *rvalue, + void *data, + void *userdata) { + + Path *p = data; + PathSpec *s; + PathType b; + _cleanup_free_ char *k = NULL; + int r; + + assert(filename); + assert(lvalue); + assert(rvalue); + assert(data); + + if (isempty(rvalue)) { + /* Empty assignment clears list */ + path_free_specs(p); + return 0; + } + + b = path_type_from_string(lvalue); + if (b < 0) { + log_syntax(unit, LOG_ERR, filename, line, 0, "Failed to parse path type, ignoring: %s", lvalue); + return 0; + } + + r = unit_full_printf(UNIT(p), rvalue, &k); + if (r < 0) { + log_syntax(unit, LOG_ERR, filename, line, r, "Failed to resolve unit specifiers on %s. Ignoring.", rvalue); + return 0; + } + + if (!path_is_absolute(k)) { + log_syntax(unit, LOG_ERR, filename, line, 0, "Path is not absolute, ignoring: %s", k); + return 0; + } + + s = new0(PathSpec, 1); + if (!s) + return log_oom(); + + s->unit = UNIT(p); + s->path = path_kill_slashes(k); + k = NULL; + s->type = b; + s->inotify_fd = -1; + + LIST_PREPEND(spec, p->specs, s); + + return 0; +} + +int config_parse_socket_service( + const char *unit, + const char *filename, + unsigned line, + const char *section, + unsigned section_line, + const char *lvalue, + int ltype, + const char *rvalue, + void *data, + void *userdata) { + + _cleanup_(sd_bus_error_free) sd_bus_error error = SD_BUS_ERROR_NULL; + _cleanup_free_ char *p = NULL; + Socket *s = data; + Unit *x; + int r; + + assert(filename); + assert(lvalue); + assert(rvalue); + assert(data); + + r = unit_name_printf(UNIT(s), rvalue, &p); + if (r < 0) { + log_syntax(unit, LOG_ERR, filename, line, r, "Failed to resolve specifiers, ignoring: %s", rvalue); + return 0; + } + + if (!endswith(p, ".service")) { + log_syntax(unit, LOG_ERR, filename, line, 0, "Unit must be of type service, ignoring: %s", rvalue); + return 0; + } + + r = manager_load_unit(UNIT(s)->manager, p, NULL, &error, &x); + if (r < 0) { + log_syntax(unit, LOG_ERR, filename, line, r, "Failed to load unit %s, ignoring: %s", rvalue, bus_error_message(&error, r)); + return 0; + } + + unit_ref_set(&s->service, x); + + return 0; +} + +int config_parse_fdname( + const char *unit, + const char *filename, + unsigned line, + const char *section, + unsigned section_line, + const char *lvalue, + int ltype, + const char *rvalue, + void *data, + void *userdata) { + + _cleanup_free_ char *p = NULL; + Socket *s = data; + int r; + + assert(filename); + assert(lvalue); + assert(rvalue); + assert(data); + + if (isempty(rvalue)) { + s->fdname = mfree(s->fdname); + return 0; + } + + r = unit_name_printf(UNIT(s), rvalue, &p); + if (r < 0) { + log_syntax(unit, LOG_ERR, filename, line, r, "Failed to resolve specifiers, ignoring: %s", rvalue); + return 0; + } + + if (!fdname_is_valid(p)) { + log_syntax(unit, LOG_ERR, filename, line, 0, "Invalid file descriptor name, ignoring: %s", p); + return 0; + } + + free(s->fdname); + s->fdname = p; + p = NULL; + + return 0; +} + +int config_parse_service_sockets( + const char *unit, + const char *filename, + unsigned line, + const char *section, + unsigned section_line, + const char *lvalue, + int ltype, + const char *rvalue, + void *data, + void *userdata) { + + Service *s = data; + const char *p; + int r; + + assert(filename); + assert(lvalue); + assert(rvalue); + assert(data); + + p = rvalue; + for (;;) { + _cleanup_free_ char *word = NULL, *k = NULL; + + r = extract_first_word(&p, &word, NULL, 0); + if (r == 0) + break; + if (r == -ENOMEM) + return log_oom(); + if (r < 0) { + log_syntax(unit, LOG_ERR, filename, line, r, "Trailing garbage in sockets, ignoring: %s", rvalue); + break; + } + + r = unit_name_printf(UNIT(s), word, &k); + if (r < 0) { + log_syntax(unit, LOG_ERR, filename, line, r, "Failed to resolve specifiers, ignoring: %m"); + continue; + } + + if (!endswith(k, ".socket")) { + log_syntax(unit, LOG_ERR, filename, line, 0, "Unit must be of type socket, ignoring: %s", k); + continue; + } + + r = unit_add_two_dependencies_by_name(UNIT(s), UNIT_WANTS, UNIT_AFTER, k, NULL, true); + if (r < 0) + log_syntax(unit, LOG_ERR, filename, line, r, "Failed to add dependency on %s, ignoring: %m", k); + + r = unit_add_dependency_by_name(UNIT(s), UNIT_TRIGGERED_BY, k, NULL, true); + if (r < 0) + log_syntax(unit, LOG_ERR, filename, line, r, "Failed to add dependency on %s, ignoring: %m", k); + } + + return 0; +} + +int config_parse_bus_name( + const char *unit, + const char *filename, + unsigned line, + const char *section, + unsigned section_line, + const char *lvalue, + int ltype, + const char *rvalue, + void *data, + void *userdata) { + + _cleanup_free_ char *k = NULL; + Unit *u = userdata; + int r; + + assert(filename); + assert(lvalue); + assert(rvalue); + assert(u); + + r = unit_full_printf(u, rvalue, &k); + if (r < 0) { + log_syntax(unit, LOG_ERR, filename, line, r, "Failed to resolve unit specifiers on %s, ignoring: %m", rvalue); + return 0; + } + + if (!service_name_is_valid(k)) { + log_syntax(unit, LOG_ERR, filename, line, 0, "Invalid bus name %s, ignoring.", k); + return 0; + } + + return config_parse_string(unit, filename, line, section, section_line, lvalue, ltype, k, data, userdata); +} + +int config_parse_service_timeout( + const char *unit, + const char *filename, + unsigned line, + const char *section, + unsigned section_line, + const char *lvalue, + int ltype, + const char *rvalue, + void *data, + void *userdata) { + + Service *s = userdata; + usec_t usec; + int r; + + assert(filename); + assert(lvalue); + assert(rvalue); + assert(s); + + /* This is called for three cases: TimeoutSec=, TimeoutStopSec= and TimeoutStartSec=. */ + + r = parse_sec(rvalue, &usec); + if (r < 0) { + log_syntax(unit, LOG_ERR, filename, line, r, "Failed to parse %s= parameter, ignoring: %s", lvalue, rvalue); + return 0; + } + + /* Traditionally, these options accepted 0 to disable the timeouts. However, a timeout of 0 suggests it happens + * immediately, hence fix this to become USEC_INFINITY instead. This is in-line with how we internally handle + * all other timeouts. */ + if (usec <= 0) + usec = USEC_INFINITY; + + if (!streq(lvalue, "TimeoutStopSec")) { + s->start_timeout_defined = true; + s->timeout_start_usec = usec; + } + + if (!streq(lvalue, "TimeoutStartSec")) + s->timeout_stop_usec = usec; + + return 0; +} + +int config_parse_sec_fix_0( + const char *unit, + const char *filename, + unsigned line, + const char *section, + unsigned section_line, + const char *lvalue, + int ltype, + const char *rvalue, + void *data, + void *userdata) { + + usec_t *usec = data; + int r; + + assert(filename); + assert(lvalue); + assert(rvalue); + assert(usec); + + /* This is pretty much like config_parse_sec(), except that this treats a time of 0 as infinity, for + * compatibility with older versions of systemd where 0 instead of infinity was used as indicator to turn off a + * timeout. */ + + r = parse_sec(rvalue, usec); + if (r < 0) { + log_syntax(unit, LOG_ERR, filename, line, r, "Failed to parse %s= parameter, ignoring: %s", lvalue, rvalue); + return 0; + } + + if (*usec <= 0) + *usec = USEC_INFINITY; + + return 0; +} + +int config_parse_busname_service( + const char *unit, + const char *filename, + unsigned line, + const char *section, + unsigned section_line, + const char *lvalue, + int ltype, + const char *rvalue, + void *data, + void *userdata) { + + _cleanup_(sd_bus_error_free) sd_bus_error error = SD_BUS_ERROR_NULL; + BusName *n = data; + int r; + Unit *x; + _cleanup_free_ char *p = NULL; + + assert(filename); + assert(lvalue); + assert(rvalue); + assert(data); + + r = unit_name_printf(UNIT(n), rvalue, &p); + if (r < 0) { + log_syntax(unit, LOG_ERR, filename, line, r, "Failed to resolve specifiers, ignoring: %s", rvalue); + return 0; + } + + if (!endswith(p, ".service")) { + log_syntax(unit, LOG_ERR, filename, line, 0, "Unit must be of type service, ignoring: %s", rvalue); + return 0; + } + + r = manager_load_unit(UNIT(n)->manager, p, NULL, &error, &x); + if (r < 0) { + log_syntax(unit, LOG_ERR, filename, line, r, "Failed to load unit %s, ignoring: %s", rvalue, bus_error_message(&error, r)); + return 0; + } + + unit_ref_set(&n->service, x); + + return 0; +} + +DEFINE_CONFIG_PARSE_ENUM(config_parse_bus_policy_world, bus_policy_access, BusPolicyAccess, "Failed to parse bus name policy access"); + +int config_parse_bus_policy( + const char *unit, + const char *filename, + unsigned line, + const char *section, + unsigned section_line, + const char *lvalue, + int ltype, + const char *rvalue, + void *data, + void *userdata) { + + _cleanup_free_ BusNamePolicy *p = NULL; + _cleanup_free_ char *id_str = NULL; + BusName *busname = data; + char *access_str; + + assert(filename); + assert(lvalue); + assert(rvalue); + assert(data); + + p = new0(BusNamePolicy, 1); + if (!p) + return log_oom(); + + if (streq(lvalue, "AllowUser")) + p->type = BUSNAME_POLICY_TYPE_USER; + else if (streq(lvalue, "AllowGroup")) + p->type = BUSNAME_POLICY_TYPE_GROUP; + else + assert_not_reached("Unknown lvalue"); + + id_str = strdup(rvalue); + if (!id_str) + return log_oom(); + + access_str = strpbrk(id_str, WHITESPACE); + if (!access_str) { + log_syntax(unit, LOG_ERR, filename, line, 0, "Invalid busname policy value '%s'", rvalue); + return 0; + } + + *access_str = '\0'; + access_str++; + access_str += strspn(access_str, WHITESPACE); + + p->access = bus_policy_access_from_string(access_str); + if (p->access < 0) { + log_syntax(unit, LOG_ERR, filename, line, 0, "Invalid busname policy access type '%s'", access_str); + return 0; + } + + p->name = id_str; + id_str = NULL; + + LIST_PREPEND(policy, busname->policy, p); + p = NULL; + + return 0; +} + +int config_parse_working_directory( + const char *unit, + const char *filename, + unsigned line, + const char *section, + unsigned section_line, + const char *lvalue, + int ltype, + const char *rvalue, + void *data, + void *userdata) { + + ExecContext *c = data; + Unit *u = userdata; + bool missing_ok; + int r; + + assert(filename); + assert(lvalue); + assert(rvalue); + assert(c); + assert(u); + + if (rvalue[0] == '-') { + missing_ok = true; + rvalue++; + } else + missing_ok = false; + + if (streq(rvalue, "~")) { + c->working_directory_home = true; + c->working_directory = mfree(c->working_directory); + } else { + _cleanup_free_ char *k = NULL; + + r = unit_full_printf(u, rvalue, &k); + if (r < 0) { + log_syntax(unit, LOG_ERR, filename, line, r, "Failed to resolve unit specifiers in working directory path '%s', ignoring: %m", rvalue); + return 0; + } + + path_kill_slashes(k); + + if (!utf8_is_valid(k)) { + log_syntax_invalid_utf8(unit, LOG_ERR, filename, line, rvalue); + return 0; + } + + if (!path_is_absolute(k)) { + log_syntax(unit, LOG_ERR, filename, line, 0, "Working directory path '%s' is not absolute, ignoring.", rvalue); + return 0; + } + + free(c->working_directory); + c->working_directory = k; + k = NULL; + + c->working_directory_home = false; + } + + c->working_directory_missing_ok = missing_ok; + return 0; +} + +int config_parse_unit_env_file(const char *unit, + const char *filename, + unsigned line, + const char *section, + unsigned section_line, + const char *lvalue, + int ltype, + const char *rvalue, + void *data, + void *userdata) { + + char ***env = data; + Unit *u = userdata; + _cleanup_free_ char *n = NULL; + int r; + + assert(filename); + assert(lvalue); + assert(rvalue); + assert(data); + + if (isempty(rvalue)) { + /* Empty assignment frees the list */ + *env = strv_free(*env); + return 0; + } + + r = unit_full_printf(u, rvalue, &n); + if (r < 0) { + log_syntax(unit, LOG_ERR, filename, line, r, "Failed to resolve specifiers, ignoring: %s", rvalue); + return 0; + } + + if (!path_is_absolute(n[0] == '-' ? n + 1 : n)) { + log_syntax(unit, LOG_ERR, filename, line, 0, "Path '%s' is not absolute, ignoring.", n); + return 0; + } + + r = strv_extend(env, n); + if (r < 0) + return log_oom(); + + return 0; +} + +int config_parse_environ(const char *unit, + const char *filename, + unsigned line, + const char *section, + unsigned section_line, + const char *lvalue, + int ltype, + const char *rvalue, + void *data, + void *userdata) { + + Unit *u = userdata; + char*** env = data; + const char *word, *state; + size_t l; + _cleanup_free_ char *k = NULL; + int r; + + assert(filename); + assert(lvalue); + assert(rvalue); + assert(data); + + if (isempty(rvalue)) { + /* Empty assignment resets the list */ + *env = strv_free(*env); + return 0; + } + + if (u) { + r = unit_full_printf(u, rvalue, &k); + if (r < 0) { + log_syntax(unit, LOG_ERR, filename, line, r, "Failed to resolve specifiers, ignoring: %s", rvalue); + return 0; + } + } + + if (!k) { + k = strdup(rvalue); + if (!k) + return log_oom(); + } + + FOREACH_WORD_QUOTED(word, l, k, state) { + _cleanup_free_ char *n = NULL; + char **x; + + r = cunescape_length(word, l, 0, &n); + if (r < 0) { + log_syntax(unit, LOG_ERR, filename, line, r, "Couldn't unescape assignment, ignoring: %s", rvalue); + continue; + } + + if (!env_assignment_is_valid(n)) { + log_syntax(unit, LOG_ERR, filename, line, 0, "Invalid environment assignment, ignoring: %s", rvalue); + continue; + } + + x = strv_env_set(*env, n); + if (!x) + return log_oom(); + + strv_free(*env); + *env = x; + } + if (!isempty(state)) + log_syntax(unit, LOG_ERR, filename, line, 0, "Trailing garbage, ignoring."); + + return 0; +} + +int config_parse_pass_environ(const char *unit, + const char *filename, + unsigned line, + const char *section, + unsigned section_line, + const char *lvalue, + int ltype, + const char *rvalue, + void *data, + void *userdata) { + + const char *whole_rvalue = rvalue; + char*** passenv = data; + _cleanup_strv_free_ char **n = NULL; + size_t nlen = 0, nbufsize = 0; + int r; + + assert(filename); + assert(lvalue); + assert(rvalue); + assert(data); + + if (isempty(rvalue)) { + /* Empty assignment resets the list */ + *passenv = strv_free(*passenv); + return 0; + } + + for (;;) { + _cleanup_free_ char *word = NULL; + + r = extract_first_word(&rvalue, &word, WHITESPACE, EXTRACT_QUOTES); + if (r == 0) + break; + if (r == -ENOMEM) + return log_oom(); + if (r < 0) { + log_syntax(unit, LOG_ERR, filename, line, r, + "Trailing garbage in %s, ignoring: %s", lvalue, whole_rvalue); + break; + } + + if (!env_name_is_valid(word)) { + log_syntax(unit, LOG_ERR, filename, line, EINVAL, + "Invalid environment name for %s, ignoring: %s", lvalue, word); + continue; + } + + if (!GREEDY_REALLOC(n, nbufsize, nlen + 2)) + return log_oom(); + n[nlen++] = word; + n[nlen] = NULL; + word = NULL; + } + + if (n) { + r = strv_extend_strv(passenv, n, true); + if (r < 0) + return r; + } + + return 0; +} + +int config_parse_ip_tos(const char *unit, + const char *filename, + unsigned line, + const char *section, + unsigned section_line, + const char *lvalue, + int ltype, + const char *rvalue, + void *data, + void *userdata) { + + int *ip_tos = data, x; + + assert(filename); + assert(lvalue); + assert(rvalue); + assert(data); + + x = ip_tos_from_string(rvalue); + if (x < 0) { + log_syntax(unit, LOG_ERR, filename, line, 0, "Failed to parse IP TOS value, ignoring: %s", rvalue); + return 0; + } + + *ip_tos = x; + return 0; +} + +int config_parse_unit_condition_path( + const char *unit, + const char *filename, + unsigned line, + const char *section, + unsigned section_line, + const char *lvalue, + int ltype, + const char *rvalue, + void *data, + void *userdata) { + + _cleanup_free_ char *p = NULL; + Condition **list = data, *c; + ConditionType t = ltype; + bool trigger, negate; + Unit *u = userdata; + int r; + + assert(filename); + assert(lvalue); + assert(rvalue); + assert(data); + + if (isempty(rvalue)) { + /* Empty assignment resets the list */ + *list = condition_free_list(*list); + return 0; + } + + trigger = rvalue[0] == '|'; + if (trigger) + rvalue++; + + negate = rvalue[0] == '!'; + if (negate) + rvalue++; + + r = unit_full_printf(u, rvalue, &p); + if (r < 0) { + log_syntax(unit, LOG_ERR, filename, line, r, "Failed to resolve specifiers, ignoring: %s", rvalue); + return 0; + } + + if (!path_is_absolute(p)) { + log_syntax(unit, LOG_ERR, filename, line, 0, "Path in condition not absolute, ignoring: %s", p); + return 0; + } + + c = condition_new(t, p, trigger, negate); + if (!c) + return log_oom(); + + LIST_PREPEND(conditions, *list, c); + return 0; +} + +int config_parse_unit_condition_string( + const char *unit, + const char *filename, + unsigned line, + const char *section, + unsigned section_line, + const char *lvalue, + int ltype, + const char *rvalue, + void *data, + void *userdata) { + + _cleanup_free_ char *s = NULL; + Condition **list = data, *c; + ConditionType t = ltype; + bool trigger, negate; + Unit *u = userdata; + int r; + + assert(filename); + assert(lvalue); + assert(rvalue); + assert(data); + + if (isempty(rvalue)) { + /* Empty assignment resets the list */ + *list = condition_free_list(*list); + return 0; + } + + trigger = rvalue[0] == '|'; + if (trigger) + rvalue++; + + negate = rvalue[0] == '!'; + if (negate) + rvalue++; + + r = unit_full_printf(u, rvalue, &s); + if (r < 0) { + log_syntax(unit, LOG_ERR, filename, line, r, "Failed to resolve specifiers, ignoring: %s", rvalue); + return 0; + } + + c = condition_new(t, s, trigger, negate); + if (!c) + return log_oom(); + + LIST_PREPEND(conditions, *list, c); + return 0; +} + +int config_parse_unit_condition_null( + const char *unit, + const char *filename, + unsigned line, + const char *section, + unsigned section_line, + const char *lvalue, + int ltype, + const char *rvalue, + void *data, + void *userdata) { + + Condition **list = data, *c; + bool trigger, negate; + int b; + + assert(filename); + assert(lvalue); + assert(rvalue); + assert(data); + + if (isempty(rvalue)) { + /* Empty assignment resets the list */ + *list = condition_free_list(*list); + return 0; + } + + trigger = rvalue[0] == '|'; + if (trigger) + rvalue++; + + negate = rvalue[0] == '!'; + if (negate) + rvalue++; + + b = parse_boolean(rvalue); + if (b < 0) { + log_syntax(unit, LOG_ERR, filename, line, b, "Failed to parse boolean value in condition, ignoring: %s", rvalue); + return 0; + } + + if (!b) + negate = !negate; + + c = condition_new(CONDITION_NULL, NULL, trigger, negate); + if (!c) + return log_oom(); + + LIST_PREPEND(conditions, *list, c); + return 0; +} + +DEFINE_CONFIG_PARSE_ENUM(config_parse_notify_access, notify_access, NotifyAccess, "Failed to parse notify access specifier"); +DEFINE_CONFIG_PARSE_ENUM(config_parse_failure_action, failure_action, FailureAction, "Failed to parse failure action specifier"); + +int config_parse_unit_requires_mounts_for( + const char *unit, + const char *filename, + unsigned line, + const char *section, + unsigned section_line, + const char *lvalue, + int ltype, + const char *rvalue, + void *data, + void *userdata) { + + Unit *u = userdata; + const char *word, *state; + size_t l; + + assert(filename); + assert(lvalue); + assert(rvalue); + assert(data); + + FOREACH_WORD_QUOTED(word, l, rvalue, state) { + int r; + _cleanup_free_ char *n; + + n = strndup(word, l); + if (!n) + return log_oom(); + + if (!utf8_is_valid(n)) { + log_syntax_invalid_utf8(unit, LOG_ERR, filename, line, rvalue); + continue; + } + + r = unit_require_mounts_for(u, n); + if (r < 0) { + log_syntax(unit, LOG_ERR, filename, line, r, "Failed to add required mount for, ignoring: %s", rvalue); + continue; + } + } + if (!isempty(state)) + log_syntax(unit, LOG_ERR, filename, line, 0, "Trailing garbage, ignoring."); + + return 0; +} + +int config_parse_documentation(const char *unit, + const char *filename, + unsigned line, + const char *section, + unsigned section_line, + const char *lvalue, + int ltype, + const char *rvalue, + void *data, + void *userdata) { + + Unit *u = userdata; + int r; + char **a, **b; + + assert(filename); + assert(lvalue); + assert(rvalue); + assert(u); + + if (isempty(rvalue)) { + /* Empty assignment resets the list */ + u->documentation = strv_free(u->documentation); + return 0; + } + + r = config_parse_unit_strv_printf(unit, filename, line, section, section_line, lvalue, ltype, + rvalue, data, userdata); + if (r < 0) + return r; + + for (a = b = u->documentation; a && *a; a++) { + + if (documentation_url_is_valid(*a)) + *(b++) = *a; + else { + log_syntax(unit, LOG_ERR, filename, line, 0, "Invalid URL, ignoring: %s", *a); + free(*a); + } + } + if (b) + *b = NULL; + + return r; +} + +#ifdef HAVE_SECCOMP +static int syscall_filter_parse_one( + const char *unit, + const char *filename, + unsigned line, + ExecContext *c, + bool invert, + const char *t, + bool warn) { + int r; + + if (*t == '@') { + const SystemCallFilterSet *set; + + for (set = syscall_filter_sets; set->set_name; set++) + if (streq(set->set_name, t)) { + const char *sys; + + NULSTR_FOREACH(sys, set->value) { + r = syscall_filter_parse_one(unit, filename, line, c, invert, sys, false); + if (r < 0) + return r; + } + break; + } + } else { + int id; + + id = seccomp_syscall_resolve_name(t); + if (id == __NR_SCMP_ERROR) { + if (warn) + log_syntax(unit, LOG_ERR, filename, line, 0, "Failed to parse system call, ignoring: %s", t); + return 0; + } + + /* If we previously wanted to forbid a syscall and now + * we want to allow it, then remove it from the list + */ + if (!invert == c->syscall_whitelist) { + r = set_put(c->syscall_filter, INT_TO_PTR(id + 1)); + if (r == 0) + return 0; + if (r < 0) + return log_oom(); + } else + set_remove(c->syscall_filter, INT_TO_PTR(id + 1)); + } + return 0; +} + +int config_parse_syscall_filter( + const char *unit, + const char *filename, + unsigned line, + const char *section, + unsigned section_line, + const char *lvalue, + int ltype, + const char *rvalue, + void *data, + void *userdata) { + + ExecContext *c = data; + Unit *u = userdata; + bool invert = false; + const char *word, *state; + size_t l; + int r; + + assert(filename); + assert(lvalue); + assert(rvalue); + assert(u); + + if (isempty(rvalue)) { + /* Empty assignment resets the list */ + c->syscall_filter = set_free(c->syscall_filter); + c->syscall_whitelist = false; + return 0; + } + + if (rvalue[0] == '~') { + invert = true; + rvalue++; + } + + if (!c->syscall_filter) { + c->syscall_filter = set_new(NULL); + if (!c->syscall_filter) + return log_oom(); + + if (invert) + /* Allow everything but the ones listed */ + c->syscall_whitelist = false; + else { + /* Allow nothing but the ones listed */ + c->syscall_whitelist = true; + + /* Accept default syscalls if we are on a whitelist */ + r = syscall_filter_parse_one(unit, filename, line, c, false, "@default", false); + if (r < 0) + return r; + } + } + + FOREACH_WORD_QUOTED(word, l, rvalue, state) { + _cleanup_free_ char *t = NULL; + + t = strndup(word, l); + if (!t) + return log_oom(); + + r = syscall_filter_parse_one(unit, filename, line, c, invert, t, true); + if (r < 0) + return r; + } + if (!isempty(state)) + log_syntax(unit, LOG_ERR, filename, line, 0, "Trailing garbage, ignoring."); + + /* Turn on NNP, but only if it wasn't configured explicitly + * before, and only if we are in user mode. */ + if (!c->no_new_privileges_set && MANAGER_IS_USER(u->manager)) + c->no_new_privileges = true; + + return 0; +} + +int config_parse_syscall_archs( + const char *unit, + const char *filename, + unsigned line, + const char *section, + unsigned section_line, + const char *lvalue, + int ltype, + const char *rvalue, + void *data, + void *userdata) { + + Set **archs = data; + const char *word, *state; + size_t l; + int r; + + if (isempty(rvalue)) { + *archs = set_free(*archs); + return 0; + } + + r = set_ensure_allocated(archs, NULL); + if (r < 0) + return log_oom(); + + FOREACH_WORD_QUOTED(word, l, rvalue, state) { + _cleanup_free_ char *t = NULL; + uint32_t a; + + t = strndup(word, l); + if (!t) + return log_oom(); + + r = seccomp_arch_from_string(t, &a); + if (r < 0) { + log_syntax(unit, LOG_ERR, filename, line, 0, "Failed to parse system call architecture, ignoring: %s", t); + continue; + } + + r = set_put(*archs, UINT32_TO_PTR(a + 1)); + if (r == 0) + continue; + if (r < 0) + return log_oom(); + } + if (!isempty(state)) + log_syntax(unit, LOG_ERR, filename, line, 0, "Trailing garbage, ignoring."); + + return 0; +} + +int config_parse_syscall_errno( + const char *unit, + const char *filename, + unsigned line, + const char *section, + unsigned section_line, + const char *lvalue, + int ltype, + const char *rvalue, + void *data, + void *userdata) { + + ExecContext *c = data; + int e; + + assert(filename); + assert(lvalue); + assert(rvalue); + + if (isempty(rvalue)) { + /* Empty assignment resets to KILL */ + c->syscall_errno = 0; + return 0; + } + + e = errno_from_name(rvalue); + if (e < 0) { + log_syntax(unit, LOG_ERR, filename, line, 0, "Failed to parse error number, ignoring: %s", rvalue); + return 0; + } + + c->syscall_errno = e; + return 0; +} + +int config_parse_address_families( + const char *unit, + const char *filename, + unsigned line, + const char *section, + unsigned section_line, + const char *lvalue, + int ltype, + const char *rvalue, + void *data, + void *userdata) { + + ExecContext *c = data; + bool invert = false; + const char *word, *state; + size_t l; + int r; + + assert(filename); + assert(lvalue); + assert(rvalue); + + if (isempty(rvalue)) { + /* Empty assignment resets the list */ + c->address_families = set_free(c->address_families); + c->address_families_whitelist = false; + return 0; + } + + if (rvalue[0] == '~') { + invert = true; + rvalue++; + } + + if (!c->address_families) { + c->address_families = set_new(NULL); + if (!c->address_families) + return log_oom(); + + c->address_families_whitelist = !invert; + } + + FOREACH_WORD_QUOTED(word, l, rvalue, state) { + _cleanup_free_ char *t = NULL; + int af; + + t = strndup(word, l); + if (!t) + return log_oom(); + + af = af_from_name(t); + if (af <= 0) { + log_syntax(unit, LOG_ERR, filename, line, 0, "Failed to parse address family, ignoring: %s", t); + continue; + } + + /* If we previously wanted to forbid an address family and now + * we want to allow it, then remove it from the list + */ + if (!invert == c->address_families_whitelist) { + r = set_put(c->address_families, INT_TO_PTR(af)); + if (r == 0) + continue; + if (r < 0) + return log_oom(); + } else + set_remove(c->address_families, INT_TO_PTR(af)); + } + if (!isempty(state)) + log_syntax(unit, LOG_ERR, filename, line, 0, "Trailing garbage, ignoring."); + + return 0; +} +#endif + +int config_parse_unit_slice( + const char *unit, + const char *filename, + unsigned line, + const char *section, + unsigned section_line, + const char *lvalue, + int ltype, + const char *rvalue, + void *data, + void *userdata) { + + _cleanup_free_ char *k = NULL; + Unit *u = userdata, *slice = NULL; + int r; + + assert(filename); + assert(lvalue); + assert(rvalue); + assert(u); + + r = unit_name_printf(u, rvalue, &k); + if (r < 0) { + log_syntax(unit, LOG_ERR, filename, line, r, "Failed to resolve unit specifiers on %s. Ignoring.", rvalue); + return 0; + } + + r = manager_load_unit(u->manager, k, NULL, NULL, &slice); + if (r < 0) { + log_syntax(unit, LOG_ERR, filename, line, r, "Failed to load slice unit %s. Ignoring.", k); + return 0; + } + + r = unit_set_slice(u, slice); + if (r < 0) { + log_syntax(unit, LOG_ERR, filename, line, r, "Failed to assign slice %s to unit %s. Ignoring.", slice->id, u->id); + return 0; + } + + return 0; +} + +DEFINE_CONFIG_PARSE_ENUM(config_parse_device_policy, cgroup_device_policy, CGroupDevicePolicy, "Failed to parse device policy"); + +int config_parse_cpu_shares( + const char *unit, + const char *filename, + unsigned line, + const char *section, + unsigned section_line, + const char *lvalue, + int ltype, + const char *rvalue, + void *data, + void *userdata) { + + uint64_t *shares = data; + int r; + + assert(filename); + assert(lvalue); + assert(rvalue); + + r = cg_cpu_shares_parse(rvalue, shares); + if (r < 0) { + log_syntax(unit, LOG_ERR, filename, line, r, "CPU shares '%s' invalid. Ignoring.", rvalue); + return 0; + } + + return 0; +} + +int config_parse_cpu_quota( + const char *unit, + const char *filename, + unsigned line, + const char *section, + unsigned section_line, + const char *lvalue, + int ltype, + const char *rvalue, + void *data, + void *userdata) { + + CGroupContext *c = data; + int r; + + assert(filename); + assert(lvalue); + assert(rvalue); + + if (isempty(rvalue)) { + c->cpu_quota_per_sec_usec = USEC_INFINITY; + return 0; + } + + r = parse_percent(rvalue); + if (r <= 0) { + log_syntax(unit, LOG_ERR, filename, line, r, "CPU quota '%s' invalid. Ignoring.", rvalue); + return 0; + } + + c->cpu_quota_per_sec_usec = ((usec_t) r * USEC_PER_SEC) / 100U; + return 0; +} + +int config_parse_memory_limit( + const char *unit, + const char *filename, + unsigned line, + const char *section, + unsigned section_line, + const char *lvalue, + int ltype, + const char *rvalue, + void *data, + void *userdata) { + + CGroupContext *c = data; + uint64_t bytes = CGROUP_LIMIT_MAX; + int r; + + if (!isempty(rvalue) && !streq(rvalue, "infinity")) { + + r = parse_percent(rvalue); + if (r < 0) { + r = parse_size(rvalue, 1024, &bytes); + if (r < 0) { + log_syntax(unit, LOG_ERR, filename, line, r, "Memory limit '%s' invalid. Ignoring.", rvalue); + return 0; + } + } else + bytes = physical_memory_scale(r, 100U); + + if (bytes <= 0 || bytes >= UINT64_MAX) { + log_syntax(unit, LOG_ERR, filename, line, 0, "Memory limit '%s' out of range. Ignoring.", rvalue); + return 0; + } + } + + if (streq(lvalue, "MemoryLow")) + c->memory_low = bytes; + else if (streq(lvalue, "MemoryHigh")) + c->memory_high = bytes; + else if (streq(lvalue, "MemoryMax")) + c->memory_max = bytes; + else + c->memory_limit = bytes; + + return 0; +} + +int config_parse_tasks_max( + const char *unit, + const char *filename, + unsigned line, + const char *section, + unsigned section_line, + const char *lvalue, + int ltype, + const char *rvalue, + void *data, + void *userdata) { + + uint64_t *tasks_max = data, u; + int r; + + if (isempty(rvalue) || streq(rvalue, "infinity")) { + *tasks_max = (uint64_t) -1; + return 0; + } + + r = parse_percent(rvalue); + if (r < 0) { + r = safe_atou64(rvalue, &u); + if (r < 0) { + log_syntax(unit, LOG_ERR, filename, line, r, "Maximum tasks value '%s' invalid. Ignoring.", rvalue); + return 0; + } + } else + u = system_tasks_max_scale(r, 100U); + + if (u <= 0 || u >= UINT64_MAX) { + log_syntax(unit, LOG_ERR, filename, line, 0, "Maximum tasks value '%s' out of range. Ignoring.", rvalue); + return 0; + } + + *tasks_max = u; + return 0; +} + +int config_parse_device_allow( + const char *unit, + const char *filename, + unsigned line, + const char *section, + unsigned section_line, + const char *lvalue, + int ltype, + const char *rvalue, + void *data, + void *userdata) { + + _cleanup_free_ char *path = NULL, *t = NULL; + CGroupContext *c = data; + CGroupDeviceAllow *a; + const char *m = NULL; + size_t n; + int r; + + if (isempty(rvalue)) { + while (c->device_allow) + cgroup_context_free_device_allow(c, c->device_allow); + + return 0; + } + + r = unit_full_printf(userdata, rvalue, &t); + if(r < 0) { + log_syntax(unit, LOG_WARNING, filename, line, r, + "Failed to resolve specifiers in %s, ignoring: %m", + rvalue); + } + + n = strcspn(t, WHITESPACE); + + path = strndup(t, n); + if (!path) + return log_oom(); + + if (!startswith(path, "/dev/") && + !startswith(path, "block-") && + !startswith(path, "char-")) { + log_syntax(unit, LOG_ERR, filename, line, 0, "Invalid device node path '%s'. Ignoring.", path); + return 0; + } + + m = t + n + strspn(t + n, WHITESPACE); + if (isempty(m)) + m = "rwm"; + + if (!in_charset(m, "rwm")) { + log_syntax(unit, LOG_ERR, filename, line, 0, "Invalid device rights '%s'. Ignoring.", m); + return 0; + } + + a = new0(CGroupDeviceAllow, 1); + if (!a) + return log_oom(); + + a->path = path; + path = NULL; + a->r = !!strchr(m, 'r'); + a->w = !!strchr(m, 'w'); + a->m = !!strchr(m, 'm'); + + LIST_PREPEND(device_allow, c->device_allow, a); + return 0; +} + +int config_parse_io_weight( + const char *unit, + const char *filename, + unsigned line, + const char *section, + unsigned section_line, + const char *lvalue, + int ltype, + const char *rvalue, + void *data, + void *userdata) { + + uint64_t *weight = data; + int r; + + assert(filename); + assert(lvalue); + assert(rvalue); + + r = cg_weight_parse(rvalue, weight); + if (r < 0) { + log_syntax(unit, LOG_ERR, filename, line, r, "IO weight '%s' invalid. Ignoring.", rvalue); + return 0; + } + + return 0; +} + +int config_parse_io_device_weight( + const char *unit, + const char *filename, + unsigned line, + const char *section, + unsigned section_line, + const char *lvalue, + int ltype, + const char *rvalue, + void *data, + void *userdata) { + + _cleanup_free_ char *path = NULL; + CGroupIODeviceWeight *w; + CGroupContext *c = data; + const char *weight; + uint64_t u; + size_t n; + int r; + + assert(filename); + assert(lvalue); + assert(rvalue); + + if (isempty(rvalue)) { + while (c->io_device_weights) + cgroup_context_free_io_device_weight(c, c->io_device_weights); + + return 0; + } + + n = strcspn(rvalue, WHITESPACE); + weight = rvalue + n; + weight += strspn(weight, WHITESPACE); + + if (isempty(weight)) { + log_syntax(unit, LOG_ERR, filename, line, 0, "Expected block device and device weight. Ignoring."); + return 0; + } + + path = strndup(rvalue, n); + if (!path) + return log_oom(); + + if (!path_startswith(path, "/dev")) { + log_syntax(unit, LOG_ERR, filename, line, 0, "Invalid device node path '%s'. Ignoring.", path); + return 0; + } + + r = cg_weight_parse(weight, &u); + if (r < 0) { + log_syntax(unit, LOG_ERR, filename, line, r, "IO weight '%s' invalid. Ignoring.", weight); + return 0; + } + + assert(u != CGROUP_WEIGHT_INVALID); + + w = new0(CGroupIODeviceWeight, 1); + if (!w) + return log_oom(); + + w->path = path; + path = NULL; + + w->weight = u; + + LIST_PREPEND(device_weights, c->io_device_weights, w); + return 0; +} + +int config_parse_io_limit( + const char *unit, + const char *filename, + unsigned line, + const char *section, + unsigned section_line, + const char *lvalue, + int ltype, + const char *rvalue, + void *data, + void *userdata) { + + _cleanup_free_ char *path = NULL; + CGroupIODeviceLimit *l = NULL, *t; + CGroupContext *c = data; + CGroupIOLimitType type; + const char *limit; + uint64_t num; + size_t n; + int r; + + assert(filename); + assert(lvalue); + assert(rvalue); + + type = cgroup_io_limit_type_from_string(lvalue); + assert(type >= 0); + + if (isempty(rvalue)) { + LIST_FOREACH(device_limits, l, c->io_device_limits) + l->limits[type] = cgroup_io_limit_defaults[type]; + return 0; + } + + n = strcspn(rvalue, WHITESPACE); + limit = rvalue + n; + limit += strspn(limit, WHITESPACE); + + if (!*limit) { + log_syntax(unit, LOG_ERR, filename, line, 0, "Expected space separated pair of device node and bandwidth. Ignoring."); + return 0; + } + + path = strndup(rvalue, n); + if (!path) + return log_oom(); + + if (!path_startswith(path, "/dev")) { + log_syntax(unit, LOG_ERR, filename, line, 0, "Invalid device node path '%s'. Ignoring.", path); + return 0; + } + + if (streq("infinity", limit)) { + num = CGROUP_LIMIT_MAX; + } else { + r = parse_size(limit, 1000, &num); + if (r < 0 || num <= 0) { + log_syntax(unit, LOG_ERR, filename, line, r, "IO Limit '%s' invalid. Ignoring.", rvalue); + return 0; + } + } + + LIST_FOREACH(device_limits, t, c->io_device_limits) { + if (path_equal(path, t->path)) { + l = t; + break; + } + } + + if (!l) { + CGroupIOLimitType ttype; + + l = new0(CGroupIODeviceLimit, 1); + if (!l) + return log_oom(); + + l->path = path; + path = NULL; + for (ttype = 0; ttype < _CGROUP_IO_LIMIT_TYPE_MAX; ttype++) + l->limits[ttype] = cgroup_io_limit_defaults[ttype]; + + LIST_PREPEND(device_limits, c->io_device_limits, l); + } + + l->limits[type] = num; + + return 0; +} + +int config_parse_blockio_weight( + const char *unit, + const char *filename, + unsigned line, + const char *section, + unsigned section_line, + const char *lvalue, + int ltype, + const char *rvalue, + void *data, + void *userdata) { + + uint64_t *weight = data; + int r; + + assert(filename); + assert(lvalue); + assert(rvalue); + + r = cg_blkio_weight_parse(rvalue, weight); + if (r < 0) { + log_syntax(unit, LOG_ERR, filename, line, r, "Block IO weight '%s' invalid. Ignoring.", rvalue); + return 0; + } + + return 0; +} + +int config_parse_blockio_device_weight( + const char *unit, + const char *filename, + unsigned line, + const char *section, + unsigned section_line, + const char *lvalue, + int ltype, + const char *rvalue, + void *data, + void *userdata) { + + _cleanup_free_ char *path = NULL; + CGroupBlockIODeviceWeight *w; + CGroupContext *c = data; + const char *weight; + uint64_t u; + size_t n; + int r; + + assert(filename); + assert(lvalue); + assert(rvalue); + + if (isempty(rvalue)) { + while (c->blockio_device_weights) + cgroup_context_free_blockio_device_weight(c, c->blockio_device_weights); + + return 0; + } + + n = strcspn(rvalue, WHITESPACE); + weight = rvalue + n; + weight += strspn(weight, WHITESPACE); + + if (isempty(weight)) { + log_syntax(unit, LOG_ERR, filename, line, 0, "Expected block device and device weight. Ignoring."); + return 0; + } + + path = strndup(rvalue, n); + if (!path) + return log_oom(); + + if (!path_startswith(path, "/dev")) { + log_syntax(unit, LOG_ERR, filename, line, 0, "Invalid device node path '%s'. Ignoring.", path); + return 0; + } + + r = cg_blkio_weight_parse(weight, &u); + if (r < 0) { + log_syntax(unit, LOG_ERR, filename, line, r, "Block IO weight '%s' invalid. Ignoring.", weight); + return 0; + } + + assert(u != CGROUP_BLKIO_WEIGHT_INVALID); + + w = new0(CGroupBlockIODeviceWeight, 1); + if (!w) + return log_oom(); + + w->path = path; + path = NULL; + + w->weight = u; + + LIST_PREPEND(device_weights, c->blockio_device_weights, w); + return 0; +} + +int config_parse_blockio_bandwidth( + const char *unit, + const char *filename, + unsigned line, + const char *section, + unsigned section_line, + const char *lvalue, + int ltype, + const char *rvalue, + void *data, + void *userdata) { + + _cleanup_free_ char *path = NULL; + CGroupBlockIODeviceBandwidth *b = NULL, *t; + CGroupContext *c = data; + const char *bandwidth; + uint64_t bytes; + bool read; + size_t n; + int r; + + assert(filename); + assert(lvalue); + assert(rvalue); + + read = streq("BlockIOReadBandwidth", lvalue); + + if (isempty(rvalue)) { + LIST_FOREACH(device_bandwidths, b, c->blockio_device_bandwidths) { + b->rbps = CGROUP_LIMIT_MAX; + b->wbps = CGROUP_LIMIT_MAX; + } + return 0; + } + + n = strcspn(rvalue, WHITESPACE); + bandwidth = rvalue + n; + bandwidth += strspn(bandwidth, WHITESPACE); + + if (!*bandwidth) { + log_syntax(unit, LOG_ERR, filename, line, 0, "Expected space separated pair of device node and bandwidth. Ignoring."); + return 0; + } + + path = strndup(rvalue, n); + if (!path) + return log_oom(); + + if (!path_startswith(path, "/dev")) { + log_syntax(unit, LOG_ERR, filename, line, 0, "Invalid device node path '%s'. Ignoring.", path); + return 0; + } + + r = parse_size(bandwidth, 1000, &bytes); + if (r < 0 || bytes <= 0) { + log_syntax(unit, LOG_ERR, filename, line, r, "Block IO Bandwidth '%s' invalid. Ignoring.", rvalue); + return 0; + } + + LIST_FOREACH(device_bandwidths, t, c->blockio_device_bandwidths) { + if (path_equal(path, t->path)) { + b = t; + break; + } + } + + if (!t) { + b = new0(CGroupBlockIODeviceBandwidth, 1); + if (!b) + return log_oom(); + + b->path = path; + path = NULL; + b->rbps = CGROUP_LIMIT_MAX; + b->wbps = CGROUP_LIMIT_MAX; + + LIST_PREPEND(device_bandwidths, c->blockio_device_bandwidths, b); + } + + if (read) + b->rbps = bytes; + else + b->wbps = bytes; + + return 0; +} + +DEFINE_CONFIG_PARSE_ENUM(config_parse_job_mode, job_mode, JobMode, "Failed to parse job mode"); + +int config_parse_job_mode_isolate( + const char *unit, + const char *filename, + unsigned line, + const char *section, + unsigned section_line, + const char *lvalue, + int ltype, + const char *rvalue, + void *data, + void *userdata) { + + JobMode *m = data; + int r; + + assert(filename); + assert(lvalue); + assert(rvalue); + + r = parse_boolean(rvalue); + if (r < 0) { + log_syntax(unit, LOG_ERR, filename, line, r, "Failed to parse boolean, ignoring: %s", rvalue); + return 0; + } + + *m = r ? JOB_ISOLATE : JOB_REPLACE; + return 0; +} + +int config_parse_runtime_directory( + const char *unit, + const char *filename, + unsigned line, + const char *section, + unsigned section_line, + const char *lvalue, + int ltype, + const char *rvalue, + void *data, + void *userdata) { + + char***rt = data; + Unit *u = userdata; + const char *word, *state; + size_t l; + int r; + + assert(filename); + assert(lvalue); + assert(rvalue); + assert(data); + + if (isempty(rvalue)) { + /* Empty assignment resets the list */ + *rt = strv_free(*rt); + return 0; + } + + FOREACH_WORD_QUOTED(word, l, rvalue, state) { + _cleanup_free_ char *t = NULL, *n = NULL; + + t = strndup(word, l); + if (!t) + return log_oom(); + + r = unit_name_printf(u, t, &n); + if (r < 0) { + log_syntax(unit, LOG_ERR, filename, line, r, "Failed to resolve specifiers, ignoring: %m"); + continue; + } + + if (!filename_is_valid(n)) { + log_syntax(unit, LOG_ERR, filename, line, 0, "Runtime directory is not valid, ignoring assignment: %s", rvalue); + continue; + } + + r = strv_push(rt, n); + if (r < 0) + return log_oom(); + + n = NULL; + } + if (!isempty(state)) + log_syntax(unit, LOG_ERR, filename, line, 0, "Trailing garbage, ignoring."); + + return 0; +} + +int config_parse_set_status( + const char *unit, + const char *filename, + unsigned line, + const char *section, + unsigned section_line, + const char *lvalue, + int ltype, + const char *rvalue, + void *data, + void *userdata) { + + size_t l; + const char *word, *state; + int r; + ExitStatusSet *status_set = data; + + assert(filename); + assert(lvalue); + assert(rvalue); + assert(data); + + /* Empty assignment resets the list */ + if (isempty(rvalue)) { + exit_status_set_free(status_set); + return 0; + } + + FOREACH_WORD(word, l, rvalue, state) { + _cleanup_free_ char *temp; + int val; + Set **set; + + temp = strndup(word, l); + if (!temp) + return log_oom(); + + r = safe_atoi(temp, &val); + if (r < 0) { + val = signal_from_string_try_harder(temp); + + if (val <= 0) { + log_syntax(unit, LOG_ERR, filename, line, 0, "Failed to parse value, ignoring: %s", word); + continue; + } + set = &status_set->signal; + } else { + if (val < 0 || val > 255) { + log_syntax(unit, LOG_ERR, filename, line, 0, "Value %d is outside range 0-255, ignoring", val); + continue; + } + set = &status_set->status; + } + + r = set_ensure_allocated(set, NULL); + if (r < 0) + return log_oom(); + + r = set_put(*set, INT_TO_PTR(val)); + if (r < 0) { + log_syntax(unit, LOG_ERR, filename, line, r, "Unable to store: %s", word); + return r; + } + } + if (!isempty(state)) + log_syntax(unit, LOG_ERR, filename, line, 0, "Trailing garbage, ignoring."); + + return 0; +} + +int config_parse_namespace_path_strv( + const char *unit, + const char *filename, + unsigned line, + const char *section, + unsigned section_line, + const char *lvalue, + int ltype, + const char *rvalue, + void *data, + void *userdata) { + + char*** sv = data; + const char *prev; + const char *cur; + int r; + + assert(filename); + assert(lvalue); + assert(rvalue); + assert(data); + + if (isempty(rvalue)) { + /* Empty assignment resets the list */ + *sv = strv_free(*sv); + return 0; + } + + prev = cur = rvalue; + for (;;) { + _cleanup_free_ char *word = NULL; + int offset; + + r = extract_first_word(&cur, &word, NULL, EXTRACT_QUOTES); + if (r == 0) + break; + if (r == -ENOMEM) + return log_oom(); + if (r < 0) { + log_syntax(unit, LOG_ERR, filename, line, r, "Trailing garbage, ignoring: %s", prev); + return 0; + } + + if (!utf8_is_valid(word)) { + log_syntax_invalid_utf8(unit, LOG_ERR, filename, line, word); + prev = cur; + continue; + } + + offset = word[0] == '-'; + if (!path_is_absolute(word + offset)) { + log_syntax(unit, LOG_ERR, filename, line, 0, "Not an absolute path, ignoring: %s", word); + prev = cur; + continue; + } + + path_kill_slashes(word + offset); + + r = strv_push(sv, word); + if (r < 0) + return log_oom(); + + prev = cur; + word = NULL; + } + + return 0; +} + +int config_parse_no_new_privileges( + const char* unit, + const char *filename, + unsigned line, + const char *section, + unsigned section_line, + const char *lvalue, + int ltype, + const char *rvalue, + void *data, + void *userdata) { + + ExecContext *c = data; + int k; + + assert(filename); + assert(lvalue); + assert(rvalue); + assert(data); + + k = parse_boolean(rvalue); + if (k < 0) { + log_syntax(unit, LOG_ERR, filename, line, k, "Failed to parse boolean value, ignoring: %s", rvalue); + return 0; + } + + c->no_new_privileges = !!k; + c->no_new_privileges_set = true; + + return 0; +} + +int config_parse_protect_home( + const char* unit, + const char *filename, + unsigned line, + const char *section, + unsigned section_line, + const char *lvalue, + int ltype, + const char *rvalue, + void *data, + void *userdata) { + + ExecContext *c = data; + int k; + + assert(filename); + assert(lvalue); + assert(rvalue); + assert(data); + + /* Our enum shall be a superset of booleans, hence first try + * to parse as boolean, and then as enum */ + + k = parse_boolean(rvalue); + if (k > 0) + c->protect_home = PROTECT_HOME_YES; + else if (k == 0) + c->protect_home = PROTECT_HOME_NO; + else { + ProtectHome h; + + h = protect_home_from_string(rvalue); + if (h < 0) { + log_syntax(unit, LOG_ERR, filename, line, 0, "Failed to parse protect home value, ignoring: %s", rvalue); + return 0; + } + + c->protect_home = h; + } + + return 0; +} + +int config_parse_protect_system( + const char* unit, + const char *filename, + unsigned line, + const char *section, + unsigned section_line, + const char *lvalue, + int ltype, + const char *rvalue, + void *data, + void *userdata) { + + ExecContext *c = data; + int k; + + assert(filename); + assert(lvalue); + assert(rvalue); + assert(data); + + /* Our enum shall be a superset of booleans, hence first try + * to parse as boolean, and then as enum */ + + k = parse_boolean(rvalue); + if (k > 0) + c->protect_system = PROTECT_SYSTEM_YES; + else if (k == 0) + c->protect_system = PROTECT_SYSTEM_NO; + else { + ProtectSystem s; + + s = protect_system_from_string(rvalue); + if (s < 0) { + log_syntax(unit, LOG_ERR, filename, line, 0, "Failed to parse protect system value, ignoring: %s", rvalue); + return 0; + } + + c->protect_system = s; + } + + return 0; +} + +#define FOLLOW_MAX 8 + +static int open_follow(char **filename, FILE **_f, Set *names, char **_final) { + char *id = NULL; + unsigned c = 0; + int fd, r; + FILE *f; + + assert(filename); + assert(*filename); + assert(_f); + assert(names); + + /* This will update the filename pointer if the loaded file is + * reached by a symlink. The old string will be freed. */ + + for (;;) { + char *target, *name; + + if (c++ >= FOLLOW_MAX) + return -ELOOP; + + path_kill_slashes(*filename); + + /* Add the file name we are currently looking at to + * the names of this unit, but only if it is a valid + * unit name. */ + name = basename(*filename); + if (unit_name_is_valid(name, UNIT_NAME_ANY)) { + + id = set_get(names, name); + if (!id) { + id = strdup(name); + if (!id) + return -ENOMEM; + + r = set_consume(names, id); + if (r < 0) + return r; + } + } + + /* Try to open the file name, but don't if its a symlink */ + fd = open(*filename, O_RDONLY|O_CLOEXEC|O_NOCTTY|O_NOFOLLOW); + if (fd >= 0) + break; + + if (errno != ELOOP) + return -errno; + + /* Hmm, so this is a symlink. Let's read the name, and follow it manually */ + r = readlink_and_make_absolute(*filename, &target); + if (r < 0) + return r; + + free(*filename); + *filename = target; + } + + f = fdopen(fd, "re"); + if (!f) { + safe_close(fd); + return -errno; + } + + *_f = f; + *_final = id; + + return 0; +} + +static int merge_by_names(Unit **u, Set *names, const char *id) { + char *k; + int r; + + assert(u); + assert(*u); + assert(names); + + /* Let's try to add in all symlink names we found */ + while ((k = set_steal_first(names))) { + + /* First try to merge in the other name into our + * unit */ + r = unit_merge_by_name(*u, k); + if (r < 0) { + Unit *other; + + /* Hmm, we couldn't merge the other unit into + * ours? Then let's try it the other way + * round */ + + /* If the symlink name we are looking at is unit template, then + we must search for instance of this template */ + if (unit_name_is_valid(k, UNIT_NAME_TEMPLATE) && (*u)->instance) { + _cleanup_free_ char *instance = NULL; + + r = unit_name_replace_instance(k, (*u)->instance, &instance); + if (r < 0) + return r; + + other = manager_get_unit((*u)->manager, instance); + } else + other = manager_get_unit((*u)->manager, k); + + free(k); + + if (other) { + r = unit_merge(other, *u); + if (r >= 0) { + *u = other; + return merge_by_names(u, names, NULL); + } + } + + return r; + } + + if (id == k) + unit_choose_id(*u, id); + + free(k); + } + + return 0; +} + +static int load_from_path(Unit *u, const char *path) { + _cleanup_set_free_free_ Set *symlink_names = NULL; + _cleanup_fclose_ FILE *f = NULL; + _cleanup_free_ char *filename = NULL; + char *id = NULL; + Unit *merged; + struct stat st; + int r; + + assert(u); + assert(path); + + symlink_names = set_new(&string_hash_ops); + if (!symlink_names) + return -ENOMEM; + + if (path_is_absolute(path)) { + + filename = strdup(path); + if (!filename) + return -ENOMEM; + + r = open_follow(&filename, &f, symlink_names, &id); + if (r < 0) { + filename = mfree(filename); + if (r != -ENOENT) + return r; + } + + } else { + char **p; + + STRV_FOREACH(p, u->manager->lookup_paths.search_path) { + + /* Instead of opening the path right away, we manually + * follow all symlinks and add their name to our unit + * name set while doing so */ + filename = path_make_absolute(path, *p); + if (!filename) + return -ENOMEM; + + if (u->manager->unit_path_cache && + !set_get(u->manager->unit_path_cache, filename)) + r = -ENOENT; + else + r = open_follow(&filename, &f, symlink_names, &id); + if (r >= 0) + break; + filename = mfree(filename); + + /* ENOENT means that the file is missing or is a dangling symlink. + * ENOTDIR means that one of paths we expect to be is a directory + * is not a directory, we should just ignore that. + * EACCES means that the directory or file permissions are wrong. + */ + if (r == -EACCES) + log_debug_errno(r, "Cannot access \"%s\": %m", filename); + else if (!IN_SET(r, -ENOENT, -ENOTDIR)) + return r; + + /* Empty the symlink names for the next run */ + set_clear_free(symlink_names); + } + } + + if (!filename) + /* Hmm, no suitable file found? */ + return 0; + + if (!unit_type_may_alias(u->type) && set_size(symlink_names) > 1) { + log_unit_warning(u, "Unit type of %s does not support alias names, refusing loading via symlink.", u->id); + return -ELOOP; + } + + merged = u; + r = merge_by_names(&merged, symlink_names, id); + if (r < 0) + return r; + + if (merged != u) { + u->load_state = UNIT_MERGED; + return 0; + } + + if (fstat(fileno(f), &st) < 0) + return -errno; + + if (null_or_empty(&st)) { + u->load_state = UNIT_MASKED; + u->fragment_mtime = 0; + } else { + u->load_state = UNIT_LOADED; + u->fragment_mtime = timespec_load(&st.st_mtim); + + /* Now, parse the file contents */ + r = config_parse(u->id, filename, f, + UNIT_VTABLE(u)->sections, + config_item_perf_lookup, load_fragment_gperf_lookup, + false, true, false, u); + if (r < 0) + return r; + } + + free(u->fragment_path); + u->fragment_path = filename; + filename = NULL; + + if (u->source_path) { + if (stat(u->source_path, &st) >= 0) + u->source_mtime = timespec_load(&st.st_mtim); + else + u->source_mtime = 0; + } + + return 0; +} + +int unit_load_fragment(Unit *u) { + int r; + Iterator i; + const char *t; + + assert(u); + assert(u->load_state == UNIT_STUB); + assert(u->id); + + if (u->transient) { + u->load_state = UNIT_LOADED; + return 0; + } + + /* First, try to find the unit under its id. We always look + * for unit files in the default directories, to make it easy + * to override things by placing things in /etc/systemd/system */ + r = load_from_path(u, u->id); + if (r < 0) + return r; + + /* Try to find an alias we can load this with */ + if (u->load_state == UNIT_STUB) { + SET_FOREACH(t, u->names, i) { + + if (t == u->id) + continue; + + r = load_from_path(u, t); + if (r < 0) + return r; + + if (u->load_state != UNIT_STUB) + break; + } + } + + /* And now, try looking for it under the suggested (originally linked) path */ + if (u->load_state == UNIT_STUB && u->fragment_path) { + + r = load_from_path(u, u->fragment_path); + if (r < 0) + return r; + + if (u->load_state == UNIT_STUB) + /* Hmm, this didn't work? Then let's get rid + * of the fragment path stored for us, so that + * we don't point to an invalid location. */ + u->fragment_path = mfree(u->fragment_path); + } + + /* Look for a template */ + if (u->load_state == UNIT_STUB && u->instance) { + _cleanup_free_ char *k = NULL; + + r = unit_name_template(u->id, &k); + if (r < 0) + return r; + + r = load_from_path(u, k); + if (r < 0) + return r; + + if (u->load_state == UNIT_STUB) { + SET_FOREACH(t, u->names, i) { + _cleanup_free_ char *z = NULL; + + if (t == u->id) + continue; + + r = unit_name_template(t, &z); + if (r < 0) + return r; + + r = load_from_path(u, z); + if (r < 0) + return r; + + if (u->load_state != UNIT_STUB) + break; + } + } + } + + return 0; +} + +void unit_dump_config_items(FILE *f) { + static const struct { + const ConfigParserCallback callback; + const char *rvalue; + } table[] = { +#if !defined(HAVE_SYSV_COMPAT) || !defined(HAVE_SECCOMP) || !defined(HAVE_PAM) || !defined(HAVE_SELINUX) || !defined(HAVE_SMACK) || !defined(HAVE_APPARMOR) + { config_parse_warn_compat, "NOTSUPPORTED" }, +#endif + { config_parse_int, "INTEGER" }, + { config_parse_unsigned, "UNSIGNED" }, + { config_parse_iec_size, "SIZE" }, + { config_parse_iec_uint64, "SIZE" }, + { config_parse_si_size, "SIZE" }, + { config_parse_bool, "BOOLEAN" }, + { config_parse_string, "STRING" }, + { config_parse_path, "PATH" }, + { config_parse_unit_path_printf, "PATH" }, + { config_parse_strv, "STRING [...]" }, + { config_parse_exec_nice, "NICE" }, + { config_parse_exec_oom_score_adjust, "OOMSCOREADJUST" }, + { config_parse_exec_io_class, "IOCLASS" }, + { config_parse_exec_io_priority, "IOPRIORITY" }, + { config_parse_exec_cpu_sched_policy, "CPUSCHEDPOLICY" }, + { config_parse_exec_cpu_sched_prio, "CPUSCHEDPRIO" }, + { config_parse_exec_cpu_affinity, "CPUAFFINITY" }, + { config_parse_mode, "MODE" }, + { config_parse_unit_env_file, "FILE" }, + { config_parse_output, "OUTPUT" }, + { config_parse_input, "INPUT" }, + { config_parse_log_facility, "FACILITY" }, + { config_parse_log_level, "LEVEL" }, + { config_parse_exec_secure_bits, "SECUREBITS" }, + { config_parse_capability_set, "BOUNDINGSET" }, + { config_parse_limit, "LIMIT" }, + { config_parse_unit_deps, "UNIT [...]" }, + { config_parse_exec, "PATH [ARGUMENT [...]]" }, + { config_parse_service_type, "SERVICETYPE" }, + { config_parse_service_restart, "SERVICERESTART" }, +#ifdef HAVE_SYSV_COMPAT + { config_parse_sysv_priority, "SYSVPRIORITY" }, +#endif + { config_parse_kill_mode, "KILLMODE" }, + { config_parse_signal, "SIGNAL" }, + { config_parse_socket_listen, "SOCKET [...]" }, + { config_parse_socket_bind, "SOCKETBIND" }, + { config_parse_socket_bindtodevice, "NETWORKINTERFACE" }, + { config_parse_sec, "SECONDS" }, + { config_parse_nsec, "NANOSECONDS" }, + { config_parse_namespace_path_strv, "PATH [...]" }, + { config_parse_unit_requires_mounts_for, "PATH [...]" }, + { config_parse_exec_mount_flags, "MOUNTFLAG [...]" }, + { config_parse_unit_string_printf, "STRING" }, + { config_parse_trigger_unit, "UNIT" }, + { config_parse_timer, "TIMER" }, + { config_parse_path_spec, "PATH" }, + { config_parse_notify_access, "ACCESS" }, + { config_parse_ip_tos, "TOS" }, + { config_parse_unit_condition_path, "CONDITION" }, + { config_parse_unit_condition_string, "CONDITION" }, + { config_parse_unit_condition_null, "CONDITION" }, + { config_parse_unit_slice, "SLICE" }, + { config_parse_documentation, "URL" }, + { config_parse_service_timeout, "SECONDS" }, + { config_parse_failure_action, "ACTION" }, + { config_parse_set_status, "STATUS" }, + { config_parse_service_sockets, "SOCKETS" }, + { config_parse_environ, "ENVIRON" }, +#ifdef HAVE_SECCOMP + { config_parse_syscall_filter, "SYSCALLS" }, + { config_parse_syscall_archs, "ARCHS" }, + { config_parse_syscall_errno, "ERRNO" }, + { config_parse_address_families, "FAMILIES" }, +#endif + { config_parse_cpu_shares, "SHARES" }, + { config_parse_memory_limit, "LIMIT" }, + { config_parse_device_allow, "DEVICE" }, + { config_parse_device_policy, "POLICY" }, + { config_parse_io_limit, "LIMIT" }, + { config_parse_io_weight, "WEIGHT" }, + { config_parse_io_device_weight, "DEVICEWEIGHT" }, + { config_parse_blockio_bandwidth, "BANDWIDTH" }, + { config_parse_blockio_weight, "WEIGHT" }, + { config_parse_blockio_device_weight, "DEVICEWEIGHT" }, + { config_parse_long, "LONG" }, + { config_parse_socket_service, "SERVICE" }, +#ifdef HAVE_SELINUX + { config_parse_exec_selinux_context, "LABEL" }, +#endif + { config_parse_job_mode, "MODE" }, + { config_parse_job_mode_isolate, "BOOLEAN" }, + { config_parse_personality, "PERSONALITY" }, + }; + + const char *prev = NULL; + const char *i; + + assert(f); + + NULSTR_FOREACH(i, load_fragment_gperf_nulstr) { + const char *rvalue = "OTHER", *lvalue; + unsigned j; + size_t prefix_len; + const char *dot; + const ConfigPerfItem *p; + + assert_se(p = load_fragment_gperf_lookup(i, strlen(i))); + + dot = strchr(i, '.'); + lvalue = dot ? dot + 1 : i; + prefix_len = dot-i; + + if (dot) + if (!prev || !strneq(prev, i, prefix_len+1)) { + if (prev) + fputc('\n', f); + + fprintf(f, "[%.*s]\n", (int) prefix_len, i); + } + + for (j = 0; j < ELEMENTSOF(table); j++) + if (p->parse == table[j].callback) { + rvalue = table[j].rvalue; + break; + } + + fprintf(f, "%s=%s\n", lvalue, rvalue); + prev = i; + } +} diff --git a/src/grp-system/libcore/src/locale-setup.c b/src/grp-system/libcore/src/locale-setup.c new file mode 100644 index 0000000000..1794b474b8 --- /dev/null +++ b/src/grp-system/libcore/src/locale-setup.c @@ -0,0 +1,125 @@ +/*** + This file is part of systemd. + + Copyright 2010 Lennart Poettering + + systemd is free software; you can redistribute it and/or modify it + under the terms of the GNU Lesser General Public License as published by + the Free Software Foundation; either version 2.1 of the License, or + (at your option) any later version. + + systemd is distributed in the hope that it will be useful, but + WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + Lesser General Public License for more details. + + You should have received a copy of the GNU Lesser General Public License + along with systemd; If not, see . +***/ + +#include +#include + +#include "systemd-basic/env-util.h" +#include "systemd-basic/fileio.h" +#include "systemd-basic/locale-util.h" +#include "systemd-basic/string-util.h" +#include "systemd-basic/strv.h" +#include "systemd-basic/util.h" +#include "systemd-basic/virt.h" + +#include "locale-setup.h" + +int locale_setup(char ***environment) { + char **add; + char *variables[_VARIABLE_LC_MAX] = {}; + int r = 0, i; + + if (detect_container() <= 0) { + r = parse_env_file("/proc/cmdline", WHITESPACE, + "locale.LANG", &variables[VARIABLE_LANG], + "locale.LANGUAGE", &variables[VARIABLE_LANGUAGE], + "locale.LC_CTYPE", &variables[VARIABLE_LC_CTYPE], + "locale.LC_NUMERIC", &variables[VARIABLE_LC_NUMERIC], + "locale.LC_TIME", &variables[VARIABLE_LC_TIME], + "locale.LC_COLLATE", &variables[VARIABLE_LC_COLLATE], + "locale.LC_MONETARY", &variables[VARIABLE_LC_MONETARY], + "locale.LC_MESSAGES", &variables[VARIABLE_LC_MESSAGES], + "locale.LC_PAPER", &variables[VARIABLE_LC_PAPER], + "locale.LC_NAME", &variables[VARIABLE_LC_NAME], + "locale.LC_ADDRESS", &variables[VARIABLE_LC_ADDRESS], + "locale.LC_TELEPHONE", &variables[VARIABLE_LC_TELEPHONE], + "locale.LC_MEASUREMENT", &variables[VARIABLE_LC_MEASUREMENT], + "locale.LC_IDENTIFICATION", &variables[VARIABLE_LC_IDENTIFICATION], + NULL); + + if (r < 0 && r != -ENOENT) + log_warning_errno(r, "Failed to read /proc/cmdline: %m"); + } + + /* Hmm, nothing set on the kernel cmd line? Then let's + * try /etc/locale.conf */ + if (r <= 0) { + r = parse_env_file("/etc/locale.conf", NEWLINE, + "LANG", &variables[VARIABLE_LANG], + "LANGUAGE", &variables[VARIABLE_LANGUAGE], + "LC_CTYPE", &variables[VARIABLE_LC_CTYPE], + "LC_NUMERIC", &variables[VARIABLE_LC_NUMERIC], + "LC_TIME", &variables[VARIABLE_LC_TIME], + "LC_COLLATE", &variables[VARIABLE_LC_COLLATE], + "LC_MONETARY", &variables[VARIABLE_LC_MONETARY], + "LC_MESSAGES", &variables[VARIABLE_LC_MESSAGES], + "LC_PAPER", &variables[VARIABLE_LC_PAPER], + "LC_NAME", &variables[VARIABLE_LC_NAME], + "LC_ADDRESS", &variables[VARIABLE_LC_ADDRESS], + "LC_TELEPHONE", &variables[VARIABLE_LC_TELEPHONE], + "LC_MEASUREMENT", &variables[VARIABLE_LC_MEASUREMENT], + "LC_IDENTIFICATION", &variables[VARIABLE_LC_IDENTIFICATION], + NULL); + + if (r < 0 && r != -ENOENT) + log_warning_errno(r, "Failed to read /etc/locale.conf: %m"); + } + + add = NULL; + for (i = 0; i < _VARIABLE_LC_MAX; i++) { + char *s; + + if (!variables[i]) + continue; + + s = strjoin(locale_variable_to_string(i), "=", variables[i], NULL); + if (!s) { + r = -ENOMEM; + goto finish; + } + + if (strv_consume(&add, s) < 0) { + r = -ENOMEM; + goto finish; + } + } + + if (!strv_isempty(add)) { + char **e; + + e = strv_env_merge(2, *environment, add); + if (!e) { + r = -ENOMEM; + goto finish; + } + + strv_free(*environment); + *environment = e; + } + + r = 0; + +finish: + strv_free(add); + + for (i = 0; i < _VARIABLE_LC_MAX; i++) + free(variables[i]); + + return r; +} diff --git a/src/grp-system/libcore/src/locale-setup.h b/src/grp-system/libcore/src/locale-setup.h new file mode 100644 index 0000000000..3b97497afe --- /dev/null +++ b/src/grp-system/libcore/src/locale-setup.h @@ -0,0 +1,22 @@ +#pragma once + +/*** + This file is part of systemd. + + Copyright 2010 Lennart Poettering + + systemd is free software; you can redistribute it and/or modify it + under the terms of the GNU Lesser General Public License as published by + the Free Software Foundation; either version 2.1 of the License, or + (at your option) any later version. + + systemd is distributed in the hope that it will be useful, but + WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + Lesser General Public License for more details. + + You should have received a copy of the GNU Lesser General Public License + along with systemd; If not, see . +***/ + +int locale_setup(char ***environment); diff --git a/src/grp-system/libcore/src/loopback-setup.c b/src/grp-system/libcore/src/loopback-setup.c new file mode 100644 index 0000000000..8eb6251bd7 --- /dev/null +++ b/src/grp-system/libcore/src/loopback-setup.c @@ -0,0 +1,89 @@ +/*** + This file is part of systemd. + + Copyright 2010 Lennart Poettering + + systemd is free software; you can redistribute it and/or modify it + under the terms of the GNU Lesser General Public License as published by + the Free Software Foundation; either version 2.1 of the License, or + (at your option) any later version. + + systemd is distributed in the hope that it will be useful, but + WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + Lesser General Public License for more details. + + You should have received a copy of the GNU Lesser General Public License + along with systemd; If not, see . +***/ + +#include +#include + +#include "core/loopback-setup.h" +#include "sd-netlink/netlink-util.h" +#include "systemd-basic/missing.h" +#include "systemd-staging/sd-netlink.h" + +static int start_loopback(sd_netlink *rtnl) { + _cleanup_(sd_netlink_message_unrefp) sd_netlink_message *req = NULL; + int r; + + r = sd_rtnl_message_new_link(rtnl, &req, RTM_SETLINK, LOOPBACK_IFINDEX); + if (r < 0) + return r; + + r = sd_rtnl_message_link_set_flags(req, IFF_UP, IFF_UP); + if (r < 0) + return r; + + r = sd_netlink_call(rtnl, req, 0, NULL); + if (r < 0) + return r; + + return 0; +} + +static bool check_loopback(sd_netlink *rtnl) { + _cleanup_(sd_netlink_message_unrefp) sd_netlink_message *req = NULL, *reply = NULL; + unsigned flags; + int r; + + r = sd_rtnl_message_new_link(rtnl, &req, RTM_GETLINK, LOOPBACK_IFINDEX); + if (r < 0) + return false; + + r = sd_netlink_call(rtnl, req, 0, &reply); + if (r < 0) + return false; + + r = sd_rtnl_message_link_get_flags(reply, &flags); + if (r < 0) + return false; + + return flags & IFF_UP; +} + +int loopback_setup(void) { + _cleanup_(sd_netlink_unrefp) sd_netlink *rtnl = NULL; + int r; + + r = sd_netlink_open(&rtnl); + if (r < 0) + return r; + + r = start_loopback(rtnl); + if (r < 0) { + + /* If we lack the permissions to configure the + * loopback device, but we find it to be already + * configured, let's exit cleanly, in order to + * supported unprivileged containers. */ + if (r == -EPERM && check_loopback(rtnl)) + return 0; + + return log_warning_errno(r, "Failed to configure loopback device: %m"); + } + + return 0; +} diff --git a/src/grp-system/libcore/src/machine-id-setup.c b/src/grp-system/libcore/src/machine-id-setup.c new file mode 100644 index 0000000000..a8c6ae046c --- /dev/null +++ b/src/grp-system/libcore/src/machine-id-setup.c @@ -0,0 +1,258 @@ +/*** + This file is part of systemd. + + Copyright 2010 Lennart Poettering + + systemd is free software; you can redistribute it and/or modify it + under the terms of the GNU Lesser General Public License as published by + the Free Software Foundation; either version 2.1 of the License, or + (at your option) any later version. + + systemd is distributed in the hope that it will be useful, but + WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + Lesser General Public License for more details. + + You should have received a copy of the GNU Lesser General Public License + along with systemd; If not, see . +***/ + +#include +#include +#include +#include + +#include + +#include "core/machine-id-setup.h" +#include "sd-id128/id128-util.h" +#include "systemd-basic/alloc-util.h" +#include "systemd-basic/fd-util.h" +#include "systemd-basic/fs-util.h" +#include "systemd-basic/log.h" +#include "systemd-basic/macro.h" +#include "systemd-basic/mkdir.h" +#include "systemd-basic/mount-util.h" +#include "systemd-basic/path-util.h" +#include "systemd-basic/process-util.h" +#include "systemd-basic/stat-util.h" +#include "systemd-basic/string-util.h" +#include "systemd-basic/umask-util.h" +#include "systemd-basic/util.h" +#include "systemd-basic/virt.h" + +static int generate_machine_id(const char *root, sd_id128_t *ret) { + const char *dbus_machine_id; + _cleanup_close_ int fd = -1; + int r; + + assert(ret); + + /* First, try reading the D-Bus machine id, unless it is a symlink */ + dbus_machine_id = prefix_roota(root, "/var/lib/dbus/machine-id"); + fd = open(dbus_machine_id, O_RDONLY|O_CLOEXEC|O_NOCTTY|O_NOFOLLOW); + if (fd >= 0) { + if (id128_read_fd(fd, ID128_PLAIN, ret) >= 0) { + log_info("Initializing machine ID from D-Bus machine ID."); + return 0; + } + + fd = safe_close(fd); + } + + if (isempty(root)) { + /* If that didn't work, see if we are running in a container, + * and a machine ID was passed in via $container_uuid the way + * libvirt/LXC does it */ + + if (detect_container() > 0) { + _cleanup_free_ char *e = NULL; + + if (getenv_for_pid(1, "container_uuid", &e) > 0 && + sd_id128_from_string(e, ret) >= 0) { + log_info("Initializing machine ID from container UUID."); + return 0; + } + + } else if (detect_vm() == VIRTUALIZATION_KVM) { + + /* If we are not running in a container, see if we are + * running in qemu/kvm and a machine ID was passed in + * via -uuid on the qemu/kvm command line */ + + if (id128_read("/sys/class/dmi/id/product_uuid", ID128_UUID, ret) >= 0) { + log_info("Initializing machine ID from KVM UUID."); + return 0; + } + } + } + + /* If that didn't work, generate a random machine id */ + r = sd_id128_randomize(ret); + if (r < 0) + return log_error_errno(r, "Failed to generate randomized : %m"); + + log_info("Initializing machine ID from random generator."); + return 0; +} + +int machine_id_setup(const char *root, sd_id128_t machine_id, sd_id128_t *ret) { + const char *etc_machine_id, *run_machine_id; + _cleanup_close_ int fd = -1; + bool writable; + int r; + + etc_machine_id = prefix_roota(root, "/etc/machine-id"); + + RUN_WITH_UMASK(0000) { + /* We create this 0444, to indicate that this isn't really + * something you should ever modify. Of course, since the file + * will be owned by root it doesn't matter much, but maybe + * people look. */ + + (void) mkdir_parents(etc_machine_id, 0755); + fd = open(etc_machine_id, O_RDWR|O_CREAT|O_CLOEXEC|O_NOCTTY, 0444); + if (fd < 0) { + int old_errno = errno; + + fd = open(etc_machine_id, O_RDONLY|O_CLOEXEC|O_NOCTTY); + if (fd < 0) { + if (old_errno == EROFS && errno == ENOENT) + log_error_errno(errno, + "System cannot boot: Missing /etc/machine-id and /etc is mounted read-only.\n" + "Booting up is supported only when:\n" + "1) /etc/machine-id exists and is populated.\n" + "2) /etc/machine-id exists and is empty.\n" + "3) /etc/machine-id is missing and /etc is writable.\n"); + else + log_error_errno(errno, "Cannot open %s: %m", etc_machine_id); + + return -errno; + } + + writable = false; + } else + writable = true; + } + + /* A we got a valid machine ID argument, that's what counts */ + if (sd_id128_is_null(machine_id)) { + + /* Try to read any existing machine ID */ + if (id128_read_fd(fd, ID128_PLAIN, ret) >= 0) + return 0; + + /* Hmm, so, the id currently stored is not useful, then let's generate one */ + r = generate_machine_id(root, &machine_id); + if (r < 0) + return r; + + if (lseek(fd, 0, SEEK_SET) == (off_t) -1) + return log_error_errno(errno, "Failed to seek: %m"); + } + + if (writable) + if (id128_write_fd(fd, ID128_PLAIN, machine_id, true) >= 0) + goto finish; + + fd = safe_close(fd); + + /* Hmm, we couldn't write it? So let's write it to /run/machine-id as a replacement */ + + run_machine_id = prefix_roota(root, "/run/machine-id"); + + RUN_WITH_UMASK(0022) + r = id128_write(run_machine_id, ID128_PLAIN, machine_id, false); + if (r < 0) { + (void) unlink(run_machine_id); + return log_error_errno(r, "Cannot write %s: %m", run_machine_id); + } + + /* And now, let's mount it over */ + if (mount(run_machine_id, etc_machine_id, NULL, MS_BIND, NULL) < 0) { + (void) unlink_noerrno(run_machine_id); + return log_error_errno(errno, "Failed to mount %s: %m", etc_machine_id); + } + + log_info("Installed transient %s file.", etc_machine_id); + + /* Mark the mount read-only */ + if (mount(NULL, etc_machine_id, NULL, MS_BIND|MS_RDONLY|MS_REMOUNT, NULL) < 0) + log_warning_errno(errno, "Failed to make transient %s read-only, ignoring: %m", etc_machine_id); + +finish: + if (ret) + *ret = machine_id; + + return 0; +} + +int machine_id_commit(const char *root) { + _cleanup_close_ int fd = -1, initial_mntns_fd = -1; + const char *etc_machine_id; + sd_id128_t id; + int r; + + /* Replaces a tmpfs bind mount of /etc/machine-id by a proper file, atomically. For this, the umount is removed + * in a mount namespace, a new file is created at the right place. Afterwards the mount is also removed in the + * original mount namespace, thus revealing the file that was just created. */ + + etc_machine_id = prefix_roota(root, "/etc/machine-id"); + + r = path_is_mount_point(etc_machine_id, 0); + if (r < 0) + return log_error_errno(r, "Failed to determine whether %s is a mount point: %m", etc_machine_id); + if (r == 0) { + log_debug("%s is not a mount point. Nothing to do.", etc_machine_id); + return 0; + } + + /* Read existing machine-id */ + fd = open(etc_machine_id, O_RDONLY|O_CLOEXEC|O_NOCTTY); + if (fd < 0) + return log_error_errno(errno, "Cannot open %s: %m", etc_machine_id); + + r = fd_is_temporary_fs(fd); + if (r < 0) + return log_error_errno(r, "Failed to determine whether %s is on a temporary file system: %m", etc_machine_id); + if (r == 0) { + log_error("%s is not on a temporary file system.", etc_machine_id); + return -EROFS; + } + + r = id128_read_fd(fd, ID128_PLAIN, &id); + if (r < 0) + return log_error_errno(r, "We didn't find a valid machine ID in %s.", etc_machine_id); + + fd = safe_close(fd); + + /* Store current mount namespace */ + r = namespace_open(0, NULL, &initial_mntns_fd, NULL, NULL, NULL); + if (r < 0) + return log_error_errno(r, "Can't fetch current mount namespace: %m"); + + /* Switch to a new mount namespace, isolate ourself and unmount etc_machine_id in our new namespace */ + if (unshare(CLONE_NEWNS) < 0) + return log_error_errno(errno, "Failed to enter new namespace: %m"); + + if (mount(NULL, "/", NULL, MS_SLAVE | MS_REC, NULL) < 0) + return log_error_errno(errno, "Couldn't make-rslave / mountpoint in our private namespace: %m"); + + if (umount(etc_machine_id) < 0) + return log_error_errno(errno, "Failed to unmount transient %s file in our private namespace: %m", etc_machine_id); + + /* Update a persistent version of etc_machine_id */ + r = id128_write(etc_machine_id, ID128_PLAIN, id, true); + if (r < 0) + return log_error_errno(r, "Cannot write %s. This is mandatory to get a persistent machine ID: %m", etc_machine_id); + + /* Return to initial namespace and proceed a lazy tmpfs unmount */ + r = namespace_enter(-1, initial_mntns_fd, -1, -1, -1); + if (r < 0) + return log_warning_errno(r, "Failed to switch back to initial mount namespace: %m.\nWe'll keep transient %s file until next reboot.", etc_machine_id); + + if (umount2(etc_machine_id, MNT_DETACH) < 0) + return log_warning_errno(errno, "Failed to unmount transient %s file: %m.\nWe keep that mount until next reboot.", etc_machine_id); + + return 0; +} diff --git a/src/grp-system/libcore/src/manager.c b/src/grp-system/libcore/src/manager.c new file mode 100644 index 0000000000..b769c33b20 --- /dev/null +++ b/src/grp-system/libcore/src/manager.c @@ -0,0 +1,3149 @@ +/*** + This file is part of systemd. + + Copyright 2010 Lennart Poettering + + systemd is free software; you can redistribute it and/or modify it + under the terms of the GNU Lesser General Public License as published by + the Free Software Foundation; either version 2.1 of the License, or + (at your option) any later version. + + systemd is distributed in the hope that it will be useful, but + WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + Lesser General Public License for more details. + + You should have received a copy of the GNU Lesser General Public License + along with systemd; If not, see . +***/ + +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include + +#include + +#ifdef HAVE_AUDIT +#include +#endif + +#include +#include + +#include "core/dbus-manager.h" +#include "core/manager.h" +#include "sd-bus/bus-common-errors.h" +#include "sd-bus/bus-error.h" +#include "sd-bus/bus-kernel.h" +#include "sd-bus/bus-util.h" +#include "systemd-basic/alloc-util.h" +#include "systemd-basic/dirent-util.h" +#include "systemd-basic/env-util.h" +#include "systemd-basic/escape.h" +#include "systemd-basic/exit-status.h" +#include "systemd-basic/fd-util.h" +#include "systemd-basic/fileio.h" +#include "systemd-basic/fs-util.h" +#include "systemd-basic/hashmap.h" +#include "systemd-basic/io-util.h" +#include "systemd-basic/log.h" +#include "systemd-basic/macro.h" +#include "systemd-basic/missing.h" +#include "systemd-basic/mkdir.h" +#include "systemd-basic/parse-util.h" +#include "systemd-basic/path-util.h" +#include "systemd-basic/process-util.h" +#include "systemd-basic/ratelimit.h" +#include "systemd-basic/rm-rf.h" +#include "systemd-basic/signal-util.h" +#include "systemd-basic/special.h" +#include "systemd-basic/stat-util.h" +#include "systemd-basic/string-table.h" +#include "systemd-basic/string-util.h" +#include "systemd-basic/strv.h" +#include "systemd-basic/terminal-util.h" +#include "systemd-basic/time-util.h" +#include "systemd-basic/umask-util.h" +#include "systemd-basic/unit-name.h" +#include "systemd-basic/util.h" +#include "systemd-basic/virt.h" +#include "systemd-shared/boot-timestamps.h" +#include "systemd-shared/path-lookup.h" +#include "systemd-shared/watchdog.h" + +#include "audit-fd.h" +#include "dbus-job.h" +#include "dbus-unit.h" +#include "dbus.h" +#include "locale-setup.h" +#include "transaction.h" + +#define NOTIFY_RCVBUF_SIZE (8*1024*1024) +#define CGROUPS_AGENT_RCVBUF_SIZE (8*1024*1024) + +/* Initial delay and the interval for printing status messages about running jobs */ +#define JOBS_IN_PROGRESS_WAIT_USEC (5*USEC_PER_SEC) +#define JOBS_IN_PROGRESS_PERIOD_USEC (USEC_PER_SEC / 3) +#define JOBS_IN_PROGRESS_PERIOD_DIVISOR 3 + +static int manager_dispatch_notify_fd(sd_event_source *source, int fd, uint32_t revents, void *userdata); +static int manager_dispatch_cgroups_agent_fd(sd_event_source *source, int fd, uint32_t revents, void *userdata); +static int manager_dispatch_signal_fd(sd_event_source *source, int fd, uint32_t revents, void *userdata); +static int manager_dispatch_time_change_fd(sd_event_source *source, int fd, uint32_t revents, void *userdata); +static int manager_dispatch_idle_pipe_fd(sd_event_source *source, int fd, uint32_t revents, void *userdata); +static int manager_dispatch_jobs_in_progress(sd_event_source *source, usec_t usec, void *userdata); +static int manager_dispatch_run_queue(sd_event_source *source, void *userdata); +static int manager_run_generators(Manager *m); + +static void manager_watch_jobs_in_progress(Manager *m) { + usec_t next; + int r; + + assert(m); + + if (m->jobs_in_progress_event_source) + return; + + next = now(CLOCK_MONOTONIC) + JOBS_IN_PROGRESS_WAIT_USEC; + r = sd_event_add_time( + m->event, + &m->jobs_in_progress_event_source, + CLOCK_MONOTONIC, + next, 0, + manager_dispatch_jobs_in_progress, m); + if (r < 0) + return; + + (void) sd_event_source_set_description(m->jobs_in_progress_event_source, "manager-jobs-in-progress"); +} + +#define CYLON_BUFFER_EXTRA (2*(sizeof(ANSI_RED)-1) + sizeof(ANSI_HIGHLIGHT_RED)-1 + 2*(sizeof(ANSI_NORMAL)-1)) + +static void draw_cylon(char buffer[], size_t buflen, unsigned width, unsigned pos) { + char *p = buffer; + + assert(buflen >= CYLON_BUFFER_EXTRA + width + 1); + assert(pos <= width+1); /* 0 or width+1 mean that the center light is behind the corner */ + + if (pos > 1) { + if (pos > 2) + p = mempset(p, ' ', pos-2); + if (log_get_show_color()) + p = stpcpy(p, ANSI_RED); + *p++ = '*'; + } + + if (pos > 0 && pos <= width) { + if (log_get_show_color()) + p = stpcpy(p, ANSI_HIGHLIGHT_RED); + *p++ = '*'; + } + + if (log_get_show_color()) + p = stpcpy(p, ANSI_NORMAL); + + if (pos < width) { + if (log_get_show_color()) + p = stpcpy(p, ANSI_RED); + *p++ = '*'; + if (pos < width-1) + p = mempset(p, ' ', width-1-pos); + if (log_get_show_color()) + strcpy(p, ANSI_NORMAL); + } +} + +void manager_flip_auto_status(Manager *m, bool enable) { + assert(m); + + if (enable) { + if (m->show_status == SHOW_STATUS_AUTO) + manager_set_show_status(m, SHOW_STATUS_TEMPORARY); + } else { + if (m->show_status == SHOW_STATUS_TEMPORARY) + manager_set_show_status(m, SHOW_STATUS_AUTO); + } +} + +static void manager_print_jobs_in_progress(Manager *m) { + _cleanup_free_ char *job_of_n = NULL; + Iterator i; + Job *j; + unsigned counter = 0, print_nr; + char cylon[6 + CYLON_BUFFER_EXTRA + 1]; + unsigned cylon_pos; + char time[FORMAT_TIMESPAN_MAX], limit[FORMAT_TIMESPAN_MAX] = "no limit"; + uint64_t x; + + assert(m); + assert(m->n_running_jobs > 0); + + manager_flip_auto_status(m, true); + + print_nr = (m->jobs_in_progress_iteration / JOBS_IN_PROGRESS_PERIOD_DIVISOR) % m->n_running_jobs; + + HASHMAP_FOREACH(j, m->jobs, i) + if (j->state == JOB_RUNNING && counter++ == print_nr) + break; + + /* m->n_running_jobs must be consistent with the contents of m->jobs, + * so the above loop must have succeeded in finding j. */ + assert(counter == print_nr + 1); + assert(j); + + cylon_pos = m->jobs_in_progress_iteration % 14; + if (cylon_pos >= 8) + cylon_pos = 14 - cylon_pos; + draw_cylon(cylon, sizeof(cylon), 6, cylon_pos); + + m->jobs_in_progress_iteration++; + + if (m->n_running_jobs > 1) { + if (asprintf(&job_of_n, "(%u of %u) ", counter, m->n_running_jobs) < 0) + job_of_n = NULL; + } + + format_timespan(time, sizeof(time), now(CLOCK_MONOTONIC) - j->begin_usec, 1*USEC_PER_SEC); + if (job_get_timeout(j, &x) > 0) + format_timespan(limit, sizeof(limit), x - j->begin_usec, 1*USEC_PER_SEC); + + manager_status_printf(m, STATUS_TYPE_EPHEMERAL, cylon, + "%sA %s job is running for %s (%s / %s)", + strempty(job_of_n), + job_type_to_string(j->type), + unit_description(j->unit), + time, limit); +} + +static int have_ask_password(void) { + _cleanup_closedir_ DIR *dir; + + dir = opendir("/run/systemd/ask-password"); + if (!dir) { + if (errno == ENOENT) + return false; + else + return -errno; + } + + for (;;) { + struct dirent *de; + + errno = 0; + de = readdir(dir); + if (!de && errno > 0) + return -errno; + if (!de) + return false; + + if (startswith(de->d_name, "ask.")) + return true; + } +} + +static int manager_dispatch_ask_password_fd(sd_event_source *source, + int fd, uint32_t revents, void *userdata) { + Manager *m = userdata; + + assert(m); + + flush_fd(fd); + + m->have_ask_password = have_ask_password(); + if (m->have_ask_password < 0) + /* Log error but continue. Negative have_ask_password + * is treated as unknown status. */ + log_error_errno(m->have_ask_password, "Failed to list /run/systemd/ask-password: %m"); + + return 0; +} + +static void manager_close_ask_password(Manager *m) { + assert(m); + + m->ask_password_event_source = sd_event_source_unref(m->ask_password_event_source); + m->ask_password_inotify_fd = safe_close(m->ask_password_inotify_fd); + m->have_ask_password = -EINVAL; +} + +static int manager_check_ask_password(Manager *m) { + int r; + + assert(m); + + if (!m->ask_password_event_source) { + assert(m->ask_password_inotify_fd < 0); + + mkdir_p_label("/run/systemd/ask-password", 0755); + + m->ask_password_inotify_fd = inotify_init1(IN_NONBLOCK|IN_CLOEXEC); + if (m->ask_password_inotify_fd < 0) + return log_error_errno(errno, "inotify_init1() failed: %m"); + + if (inotify_add_watch(m->ask_password_inotify_fd, "/run/systemd/ask-password", IN_CREATE|IN_DELETE|IN_MOVE) < 0) { + log_error_errno(errno, "Failed to add watch on /run/systemd/ask-password: %m"); + manager_close_ask_password(m); + return -errno; + } + + r = sd_event_add_io(m->event, &m->ask_password_event_source, + m->ask_password_inotify_fd, EPOLLIN, + manager_dispatch_ask_password_fd, m); + if (r < 0) { + log_error_errno(errno, "Failed to add event source for /run/systemd/ask-password: %m"); + manager_close_ask_password(m); + return -errno; + } + + (void) sd_event_source_set_description(m->ask_password_event_source, "manager-ask-password"); + + /* Queries might have been added meanwhile... */ + manager_dispatch_ask_password_fd(m->ask_password_event_source, + m->ask_password_inotify_fd, EPOLLIN, m); + } + + return m->have_ask_password; +} + +static int manager_watch_idle_pipe(Manager *m) { + int r; + + assert(m); + + if (m->idle_pipe_event_source) + return 0; + + if (m->idle_pipe[2] < 0) + return 0; + + r = sd_event_add_io(m->event, &m->idle_pipe_event_source, m->idle_pipe[2], EPOLLIN, manager_dispatch_idle_pipe_fd, m); + if (r < 0) + return log_error_errno(r, "Failed to watch idle pipe: %m"); + + (void) sd_event_source_set_description(m->idle_pipe_event_source, "manager-idle-pipe"); + + return 0; +} + +static void manager_close_idle_pipe(Manager *m) { + assert(m); + + m->idle_pipe_event_source = sd_event_source_unref(m->idle_pipe_event_source); + + safe_close_pair(m->idle_pipe); + safe_close_pair(m->idle_pipe + 2); +} + +static int manager_setup_time_change(Manager *m) { + int r; + + /* We only care for the cancellation event, hence we set the + * timeout to the latest possible value. */ + struct itimerspec its = { + .it_value.tv_sec = TIME_T_MAX, + }; + + assert(m); + assert_cc(sizeof(time_t) == sizeof(TIME_T_MAX)); + + if (m->test_run) + return 0; + + /* Uses TFD_TIMER_CANCEL_ON_SET to get notifications whenever + * CLOCK_REALTIME makes a jump relative to CLOCK_MONOTONIC */ + + m->time_change_fd = timerfd_create(CLOCK_REALTIME, TFD_NONBLOCK|TFD_CLOEXEC); + if (m->time_change_fd < 0) + return log_error_errno(errno, "Failed to create timerfd: %m"); + + if (timerfd_settime(m->time_change_fd, TFD_TIMER_ABSTIME|TFD_TIMER_CANCEL_ON_SET, &its, NULL) < 0) { + log_debug_errno(errno, "Failed to set up TFD_TIMER_CANCEL_ON_SET, ignoring: %m"); + m->time_change_fd = safe_close(m->time_change_fd); + return 0; + } + + r = sd_event_add_io(m->event, &m->time_change_event_source, m->time_change_fd, EPOLLIN, manager_dispatch_time_change_fd, m); + if (r < 0) + return log_error_errno(r, "Failed to create time change event source: %m"); + + (void) sd_event_source_set_description(m->time_change_event_source, "manager-time-change"); + + log_debug("Set up TFD_TIMER_CANCEL_ON_SET timerfd."); + + return 0; +} + +static int enable_special_signals(Manager *m) { + _cleanup_close_ int fd = -1; + + assert(m); + + if (m->test_run) + return 0; + + /* Enable that we get SIGINT on control-alt-del. In containers + * this will fail with EPERM (older) or EINVAL (newer), so + * ignore that. */ + if (reboot(RB_DISABLE_CAD) < 0 && errno != EPERM && errno != EINVAL) + log_warning_errno(errno, "Failed to enable ctrl-alt-del handling: %m"); + + fd = open_terminal("/dev/tty0", O_RDWR|O_NOCTTY|O_CLOEXEC); + if (fd < 0) { + /* Support systems without virtual console */ + if (fd != -ENOENT) + log_warning_errno(errno, "Failed to open /dev/tty0: %m"); + } else { + /* Enable that we get SIGWINCH on kbrequest */ + if (ioctl(fd, KDSIGACCEPT, SIGWINCH) < 0) + log_warning_errno(errno, "Failed to enable kbrequest handling: %m"); + } + + return 0; +} + +static int manager_setup_signals(Manager *m) { + struct sigaction sa = { + .sa_handler = SIG_DFL, + .sa_flags = SA_NOCLDSTOP|SA_RESTART, + }; + sigset_t mask; + int r; + + assert(m); + + assert_se(sigaction(SIGCHLD, &sa, NULL) == 0); + + /* We make liberal use of realtime signals here. On + * Linux/glibc we have 30 of them (with the exception of Linux + * on hppa, see below), between SIGRTMIN+0 ... SIGRTMIN+30 + * (aka SIGRTMAX). */ + + assert_se(sigemptyset(&mask) == 0); + sigset_add_many(&mask, + SIGCHLD, /* Child died */ + SIGTERM, /* Reexecute daemon */ + SIGHUP, /* Reload configuration */ + SIGUSR1, /* systemd/upstart: reconnect to D-Bus */ + SIGUSR2, /* systemd: dump status */ + SIGINT, /* Kernel sends us this on control-alt-del */ + SIGWINCH, /* Kernel sends us this on kbrequest (alt-arrowup) */ + SIGPWR, /* Some kernel drivers and upsd send us this on power failure */ + + SIGRTMIN+0, /* systemd: start default.target */ + SIGRTMIN+1, /* systemd: isolate rescue.target */ + SIGRTMIN+2, /* systemd: isolate emergency.target */ + SIGRTMIN+3, /* systemd: start halt.target */ + SIGRTMIN+4, /* systemd: start poweroff.target */ + SIGRTMIN+5, /* systemd: start reboot.target */ + SIGRTMIN+6, /* systemd: start kexec.target */ + + /* ... space for more special targets ... */ + + SIGRTMIN+13, /* systemd: Immediate halt */ + SIGRTMIN+14, /* systemd: Immediate poweroff */ + SIGRTMIN+15, /* systemd: Immediate reboot */ + SIGRTMIN+16, /* systemd: Immediate kexec */ + + /* ... space for more immediate system state changes ... */ + + SIGRTMIN+20, /* systemd: enable status messages */ + SIGRTMIN+21, /* systemd: disable status messages */ + SIGRTMIN+22, /* systemd: set log level to LOG_DEBUG */ + SIGRTMIN+23, /* systemd: set log level to LOG_INFO */ + SIGRTMIN+24, /* systemd: Immediate exit (--user only) */ + + /* .. one free signal here ... */ + +#if !defined(__hppa64__) && !defined(__hppa__) + /* Apparently Linux on hppa has fewer RT + * signals (SIGRTMAX is SIGRTMIN+25 there), + * hence let's not try to make use of them + * here. Since these commands are accessible + * by different means and only really a safety + * net, the missing functionality on hppa + * shouldn't matter. */ + + SIGRTMIN+26, /* systemd: set log target to journal-or-kmsg */ + SIGRTMIN+27, /* systemd: set log target to console */ + SIGRTMIN+28, /* systemd: set log target to kmsg */ + SIGRTMIN+29, /* systemd: set log target to syslog-or-kmsg (obsolete) */ + + /* ... one free signal here SIGRTMIN+30 ... */ +#endif + -1); + assert_se(sigprocmask(SIG_SETMASK, &mask, NULL) == 0); + + m->signal_fd = signalfd(-1, &mask, SFD_NONBLOCK|SFD_CLOEXEC); + if (m->signal_fd < 0) + return -errno; + + r = sd_event_add_io(m->event, &m->signal_event_source, m->signal_fd, EPOLLIN, manager_dispatch_signal_fd, m); + if (r < 0) + return r; + + (void) sd_event_source_set_description(m->signal_event_source, "manager-signal"); + + /* Process signals a bit earlier than the rest of things, but later than notify_fd processing, so that the + * notify processing can still figure out to which process/service a message belongs, before we reap the + * process. Also, process this before handling cgroup notifications, so that we always collect child exit + * status information before detecting that there's no process in a cgroup. */ + r = sd_event_source_set_priority(m->signal_event_source, SD_EVENT_PRIORITY_NORMAL-6); + if (r < 0) + return r; + + if (MANAGER_IS_SYSTEM(m)) + return enable_special_signals(m); + + return 0; +} + +static void manager_clean_environment(Manager *m) { + assert(m); + + /* Let's remove some environment variables that we + * need ourselves to communicate with our clients */ + strv_env_unset_many( + m->environment, + "NOTIFY_SOCKET", + "MAINPID", + "MANAGERPID", + "LISTEN_PID", + "LISTEN_FDS", + "LISTEN_FDNAMES", + "WATCHDOG_PID", + "WATCHDOG_USEC", + NULL); +} + +static int manager_default_environment(Manager *m) { + assert(m); + + if (MANAGER_IS_SYSTEM(m)) { + /* The system manager always starts with a clean + * environment for its children. It does not import + * the kernel or the parents exported variables. + * + * The initial passed environ is untouched to keep + * /proc/self/environ valid; it is used for tagging + * the init process inside containers. */ + m->environment = strv_new("PATH=" DEFAULT_PATH, + NULL); + + /* Import locale variables LC_*= from configuration */ + locale_setup(&m->environment); + } else { + /* The user manager passes its own environment + * along to its children. */ + m->environment = strv_copy(environ); + } + + if (!m->environment) + return -ENOMEM; + + manager_clean_environment(m); + strv_sort(m->environment); + + return 0; +} + + +int manager_new(UnitFileScope scope, bool test_run, Manager **_m) { + Manager *m; + int r; + + assert(_m); + assert(IN_SET(scope, UNIT_FILE_SYSTEM, UNIT_FILE_USER)); + + m = new0(Manager, 1); + if (!m) + return -ENOMEM; + + m->unit_file_scope = scope; + m->exit_code = _MANAGER_EXIT_CODE_INVALID; + m->default_timer_accuracy_usec = USEC_PER_MINUTE; + m->default_tasks_accounting = true; + m->default_tasks_max = UINT64_MAX; + +#ifdef ENABLE_EFI + if (MANAGER_IS_SYSTEM(m) && detect_container() <= 0) + boot_timestamps(&m->userspace_timestamp, &m->firmware_timestamp, &m->loader_timestamp); +#endif + + /* Prepare log fields we can use for structured logging */ + if (MANAGER_IS_SYSTEM(m)) { + m->unit_log_field = "UNIT="; + m->unit_log_format_string = "UNIT=%s"; + } else { + m->unit_log_field = "USER_UNIT="; + m->unit_log_format_string = "USER_UNIT=%s"; + } + + m->idle_pipe[0] = m->idle_pipe[1] = m->idle_pipe[2] = m->idle_pipe[3] = -1; + + m->pin_cgroupfs_fd = m->notify_fd = m->cgroups_agent_fd = m->signal_fd = m->time_change_fd = + m->dev_autofs_fd = m->private_listen_fd = m->kdbus_fd = m->cgroup_inotify_fd = + m->ask_password_inotify_fd = -1; + + m->current_job_id = 1; /* start as id #1, so that we can leave #0 around as "null-like" value */ + + m->have_ask_password = -EINVAL; /* we don't know */ + m->first_boot = -1; + + m->test_run = test_run; + + /* Reboot immediately if the user hits C-A-D more often than 7x per 2s */ + RATELIMIT_INIT(m->ctrl_alt_del_ratelimit, 2 * USEC_PER_SEC, 7); + + r = manager_default_environment(m); + if (r < 0) + goto fail; + + r = hashmap_ensure_allocated(&m->units, &string_hash_ops); + if (r < 0) + goto fail; + + r = hashmap_ensure_allocated(&m->jobs, NULL); + if (r < 0) + goto fail; + + r = hashmap_ensure_allocated(&m->cgroup_unit, &string_hash_ops); + if (r < 0) + goto fail; + + r = hashmap_ensure_allocated(&m->watch_bus, &string_hash_ops); + if (r < 0) + goto fail; + + r = sd_event_default(&m->event); + if (r < 0) + goto fail; + + r = sd_event_add_defer(m->event, &m->run_queue_event_source, manager_dispatch_run_queue, m); + if (r < 0) + goto fail; + + r = sd_event_source_set_priority(m->run_queue_event_source, SD_EVENT_PRIORITY_IDLE); + if (r < 0) + goto fail; + + r = sd_event_source_set_enabled(m->run_queue_event_source, SD_EVENT_OFF); + if (r < 0) + goto fail; + + (void) sd_event_source_set_description(m->run_queue_event_source, "manager-run-queue"); + + r = manager_setup_signals(m); + if (r < 0) + goto fail; + + r = manager_setup_cgroup(m); + if (r < 0) + goto fail; + + r = manager_setup_time_change(m); + if (r < 0) + goto fail; + + m->udev = udev_new(); + if (!m->udev) { + r = -ENOMEM; + goto fail; + } + + /* Note that we set up neither kdbus, nor the notify fd + * here. We do that after deserialization, since they might + * have gotten serialized across the reexec. */ + + m->taint_usr = dir_is_empty("/usr") > 0; + + *_m = m; + return 0; + +fail: + manager_free(m); + return r; +} + +static int manager_setup_notify(Manager *m) { + int r; + + if (m->test_run) + return 0; + + if (m->notify_fd < 0) { + _cleanup_close_ int fd = -1; + union sockaddr_union sa = { + .sa.sa_family = AF_UNIX, + }; + static const int one = 1; + const char *e; + + /* First free all secondary fields */ + m->notify_socket = mfree(m->notify_socket); + m->notify_event_source = sd_event_source_unref(m->notify_event_source); + + fd = socket(AF_UNIX, SOCK_DGRAM|SOCK_CLOEXEC|SOCK_NONBLOCK, 0); + if (fd < 0) + return log_error_errno(errno, "Failed to allocate notification socket: %m"); + + fd_inc_rcvbuf(fd, NOTIFY_RCVBUF_SIZE); + + e = manager_get_runtime_prefix(m); + if (!e) { + log_error("Failed to determine runtime prefix."); + return -EINVAL; + } + + m->notify_socket = strappend(e, "/systemd/notify"); + if (!m->notify_socket) + return log_oom(); + + (void) mkdir_parents_label(m->notify_socket, 0755); + (void) unlink(m->notify_socket); + + strncpy(sa.un.sun_path, m->notify_socket, sizeof(sa.un.sun_path)-1); + r = bind(fd, &sa.sa, SOCKADDR_UN_LEN(sa.un)); + if (r < 0) + return log_error_errno(errno, "bind(%s) failed: %m", sa.un.sun_path); + + r = setsockopt(fd, SOL_SOCKET, SO_PASSCRED, &one, sizeof(one)); + if (r < 0) + return log_error_errno(errno, "SO_PASSCRED failed: %m"); + + m->notify_fd = fd; + fd = -1; + + log_debug("Using notification socket %s", m->notify_socket); + } + + if (!m->notify_event_source) { + r = sd_event_add_io(m->event, &m->notify_event_source, m->notify_fd, EPOLLIN, manager_dispatch_notify_fd, m); + if (r < 0) + return log_error_errno(r, "Failed to allocate notify event source: %m"); + + /* Process notification messages a bit earlier than SIGCHLD, so that we can still identify to which + * service an exit message belongs. */ + r = sd_event_source_set_priority(m->notify_event_source, SD_EVENT_PRIORITY_NORMAL-7); + if (r < 0) + return log_error_errno(r, "Failed to set priority of notify event source: %m"); + + (void) sd_event_source_set_description(m->notify_event_source, "manager-notify"); + } + + return 0; +} + +static int manager_setup_cgroups_agent(Manager *m) { + + static const union sockaddr_union sa = { + .un.sun_family = AF_UNIX, + .un.sun_path = "/run/systemd/cgroups-agent", + }; + int r; + + /* This creates a listening socket we receive cgroups agent messages on. We do not use D-Bus for delivering + * these messages from the cgroups agent binary to PID 1, as the cgroups agent binary is very short-living, and + * each instance of it needs a new D-Bus connection. Since D-Bus connections are SOCK_STREAM/AF_UNIX, on + * overloaded systems the backlog of the D-Bus socket becomes relevant, as not more than the configured number + * of D-Bus connections may be queued until the kernel will start dropping further incoming connections, + * possibly resulting in lost cgroups agent messages. To avoid this, we'll use a private SOCK_DGRAM/AF_UNIX + * socket, where no backlog is relevant as communication may take place without an actual connect() cycle, and + * we thus won't lose messages. + * + * Note that PID 1 will forward the agent message to system bus, so that the user systemd instance may listen + * to it. The system instance hence listens on this special socket, but the user instances listen on the system + * bus for these messages. */ + + if (m->test_run) + return 0; + + if (!MANAGER_IS_SYSTEM(m)) + return 0; + + if (cg_unified() > 0) /* We don't need this anymore on the unified hierarchy */ + return 0; + + if (m->cgroups_agent_fd < 0) { + _cleanup_close_ int fd = -1; + + /* First free all secondary fields */ + m->cgroups_agent_event_source = sd_event_source_unref(m->cgroups_agent_event_source); + + fd = socket(AF_UNIX, SOCK_DGRAM|SOCK_CLOEXEC|SOCK_NONBLOCK, 0); + if (fd < 0) + return log_error_errno(errno, "Failed to allocate cgroups agent socket: %m"); + + fd_inc_rcvbuf(fd, CGROUPS_AGENT_RCVBUF_SIZE); + + (void) unlink(sa.un.sun_path); + + /* Only allow root to connect to this socket */ + RUN_WITH_UMASK(0077) + r = bind(fd, &sa.sa, SOCKADDR_UN_LEN(sa.un)); + if (r < 0) + return log_error_errno(errno, "bind(%s) failed: %m", sa.un.sun_path); + + m->cgroups_agent_fd = fd; + fd = -1; + } + + if (!m->cgroups_agent_event_source) { + r = sd_event_add_io(m->event, &m->cgroups_agent_event_source, m->cgroups_agent_fd, EPOLLIN, manager_dispatch_cgroups_agent_fd, m); + if (r < 0) + return log_error_errno(r, "Failed to allocate cgroups agent event source: %m"); + + /* Process cgroups notifications early, but after having processed service notification messages or + * SIGCHLD signals, so that a cgroup running empty is always just the last safety net of notification, + * and we collected the metadata the notification and SIGCHLD stuff offers first. Also see handling of + * cgroup inotify for the unified cgroup stuff. */ + r = sd_event_source_set_priority(m->cgroups_agent_event_source, SD_EVENT_PRIORITY_NORMAL-5); + if (r < 0) + return log_error_errno(r, "Failed to set priority of cgroups agent event source: %m"); + + (void) sd_event_source_set_description(m->cgroups_agent_event_source, "manager-cgroups-agent"); + } + + return 0; +} + +static int manager_connect_bus(Manager *m, bool reexecuting) { + bool try_bus_connect; + + assert(m); + + if (m->test_run) + return 0; + + try_bus_connect = + m->kdbus_fd >= 0 || + reexecuting || + (MANAGER_IS_USER(m) && getenv("DBUS_SESSION_BUS_ADDRESS")); + + /* Try to connect to the buses, if possible. */ + return bus_init(m, try_bus_connect); +} + +static unsigned manager_dispatch_cleanup_queue(Manager *m) { + Unit *u; + unsigned n = 0; + + assert(m); + + while ((u = m->cleanup_queue)) { + assert(u->in_cleanup_queue); + + unit_free(u); + n++; + } + + return n; +} + +enum { + GC_OFFSET_IN_PATH, /* This one is on the path we were traveling */ + GC_OFFSET_UNSURE, /* No clue */ + GC_OFFSET_GOOD, /* We still need this unit */ + GC_OFFSET_BAD, /* We don't need this unit anymore */ + _GC_OFFSET_MAX +}; + +static void unit_gc_mark_good(Unit *u, unsigned gc_marker) +{ + Iterator i; + Unit *other; + + u->gc_marker = gc_marker + GC_OFFSET_GOOD; + + /* Recursively mark referenced units as GOOD as well */ + SET_FOREACH(other, u->dependencies[UNIT_REFERENCES], i) + if (other->gc_marker == gc_marker + GC_OFFSET_UNSURE) + unit_gc_mark_good(other, gc_marker); +} + +static void unit_gc_sweep(Unit *u, unsigned gc_marker) { + Iterator i; + Unit *other; + bool is_bad; + + assert(u); + + if (u->gc_marker == gc_marker + GC_OFFSET_GOOD || + u->gc_marker == gc_marker + GC_OFFSET_BAD || + u->gc_marker == gc_marker + GC_OFFSET_UNSURE || + u->gc_marker == gc_marker + GC_OFFSET_IN_PATH) + return; + + if (u->in_cleanup_queue) + goto bad; + + if (unit_check_gc(u)) + goto good; + + u->gc_marker = gc_marker + GC_OFFSET_IN_PATH; + + is_bad = true; + + SET_FOREACH(other, u->dependencies[UNIT_REFERENCED_BY], i) { + unit_gc_sweep(other, gc_marker); + + if (other->gc_marker == gc_marker + GC_OFFSET_GOOD) + goto good; + + if (other->gc_marker != gc_marker + GC_OFFSET_BAD) + is_bad = false; + } + + if (is_bad) + goto bad; + + /* We were unable to find anything out about this entry, so + * let's investigate it later */ + u->gc_marker = gc_marker + GC_OFFSET_UNSURE; + unit_add_to_gc_queue(u); + return; + +bad: + /* We definitely know that this one is not useful anymore, so + * let's mark it for deletion */ + u->gc_marker = gc_marker + GC_OFFSET_BAD; + unit_add_to_cleanup_queue(u); + return; + +good: + unit_gc_mark_good(u, gc_marker); +} + +static unsigned manager_dispatch_gc_queue(Manager *m) { + Unit *u; + unsigned n = 0; + unsigned gc_marker; + + assert(m); + + /* log_debug("Running GC..."); */ + + m->gc_marker += _GC_OFFSET_MAX; + if (m->gc_marker + _GC_OFFSET_MAX <= _GC_OFFSET_MAX) + m->gc_marker = 1; + + gc_marker = m->gc_marker; + + while ((u = m->gc_queue)) { + assert(u->in_gc_queue); + + unit_gc_sweep(u, gc_marker); + + LIST_REMOVE(gc_queue, m->gc_queue, u); + u->in_gc_queue = false; + + n++; + + if (u->gc_marker == gc_marker + GC_OFFSET_BAD || + u->gc_marker == gc_marker + GC_OFFSET_UNSURE) { + if (u->id) + log_unit_debug(u, "Collecting."); + u->gc_marker = gc_marker + GC_OFFSET_BAD; + unit_add_to_cleanup_queue(u); + } + } + + m->n_in_gc_queue = 0; + + return n; +} + +static void manager_clear_jobs_and_units(Manager *m) { + Unit *u; + + assert(m); + + while ((u = hashmap_first(m->units))) + unit_free(u); + + manager_dispatch_cleanup_queue(m); + + assert(!m->load_queue); + assert(!m->run_queue); + assert(!m->dbus_unit_queue); + assert(!m->dbus_job_queue); + assert(!m->cleanup_queue); + assert(!m->gc_queue); + + assert(hashmap_isempty(m->jobs)); + assert(hashmap_isempty(m->units)); + + m->n_on_console = 0; + m->n_running_jobs = 0; +} + +Manager* manager_free(Manager *m) { + UnitType c; + int i; + + if (!m) + return NULL; + + manager_clear_jobs_and_units(m); + + for (c = 0; c < _UNIT_TYPE_MAX; c++) + if (unit_vtable[c]->shutdown) + unit_vtable[c]->shutdown(m); + + /* If we reexecute ourselves, we keep the root cgroup + * around */ + manager_shutdown_cgroup(m, m->exit_code != MANAGER_REEXECUTE); + + lookup_paths_flush_generator(&m->lookup_paths); + + bus_done(m); + + hashmap_free(m->units); + hashmap_free(m->jobs); + hashmap_free(m->watch_pids1); + hashmap_free(m->watch_pids2); + hashmap_free(m->watch_bus); + + set_free(m->startup_units); + set_free(m->failed_units); + + sd_event_source_unref(m->signal_event_source); + sd_event_source_unref(m->notify_event_source); + sd_event_source_unref(m->cgroups_agent_event_source); + sd_event_source_unref(m->time_change_event_source); + sd_event_source_unref(m->jobs_in_progress_event_source); + sd_event_source_unref(m->run_queue_event_source); + + safe_close(m->signal_fd); + safe_close(m->notify_fd); + safe_close(m->cgroups_agent_fd); + safe_close(m->time_change_fd); + safe_close(m->kdbus_fd); + + manager_close_ask_password(m); + + manager_close_idle_pipe(m); + + udev_unref(m->udev); + sd_event_unref(m->event); + + free(m->notify_socket); + + lookup_paths_free(&m->lookup_paths); + strv_free(m->environment); + + hashmap_free(m->cgroup_unit); + set_free_free(m->unit_path_cache); + + free(m->switch_root); + free(m->switch_root_init); + + for (i = 0; i < _RLIMIT_MAX; i++) + m->rlimit[i] = mfree(m->rlimit[i]); + + assert(hashmap_isempty(m->units_requiring_mounts_for)); + hashmap_free(m->units_requiring_mounts_for); + + free(m); + return NULL; +} + +void manager_enumerate(Manager *m) { + UnitType c; + + assert(m); + + /* Let's ask every type to load all units from disk/kernel + * that it might know */ + for (c = 0; c < _UNIT_TYPE_MAX; c++) { + if (!unit_type_supported(c)) { + log_debug("Unit type .%s is not supported on this system.", unit_type_to_string(c)); + continue; + } + + if (!unit_vtable[c]->enumerate) + continue; + + unit_vtable[c]->enumerate(m); + } + + manager_dispatch_load_queue(m); +} + +static void manager_coldplug(Manager *m) { + Iterator i; + Unit *u; + char *k; + int r; + + assert(m); + + /* Then, let's set up their initial state. */ + HASHMAP_FOREACH_KEY(u, k, m->units, i) { + + /* ignore aliases */ + if (u->id != k) + continue; + + r = unit_coldplug(u); + if (r < 0) + log_warning_errno(r, "We couldn't coldplug %s, proceeding anyway: %m", u->id); + } +} + +static void manager_build_unit_path_cache(Manager *m) { + char **i; + int r; + + assert(m); + + set_free_free(m->unit_path_cache); + + m->unit_path_cache = set_new(&string_hash_ops); + if (!m->unit_path_cache) { + r = -ENOMEM; + goto fail; + } + + /* This simply builds a list of files we know exist, so that + * we don't always have to go to disk */ + + STRV_FOREACH(i, m->lookup_paths.search_path) { + _cleanup_closedir_ DIR *d = NULL; + struct dirent *de; + + d = opendir(*i); + if (!d) { + if (errno != ENOENT) + log_warning_errno(errno, "Failed to open directory %s, ignoring: %m", *i); + continue; + } + + FOREACH_DIRENT(de, d, r = -errno; goto fail) { + char *p; + + p = strjoin(streq(*i, "/") ? "" : *i, "/", de->d_name, NULL); + if (!p) { + r = -ENOMEM; + goto fail; + } + + r = set_consume(m->unit_path_cache, p); + if (r < 0) + goto fail; + } + } + + return; + +fail: + log_warning_errno(r, "Failed to build unit path cache, proceeding without: %m"); + m->unit_path_cache = set_free_free(m->unit_path_cache); +} + +static void manager_distribute_fds(Manager *m, FDSet *fds) { + Iterator i; + Unit *u; + + assert(m); + + HASHMAP_FOREACH(u, m->units, i) { + + if (fdset_size(fds) <= 0) + break; + + if (!UNIT_VTABLE(u)->distribute_fds) + continue; + + UNIT_VTABLE(u)->distribute_fds(u, fds); + } +} + +int manager_startup(Manager *m, FILE *serialization, FDSet *fds) { + int r, q; + + assert(m); + + r = lookup_paths_init(&m->lookup_paths, m->unit_file_scope, 0, NULL); + if (r < 0) + return r; + + /* Make sure the transient directory always exists, so that it remains in the search path */ + r = mkdir_p_label(m->lookup_paths.transient, 0755); + if (r < 0) + return r; + + dual_timestamp_get(&m->generators_start_timestamp); + r = manager_run_generators(m); + dual_timestamp_get(&m->generators_finish_timestamp); + if (r < 0) + return r; + + lookup_paths_reduce(&m->lookup_paths); + manager_build_unit_path_cache(m); + + /* If we will deserialize make sure that during enumeration + * this is already known, so we increase the counter here + * already */ + if (serialization) + m->n_reloading++; + + /* First, enumerate what we can from all config files */ + dual_timestamp_get(&m->units_load_start_timestamp); + manager_enumerate(m); + dual_timestamp_get(&m->units_load_finish_timestamp); + + /* Second, deserialize if there is something to deserialize */ + if (serialization) + r = manager_deserialize(m, serialization, fds); + + /* Any fds left? Find some unit which wants them. This is + * useful to allow container managers to pass some file + * descriptors to us pre-initialized. This enables + * socket-based activation of entire containers. */ + manager_distribute_fds(m, fds); + + /* We might have deserialized the notify fd, but if we didn't + * then let's create the bus now */ + q = manager_setup_notify(m); + if (q < 0 && r == 0) + r = q; + + q = manager_setup_cgroups_agent(m); + if (q < 0 && r == 0) + r = q; + + /* We might have deserialized the kdbus control fd, but if we + * didn't, then let's create the bus now. */ + manager_connect_bus(m, !!serialization); + bus_track_coldplug(m, &m->subscribed, &m->deserialized_subscribed); + + /* Third, fire things up! */ + manager_coldplug(m); + + if (serialization) { + assert(m->n_reloading > 0); + m->n_reloading--; + + /* Let's wait for the UnitNew/JobNew messages being + * sent, before we notify that the reload is + * finished */ + m->send_reloading_done = true; + } + + return r; +} + +int manager_add_job(Manager *m, JobType type, Unit *unit, JobMode mode, sd_bus_error *e, Job **_ret) { + int r; + Transaction *tr; + + assert(m); + assert(type < _JOB_TYPE_MAX); + assert(unit); + assert(mode < _JOB_MODE_MAX); + + if (mode == JOB_ISOLATE && type != JOB_START) + return sd_bus_error_setf(e, SD_BUS_ERROR_INVALID_ARGS, "Isolate is only valid for start."); + + if (mode == JOB_ISOLATE && !unit->allow_isolate) + return sd_bus_error_setf(e, BUS_ERROR_NO_ISOLATION, "Operation refused, unit may not be isolated."); + + log_unit_debug(unit, "Trying to enqueue job %s/%s/%s", unit->id, job_type_to_string(type), job_mode_to_string(mode)); + + type = job_type_collapse(type, unit); + + tr = transaction_new(mode == JOB_REPLACE_IRREVERSIBLY); + if (!tr) + return -ENOMEM; + + r = transaction_add_job_and_dependencies(tr, type, unit, NULL, true, false, + mode == JOB_IGNORE_DEPENDENCIES || mode == JOB_IGNORE_REQUIREMENTS, + mode == JOB_IGNORE_DEPENDENCIES, e); + if (r < 0) + goto tr_abort; + + if (mode == JOB_ISOLATE) { + r = transaction_add_isolate_jobs(tr, m); + if (r < 0) + goto tr_abort; + } + + r = transaction_activate(tr, m, mode, e); + if (r < 0) + goto tr_abort; + + log_unit_debug(unit, + "Enqueued job %s/%s as %u", unit->id, + job_type_to_string(type), (unsigned) tr->anchor_job->id); + + if (_ret) + *_ret = tr->anchor_job; + + transaction_free(tr); + return 0; + +tr_abort: + transaction_abort(tr); + transaction_free(tr); + return r; +} + +int manager_add_job_by_name(Manager *m, JobType type, const char *name, JobMode mode, sd_bus_error *e, Job **ret) { + Unit *unit; + int r; + + assert(m); + assert(type < _JOB_TYPE_MAX); + assert(name); + assert(mode < _JOB_MODE_MAX); + + r = manager_load_unit(m, name, NULL, NULL, &unit); + if (r < 0) + return r; + + return manager_add_job(m, type, unit, mode, e, ret); +} + +int manager_add_job_by_name_and_warn(Manager *m, JobType type, const char *name, JobMode mode, Job **ret) { + _cleanup_(sd_bus_error_free) sd_bus_error error = SD_BUS_ERROR_NULL; + int r; + + assert(m); + assert(type < _JOB_TYPE_MAX); + assert(name); + assert(mode < _JOB_MODE_MAX); + + r = manager_add_job_by_name(m, type, name, mode, &error, ret); + if (r < 0) + return log_warning_errno(r, "Failed to enqueue %s job for %s: %s", job_mode_to_string(mode), name, bus_error_message(&error, r)); + + return r; +} + +Job *manager_get_job(Manager *m, uint32_t id) { + assert(m); + + return hashmap_get(m->jobs, UINT32_TO_PTR(id)); +} + +Unit *manager_get_unit(Manager *m, const char *name) { + assert(m); + assert(name); + + return hashmap_get(m->units, name); +} + +unsigned manager_dispatch_load_queue(Manager *m) { + Unit *u; + unsigned n = 0; + + assert(m); + + /* Make sure we are not run recursively */ + if (m->dispatching_load_queue) + return 0; + + m->dispatching_load_queue = true; + + /* Dispatches the load queue. Takes a unit from the queue and + * tries to load its data until the queue is empty */ + + while ((u = m->load_queue)) { + assert(u->in_load_queue); + + unit_load(u); + n++; + } + + m->dispatching_load_queue = false; + return n; +} + +int manager_load_unit_prepare( + Manager *m, + const char *name, + const char *path, + sd_bus_error *e, + Unit **_ret) { + + Unit *ret; + UnitType t; + int r; + + assert(m); + assert(name || path); + + /* This will prepare the unit for loading, but not actually + * load anything from disk. */ + + if (path && !is_path(path)) + return sd_bus_error_setf(e, SD_BUS_ERROR_INVALID_ARGS, "Path %s is not absolute.", path); + + if (!name) + name = basename(path); + + t = unit_name_to_type(name); + + if (t == _UNIT_TYPE_INVALID || !unit_name_is_valid(name, UNIT_NAME_PLAIN|UNIT_NAME_INSTANCE)) { + if (unit_name_is_valid(name, UNIT_NAME_TEMPLATE)) + return sd_bus_error_setf(e, SD_BUS_ERROR_INVALID_ARGS, "Unit name %s is missing the instance name.", name); + + return sd_bus_error_setf(e, SD_BUS_ERROR_INVALID_ARGS, "Unit name %s is not valid.", name); + } + + ret = manager_get_unit(m, name); + if (ret) { + *_ret = ret; + return 1; + } + + ret = unit_new(m, unit_vtable[t]->object_size); + if (!ret) + return -ENOMEM; + + if (path) { + ret->fragment_path = strdup(path); + if (!ret->fragment_path) { + unit_free(ret); + return -ENOMEM; + } + } + + r = unit_add_name(ret, name); + if (r < 0) { + unit_free(ret); + return r; + } + + unit_add_to_load_queue(ret); + unit_add_to_dbus_queue(ret); + unit_add_to_gc_queue(ret); + + if (_ret) + *_ret = ret; + + return 0; +} + +int manager_load_unit( + Manager *m, + const char *name, + const char *path, + sd_bus_error *e, + Unit **_ret) { + + int r; + + assert(m); + + /* This will load the service information files, but not actually + * start any services or anything. */ + + r = manager_load_unit_prepare(m, name, path, e, _ret); + if (r != 0) + return r; + + manager_dispatch_load_queue(m); + + if (_ret) + *_ret = unit_follow_merge(*_ret); + + return 0; +} + +void manager_dump_jobs(Manager *s, FILE *f, const char *prefix) { + Iterator i; + Job *j; + + assert(s); + assert(f); + + HASHMAP_FOREACH(j, s->jobs, i) + job_dump(j, f, prefix); +} + +void manager_dump_units(Manager *s, FILE *f, const char *prefix) { + Iterator i; + Unit *u; + const char *t; + + assert(s); + assert(f); + + HASHMAP_FOREACH_KEY(u, t, s->units, i) + if (u->id == t) + unit_dump(u, f, prefix); +} + +void manager_clear_jobs(Manager *m) { + Job *j; + + assert(m); + + while ((j = hashmap_first(m->jobs))) + /* No need to recurse. We're cancelling all jobs. */ + job_finish_and_invalidate(j, JOB_CANCELED, false, false); +} + +static int manager_dispatch_run_queue(sd_event_source *source, void *userdata) { + Manager *m = userdata; + Job *j; + + assert(source); + assert(m); + + while ((j = m->run_queue)) { + assert(j->installed); + assert(j->in_run_queue); + + job_run_and_invalidate(j); + } + + if (m->n_running_jobs > 0) + manager_watch_jobs_in_progress(m); + + if (m->n_on_console > 0) + manager_watch_idle_pipe(m); + + return 1; +} + +static unsigned manager_dispatch_dbus_queue(Manager *m) { + Job *j; + Unit *u; + unsigned n = 0; + + assert(m); + + if (m->dispatching_dbus_queue) + return 0; + + m->dispatching_dbus_queue = true; + + while ((u = m->dbus_unit_queue)) { + assert(u->in_dbus_queue); + + bus_unit_send_change_signal(u); + n++; + } + + while ((j = m->dbus_job_queue)) { + assert(j->in_dbus_queue); + + bus_job_send_change_signal(j); + n++; + } + + m->dispatching_dbus_queue = false; + + if (m->send_reloading_done) { + m->send_reloading_done = false; + + bus_manager_send_reloading(m, false); + } + + if (m->queued_message) + bus_send_queued_message(m); + + return n; +} + +static int manager_dispatch_cgroups_agent_fd(sd_event_source *source, int fd, uint32_t revents, void *userdata) { + Manager *m = userdata; + char buf[PATH_MAX+1]; + ssize_t n; + + n = recv(fd, buf, sizeof(buf), 0); + if (n < 0) + return log_error_errno(errno, "Failed to read cgroups agent message: %m"); + if (n == 0) { + log_error("Got zero-length cgroups agent message, ignoring."); + return 0; + } + if ((size_t) n >= sizeof(buf)) { + log_error("Got overly long cgroups agent message, ignoring."); + return 0; + } + + if (memchr(buf, 0, n)) { + log_error("Got cgroups agent message with embedded NUL byte, ignoring."); + return 0; + } + buf[n] = 0; + + manager_notify_cgroup_empty(m, buf); + bus_forward_agent_released(m, buf); + + return 0; +} + +static void manager_invoke_notify_message(Manager *m, Unit *u, pid_t pid, const char *buf, FDSet *fds) { + _cleanup_strv_free_ char **tags = NULL; + + assert(m); + assert(u); + assert(buf); + + tags = strv_split(buf, "\n\r"); + if (!tags) { + log_oom(); + return; + } + + if (UNIT_VTABLE(u)->notify_message) + UNIT_VTABLE(u)->notify_message(u, pid, tags, fds); + else + log_unit_debug(u, "Got notification message for unit. Ignoring."); +} + +static int manager_dispatch_notify_fd(sd_event_source *source, int fd, uint32_t revents, void *userdata) { + + _cleanup_fdset_free_ FDSet *fds = NULL; + Manager *m = userdata; + char buf[NOTIFY_BUFFER_MAX+1]; + struct iovec iovec = { + .iov_base = buf, + .iov_len = sizeof(buf)-1, + }; + union { + struct cmsghdr cmsghdr; + uint8_t buf[CMSG_SPACE(sizeof(struct ucred)) + + CMSG_SPACE(sizeof(int) * NOTIFY_FD_MAX)]; + } control = {}; + struct msghdr msghdr = { + .msg_iov = &iovec, + .msg_iovlen = 1, + .msg_control = &control, + .msg_controllen = sizeof(control), + }; + + struct cmsghdr *cmsg; + struct ucred *ucred = NULL; + bool found = false; + Unit *u1, *u2, *u3; + int r, *fd_array = NULL; + unsigned n_fds = 0; + ssize_t n; + + assert(m); + assert(m->notify_fd == fd); + + if (revents != EPOLLIN) { + log_warning("Got unexpected poll event for notify fd."); + return 0; + } + + n = recvmsg(m->notify_fd, &msghdr, MSG_DONTWAIT|MSG_CMSG_CLOEXEC); + if (n < 0) { + if (!IN_SET(errno, EAGAIN, EINTR)) + log_error("Failed to receive notification message: %m"); + + /* It's not an option to return an error here since it + * would disable the notification handler entirely. Services + * wouldn't be able to send the WATCHDOG message for + * example... */ + return 0; + } + + CMSG_FOREACH(cmsg, &msghdr) { + if (cmsg->cmsg_level == SOL_SOCKET && cmsg->cmsg_type == SCM_RIGHTS) { + + fd_array = (int*) CMSG_DATA(cmsg); + n_fds = (cmsg->cmsg_len - CMSG_LEN(0)) / sizeof(int); + + } else if (cmsg->cmsg_level == SOL_SOCKET && + cmsg->cmsg_type == SCM_CREDENTIALS && + cmsg->cmsg_len == CMSG_LEN(sizeof(struct ucred))) { + + ucred = (struct ucred*) CMSG_DATA(cmsg); + } + } + + if (n_fds > 0) { + assert(fd_array); + + r = fdset_new_array(&fds, fd_array, n_fds); + if (r < 0) { + close_many(fd_array, n_fds); + log_oom(); + return 0; + } + } + + if (!ucred || ucred->pid <= 0) { + log_warning("Received notify message without valid credentials. Ignoring."); + return 0; + } + + if ((size_t) n >= sizeof(buf)) { + log_warning("Received notify message exceeded maximum size. Ignoring."); + return 0; + } + + /* The message should be a string. Here we make sure it's NUL-terminated, + * but only the part until first NUL will be used anyway. */ + buf[n] = 0; + + /* Notify every unit that might be interested, but try + * to avoid notifying the same one multiple times. */ + u1 = manager_get_unit_by_pid_cgroup(m, ucred->pid); + if (u1) { + manager_invoke_notify_message(m, u1, ucred->pid, buf, fds); + found = true; + } + + u2 = hashmap_get(m->watch_pids1, PID_TO_PTR(ucred->pid)); + if (u2 && u2 != u1) { + manager_invoke_notify_message(m, u2, ucred->pid, buf, fds); + found = true; + } + + u3 = hashmap_get(m->watch_pids2, PID_TO_PTR(ucred->pid)); + if (u3 && u3 != u2 && u3 != u1) { + manager_invoke_notify_message(m, u3, ucred->pid, buf, fds); + found = true; + } + + if (!found) + log_warning("Cannot find unit for notify message of PID "PID_FMT".", ucred->pid); + + if (fdset_size(fds) > 0) + log_warning("Got auxiliary fds with notification message, closing all."); + + return 0; +} + +static void invoke_sigchld_event(Manager *m, Unit *u, const siginfo_t *si) { + uint64_t iteration; + + assert(m); + assert(u); + assert(si); + + sd_event_get_iteration(m->event, &iteration); + + log_unit_debug(u, "Child "PID_FMT" belongs to %s", si->si_pid, u->id); + + unit_unwatch_pid(u, si->si_pid); + + if (UNIT_VTABLE(u)->sigchld_event) { + if (set_size(u->pids) <= 1 || + iteration != u->sigchldgen || + unit_main_pid(u) == si->si_pid || + unit_control_pid(u) == si->si_pid) { + UNIT_VTABLE(u)->sigchld_event(u, si->si_pid, si->si_code, si->si_status); + u->sigchldgen = iteration; + } else + log_debug("%s already issued a sigchld this iteration %" PRIu64 ", skipping. Pids still being watched %d", u->id, iteration, set_size(u->pids)); + } +} + +static int manager_dispatch_sigchld(Manager *m) { + assert(m); + + for (;;) { + siginfo_t si = {}; + + /* First we call waitd() for a PID and do not reap the + * zombie. That way we can still access /proc/$PID for + * it while it is a zombie. */ + if (waitid(P_ALL, 0, &si, WEXITED|WNOHANG|WNOWAIT) < 0) { + + if (errno == ECHILD) + break; + + if (errno == EINTR) + continue; + + return -errno; + } + + if (si.si_pid <= 0) + break; + + if (si.si_code == CLD_EXITED || si.si_code == CLD_KILLED || si.si_code == CLD_DUMPED) { + _cleanup_free_ char *name = NULL; + Unit *u1, *u2, *u3; + + get_process_comm(si.si_pid, &name); + + log_debug("Child "PID_FMT" (%s) died (code=%s, status=%i/%s)", + si.si_pid, strna(name), + sigchld_code_to_string(si.si_code), + si.si_status, + strna(si.si_code == CLD_EXITED + ? exit_status_to_string(si.si_status, EXIT_STATUS_FULL) + : signal_to_string(si.si_status))); + + /* And now figure out the unit this belongs + * to, it might be multiple... */ + u1 = manager_get_unit_by_pid_cgroup(m, si.si_pid); + if (u1) + invoke_sigchld_event(m, u1, &si); + u2 = hashmap_get(m->watch_pids1, PID_TO_PTR(si.si_pid)); + if (u2 && u2 != u1) + invoke_sigchld_event(m, u2, &si); + u3 = hashmap_get(m->watch_pids2, PID_TO_PTR(si.si_pid)); + if (u3 && u3 != u2 && u3 != u1) + invoke_sigchld_event(m, u3, &si); + } + + /* And now, we actually reap the zombie. */ + if (waitid(P_PID, si.si_pid, &si, WEXITED) < 0) { + if (errno == EINTR) + continue; + + return -errno; + } + } + + return 0; +} + +static int manager_start_target(Manager *m, const char *name, JobMode mode) { + _cleanup_(sd_bus_error_free) sd_bus_error error = SD_BUS_ERROR_NULL; + int r; + + log_debug("Activating special unit %s", name); + + r = manager_add_job_by_name(m, JOB_START, name, mode, &error, NULL); + if (r < 0) + log_error("Failed to enqueue %s job: %s", name, bus_error_message(&error, r)); + + return r; +} + +static int manager_dispatch_signal_fd(sd_event_source *source, int fd, uint32_t revents, void *userdata) { + Manager *m = userdata; + ssize_t n; + struct signalfd_siginfo sfsi; + bool sigchld = false; + int r; + + assert(m); + assert(m->signal_fd == fd); + + if (revents != EPOLLIN) { + log_warning("Got unexpected events from signal file descriptor."); + return 0; + } + + for (;;) { + n = read(m->signal_fd, &sfsi, sizeof(sfsi)); + if (n != sizeof(sfsi)) { + + if (n >= 0) + return -EIO; + + if (errno == EINTR || errno == EAGAIN) + break; + + return -errno; + } + + log_received_signal(sfsi.ssi_signo == SIGCHLD || + (sfsi.ssi_signo == SIGTERM && MANAGER_IS_USER(m)) + ? LOG_DEBUG : LOG_INFO, + &sfsi); + + switch (sfsi.ssi_signo) { + + case SIGCHLD: + sigchld = true; + break; + + case SIGTERM: + if (MANAGER_IS_SYSTEM(m)) { + /* This is for compatibility with the + * original sysvinit */ + m->exit_code = MANAGER_REEXECUTE; + break; + } + + /* Fall through */ + + case SIGINT: + if (MANAGER_IS_SYSTEM(m)) { + + /* If the user presses C-A-D more than + * 7 times within 2s, we reboot + * immediately. */ + + if (ratelimit_test(&m->ctrl_alt_del_ratelimit)) + manager_start_target(m, SPECIAL_CTRL_ALT_DEL_TARGET, JOB_REPLACE_IRREVERSIBLY); + else { + log_notice("Ctrl-Alt-Del was pressed more than 7 times within 2s, rebooting immediately."); + status_printf(NULL, true, false, "Ctrl-Alt-Del was pressed more than 7 times within 2s, rebooting immediately."); + m->exit_code = MANAGER_REBOOT; + } + + break; + } + + /* Run the exit target if there is one, if not, just exit. */ + if (manager_start_target(m, SPECIAL_EXIT_TARGET, JOB_REPLACE) < 0) { + m->exit_code = MANAGER_EXIT; + return 0; + } + + break; + + case SIGWINCH: + if (MANAGER_IS_SYSTEM(m)) + manager_start_target(m, SPECIAL_KBREQUEST_TARGET, JOB_REPLACE); + + /* This is a nop on non-init */ + break; + + case SIGPWR: + if (MANAGER_IS_SYSTEM(m)) + manager_start_target(m, SPECIAL_SIGPWR_TARGET, JOB_REPLACE); + + /* This is a nop on non-init */ + break; + + case SIGUSR1: { + Unit *u; + + u = manager_get_unit(m, SPECIAL_DBUS_SERVICE); + + if (!u || UNIT_IS_ACTIVE_OR_RELOADING(unit_active_state(u))) { + log_info("Trying to reconnect to bus..."); + bus_init(m, true); + } + + if (!u || !UNIT_IS_ACTIVE_OR_ACTIVATING(unit_active_state(u))) { + log_info("Loading D-Bus service..."); + manager_start_target(m, SPECIAL_DBUS_SERVICE, JOB_REPLACE); + } + + break; + } + + case SIGUSR2: { + _cleanup_free_ char *dump = NULL; + _cleanup_fclose_ FILE *f = NULL; + size_t size; + + f = open_memstream(&dump, &size); + if (!f) { + log_warning_errno(errno, "Failed to allocate memory stream: %m"); + break; + } + + manager_dump_units(m, f, "\t"); + manager_dump_jobs(m, f, "\t"); + + r = fflush_and_check(f); + if (r < 0) { + log_warning_errno(r, "Failed to write status stream: %m"); + break; + } + + log_dump(LOG_INFO, dump); + break; + } + + case SIGHUP: + m->exit_code = MANAGER_RELOAD; + break; + + default: { + + /* Starting SIGRTMIN+0 */ + static const char * const target_table[] = { + [0] = SPECIAL_DEFAULT_TARGET, + [1] = SPECIAL_RESCUE_TARGET, + [2] = SPECIAL_EMERGENCY_TARGET, + [3] = SPECIAL_HALT_TARGET, + [4] = SPECIAL_POWEROFF_TARGET, + [5] = SPECIAL_REBOOT_TARGET, + [6] = SPECIAL_KEXEC_TARGET + }; + + /* Starting SIGRTMIN+13, so that target halt and system halt are 10 apart */ + static const ManagerExitCode code_table[] = { + [0] = MANAGER_HALT, + [1] = MANAGER_POWEROFF, + [2] = MANAGER_REBOOT, + [3] = MANAGER_KEXEC + }; + + if ((int) sfsi.ssi_signo >= SIGRTMIN+0 && + (int) sfsi.ssi_signo < SIGRTMIN+(int) ELEMENTSOF(target_table)) { + int idx = (int) sfsi.ssi_signo - SIGRTMIN; + manager_start_target(m, target_table[idx], + (idx == 1 || idx == 2) ? JOB_ISOLATE : JOB_REPLACE); + break; + } + + if ((int) sfsi.ssi_signo >= SIGRTMIN+13 && + (int) sfsi.ssi_signo < SIGRTMIN+13+(int) ELEMENTSOF(code_table)) { + m->exit_code = code_table[sfsi.ssi_signo - SIGRTMIN - 13]; + break; + } + + switch (sfsi.ssi_signo - SIGRTMIN) { + + case 20: + manager_set_show_status(m, SHOW_STATUS_YES); + break; + + case 21: + manager_set_show_status(m, SHOW_STATUS_NO); + break; + + case 22: + log_set_max_level(LOG_DEBUG); + log_info("Setting log level to debug."); + break; + + case 23: + log_set_max_level(LOG_INFO); + log_info("Setting log level to info."); + break; + + case 24: + if (MANAGER_IS_USER(m)) { + m->exit_code = MANAGER_EXIT; + return 0; + } + + /* This is a nop on init */ + break; + + case 26: + case 29: /* compatibility: used to be mapped to LOG_TARGET_SYSLOG_OR_KMSG */ + log_set_target(LOG_TARGET_JOURNAL_OR_KMSG); + log_notice("Setting log target to journal-or-kmsg."); + break; + + case 27: + log_set_target(LOG_TARGET_CONSOLE); + log_notice("Setting log target to console."); + break; + + case 28: + log_set_target(LOG_TARGET_KMSG); + log_notice("Setting log target to kmsg."); + break; + + default: + log_warning("Got unhandled signal <%s>.", signal_to_string(sfsi.ssi_signo)); + } + } + } + } + + if (sigchld) + manager_dispatch_sigchld(m); + + return 0; +} + +static int manager_dispatch_time_change_fd(sd_event_source *source, int fd, uint32_t revents, void *userdata) { + Manager *m = userdata; + Iterator i; + Unit *u; + + assert(m); + assert(m->time_change_fd == fd); + + log_struct(LOG_INFO, + LOG_MESSAGE_ID(SD_MESSAGE_TIME_CHANGE), + LOG_MESSAGE("Time has been changed"), + NULL); + + /* Restart the watch */ + m->time_change_event_source = sd_event_source_unref(m->time_change_event_source); + m->time_change_fd = safe_close(m->time_change_fd); + + manager_setup_time_change(m); + + HASHMAP_FOREACH(u, m->units, i) + if (UNIT_VTABLE(u)->time_change) + UNIT_VTABLE(u)->time_change(u); + + return 0; +} + +static int manager_dispatch_idle_pipe_fd(sd_event_source *source, int fd, uint32_t revents, void *userdata) { + Manager *m = userdata; + + assert(m); + assert(m->idle_pipe[2] == fd); + + m->no_console_output = m->n_on_console > 0; + + manager_close_idle_pipe(m); + + return 0; +} + +static int manager_dispatch_jobs_in_progress(sd_event_source *source, usec_t usec, void *userdata) { + Manager *m = userdata; + int r; + uint64_t next; + + assert(m); + assert(source); + + manager_print_jobs_in_progress(m); + + next = now(CLOCK_MONOTONIC) + JOBS_IN_PROGRESS_PERIOD_USEC; + r = sd_event_source_set_time(source, next); + if (r < 0) + return r; + + return sd_event_source_set_enabled(source, SD_EVENT_ONESHOT); +} + +int manager_loop(Manager *m) { + int r; + + RATELIMIT_DEFINE(rl, 1*USEC_PER_SEC, 50000); + + assert(m); + m->exit_code = MANAGER_OK; + + /* Release the path cache */ + m->unit_path_cache = set_free_free(m->unit_path_cache); + + manager_check_finished(m); + + /* There might still be some zombies hanging around from + * before we were exec()'ed. Let's reap them. */ + r = manager_dispatch_sigchld(m); + if (r < 0) + return r; + + while (m->exit_code == MANAGER_OK) { + usec_t wait_usec; + + if (m->runtime_watchdog > 0 && m->runtime_watchdog != USEC_INFINITY && MANAGER_IS_SYSTEM(m)) + watchdog_ping(); + + if (!ratelimit_test(&rl)) { + /* Yay, something is going seriously wrong, pause a little */ + log_warning("Looping too fast. Throttling execution a little."); + sleep(1); + } + + if (manager_dispatch_load_queue(m) > 0) + continue; + + if (manager_dispatch_gc_queue(m) > 0) + continue; + + if (manager_dispatch_cleanup_queue(m) > 0) + continue; + + if (manager_dispatch_cgroup_queue(m) > 0) + continue; + + if (manager_dispatch_dbus_queue(m) > 0) + continue; + + /* Sleep for half the watchdog time */ + if (m->runtime_watchdog > 0 && m->runtime_watchdog != USEC_INFINITY && MANAGER_IS_SYSTEM(m)) { + wait_usec = m->runtime_watchdog / 2; + if (wait_usec <= 0) + wait_usec = 1; + } else + wait_usec = USEC_INFINITY; + + r = sd_event_run(m->event, wait_usec); + if (r < 0) + return log_error_errno(r, "Failed to run event loop: %m"); + } + + return m->exit_code; +} + +int manager_load_unit_from_dbus_path(Manager *m, const char *s, sd_bus_error *e, Unit **_u) { + _cleanup_free_ char *n = NULL; + Unit *u; + int r; + + assert(m); + assert(s); + assert(_u); + + r = unit_name_from_dbus_path(s, &n); + if (r < 0) + return r; + + r = manager_load_unit(m, n, NULL, e, &u); + if (r < 0) + return r; + + *_u = u; + + return 0; +} + +int manager_get_job_from_dbus_path(Manager *m, const char *s, Job **_j) { + const char *p; + unsigned id; + Job *j; + int r; + + assert(m); + assert(s); + assert(_j); + + p = startswith(s, "/org/freedesktop/systemd1/job/"); + if (!p) + return -EINVAL; + + r = safe_atou(p, &id); + if (r < 0) + return r; + + j = manager_get_job(m, id); + if (!j) + return -ENOENT; + + *_j = j; + + return 0; +} + +void manager_send_unit_audit(Manager *m, Unit *u, int type, bool success) { + +#ifdef HAVE_AUDIT + _cleanup_free_ char *p = NULL; + const char *msg; + int audit_fd, r; + + if (!MANAGER_IS_SYSTEM(m)) + return; + + audit_fd = get_audit_fd(); + if (audit_fd < 0) + return; + + /* Don't generate audit events if the service was already + * started and we're just deserializing */ + if (MANAGER_IS_RELOADING(m)) + return; + + if (u->type != UNIT_SERVICE) + return; + + r = unit_name_to_prefix_and_instance(u->id, &p); + if (r < 0) { + log_error_errno(r, "Failed to extract prefix and instance of unit name: %m"); + return; + } + + msg = strjoina("unit=", p); + if (audit_log_user_comm_message(audit_fd, type, msg, "systemd", NULL, NULL, NULL, success) < 0) { + if (errno == EPERM) + /* We aren't allowed to send audit messages? + * Then let's not retry again. */ + close_audit_fd(); + else + log_warning_errno(errno, "Failed to send audit message: %m"); + } +#endif + +} + +void manager_send_unit_plymouth(Manager *m, Unit *u) { + static const union sockaddr_union sa = PLYMOUTH_SOCKET; + _cleanup_free_ char *message = NULL; + _cleanup_close_ int fd = -1; + int n = 0; + + /* Don't generate plymouth events if the service was already + * started and we're just deserializing */ + if (MANAGER_IS_RELOADING(m)) + return; + + if (!MANAGER_IS_SYSTEM(m)) + return; + + if (detect_container() > 0) + return; + + if (u->type != UNIT_SERVICE && + u->type != UNIT_MOUNT && + u->type != UNIT_SWAP) + return; + + /* We set SOCK_NONBLOCK here so that we rather drop the + * message then wait for plymouth */ + fd = socket(AF_UNIX, SOCK_STREAM|SOCK_CLOEXEC|SOCK_NONBLOCK, 0); + if (fd < 0) { + log_error_errno(errno, "socket() failed: %m"); + return; + } + + if (connect(fd, &sa.sa, SOCKADDR_UN_LEN(sa.un)) < 0) { + + if (!IN_SET(errno, EPIPE, EAGAIN, ENOENT, ECONNREFUSED, ECONNRESET, ECONNABORTED)) + log_error_errno(errno, "connect() failed: %m"); + return; + } + + if (asprintf(&message, "U\002%c%s%n", (int) (strlen(u->id) + 1), u->id, &n) < 0) { + log_oom(); + return; + } + + errno = 0; + if (write(fd, message, n + 1) != n + 1) + if (!IN_SET(errno, EPIPE, EAGAIN, ENOENT, ECONNREFUSED, ECONNRESET, ECONNABORTED)) + log_error_errno(errno, "Failed to write Plymouth message: %m"); +} + +int manager_open_serialization(Manager *m, FILE **_f) { + const char *path; + int fd = -1; + FILE *f; + + assert(_f); + + path = MANAGER_IS_SYSTEM(m) ? "/run/systemd" : "/tmp"; + fd = open_tmpfile_unlinkable(path, O_RDWR|O_CLOEXEC); + if (fd < 0) + return -errno; + + log_debug("Serializing state to %s", path); + + f = fdopen(fd, "w+"); + if (!f) { + safe_close(fd); + return -errno; + } + + *_f = f; + + return 0; +} + +int manager_serialize(Manager *m, FILE *f, FDSet *fds, bool switching_root) { + Iterator i; + Unit *u; + const char *t; + char **e; + int r; + + assert(m); + assert(f); + assert(fds); + + m->n_reloading++; + + fprintf(f, "current-job-id=%"PRIu32"\n", m->current_job_id); + fprintf(f, "taint-usr=%s\n", yes_no(m->taint_usr)); + fprintf(f, "n-installed-jobs=%u\n", m->n_installed_jobs); + fprintf(f, "n-failed-jobs=%u\n", m->n_failed_jobs); + + dual_timestamp_serialize(f, "firmware-timestamp", &m->firmware_timestamp); + dual_timestamp_serialize(f, "loader-timestamp", &m->loader_timestamp); + dual_timestamp_serialize(f, "kernel-timestamp", &m->kernel_timestamp); + dual_timestamp_serialize(f, "initrd-timestamp", &m->initrd_timestamp); + + if (!in_initrd()) { + dual_timestamp_serialize(f, "userspace-timestamp", &m->userspace_timestamp); + dual_timestamp_serialize(f, "finish-timestamp", &m->finish_timestamp); + dual_timestamp_serialize(f, "security-start-timestamp", &m->security_start_timestamp); + dual_timestamp_serialize(f, "security-finish-timestamp", &m->security_finish_timestamp); + dual_timestamp_serialize(f, "generators-start-timestamp", &m->generators_start_timestamp); + dual_timestamp_serialize(f, "generators-finish-timestamp", &m->generators_finish_timestamp); + dual_timestamp_serialize(f, "units-load-start-timestamp", &m->units_load_start_timestamp); + dual_timestamp_serialize(f, "units-load-finish-timestamp", &m->units_load_finish_timestamp); + } + + if (!switching_root) { + STRV_FOREACH(e, m->environment) { + _cleanup_free_ char *ce; + + ce = cescape(*e); + if (!ce) + return -ENOMEM; + + fprintf(f, "env=%s\n", *e); + } + } + + if (m->notify_fd >= 0) { + int copy; + + copy = fdset_put_dup(fds, m->notify_fd); + if (copy < 0) + return copy; + + fprintf(f, "notify-fd=%i\n", copy); + fprintf(f, "notify-socket=%s\n", m->notify_socket); + } + + if (m->cgroups_agent_fd >= 0) { + int copy; + + copy = fdset_put_dup(fds, m->cgroups_agent_fd); + if (copy < 0) + return copy; + + fprintf(f, "cgroups-agent-fd=%i\n", copy); + } + + if (m->kdbus_fd >= 0) { + int copy; + + copy = fdset_put_dup(fds, m->kdbus_fd); + if (copy < 0) + return copy; + + fprintf(f, "kdbus-fd=%i\n", copy); + } + + bus_track_serialize(m->subscribed, f); + + fputc('\n', f); + + HASHMAP_FOREACH_KEY(u, t, m->units, i) { + if (u->id != t) + continue; + + /* Start marker */ + fputs(u->id, f); + fputc('\n', f); + + r = unit_serialize(u, f, fds, !switching_root); + if (r < 0) { + m->n_reloading--; + return r; + } + } + + assert(m->n_reloading > 0); + m->n_reloading--; + + if (ferror(f)) + return -EIO; + + r = bus_fdset_add_all(m, fds); + if (r < 0) + return r; + + return 0; +} + +int manager_deserialize(Manager *m, FILE *f, FDSet *fds) { + int r = 0; + + assert(m); + assert(f); + + log_debug("Deserializing state..."); + + m->n_reloading++; + + for (;;) { + char line[LINE_MAX], *l; + + if (!fgets(line, sizeof(line), f)) { + if (feof(f)) + r = 0; + else + r = -errno; + + goto finish; + } + + char_array_0(line); + l = strstrip(line); + + if (l[0] == 0) + break; + + if (startswith(l, "current-job-id=")) { + uint32_t id; + + if (safe_atou32(l+15, &id) < 0) + log_debug("Failed to parse current job id value %s", l+15); + else + m->current_job_id = MAX(m->current_job_id, id); + + } else if (startswith(l, "n-installed-jobs=")) { + uint32_t n; + + if (safe_atou32(l+17, &n) < 0) + log_debug("Failed to parse installed jobs counter %s", l+17); + else + m->n_installed_jobs += n; + + } else if (startswith(l, "n-failed-jobs=")) { + uint32_t n; + + if (safe_atou32(l+14, &n) < 0) + log_debug("Failed to parse failed jobs counter %s", l+14); + else + m->n_failed_jobs += n; + + } else if (startswith(l, "taint-usr=")) { + int b; + + b = parse_boolean(l+10); + if (b < 0) + log_debug("Failed to parse taint /usr flag %s", l+10); + else + m->taint_usr = m->taint_usr || b; + + } else if (startswith(l, "firmware-timestamp=")) + dual_timestamp_deserialize(l+19, &m->firmware_timestamp); + else if (startswith(l, "loader-timestamp=")) + dual_timestamp_deserialize(l+17, &m->loader_timestamp); + else if (startswith(l, "kernel-timestamp=")) + dual_timestamp_deserialize(l+17, &m->kernel_timestamp); + else if (startswith(l, "initrd-timestamp=")) + dual_timestamp_deserialize(l+17, &m->initrd_timestamp); + else if (startswith(l, "userspace-timestamp=")) + dual_timestamp_deserialize(l+20, &m->userspace_timestamp); + else if (startswith(l, "finish-timestamp=")) + dual_timestamp_deserialize(l+17, &m->finish_timestamp); + else if (startswith(l, "security-start-timestamp=")) + dual_timestamp_deserialize(l+25, &m->security_start_timestamp); + else if (startswith(l, "security-finish-timestamp=")) + dual_timestamp_deserialize(l+26, &m->security_finish_timestamp); + else if (startswith(l, "generators-start-timestamp=")) + dual_timestamp_deserialize(l+27, &m->generators_start_timestamp); + else if (startswith(l, "generators-finish-timestamp=")) + dual_timestamp_deserialize(l+28, &m->generators_finish_timestamp); + else if (startswith(l, "units-load-start-timestamp=")) + dual_timestamp_deserialize(l+27, &m->units_load_start_timestamp); + else if (startswith(l, "units-load-finish-timestamp=")) + dual_timestamp_deserialize(l+28, &m->units_load_finish_timestamp); + else if (startswith(l, "env=")) { + _cleanup_free_ char *uce = NULL; + char **e; + + r = cunescape(l + 4, UNESCAPE_RELAX, &uce); + if (r < 0) + goto finish; + + e = strv_env_set(m->environment, uce); + if (!e) { + r = -ENOMEM; + goto finish; + } + + strv_free(m->environment); + m->environment = e; + + } else if (startswith(l, "notify-fd=")) { + int fd; + + if (safe_atoi(l + 10, &fd) < 0 || fd < 0 || !fdset_contains(fds, fd)) + log_debug("Failed to parse notify fd: %s", l + 10); + else { + m->notify_event_source = sd_event_source_unref(m->notify_event_source); + safe_close(m->notify_fd); + m->notify_fd = fdset_remove(fds, fd); + } + + } else if (startswith(l, "notify-socket=")) { + char *n; + + n = strdup(l+14); + if (!n) { + r = -ENOMEM; + goto finish; + } + + free(m->notify_socket); + m->notify_socket = n; + + } else if (startswith(l, "cgroups-agent-fd=")) { + int fd; + + if (safe_atoi(l + 17, &fd) < 0 || fd < 0 || !fdset_contains(fds, fd)) + log_debug("Failed to parse cgroups agent fd: %s", l + 10); + else { + m->cgroups_agent_event_source = sd_event_source_unref(m->cgroups_agent_event_source); + safe_close(m->cgroups_agent_fd); + m->cgroups_agent_fd = fdset_remove(fds, fd); + } + + } else if (startswith(l, "kdbus-fd=")) { + int fd; + + if (safe_atoi(l + 9, &fd) < 0 || fd < 0 || !fdset_contains(fds, fd)) + log_debug("Failed to parse kdbus fd: %s", l + 9); + else { + safe_close(m->kdbus_fd); + m->kdbus_fd = fdset_remove(fds, fd); + } + + } else { + int k; + + k = bus_track_deserialize_item(&m->deserialized_subscribed, l); + if (k < 0) + log_debug_errno(k, "Failed to deserialize bus tracker object: %m"); + else if (k == 0) + log_debug("Unknown serialization item '%s'", l); + } + } + + for (;;) { + Unit *u; + char name[UNIT_NAME_MAX+2]; + + /* Start marker */ + if (!fgets(name, sizeof(name), f)) { + if (feof(f)) + r = 0; + else + r = -errno; + + goto finish; + } + + char_array_0(name); + + r = manager_load_unit(m, strstrip(name), NULL, NULL, &u); + if (r < 0) + goto finish; + + r = unit_deserialize(u, f, fds); + if (r < 0) + goto finish; + } + +finish: + if (ferror(f)) + r = -EIO; + + assert(m->n_reloading > 0); + m->n_reloading--; + + return r; +} + +int manager_reload(Manager *m) { + int r, q; + _cleanup_fclose_ FILE *f = NULL; + _cleanup_fdset_free_ FDSet *fds = NULL; + + assert(m); + + r = manager_open_serialization(m, &f); + if (r < 0) + return r; + + m->n_reloading++; + bus_manager_send_reloading(m, true); + + fds = fdset_new(); + if (!fds) { + m->n_reloading--; + return -ENOMEM; + } + + r = manager_serialize(m, f, fds, false); + if (r < 0) { + m->n_reloading--; + return r; + } + + if (fseeko(f, 0, SEEK_SET) < 0) { + m->n_reloading--; + return -errno; + } + + /* From here on there is no way back. */ + manager_clear_jobs_and_units(m); + lookup_paths_flush_generator(&m->lookup_paths); + lookup_paths_free(&m->lookup_paths); + + q = lookup_paths_init(&m->lookup_paths, m->unit_file_scope, 0, NULL); + if (q < 0 && r >= 0) + r = q; + + /* Find new unit paths */ + q = manager_run_generators(m); + if (q < 0 && r >= 0) + r = q; + + lookup_paths_reduce(&m->lookup_paths); + manager_build_unit_path_cache(m); + + /* First, enumerate what we can from all config files */ + manager_enumerate(m); + + /* Second, deserialize our stored data */ + q = manager_deserialize(m, f, fds); + if (q < 0 && r >= 0) + r = q; + + fclose(f); + f = NULL; + + /* Re-register notify_fd as event source */ + q = manager_setup_notify(m); + if (q < 0 && r >= 0) + r = q; + + q = manager_setup_cgroups_agent(m); + if (q < 0 && r >= 0) + r = q; + + /* Third, fire things up! */ + manager_coldplug(m); + + /* Sync current state of bus names with our set of listening units */ + if (m->api_bus) + manager_sync_bus_names(m, m->api_bus); + + assert(m->n_reloading > 0); + m->n_reloading--; + + m->send_reloading_done = true; + + return r; +} + +void manager_reset_failed(Manager *m) { + Unit *u; + Iterator i; + + assert(m); + + HASHMAP_FOREACH(u, m->units, i) + unit_reset_failed(u); +} + +bool manager_unit_inactive_or_pending(Manager *m, const char *name) { + Unit *u; + + assert(m); + assert(name); + + /* Returns true if the unit is inactive or going down */ + u = manager_get_unit(m, name); + if (!u) + return true; + + return unit_inactive_or_pending(u); +} + +static void manager_notify_finished(Manager *m) { + char userspace[FORMAT_TIMESPAN_MAX], initrd[FORMAT_TIMESPAN_MAX], kernel[FORMAT_TIMESPAN_MAX], sum[FORMAT_TIMESPAN_MAX]; + usec_t firmware_usec, loader_usec, kernel_usec, initrd_usec, userspace_usec, total_usec; + + if (m->test_run) + return; + + if (MANAGER_IS_SYSTEM(m) && detect_container() <= 0) { + + /* Note that m->kernel_usec.monotonic is always at 0, + * and m->firmware_usec.monotonic and + * m->loader_usec.monotonic should be considered + * negative values. */ + + firmware_usec = m->firmware_timestamp.monotonic - m->loader_timestamp.monotonic; + loader_usec = m->loader_timestamp.monotonic - m->kernel_timestamp.monotonic; + userspace_usec = m->finish_timestamp.monotonic - m->userspace_timestamp.monotonic; + total_usec = m->firmware_timestamp.monotonic + m->finish_timestamp.monotonic; + + if (dual_timestamp_is_set(&m->initrd_timestamp)) { + + kernel_usec = m->initrd_timestamp.monotonic - m->kernel_timestamp.monotonic; + initrd_usec = m->userspace_timestamp.monotonic - m->initrd_timestamp.monotonic; + + log_struct(LOG_INFO, + LOG_MESSAGE_ID(SD_MESSAGE_STARTUP_FINISHED), + "KERNEL_USEC="USEC_FMT, kernel_usec, + "INITRD_USEC="USEC_FMT, initrd_usec, + "USERSPACE_USEC="USEC_FMT, userspace_usec, + LOG_MESSAGE("Startup finished in %s (kernel) + %s (initrd) + %s (userspace) = %s.", + format_timespan(kernel, sizeof(kernel), kernel_usec, USEC_PER_MSEC), + format_timespan(initrd, sizeof(initrd), initrd_usec, USEC_PER_MSEC), + format_timespan(userspace, sizeof(userspace), userspace_usec, USEC_PER_MSEC), + format_timespan(sum, sizeof(sum), total_usec, USEC_PER_MSEC)), + NULL); + } else { + kernel_usec = m->userspace_timestamp.monotonic - m->kernel_timestamp.monotonic; + initrd_usec = 0; + + log_struct(LOG_INFO, + LOG_MESSAGE_ID(SD_MESSAGE_STARTUP_FINISHED), + "KERNEL_USEC="USEC_FMT, kernel_usec, + "USERSPACE_USEC="USEC_FMT, userspace_usec, + LOG_MESSAGE("Startup finished in %s (kernel) + %s (userspace) = %s.", + format_timespan(kernel, sizeof(kernel), kernel_usec, USEC_PER_MSEC), + format_timespan(userspace, sizeof(userspace), userspace_usec, USEC_PER_MSEC), + format_timespan(sum, sizeof(sum), total_usec, USEC_PER_MSEC)), + NULL); + } + } else { + firmware_usec = loader_usec = initrd_usec = kernel_usec = 0; + total_usec = userspace_usec = m->finish_timestamp.monotonic - m->userspace_timestamp.monotonic; + + log_struct(LOG_INFO, + LOG_MESSAGE_ID(SD_MESSAGE_STARTUP_FINISHED), + "USERSPACE_USEC="USEC_FMT, userspace_usec, + LOG_MESSAGE("Startup finished in %s.", + format_timespan(sum, sizeof(sum), total_usec, USEC_PER_MSEC)), + NULL); + } + + bus_manager_send_finished(m, firmware_usec, loader_usec, kernel_usec, initrd_usec, userspace_usec, total_usec); + + sd_notifyf(false, + "READY=1\n" + "STATUS=Startup finished in %s.", + format_timespan(sum, sizeof(sum), total_usec, USEC_PER_MSEC)); +} + +void manager_check_finished(Manager *m) { + assert(m); + + if (MANAGER_IS_RELOADING(m)) + return; + + /* Verify that we are actually running currently. Initially + * the exit code is set to invalid, and during operation it is + * then set to MANAGER_OK */ + if (m->exit_code != MANAGER_OK) + return; + + if (hashmap_size(m->jobs) > 0) { + if (m->jobs_in_progress_event_source) + /* Ignore any failure, this is only for feedback */ + (void) sd_event_source_set_time(m->jobs_in_progress_event_source, now(CLOCK_MONOTONIC) + JOBS_IN_PROGRESS_WAIT_USEC); + + return; + } + + manager_flip_auto_status(m, false); + + /* Notify Type=idle units that we are done now */ + manager_close_idle_pipe(m); + + /* Turn off confirm spawn now */ + m->confirm_spawn = false; + + /* No need to update ask password status when we're going non-interactive */ + manager_close_ask_password(m); + + /* This is no longer the first boot */ + manager_set_first_boot(m, false); + + if (dual_timestamp_is_set(&m->finish_timestamp)) + return; + + dual_timestamp_get(&m->finish_timestamp); + + manager_notify_finished(m); + + manager_invalidate_startup_units(m); +} + +static int manager_run_generators(Manager *m) { + _cleanup_strv_free_ char **paths = NULL; + const char *argv[5]; + char **path; + int r; + + assert(m); + + if (m->test_run) + return 0; + + paths = generator_binary_paths(m->unit_file_scope); + if (!paths) + return log_oom(); + + /* Optimize by skipping the whole process by not creating output directories + * if no generators are found. */ + STRV_FOREACH(path, paths) { + if (access(*path, F_OK) >= 0) + goto found; + if (errno != ENOENT) + log_warning_errno(errno, "Failed to open generator directory %s: %m", *path); + } + + return 0; + + found: + r = lookup_paths_mkdir_generator(&m->lookup_paths); + if (r < 0) + goto finish; + + argv[0] = NULL; /* Leave this empty, execute_directory() will fill something in */ + argv[1] = m->lookup_paths.generator; + argv[2] = m->lookup_paths.generator_early; + argv[3] = m->lookup_paths.generator_late; + argv[4] = NULL; + + RUN_WITH_UMASK(0022) + execute_directories((const char* const*) paths, DEFAULT_TIMEOUT_USEC, (char**) argv); + +finish: + lookup_paths_trim_generator(&m->lookup_paths); + return r; +} + +int manager_environment_add(Manager *m, char **minus, char **plus) { + char **a = NULL, **b = NULL, **l; + assert(m); + + l = m->environment; + + if (!strv_isempty(minus)) { + a = strv_env_delete(l, 1, minus); + if (!a) + return -ENOMEM; + + l = a; + } + + if (!strv_isempty(plus)) { + b = strv_env_merge(2, l, plus); + if (!b) { + strv_free(a); + return -ENOMEM; + } + + l = b; + } + + if (m->environment != l) + strv_free(m->environment); + if (a != l) + strv_free(a); + if (b != l) + strv_free(b); + + m->environment = l; + manager_clean_environment(m); + strv_sort(m->environment); + + return 0; +} + +int manager_set_default_rlimits(Manager *m, struct rlimit **default_rlimit) { + int i; + + assert(m); + + for (i = 0; i < _RLIMIT_MAX; i++) { + m->rlimit[i] = mfree(m->rlimit[i]); + + if (!default_rlimit[i]) + continue; + + m->rlimit[i] = newdup(struct rlimit, default_rlimit[i], 1); + if (!m->rlimit[i]) + return -ENOMEM; + } + + return 0; +} + +void manager_recheck_journal(Manager *m) { + Unit *u; + + assert(m); + + if (!MANAGER_IS_SYSTEM(m)) + return; + + u = manager_get_unit(m, SPECIAL_JOURNALD_SOCKET); + if (u && SOCKET(u)->state != SOCKET_RUNNING) { + log_close_journal(); + return; + } + + u = manager_get_unit(m, SPECIAL_JOURNALD_SERVICE); + if (u && SERVICE(u)->state != SERVICE_RUNNING) { + log_close_journal(); + return; + } + + /* Hmm, OK, so the socket is fully up and the service is up + * too, then let's make use of the thing. */ + log_open(); +} + +void manager_set_show_status(Manager *m, ShowStatus mode) { + assert(m); + assert(IN_SET(mode, SHOW_STATUS_AUTO, SHOW_STATUS_NO, SHOW_STATUS_YES, SHOW_STATUS_TEMPORARY)); + + if (!MANAGER_IS_SYSTEM(m)) + return; + + if (m->show_status != mode) + log_debug("%s showing of status.", + mode == SHOW_STATUS_NO ? "Disabling" : "Enabling"); + m->show_status = mode; + + if (mode > 0) + (void) touch("/run/systemd/show-status"); + else + (void) unlink("/run/systemd/show-status"); +} + +static bool manager_get_show_status(Manager *m, StatusType type) { + assert(m); + + if (!MANAGER_IS_SYSTEM(m)) + return false; + + if (m->no_console_output) + return false; + + if (!IN_SET(manager_state(m), MANAGER_INITIALIZING, MANAGER_STARTING, MANAGER_STOPPING)) + return false; + + /* If we cannot find out the status properly, just proceed. */ + if (type != STATUS_TYPE_EMERGENCY && manager_check_ask_password(m) > 0) + return false; + + if (m->show_status > 0) + return true; + + return false; +} + +void manager_set_first_boot(Manager *m, bool b) { + assert(m); + + if (!MANAGER_IS_SYSTEM(m)) + return; + + if (m->first_boot != (int) b) { + if (b) + (void) touch("/run/systemd/first-boot"); + else + (void) unlink("/run/systemd/first-boot"); + } + + m->first_boot = b; +} + +void manager_status_printf(Manager *m, StatusType type, const char *status, const char *format, ...) { + va_list ap; + + /* If m is NULL, assume we're after shutdown and let the messages through. */ + + if (m && !manager_get_show_status(m, type)) + return; + + /* XXX We should totally drop the check for ephemeral here + * and thus effectively make 'Type=idle' pointless. */ + if (type == STATUS_TYPE_EPHEMERAL && m && m->n_on_console > 0) + return; + + va_start(ap, format); + status_vprintf(status, true, type == STATUS_TYPE_EPHEMERAL, format, ap); + va_end(ap); +} + +Set *manager_get_units_requiring_mounts_for(Manager *m, const char *path) { + char p[strlen(path)+1]; + + assert(m); + assert(path); + + strcpy(p, path); + path_kill_slashes(p); + + return hashmap_get(m->units_requiring_mounts_for, streq(p, "/") ? "" : p); +} + +const char *manager_get_runtime_prefix(Manager *m) { + assert(m); + + return MANAGER_IS_SYSTEM(m) ? + "/run" : + getenv("XDG_RUNTIME_DIR"); +} + +int manager_update_failed_units(Manager *m, Unit *u, bool failed) { + unsigned size; + int r; + + assert(m); + assert(u->manager == m); + + size = set_size(m->failed_units); + + if (failed) { + r = set_ensure_allocated(&m->failed_units, NULL); + if (r < 0) + return log_oom(); + + if (set_put(m->failed_units, u) < 0) + return log_oom(); + } else + (void) set_remove(m->failed_units, u); + + if (set_size(m->failed_units) != size) + bus_manager_send_change_signal(m); + + return 0; +} + +ManagerState manager_state(Manager *m) { + Unit *u; + + assert(m); + + /* Did we ever finish booting? If not then we are still starting up */ + if (!dual_timestamp_is_set(&m->finish_timestamp)) { + + u = manager_get_unit(m, SPECIAL_BASIC_TARGET); + if (!u || !UNIT_IS_ACTIVE_OR_RELOADING(unit_active_state(u))) + return MANAGER_INITIALIZING; + + return MANAGER_STARTING; + } + + /* Is the special shutdown target queued? If so, we are in shutdown state */ + u = manager_get_unit(m, SPECIAL_SHUTDOWN_TARGET); + if (u && u->job && IN_SET(u->job->type, JOB_START, JOB_RESTART, JOB_RELOAD_OR_START)) + return MANAGER_STOPPING; + + /* Are the rescue or emergency targets active or queued? If so we are in maintenance state */ + u = manager_get_unit(m, SPECIAL_RESCUE_TARGET); + if (u && (UNIT_IS_ACTIVE_OR_ACTIVATING(unit_active_state(u)) || + (u->job && IN_SET(u->job->type, JOB_START, JOB_RESTART, JOB_RELOAD_OR_START)))) + return MANAGER_MAINTENANCE; + + u = manager_get_unit(m, SPECIAL_EMERGENCY_TARGET); + if (u && (UNIT_IS_ACTIVE_OR_ACTIVATING(unit_active_state(u)) || + (u->job && IN_SET(u->job->type, JOB_START, JOB_RESTART, JOB_RELOAD_OR_START)))) + return MANAGER_MAINTENANCE; + + /* Are there any failed units? If so, we are in degraded mode */ + if (set_size(m->failed_units) > 0) + return MANAGER_DEGRADED; + + return MANAGER_RUNNING; +} + +static const char *const manager_state_table[_MANAGER_STATE_MAX] = { + [MANAGER_INITIALIZING] = "initializing", + [MANAGER_STARTING] = "starting", + [MANAGER_RUNNING] = "running", + [MANAGER_DEGRADED] = "degraded", + [MANAGER_MAINTENANCE] = "maintenance", + [MANAGER_STOPPING] = "stopping", +}; + +DEFINE_STRING_TABLE_LOOKUP(manager_state, ManagerState); diff --git a/src/grp-system/libcore/src/mount-setup.c b/src/grp-system/libcore/src/mount-setup.c new file mode 100644 index 0000000000..a4e72eda03 --- /dev/null +++ b/src/grp-system/libcore/src/mount-setup.c @@ -0,0 +1,419 @@ +/*** + This file is part of systemd. + + Copyright 2010 Lennart Poettering + + systemd is free software; you can redistribute it and/or modify it + under the terms of the GNU Lesser General Public License as published by + the Free Software Foundation; either version 2.1 of the License, or + (at your option) any later version. + + systemd is distributed in the hope that it will be useful, but + WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + Lesser General Public License for more details. + + You should have received a copy of the GNU Lesser General Public License + along with systemd; If not, see . +***/ + +#include +#include +#include +#include +#include + +#include "core/mount-setup.h" +#include "sd-bus/bus-util.h" +#include "systemd-basic/alloc-util.h" +#include "systemd-basic/cgroup-util.h" +#include "systemd-basic/fs-util.h" +#include "systemd-basic/label.h" +#include "systemd-basic/log.h" +#include "systemd-basic/macro.h" +#include "systemd-basic/missing.h" +#include "systemd-basic/mkdir.h" +#include "systemd-basic/mount-util.h" +#include "systemd-basic/path-util.h" +#include "systemd-basic/set.h" +#include "systemd-basic/smack-util.h" +#include "systemd-basic/strv.h" +#include "systemd-basic/user-util.h" +#include "systemd-basic/util.h" +#include "systemd-basic/virt.h" +#include "systemd-shared/dev-setup.h" +#include "systemd-shared/efivars.h" + +typedef enum MountMode { + MNT_NONE = 0, + MNT_FATAL = 1 << 0, + MNT_IN_CONTAINER = 1 << 1, +} MountMode; + +typedef struct MountPoint { + const char *what; + const char *where; + const char *type; + const char *options; + unsigned long flags; + bool (*condition_fn)(void); + MountMode mode; +} MountPoint; + +/* The first three entries we might need before SELinux is up. The + * fourth (securityfs) is needed by IMA to load a custom policy. The + * other ones we can delay until SELinux and IMA are loaded. When + * SMACK is enabled we need smackfs, too, so it's a fifth one. */ +#ifdef HAVE_SMACK +#define N_EARLY_MOUNT 5 +#else +#define N_EARLY_MOUNT 4 +#endif + +static const MountPoint mount_table[] = { + { "sysfs", "/sys", "sysfs", NULL, MS_NOSUID|MS_NOEXEC|MS_NODEV, + NULL, MNT_FATAL|MNT_IN_CONTAINER }, + { "proc", "/proc", "proc", NULL, MS_NOSUID|MS_NOEXEC|MS_NODEV, + NULL, MNT_FATAL|MNT_IN_CONTAINER }, + { "devtmpfs", "/dev", "devtmpfs", "mode=755", MS_NOSUID|MS_STRICTATIME, + NULL, MNT_FATAL|MNT_IN_CONTAINER }, + { "securityfs", "/sys/kernel/security", "securityfs", NULL, MS_NOSUID|MS_NOEXEC|MS_NODEV, + NULL, MNT_NONE }, +#ifdef HAVE_SMACK + { "smackfs", "/sys/fs/smackfs", "smackfs", "smackfsdef=*", MS_NOSUID|MS_NOEXEC|MS_NODEV, + mac_smack_use, MNT_FATAL }, + { "tmpfs", "/dev/shm", "tmpfs", "mode=1777,smackfsroot=*", MS_NOSUID|MS_NODEV|MS_STRICTATIME, + mac_smack_use, MNT_FATAL }, +#endif + { "tmpfs", "/dev/shm", "tmpfs", "mode=1777", MS_NOSUID|MS_NODEV|MS_STRICTATIME, + NULL, MNT_FATAL|MNT_IN_CONTAINER }, + { "devpts", "/dev/pts", "devpts", "mode=620,gid=" STRINGIFY(TTY_GID), MS_NOSUID|MS_NOEXEC, + NULL, MNT_IN_CONTAINER }, +#ifdef HAVE_SMACK + { "tmpfs", "/run", "tmpfs", "mode=755,smackfsroot=*", MS_NOSUID|MS_NODEV|MS_STRICTATIME, + mac_smack_use, MNT_FATAL }, +#endif + { "tmpfs", "/run", "tmpfs", "mode=755", MS_NOSUID|MS_NODEV|MS_STRICTATIME, + NULL, MNT_FATAL|MNT_IN_CONTAINER }, + { "cgroup", "/sys/fs/cgroup", "cgroup2", NULL, MS_NOSUID|MS_NOEXEC|MS_NODEV, + cg_is_unified_wanted, MNT_FATAL|MNT_IN_CONTAINER }, + { "tmpfs", "/sys/fs/cgroup", "tmpfs", "mode=755", MS_NOSUID|MS_NOEXEC|MS_NODEV|MS_STRICTATIME, + cg_is_legacy_wanted, MNT_FATAL|MNT_IN_CONTAINER }, + { "cgroup", "/sys/fs/cgroup/systemd", "cgroup", "none,name=systemd,xattr", MS_NOSUID|MS_NOEXEC|MS_NODEV, + cg_is_legacy_wanted, MNT_IN_CONTAINER }, + { "cgroup", "/sys/fs/cgroup/systemd", "cgroup", "none,name=systemd", MS_NOSUID|MS_NOEXEC|MS_NODEV, + cg_is_legacy_wanted, MNT_FATAL|MNT_IN_CONTAINER }, + { "pstore", "/sys/fs/pstore", "pstore", NULL, MS_NOSUID|MS_NOEXEC|MS_NODEV, + NULL, MNT_NONE }, +#ifdef ENABLE_EFI + { "efivarfs", "/sys/firmware/efi/efivars", "efivarfs", NULL, MS_NOSUID|MS_NOEXEC|MS_NODEV, + is_efi_boot, MNT_NONE }, +#endif +}; + +/* These are API file systems that might be mounted by other software, + * we just list them here so that we know that we should ignore them */ + +static const char ignore_paths[] = + /* SELinux file systems */ + "/sys/fs/selinux\0" + /* Container bind mounts */ + "/proc/sys\0" + "/dev/console\0" + "/proc/kmsg\0"; + +bool mount_point_is_api(const char *path) { + unsigned i; + + /* Checks if this mount point is considered "API", and hence + * should be ignored */ + + for (i = 0; i < ELEMENTSOF(mount_table); i ++) + if (path_equal(path, mount_table[i].where)) + return true; + + return path_startswith(path, "/sys/fs/cgroup/"); +} + +bool mount_point_ignore(const char *path) { + const char *i; + + NULSTR_FOREACH(i, ignore_paths) + if (path_equal(path, i)) + return true; + + return false; +} + +static int mount_one(const MountPoint *p, bool relabel) { + int r; + + assert(p); + + if (p->condition_fn && !p->condition_fn()) + return 0; + + /* Relabel first, just in case */ + if (relabel) + (void) label_fix(p->where, true, true); + + r = path_is_mount_point(p->where, AT_SYMLINK_FOLLOW); + if (r < 0 && r != -ENOENT) { + log_full_errno((p->mode & MNT_FATAL) ? LOG_ERR : LOG_DEBUG, r, "Failed to determine whether %s is a mount point: %m", p->where); + return (p->mode & MNT_FATAL) ? r : 0; + } + if (r > 0) + return 0; + + /* Skip securityfs in a container */ + if (!(p->mode & MNT_IN_CONTAINER) && detect_container() > 0) + return 0; + + /* The access mode here doesn't really matter too much, since + * the mounted file system will take precedence anyway. */ + if (relabel) + (void) mkdir_p_label(p->where, 0755); + else + (void) mkdir_p(p->where, 0755); + + log_debug("Mounting %s to %s of type %s with options %s.", + p->what, + p->where, + p->type, + strna(p->options)); + + if (mount(p->what, + p->where, + p->type, + p->flags, + p->options) < 0) { + log_full_errno((p->mode & MNT_FATAL) ? LOG_ERR : LOG_DEBUG, errno, "Failed to mount %s at %s: %m", p->type, p->where); + return (p->mode & MNT_FATAL) ? -errno : 0; + } + + /* Relabel again, since we now mounted something fresh here */ + if (relabel) + (void) label_fix(p->where, false, false); + + return 1; +} + +static int mount_points_setup(unsigned n, bool loaded_policy) { + unsigned i; + int r = 0; + + for (i = 0; i < n; i ++) { + int j; + + j = mount_one(mount_table + i, loaded_policy); + if (j != 0 && r >= 0) + r = j; + } + + return r; +} + +int mount_setup_early(void) { + assert_cc(N_EARLY_MOUNT <= ELEMENTSOF(mount_table)); + + /* Do a minimal mount of /proc and friends to enable the most + * basic stuff, such as SELinux */ + return mount_points_setup(N_EARLY_MOUNT, false); +} + +int mount_cgroup_controllers(char ***join_controllers) { + _cleanup_set_free_free_ Set *controllers = NULL; + int r; + + if (!cg_is_legacy_wanted()) + return 0; + + /* Mount all available cgroup controllers that are built into the kernel. */ + + controllers = set_new(&string_hash_ops); + if (!controllers) + return log_oom(); + + r = cg_kernel_controllers(controllers); + if (r < 0) + return log_error_errno(r, "Failed to enumerate cgroup controllers: %m"); + + for (;;) { + _cleanup_free_ char *options = NULL, *controller = NULL, *where = NULL; + MountPoint p = { + .what = "cgroup", + .type = "cgroup", + .flags = MS_NOSUID|MS_NOEXEC|MS_NODEV, + .mode = MNT_IN_CONTAINER, + }; + char ***k = NULL; + + controller = set_steal_first(controllers); + if (!controller) + break; + + if (join_controllers) + for (k = join_controllers; *k; k++) + if (strv_find(*k, controller)) + break; + + if (k && *k) { + char **i, **j; + + for (i = *k, j = *k; *i; i++) { + + if (!streq(*i, controller)) { + _cleanup_free_ char *t; + + t = set_remove(controllers, *i); + if (!t) { + free(*i); + continue; + } + } + + *(j++) = *i; + } + + *j = NULL; + + options = strv_join(*k, ","); + if (!options) + return log_oom(); + } else { + options = controller; + controller = NULL; + } + + where = strappend("/sys/fs/cgroup/", options); + if (!where) + return log_oom(); + + p.where = where; + p.options = options; + + r = mount_one(&p, true); + if (r < 0) + return r; + + if (r > 0 && k && *k) { + char **i; + + for (i = *k; *i; i++) { + _cleanup_free_ char *t = NULL; + + t = strappend("/sys/fs/cgroup/", *i); + if (!t) + return log_oom(); + + r = symlink(options, t); + if (r >= 0) { +#ifdef SMACK_RUN_LABEL + _cleanup_free_ char *src; + src = strappend("/sys/fs/cgroup/", options); + if (!src) + return log_oom(); + r = mac_smack_copy(t, src); + if (r < 0 && r != -EOPNOTSUPP) + return log_error_errno(r, "Failed to copy smack label from %s to %s: %m", src, t); +#endif + } else if (errno != EEXIST) + return log_error_errno(errno, "Failed to create symlink %s: %m", t); + } + } + } + + /* Now that we mounted everything, let's make the tmpfs the + * cgroup file systems are mounted into read-only. */ + (void) mount("tmpfs", "/sys/fs/cgroup", "tmpfs", MS_REMOUNT|MS_NOSUID|MS_NOEXEC|MS_NODEV|MS_STRICTATIME|MS_RDONLY, "mode=755"); + + return 0; +} + +#if defined(HAVE_SELINUX) || defined(HAVE_SMACK) +static int nftw_cb( + const char *fpath, + const struct stat *sb, + int tflag, + struct FTW *ftwbuf) { + + /* No need to label /dev twice in a row... */ + if (_unlikely_(ftwbuf->level == 0)) + return FTW_CONTINUE; + + label_fix(fpath, false, false); + + /* /run/initramfs is static data and big, no need to + * dynamically relabel its contents at boot... */ + if (_unlikely_(ftwbuf->level == 1 && + tflag == FTW_D && + streq(fpath, "/run/initramfs"))) + return FTW_SKIP_SUBTREE; + + return FTW_CONTINUE; +}; +#endif + +int mount_setup(bool loaded_policy) { + int r = 0; + + r = mount_points_setup(ELEMENTSOF(mount_table), loaded_policy); + + if (r < 0) + return r; + +#if defined(HAVE_SELINUX) || defined(HAVE_SMACK) + /* Nodes in devtmpfs and /run need to be manually updated for + * the appropriate labels, after mounting. The other virtual + * API file systems like /sys and /proc do not need that, they + * use the same label for all their files. */ + if (loaded_policy) { + usec_t before_relabel, after_relabel; + char timespan[FORMAT_TIMESPAN_MAX]; + + before_relabel = now(CLOCK_MONOTONIC); + + nftw("/dev", nftw_cb, 64, FTW_MOUNT|FTW_PHYS|FTW_ACTIONRETVAL); + nftw("/dev/shm", nftw_cb, 64, FTW_MOUNT|FTW_PHYS|FTW_ACTIONRETVAL); + nftw("/run", nftw_cb, 64, FTW_MOUNT|FTW_PHYS|FTW_ACTIONRETVAL); + + after_relabel = now(CLOCK_MONOTONIC); + + log_info("Relabelled /dev and /run in %s.", + format_timespan(timespan, sizeof(timespan), after_relabel - before_relabel, 0)); + } +#endif + + /* Create a few default symlinks, which are normally created + * by udevd, but some scripts might need them before we start + * udevd. */ + dev_setup(NULL, UID_INVALID, GID_INVALID); + + /* Mark the root directory as shared in regards to mount + * propagation. The kernel defaults to "private", but we think + * it makes more sense to have a default of "shared" so that + * nspawn and the container tools work out of the box. If + * specific setups need other settings they can reset the + * propagation mode to private if needed. */ + if (detect_container() <= 0) + if (mount(NULL, "/", NULL, MS_REC|MS_SHARED, NULL) < 0) + log_warning_errno(errno, "Failed to set up the root directory for shared mount propagation: %m"); + + /* Create a few directories we always want around, Note that + * sd_booted() checks for /run/systemd/system, so this mkdir + * really needs to stay for good, otherwise software that + * copied sd-daemon.c into their sources will misdetect + * systemd. */ + (void) mkdir_label("/run/systemd", 0755); + (void) mkdir_label("/run/systemd/system", 0755); + (void) mkdir_label("/run/systemd/inaccessible", 0000); + /* Set up inaccessible items */ + (void) mknod("/run/systemd/inaccessible/reg", S_IFREG | 0000, 0); + (void) mkdir_label("/run/systemd/inaccessible/dir", 0000); + (void) mknod("/run/systemd/inaccessible/chr", S_IFCHR | 0000, makedev(0, 0)); + (void) mknod("/run/systemd/inaccessible/blk", S_IFBLK | 0000, makedev(0, 0)); + (void) mkfifo("/run/systemd/inaccessible/fifo", 0000); + (void) mknod("/run/systemd/inaccessible/sock", S_IFSOCK | 0000, 0); + + return 0; +} diff --git a/src/grp-system/libcore/src/mount.c b/src/grp-system/libcore/src/mount.c new file mode 100644 index 0000000000..b0c3b6722c --- /dev/null +++ b/src/grp-system/libcore/src/mount.c @@ -0,0 +1,1883 @@ +/*** + This file is part of systemd. + + Copyright 2010 Lennart Poettering + + systemd is free software; you can redistribute it and/or modify it + under the terms of the GNU Lesser General Public License as published by + the Free Software Foundation; either version 2.1 of the License, or + (at your option) any later version. + + systemd is distributed in the hope that it will be useful, but + WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + Lesser General Public License for more details. + + You should have received a copy of the GNU Lesser General Public License + along with systemd; If not, see . +***/ + +#include +#include +#include +#include + +#include + +#include "core/manager.h" +#include "core/mount-setup.h" +#include "core/mount.h" +#include "core/unit.h" +#include "systemd-basic/alloc-util.h" +#include "systemd-basic/escape.h" +#include "systemd-basic/exit-status.h" +#include "systemd-basic/formats-util.h" +#include "systemd-basic/log.h" +#include "systemd-basic/mkdir.h" +#include "systemd-basic/mount-util.h" +#include "systemd-basic/parse-util.h" +#include "systemd-basic/path-util.h" +#include "systemd-basic/process-util.h" +#include "systemd-basic/special.h" +#include "systemd-basic/string-table.h" +#include "systemd-basic/string-util.h" +#include "systemd-basic/strv.h" +#include "systemd-basic/unit-name.h" +#include "systemd-shared/fstab-util.h" + +#include "dbus-mount.h" + +#define RETRY_UMOUNT_MAX 32 + +DEFINE_TRIVIAL_CLEANUP_FUNC(struct libmnt_table*, mnt_free_table); +DEFINE_TRIVIAL_CLEANUP_FUNC(struct libmnt_iter*, mnt_free_iter); + +static const UnitActiveState state_translation_table[_MOUNT_STATE_MAX] = { + [MOUNT_DEAD] = UNIT_INACTIVE, + [MOUNT_MOUNTING] = UNIT_ACTIVATING, + [MOUNT_MOUNTING_DONE] = UNIT_ACTIVE, + [MOUNT_MOUNTED] = UNIT_ACTIVE, + [MOUNT_REMOUNTING] = UNIT_RELOADING, + [MOUNT_UNMOUNTING] = UNIT_DEACTIVATING, + [MOUNT_MOUNTING_SIGTERM] = UNIT_DEACTIVATING, + [MOUNT_MOUNTING_SIGKILL] = UNIT_DEACTIVATING, + [MOUNT_REMOUNTING_SIGTERM] = UNIT_RELOADING, + [MOUNT_REMOUNTING_SIGKILL] = UNIT_RELOADING, + [MOUNT_UNMOUNTING_SIGTERM] = UNIT_DEACTIVATING, + [MOUNT_UNMOUNTING_SIGKILL] = UNIT_DEACTIVATING, + [MOUNT_FAILED] = UNIT_FAILED +}; + +static int mount_dispatch_timer(sd_event_source *source, usec_t usec, void *userdata); +static int mount_dispatch_io(sd_event_source *source, int fd, uint32_t revents, void *userdata); + +static bool mount_needs_network(const char *options, const char *fstype) { + if (fstab_test_option(options, "_netdev\0")) + return true; + + if (fstype && fstype_is_network(fstype)) + return true; + + return false; +} + +static bool mount_is_network(const MountParameters *p) { + assert(p); + + return mount_needs_network(p->options, p->fstype); +} + +static bool mount_is_loop(const MountParameters *p) { + assert(p); + + if (fstab_test_option(p->options, "loop\0")) + return true; + + return false; +} + +static bool mount_is_bind(const MountParameters *p) { + assert(p); + + if (fstab_test_option(p->options, "bind\0" "rbind\0")) + return true; + + if (p->fstype && STR_IN_SET(p->fstype, "bind", "rbind")) + return true; + + return false; +} + +static bool mount_is_auto(const MountParameters *p) { + assert(p); + + return !fstab_test_option(p->options, "noauto\0"); +} + +static bool mount_is_automount(const MountParameters *p) { + assert(p); + + return fstab_test_option(p->options, + "comment=systemd.automount\0" + "x-systemd.automount\0"); +} + +static bool mount_state_active(MountState state) { + return IN_SET(state, + MOUNT_MOUNTING, + MOUNT_MOUNTING_DONE, + MOUNT_REMOUNTING, + MOUNT_UNMOUNTING, + MOUNT_MOUNTING_SIGTERM, + MOUNT_MOUNTING_SIGKILL, + MOUNT_UNMOUNTING_SIGTERM, + MOUNT_UNMOUNTING_SIGKILL, + MOUNT_REMOUNTING_SIGTERM, + MOUNT_REMOUNTING_SIGKILL); +} + +static bool needs_quota(const MountParameters *p) { + assert(p); + + /* Quotas are not enabled on network filesystems, + * but we want them, for example, on storage connected via iscsi */ + if (p->fstype && fstype_is_network(p->fstype)) + return false; + + if (mount_is_bind(p)) + return false; + + return fstab_test_option(p->options, + "usrquota\0" "grpquota\0" "quota\0" "usrjquota\0" "grpjquota\0"); +} + +static void mount_init(Unit *u) { + Mount *m = MOUNT(u); + + assert(u); + assert(u->load_state == UNIT_STUB); + + m->timeout_usec = u->manager->default_timeout_start_usec; + m->directory_mode = 0755; + + if (unit_has_name(u, "-.mount")) { + /* Don't allow start/stop for root directory */ + u->refuse_manual_start = true; + u->refuse_manual_stop = true; + } else { + /* The stdio/kmsg bridge socket is on /, in order to avoid a + * dep loop, don't use kmsg logging for -.mount */ + m->exec_context.std_output = u->manager->default_std_output; + m->exec_context.std_error = u->manager->default_std_error; + } + + /* We need to make sure that /usr/bin/mount is always called + * in the same process group as us, so that the autofs kernel + * side doesn't send us another mount request while we are + * already trying to comply its last one. */ + m->exec_context.same_pgrp = true; + + m->control_command_id = _MOUNT_EXEC_COMMAND_INVALID; + + u->ignore_on_isolate = true; +} + +static int mount_arm_timer(Mount *m, usec_t usec) { + int r; + + assert(m); + + if (m->timer_event_source) { + r = sd_event_source_set_time(m->timer_event_source, usec); + if (r < 0) + return r; + + return sd_event_source_set_enabled(m->timer_event_source, SD_EVENT_ONESHOT); + } + + if (usec == USEC_INFINITY) + return 0; + + r = sd_event_add_time( + UNIT(m)->manager->event, + &m->timer_event_source, + CLOCK_MONOTONIC, + usec, 0, + mount_dispatch_timer, m); + if (r < 0) + return r; + + (void) sd_event_source_set_description(m->timer_event_source, "mount-timer"); + + return 0; +} + +static void mount_unwatch_control_pid(Mount *m) { + assert(m); + + if (m->control_pid <= 0) + return; + + unit_unwatch_pid(UNIT(m), m->control_pid); + m->control_pid = 0; +} + +static void mount_parameters_done(MountParameters *p) { + assert(p); + + free(p->what); + free(p->options); + free(p->fstype); + + p->what = p->options = p->fstype = NULL; +} + +static void mount_done(Unit *u) { + Mount *m = MOUNT(u); + + assert(m); + + m->where = mfree(m->where); + + mount_parameters_done(&m->parameters_proc_self_mountinfo); + mount_parameters_done(&m->parameters_fragment); + + m->exec_runtime = exec_runtime_unref(m->exec_runtime); + exec_command_done_array(m->exec_command, _MOUNT_EXEC_COMMAND_MAX); + m->control_command = NULL; + + mount_unwatch_control_pid(m); + + m->timer_event_source = sd_event_source_unref(m->timer_event_source); +} + +_pure_ static MountParameters* get_mount_parameters_fragment(Mount *m) { + assert(m); + + if (m->from_fragment) + return &m->parameters_fragment; + + return NULL; +} + +_pure_ static MountParameters* get_mount_parameters(Mount *m) { + assert(m); + + if (m->from_proc_self_mountinfo) + return &m->parameters_proc_self_mountinfo; + + return get_mount_parameters_fragment(m); +} + +static int mount_add_mount_links(Mount *m) { + _cleanup_free_ char *parent = NULL; + MountParameters *pm; + Unit *other; + Iterator i; + Set *s; + int r; + + assert(m); + + if (!path_equal(m->where, "/")) { + /* Adds in links to other mount points that might lie further + * up in the hierarchy */ + + parent = dirname_malloc(m->where); + if (!parent) + return -ENOMEM; + + r = unit_require_mounts_for(UNIT(m), parent); + if (r < 0) + return r; + } + + /* Adds in links to other mount points that might be needed + * for the source path (if this is a bind mount or a loop mount) to be + * available. */ + pm = get_mount_parameters_fragment(m); + if (pm && pm->what && + path_is_absolute(pm->what) && + (mount_is_bind(pm) || mount_is_loop(pm) || !mount_is_network(pm))) { + + r = unit_require_mounts_for(UNIT(m), pm->what); + if (r < 0) + return r; + } + + /* Adds in links to other units that use this path or paths + * further down in the hierarchy */ + s = manager_get_units_requiring_mounts_for(UNIT(m)->manager, m->where); + SET_FOREACH(other, s, i) { + + if (other->load_state != UNIT_LOADED) + continue; + + if (other == UNIT(m)) + continue; + + r = unit_add_dependency(other, UNIT_AFTER, UNIT(m), true); + if (r < 0) + return r; + + if (UNIT(m)->fragment_path) { + /* If we have fragment configuration, then make this dependency required */ + r = unit_add_dependency(other, UNIT_REQUIRES, UNIT(m), true); + if (r < 0) + return r; + } + } + + return 0; +} + +static int mount_add_device_links(Mount *m) { + MountParameters *p; + bool device_wants_mount = false; + int r; + + assert(m); + + p = get_mount_parameters(m); + if (!p) + return 0; + + if (!p->what) + return 0; + + if (mount_is_bind(p)) + return 0; + + if (!is_device_path(p->what)) + return 0; + + /* /dev/root is a really weird thing, it's not a real device, + * but just a path the kernel exports for the root file system + * specified on the kernel command line. Ignore it here. */ + if (path_equal(p->what, "/dev/root")) + return 0; + + if (path_equal(m->where, "/")) + return 0; + + if (mount_is_auto(p) && !mount_is_automount(p) && MANAGER_IS_SYSTEM(UNIT(m)->manager)) + device_wants_mount = true; + + r = unit_add_node_link(UNIT(m), p->what, device_wants_mount, m->from_fragment ? UNIT_BINDS_TO : UNIT_REQUIRES); + if (r < 0) + return r; + + return 0; +} + +static int mount_add_quota_links(Mount *m) { + int r; + MountParameters *p; + + assert(m); + + if (!MANAGER_IS_SYSTEM(UNIT(m)->manager)) + return 0; + + p = get_mount_parameters_fragment(m); + if (!p) + return 0; + + if (!needs_quota(p)) + return 0; + + r = unit_add_two_dependencies_by_name(UNIT(m), UNIT_BEFORE, UNIT_WANTS, SPECIAL_QUOTACHECK_SERVICE, NULL, true); + if (r < 0) + return r; + + r = unit_add_two_dependencies_by_name(UNIT(m), UNIT_BEFORE, UNIT_WANTS, SPECIAL_QUOTAON_SERVICE, NULL, true); + if (r < 0) + return r; + + return 0; +} + +static bool should_umount(Mount *m) { + MountParameters *p; + + if (PATH_IN_SET(m->where, "/", "/usr") || + path_startswith(m->where, "/run/initramfs")) + return false; + + p = get_mount_parameters(m); + if (p && fstab_test_option(p->options, "x-initrd.mount\0") && + !in_initrd()) + return false; + + return true; +} + +static int mount_add_default_dependencies(Mount *m) { + MountParameters *p; + const char *after; + int r; + + assert(m); + + if (!UNIT(m)->default_dependencies) + return 0; + + if (!MANAGER_IS_SYSTEM(UNIT(m)->manager)) + return 0; + + /* We do not add any default dependencies to /, /usr or + * /run/initramfs/, since they are guaranteed to stay + * mounted the whole time, since our system is on it. + * Also, don't bother with anything mounted below virtual + * file systems, it's also going to be virtual, and hence + * not worth the effort. */ + if (PATH_IN_SET(m->where, "/", "/usr") || + path_startswith(m->where, "/run/initramfs") || + path_startswith(m->where, "/proc") || + path_startswith(m->where, "/sys") || + path_startswith(m->where, "/dev")) + return 0; + + p = get_mount_parameters(m); + if (!p) + return 0; + + if (mount_is_network(p)) { + /* We order ourselves after network.target. This is + * primarily useful at shutdown: services that take + * down the network should order themselves before + * network.target, so that they are shut down only + * after this mount unit is stopped. */ + + r = unit_add_dependency_by_name(UNIT(m), UNIT_AFTER, SPECIAL_NETWORK_TARGET, NULL, true); + if (r < 0) + return r; + + /* We pull in network-online.target, and order + * ourselves after it. This is useful at start-up to + * actively pull in tools that want to be started + * before we start mounting network file systems, and + * whose purpose it is to delay this until the network + * is "up". */ + + r = unit_add_two_dependencies_by_name(UNIT(m), UNIT_WANTS, UNIT_AFTER, SPECIAL_NETWORK_ONLINE_TARGET, NULL, true); + if (r < 0) + return r; + + after = SPECIAL_REMOTE_FS_PRE_TARGET; + } else + after = SPECIAL_LOCAL_FS_PRE_TARGET; + + r = unit_add_dependency_by_name(UNIT(m), UNIT_AFTER, after, NULL, true); + if (r < 0) + return r; + + if (should_umount(m)) { + r = unit_add_two_dependencies_by_name(UNIT(m), UNIT_BEFORE, UNIT_CONFLICTS, SPECIAL_UMOUNT_TARGET, NULL, true); + if (r < 0) + return r; + } + + return 0; +} + +static int mount_verify(Mount *m) { + _cleanup_free_ char *e = NULL; + int r; + + assert(m); + + if (UNIT(m)->load_state != UNIT_LOADED) + return 0; + + if (!m->from_fragment && !m->from_proc_self_mountinfo) + return -ENOENT; + + r = unit_name_from_path(m->where, ".mount", &e); + if (r < 0) + return log_unit_error_errno(UNIT(m), r, "Failed to generate unit name from mount path: %m"); + + if (!unit_has_name(UNIT(m), e)) { + log_unit_error(UNIT(m), "Where= setting doesn't match unit name. Refusing."); + return -EINVAL; + } + + if (mount_point_is_api(m->where) || mount_point_ignore(m->where)) { + log_unit_error(UNIT(m), "Cannot create mount unit for API file system %s. Refusing.", m->where); + return -EINVAL; + } + + if (UNIT(m)->fragment_path && !m->parameters_fragment.what) { + log_unit_error(UNIT(m), "What= setting is missing. Refusing."); + return -EBADMSG; + } + + if (m->exec_context.pam_name && m->kill_context.kill_mode != KILL_CONTROL_GROUP) { + log_unit_error(UNIT(m), "Unit has PAM enabled. Kill mode must be set to control-group'. Refusing."); + return -EINVAL; + } + + return 0; +} + +static int mount_add_extras(Mount *m) { + Unit *u = UNIT(m); + int r; + + assert(m); + + if (u->fragment_path) + m->from_fragment = true; + + if (!m->where) { + r = unit_name_to_path(u->id, &m->where); + if (r < 0) + return r; + } + + path_kill_slashes(m->where); + + if (!u->description) { + r = unit_set_description(u, m->where); + if (r < 0) + return r; + } + + r = mount_add_device_links(m); + if (r < 0) + return r; + + r = mount_add_mount_links(m); + if (r < 0) + return r; + + r = mount_add_quota_links(m); + if (r < 0) + return r; + + r = unit_patch_contexts(u); + if (r < 0) + return r; + + r = unit_add_exec_dependencies(u, &m->exec_context); + if (r < 0) + return r; + + r = unit_set_default_slice(u); + if (r < 0) + return r; + + r = mount_add_default_dependencies(m); + if (r < 0) + return r; + + return 0; +} + +static int mount_load(Unit *u) { + Mount *m = MOUNT(u); + int r; + + assert(u); + assert(u->load_state == UNIT_STUB); + + if (m->from_proc_self_mountinfo) + r = unit_load_fragment_and_dropin_optional(u); + else + r = unit_load_fragment_and_dropin(u); + + if (r < 0) + return r; + + /* This is a new unit? Then let's add in some extras */ + if (u->load_state == UNIT_LOADED) { + r = mount_add_extras(m); + if (r < 0) + return r; + } + + return mount_verify(m); +} + +static void mount_set_state(Mount *m, MountState state) { + MountState old_state; + assert(m); + + old_state = m->state; + m->state = state; + + if (!mount_state_active(state)) { + m->timer_event_source = sd_event_source_unref(m->timer_event_source); + mount_unwatch_control_pid(m); + m->control_command = NULL; + m->control_command_id = _MOUNT_EXEC_COMMAND_INVALID; + } + + if (state != old_state) + log_unit_debug(UNIT(m), "Changed %s -> %s", mount_state_to_string(old_state), mount_state_to_string(state)); + + unit_notify(UNIT(m), state_translation_table[old_state], state_translation_table[state], m->reload_result == MOUNT_SUCCESS); + m->reload_result = MOUNT_SUCCESS; +} + +static int mount_coldplug(Unit *u) { + Mount *m = MOUNT(u); + MountState new_state = MOUNT_DEAD; + int r; + + assert(m); + assert(m->state == MOUNT_DEAD); + + if (m->deserialized_state != m->state) + new_state = m->deserialized_state; + else if (m->from_proc_self_mountinfo) + new_state = MOUNT_MOUNTED; + + if (new_state == m->state) + return 0; + + if (m->control_pid > 0 && + pid_is_unwaited(m->control_pid) && + mount_state_active(new_state)) { + + r = unit_watch_pid(UNIT(m), m->control_pid); + if (r < 0) + return r; + + r = mount_arm_timer(m, usec_add(u->state_change_timestamp.monotonic, m->timeout_usec)); + if (r < 0) + return r; + } + + mount_set_state(m, new_state); + return 0; +} + +static void mount_dump(Unit *u, FILE *f, const char *prefix) { + Mount *m = MOUNT(u); + MountParameters *p; + + assert(m); + assert(f); + + p = get_mount_parameters(m); + + fprintf(f, + "%sMount State: %s\n" + "%sResult: %s\n" + "%sWhere: %s\n" + "%sWhat: %s\n" + "%sFile System Type: %s\n" + "%sOptions: %s\n" + "%sFrom /proc/self/mountinfo: %s\n" + "%sFrom fragment: %s\n" + "%sDirectoryMode: %04o\n", + prefix, mount_state_to_string(m->state), + prefix, mount_result_to_string(m->result), + prefix, m->where, + prefix, p ? strna(p->what) : "n/a", + prefix, p ? strna(p->fstype) : "n/a", + prefix, p ? strna(p->options) : "n/a", + prefix, yes_no(m->from_proc_self_mountinfo), + prefix, yes_no(m->from_fragment), + prefix, m->directory_mode); + + if (m->control_pid > 0) + fprintf(f, + "%sControl PID: "PID_FMT"\n", + prefix, m->control_pid); + + exec_context_dump(&m->exec_context, f, prefix); + kill_context_dump(&m->kill_context, f, prefix); +} + +static int mount_spawn(Mount *m, ExecCommand *c, pid_t *_pid) { + pid_t pid; + int r; + ExecParameters exec_params = { + .apply_permissions = true, + .apply_chroot = true, + .apply_tty_stdin = true, + .stdin_fd = -1, + .stdout_fd = -1, + .stderr_fd = -1, + }; + + assert(m); + assert(c); + assert(_pid); + + (void) unit_realize_cgroup(UNIT(m)); + if (m->reset_cpu_usage) { + (void) unit_reset_cpu_usage(UNIT(m)); + m->reset_cpu_usage = false; + } + + r = unit_setup_exec_runtime(UNIT(m)); + if (r < 0) + return r; + + r = mount_arm_timer(m, usec_add(now(CLOCK_MONOTONIC), m->timeout_usec)); + if (r < 0) + return r; + + exec_params.environment = UNIT(m)->manager->environment; + exec_params.confirm_spawn = UNIT(m)->manager->confirm_spawn; + exec_params.cgroup_supported = UNIT(m)->manager->cgroup_supported; + exec_params.cgroup_path = UNIT(m)->cgroup_path; + exec_params.cgroup_delegate = m->cgroup_context.delegate; + exec_params.runtime_prefix = manager_get_runtime_prefix(UNIT(m)->manager); + + r = exec_spawn(UNIT(m), + c, + &m->exec_context, + &exec_params, + m->exec_runtime, + &pid); + if (r < 0) + return r; + + r = unit_watch_pid(UNIT(m), pid); + if (r < 0) + /* FIXME: we need to do something here */ + return r; + + *_pid = pid; + + return 0; +} + +static void mount_enter_dead(Mount *m, MountResult f) { + assert(m); + + if (f != MOUNT_SUCCESS) + m->result = f; + + exec_runtime_destroy(m->exec_runtime); + m->exec_runtime = exec_runtime_unref(m->exec_runtime); + + exec_context_destroy_runtime_directory(&m->exec_context, manager_get_runtime_prefix(UNIT(m)->manager)); + + mount_set_state(m, m->result != MOUNT_SUCCESS ? MOUNT_FAILED : MOUNT_DEAD); +} + +static void mount_enter_mounted(Mount *m, MountResult f) { + assert(m); + + if (f != MOUNT_SUCCESS) + m->result = f; + + mount_set_state(m, MOUNT_MOUNTED); +} + +static void mount_enter_signal(Mount *m, MountState state, MountResult f) { + int r; + + assert(m); + + if (f != MOUNT_SUCCESS) + m->result = f; + + r = unit_kill_context( + UNIT(m), + &m->kill_context, + (state != MOUNT_MOUNTING_SIGTERM && state != MOUNT_UNMOUNTING_SIGTERM && state != MOUNT_REMOUNTING_SIGTERM) ? + KILL_KILL : KILL_TERMINATE, + -1, + m->control_pid, + false); + if (r < 0) + goto fail; + + if (r > 0) { + r = mount_arm_timer(m, usec_add(now(CLOCK_MONOTONIC), m->timeout_usec)); + if (r < 0) + goto fail; + + mount_set_state(m, state); + } else if (state == MOUNT_REMOUNTING_SIGTERM) + mount_enter_signal(m, MOUNT_REMOUNTING_SIGKILL, MOUNT_SUCCESS); + else if (state == MOUNT_REMOUNTING_SIGKILL) + mount_enter_mounted(m, MOUNT_SUCCESS); + else if (state == MOUNT_MOUNTING_SIGTERM) + mount_enter_signal(m, MOUNT_MOUNTING_SIGKILL, MOUNT_SUCCESS); + else if (state == MOUNT_UNMOUNTING_SIGTERM) + mount_enter_signal(m, MOUNT_UNMOUNTING_SIGKILL, MOUNT_SUCCESS); + else + mount_enter_dead(m, MOUNT_SUCCESS); + + return; + +fail: + log_unit_warning_errno(UNIT(m), r, "Failed to kill processes: %m"); + + if (state == MOUNT_REMOUNTING_SIGTERM || state == MOUNT_REMOUNTING_SIGKILL) + mount_enter_mounted(m, MOUNT_FAILURE_RESOURCES); + else + mount_enter_dead(m, MOUNT_FAILURE_RESOURCES); +} + +static void mount_enter_unmounting(Mount *m) { + int r; + + assert(m); + + /* Start counting our attempts */ + if (!IN_SET(m->state, + MOUNT_UNMOUNTING, + MOUNT_UNMOUNTING_SIGTERM, + MOUNT_UNMOUNTING_SIGKILL)) + m->n_retry_umount = 0; + + m->control_command_id = MOUNT_EXEC_UNMOUNT; + m->control_command = m->exec_command + MOUNT_EXEC_UNMOUNT; + + r = exec_command_set(m->control_command, UMOUNT_PATH, m->where, NULL); + if (r < 0) + goto fail; + + mount_unwatch_control_pid(m); + + r = mount_spawn(m, m->control_command, &m->control_pid); + if (r < 0) + goto fail; + + mount_set_state(m, MOUNT_UNMOUNTING); + + return; + +fail: + log_unit_warning_errno(UNIT(m), r, "Failed to run 'umount' task: %m"); + mount_enter_mounted(m, MOUNT_FAILURE_RESOURCES); +} + +static int mount_get_opts(Mount *m, char **ret) { + return fstab_filter_options(m->parameters_fragment.options, + "nofail\0" "noauto\0" "auto\0", NULL, NULL, ret); +} + +static void mount_enter_mounting(Mount *m) { + int r; + MountParameters *p; + + assert(m); + + m->control_command_id = MOUNT_EXEC_MOUNT; + m->control_command = m->exec_command + MOUNT_EXEC_MOUNT; + + r = unit_fail_if_symlink(UNIT(m), m->where); + if (r < 0) + goto fail; + + (void) mkdir_p_label(m->where, m->directory_mode); + + unit_warn_if_dir_nonempty(UNIT(m), m->where); + + /* Create the source directory for bind-mounts if needed */ + p = get_mount_parameters_fragment(m); + if (p && mount_is_bind(p)) + (void) mkdir_p_label(p->what, m->directory_mode); + + if (m->from_fragment) { + _cleanup_free_ char *opts = NULL; + + r = mount_get_opts(m, &opts); + if (r < 0) + goto fail; + + r = exec_command_set(m->control_command, MOUNT_PATH, + m->parameters_fragment.what, m->where, NULL); + if (r >= 0 && m->sloppy_options) + r = exec_command_append(m->control_command, "-s", NULL); + if (r >= 0 && m->parameters_fragment.fstype) + r = exec_command_append(m->control_command, "-t", m->parameters_fragment.fstype, NULL); + if (r >= 0 && !isempty(opts)) + r = exec_command_append(m->control_command, "-o", opts, NULL); + } else + r = -ENOENT; + + if (r < 0) + goto fail; + + mount_unwatch_control_pid(m); + + r = mount_spawn(m, m->control_command, &m->control_pid); + if (r < 0) + goto fail; + + mount_set_state(m, MOUNT_MOUNTING); + + return; + +fail: + log_unit_warning_errno(UNIT(m), r, "Failed to run 'mount' task: %m"); + mount_enter_dead(m, MOUNT_FAILURE_RESOURCES); +} + +static void mount_enter_remounting(Mount *m) { + int r; + + assert(m); + + m->control_command_id = MOUNT_EXEC_REMOUNT; + m->control_command = m->exec_command + MOUNT_EXEC_REMOUNT; + + if (m->from_fragment) { + const char *o; + + if (m->parameters_fragment.options) + o = strjoina("remount,", m->parameters_fragment.options); + else + o = "remount"; + + r = exec_command_set(m->control_command, MOUNT_PATH, + m->parameters_fragment.what, m->where, + "-o", o, NULL); + if (r >= 0 && m->sloppy_options) + r = exec_command_append(m->control_command, "-s", NULL); + if (r >= 0 && m->parameters_fragment.fstype) + r = exec_command_append(m->control_command, "-t", m->parameters_fragment.fstype, NULL); + } else + r = -ENOENT; + + if (r < 0) + goto fail; + + mount_unwatch_control_pid(m); + + r = mount_spawn(m, m->control_command, &m->control_pid); + if (r < 0) + goto fail; + + mount_set_state(m, MOUNT_REMOUNTING); + + return; + +fail: + log_unit_warning_errno(UNIT(m), r, "Failed to run 'remount' task: %m"); + m->reload_result = MOUNT_FAILURE_RESOURCES; + mount_enter_mounted(m, MOUNT_SUCCESS); +} + +static int mount_start(Unit *u) { + Mount *m = MOUNT(u); + int r; + + assert(m); + + /* We cannot fulfill this request right now, try again later + * please! */ + if (m->state == MOUNT_UNMOUNTING || + m->state == MOUNT_UNMOUNTING_SIGTERM || + m->state == MOUNT_UNMOUNTING_SIGKILL || + m->state == MOUNT_MOUNTING_SIGTERM || + m->state == MOUNT_MOUNTING_SIGKILL) + return -EAGAIN; + + /* Already on it! */ + if (m->state == MOUNT_MOUNTING) + return 0; + + assert(m->state == MOUNT_DEAD || m->state == MOUNT_FAILED); + + r = unit_start_limit_test(u); + if (r < 0) { + mount_enter_dead(m, MOUNT_FAILURE_START_LIMIT_HIT); + return r; + } + + m->result = MOUNT_SUCCESS; + m->reload_result = MOUNT_SUCCESS; + m->reset_cpu_usage = true; + + mount_enter_mounting(m); + return 1; +} + +static int mount_stop(Unit *u) { + Mount *m = MOUNT(u); + + assert(m); + + /* Already on it */ + if (m->state == MOUNT_UNMOUNTING || + m->state == MOUNT_UNMOUNTING_SIGKILL || + m->state == MOUNT_UNMOUNTING_SIGTERM || + m->state == MOUNT_MOUNTING_SIGTERM || + m->state == MOUNT_MOUNTING_SIGKILL) + return 0; + + assert(m->state == MOUNT_MOUNTING || + m->state == MOUNT_MOUNTING_DONE || + m->state == MOUNT_MOUNTED || + m->state == MOUNT_REMOUNTING || + m->state == MOUNT_REMOUNTING_SIGTERM || + m->state == MOUNT_REMOUNTING_SIGKILL); + + mount_enter_unmounting(m); + return 1; +} + +static int mount_reload(Unit *u) { + Mount *m = MOUNT(u); + + assert(m); + + if (m->state == MOUNT_MOUNTING_DONE) + return -EAGAIN; + + assert(m->state == MOUNT_MOUNTED); + + mount_enter_remounting(m); + return 1; +} + +static int mount_serialize(Unit *u, FILE *f, FDSet *fds) { + Mount *m = MOUNT(u); + + assert(m); + assert(f); + assert(fds); + + unit_serialize_item(u, f, "state", mount_state_to_string(m->state)); + unit_serialize_item(u, f, "result", mount_result_to_string(m->result)); + unit_serialize_item(u, f, "reload-result", mount_result_to_string(m->reload_result)); + + if (m->control_pid > 0) + unit_serialize_item_format(u, f, "control-pid", PID_FMT, m->control_pid); + + if (m->control_command_id >= 0) + unit_serialize_item(u, f, "control-command", mount_exec_command_to_string(m->control_command_id)); + + return 0; +} + +static int mount_deserialize_item(Unit *u, const char *key, const char *value, FDSet *fds) { + Mount *m = MOUNT(u); + + assert(u); + assert(key); + assert(value); + assert(fds); + + if (streq(key, "state")) { + MountState state; + + if ((state = mount_state_from_string(value)) < 0) + log_unit_debug(u, "Failed to parse state value: %s", value); + else + m->deserialized_state = state; + } else if (streq(key, "result")) { + MountResult f; + + f = mount_result_from_string(value); + if (f < 0) + log_unit_debug(u, "Failed to parse result value: %s", value); + else if (f != MOUNT_SUCCESS) + m->result = f; + + } else if (streq(key, "reload-result")) { + MountResult f; + + f = mount_result_from_string(value); + if (f < 0) + log_unit_debug(u, "Failed to parse reload result value: %s", value); + else if (f != MOUNT_SUCCESS) + m->reload_result = f; + + } else if (streq(key, "control-pid")) { + pid_t pid; + + if (parse_pid(value, &pid) < 0) + log_unit_debug(u, "Failed to parse control-pid value: %s", value); + else + m->control_pid = pid; + } else if (streq(key, "control-command")) { + MountExecCommand id; + + id = mount_exec_command_from_string(value); + if (id < 0) + log_unit_debug(u, "Failed to parse exec-command value: %s", value); + else { + m->control_command_id = id; + m->control_command = m->exec_command + id; + } + } else + log_unit_debug(u, "Unknown serialization key: %s", key); + + return 0; +} + +_pure_ static UnitActiveState mount_active_state(Unit *u) { + assert(u); + + return state_translation_table[MOUNT(u)->state]; +} + +_pure_ static const char *mount_sub_state_to_string(Unit *u) { + assert(u); + + return mount_state_to_string(MOUNT(u)->state); +} + +_pure_ static bool mount_check_gc(Unit *u) { + Mount *m = MOUNT(u); + + assert(m); + + return m->from_proc_self_mountinfo; +} + +static void mount_sigchld_event(Unit *u, pid_t pid, int code, int status) { + Mount *m = MOUNT(u); + MountResult f; + + assert(m); + assert(pid >= 0); + + if (pid != m->control_pid) + return; + + m->control_pid = 0; + + if (is_clean_exit(code, status, NULL)) + f = MOUNT_SUCCESS; + else if (code == CLD_EXITED) + f = MOUNT_FAILURE_EXIT_CODE; + else if (code == CLD_KILLED) + f = MOUNT_FAILURE_SIGNAL; + else if (code == CLD_DUMPED) + f = MOUNT_FAILURE_CORE_DUMP; + else + assert_not_reached("Unknown code"); + + if (f != MOUNT_SUCCESS) + m->result = f; + + if (m->control_command) { + exec_status_exit(&m->control_command->exec_status, &m->exec_context, pid, code, status); + + m->control_command = NULL; + m->control_command_id = _MOUNT_EXEC_COMMAND_INVALID; + } + + log_unit_full(u, f == MOUNT_SUCCESS ? LOG_DEBUG : LOG_NOTICE, 0, + "Mount process exited, code=%s status=%i", sigchld_code_to_string(code), status); + + /* Note that mount(8) returning and the kernel sending us a + * mount table change event might happen out-of-order. If an + * operation succeed we assume the kernel will follow soon too + * and already change into the resulting state. If it fails + * we check if the kernel still knows about the mount. and + * change state accordingly. */ + + switch (m->state) { + + case MOUNT_MOUNTING: + case MOUNT_MOUNTING_DONE: + case MOUNT_MOUNTING_SIGKILL: + case MOUNT_MOUNTING_SIGTERM: + + if (f == MOUNT_SUCCESS) + mount_enter_mounted(m, f); + else if (m->from_proc_self_mountinfo) + mount_enter_mounted(m, f); + else + mount_enter_dead(m, f); + break; + + case MOUNT_REMOUNTING: + case MOUNT_REMOUNTING_SIGKILL: + case MOUNT_REMOUNTING_SIGTERM: + + m->reload_result = f; + if (m->from_proc_self_mountinfo) + mount_enter_mounted(m, MOUNT_SUCCESS); + else + mount_enter_dead(m, MOUNT_SUCCESS); + + break; + + case MOUNT_UNMOUNTING: + case MOUNT_UNMOUNTING_SIGKILL: + case MOUNT_UNMOUNTING_SIGTERM: + + if (f == MOUNT_SUCCESS) { + + if (m->from_proc_self_mountinfo) { + + /* Still a mount point? If so, let's + * try again. Most likely there were + * multiple mount points stacked on + * top of each other. Note that due to + * the io event priority logic we can + * be sure the new mountinfo is loaded + * before we process the SIGCHLD for + * the mount command. */ + + if (m->n_retry_umount < RETRY_UMOUNT_MAX) { + log_unit_debug(u, "Mount still present, trying again."); + m->n_retry_umount++; + mount_enter_unmounting(m); + } else { + log_unit_debug(u, "Mount still present after %u attempts to unmount, giving up.", m->n_retry_umount); + mount_enter_mounted(m, f); + } + } else + mount_enter_dead(m, f); + + } else if (m->from_proc_self_mountinfo) + mount_enter_mounted(m, f); + else + mount_enter_dead(m, f); + break; + + default: + assert_not_reached("Uh, control process died at wrong time."); + } + + /* Notify clients about changed exit status */ + unit_add_to_dbus_queue(u); +} + +static int mount_dispatch_timer(sd_event_source *source, usec_t usec, void *userdata) { + Mount *m = MOUNT(userdata); + + assert(m); + assert(m->timer_event_source == source); + + switch (m->state) { + + case MOUNT_MOUNTING: + case MOUNT_MOUNTING_DONE: + log_unit_warning(UNIT(m), "Mounting timed out. Stopping."); + mount_enter_signal(m, MOUNT_MOUNTING_SIGTERM, MOUNT_FAILURE_TIMEOUT); + break; + + case MOUNT_REMOUNTING: + log_unit_warning(UNIT(m), "Remounting timed out. Stopping."); + m->reload_result = MOUNT_FAILURE_TIMEOUT; + mount_enter_mounted(m, MOUNT_SUCCESS); + break; + + case MOUNT_UNMOUNTING: + log_unit_warning(UNIT(m), "Unmounting timed out. Stopping."); + mount_enter_signal(m, MOUNT_UNMOUNTING_SIGTERM, MOUNT_FAILURE_TIMEOUT); + break; + + case MOUNT_MOUNTING_SIGTERM: + if (m->kill_context.send_sigkill) { + log_unit_warning(UNIT(m), "Mounting timed out. Killing."); + mount_enter_signal(m, MOUNT_MOUNTING_SIGKILL, MOUNT_FAILURE_TIMEOUT); + } else { + log_unit_warning(UNIT(m), "Mounting timed out. Skipping SIGKILL. Ignoring."); + + if (m->from_proc_self_mountinfo) + mount_enter_mounted(m, MOUNT_FAILURE_TIMEOUT); + else + mount_enter_dead(m, MOUNT_FAILURE_TIMEOUT); + } + break; + + case MOUNT_REMOUNTING_SIGTERM: + if (m->kill_context.send_sigkill) { + log_unit_warning(UNIT(m), "Remounting timed out. Killing."); + mount_enter_signal(m, MOUNT_REMOUNTING_SIGKILL, MOUNT_FAILURE_TIMEOUT); + } else { + log_unit_warning(UNIT(m), "Remounting timed out. Skipping SIGKILL. Ignoring."); + + if (m->from_proc_self_mountinfo) + mount_enter_mounted(m, MOUNT_FAILURE_TIMEOUT); + else + mount_enter_dead(m, MOUNT_FAILURE_TIMEOUT); + } + break; + + case MOUNT_UNMOUNTING_SIGTERM: + if (m->kill_context.send_sigkill) { + log_unit_warning(UNIT(m), "Unmounting timed out. Killing."); + mount_enter_signal(m, MOUNT_UNMOUNTING_SIGKILL, MOUNT_FAILURE_TIMEOUT); + } else { + log_unit_warning(UNIT(m), "Unmounting timed out. Skipping SIGKILL. Ignoring."); + + if (m->from_proc_self_mountinfo) + mount_enter_mounted(m, MOUNT_FAILURE_TIMEOUT); + else + mount_enter_dead(m, MOUNT_FAILURE_TIMEOUT); + } + break; + + case MOUNT_MOUNTING_SIGKILL: + case MOUNT_REMOUNTING_SIGKILL: + case MOUNT_UNMOUNTING_SIGKILL: + log_unit_warning(UNIT(m),"Mount process still around after SIGKILL. Ignoring."); + + if (m->from_proc_self_mountinfo) + mount_enter_mounted(m, MOUNT_FAILURE_TIMEOUT); + else + mount_enter_dead(m, MOUNT_FAILURE_TIMEOUT); + break; + + default: + assert_not_reached("Timeout at wrong time."); + } + + return 0; +} + +static int mount_setup_unit( + Manager *m, + const char *what, + const char *where, + const char *options, + const char *fstype, + bool set_flags) { + + _cleanup_free_ char *e = NULL, *w = NULL, *o = NULL, *f = NULL; + bool load_extras = false; + MountParameters *p; + bool delete, changed = false; + Unit *u; + int r; + + assert(m); + assert(what); + assert(where); + assert(options); + assert(fstype); + + /* Ignore API mount points. They should never be referenced in + * dependencies ever. */ + if (mount_point_is_api(where) || mount_point_ignore(where)) + return 0; + + if (streq(fstype, "autofs")) + return 0; + + /* probably some kind of swap, ignore */ + if (!is_path(where)) + return 0; + + r = unit_name_from_path(where, ".mount", &e); + if (r < 0) + return r; + + u = manager_get_unit(m, e); + if (!u) { + delete = true; + + u = unit_new(m, sizeof(Mount)); + if (!u) + return log_oom(); + + r = unit_add_name(u, e); + if (r < 0) + goto fail; + + MOUNT(u)->where = strdup(where); + if (!MOUNT(u)->where) { + r = -ENOMEM; + goto fail; + } + + u->source_path = strdup("/proc/self/mountinfo"); + if (!u->source_path) { + r = -ENOMEM; + goto fail; + } + + if (MANAGER_IS_SYSTEM(m)) { + const char* target; + + target = mount_needs_network(options, fstype) ? SPECIAL_REMOTE_FS_TARGET : SPECIAL_LOCAL_FS_TARGET; + r = unit_add_dependency_by_name(u, UNIT_BEFORE, target, NULL, true); + if (r < 0) + goto fail; + + if (should_umount(MOUNT(u))) { + r = unit_add_dependency_by_name(u, UNIT_CONFLICTS, SPECIAL_UMOUNT_TARGET, NULL, true); + if (r < 0) + goto fail; + } + } + + unit_add_to_load_queue(u); + changed = true; + } else { + delete = false; + + if (!MOUNT(u)->where) { + MOUNT(u)->where = strdup(where); + if (!MOUNT(u)->where) { + r = -ENOMEM; + goto fail; + } + } + + if (MANAGER_IS_SYSTEM(m) && + mount_needs_network(options, fstype)) { + /* _netdev option may have shown up late, or on a + * remount. Add remote-fs dependencies, even though + * local-fs ones may already be there. */ + unit_add_dependency_by_name(u, UNIT_BEFORE, SPECIAL_REMOTE_FS_TARGET, NULL, true); + load_extras = true; + } + + if (u->load_state == UNIT_NOT_FOUND) { + u->load_state = UNIT_LOADED; + u->load_error = 0; + + /* Load in the extras later on, after we + * finished initialization of the unit */ + load_extras = true; + changed = true; + } + } + + w = strdup(what); + o = strdup(options); + f = strdup(fstype); + if (!w || !o || !f) { + r = -ENOMEM; + goto fail; + } + + p = &MOUNT(u)->parameters_proc_self_mountinfo; + + changed = changed || + !streq_ptr(p->options, options) || + !streq_ptr(p->what, what) || + !streq_ptr(p->fstype, fstype); + + if (set_flags) { + MOUNT(u)->is_mounted = true; + MOUNT(u)->just_mounted = !MOUNT(u)->from_proc_self_mountinfo; + MOUNT(u)->just_changed = changed; + } + + MOUNT(u)->from_proc_self_mountinfo = true; + + free(p->what); + p->what = w; + w = NULL; + + free(p->options); + p->options = o; + o = NULL; + + free(p->fstype); + p->fstype = f; + f = NULL; + + if (load_extras) { + r = mount_add_extras(MOUNT(u)); + if (r < 0) + goto fail; + } + + if (changed) + unit_add_to_dbus_queue(u); + + return 0; + +fail: + log_warning_errno(r, "Failed to set up mount unit: %m"); + + if (delete && u) + unit_free(u); + + return r; +} + +static int mount_load_proc_self_mountinfo(Manager *m, bool set_flags) { + _cleanup_(mnt_free_tablep) struct libmnt_table *t = NULL; + _cleanup_(mnt_free_iterp) struct libmnt_iter *i = NULL; + int r = 0; + + assert(m); + + t = mnt_new_table(); + if (!t) + return log_oom(); + + i = mnt_new_iter(MNT_ITER_FORWARD); + if (!i) + return log_oom(); + + r = mnt_table_parse_mtab(t, NULL); + if (r < 0) + return log_error_errno(r, "Failed to parse /proc/self/mountinfo: %m"); + + r = 0; + for (;;) { + const char *device, *path, *options, *fstype; + _cleanup_free_ char *d = NULL, *p = NULL; + struct libmnt_fs *fs; + int k; + + k = mnt_table_next_fs(t, i, &fs); + if (k == 1) + break; + if (k < 0) + return log_error_errno(k, "Failed to get next entry from /proc/self/mountinfo: %m"); + + device = mnt_fs_get_source(fs); + path = mnt_fs_get_target(fs); + options = mnt_fs_get_options(fs); + fstype = mnt_fs_get_fstype(fs); + + if (!device || !path) + continue; + + if (cunescape(device, UNESCAPE_RELAX, &d) < 0) + return log_oom(); + + if (cunescape(path, UNESCAPE_RELAX, &p) < 0) + return log_oom(); + + (void) device_found_node(m, d, true, DEVICE_FOUND_MOUNT, set_flags); + + k = mount_setup_unit(m, d, p, options, fstype, set_flags); + if (r == 0 && k < 0) + r = k; + } + + return r; +} + +static void mount_shutdown(Manager *m) { + + assert(m); + + m->mount_event_source = sd_event_source_unref(m->mount_event_source); + + mnt_unref_monitor(m->mount_monitor); + m->mount_monitor = NULL; +} + +static int mount_get_timeout(Unit *u, usec_t *timeout) { + Mount *m = MOUNT(u); + usec_t t; + int r; + + if (!m->timer_event_source) + return 0; + + r = sd_event_source_get_time(m->timer_event_source, &t); + if (r < 0) + return r; + if (t == USEC_INFINITY) + return 0; + + *timeout = t; + return 1; +} + +static void mount_enumerate(Manager *m) { + int r; + + assert(m); + + mnt_init_debug(0); + + if (!m->mount_monitor) { + int fd; + + m->mount_monitor = mnt_new_monitor(); + if (!m->mount_monitor) { + log_oom(); + goto fail; + } + + r = mnt_monitor_enable_kernel(m->mount_monitor, 1); + if (r < 0) { + log_error_errno(r, "Failed to enable watching of kernel mount events: %m"); + goto fail; + } + + r = mnt_monitor_enable_userspace(m->mount_monitor, 1, NULL); + if (r < 0) { + log_error_errno(r, "Failed to enable watching of userspace mount events: %m"); + goto fail; + } + + /* mnt_unref_monitor() will close the fd */ + fd = r = mnt_monitor_get_fd(m->mount_monitor); + if (r < 0) { + log_error_errno(r, "Failed to acquire watch file descriptor: %m"); + goto fail; + } + + r = sd_event_add_io(m->event, &m->mount_event_source, fd, EPOLLIN, mount_dispatch_io, m); + if (r < 0) { + log_error_errno(r, "Failed to watch mount file descriptor: %m"); + goto fail; + } + + r = sd_event_source_set_priority(m->mount_event_source, -10); + if (r < 0) { + log_error_errno(r, "Failed to adjust mount watch priority: %m"); + goto fail; + } + + (void) sd_event_source_set_description(m->mount_event_source, "mount-monitor-dispatch"); + } + + r = mount_load_proc_self_mountinfo(m, false); + if (r < 0) + goto fail; + + return; + +fail: + mount_shutdown(m); +} + +static int mount_dispatch_io(sd_event_source *source, int fd, uint32_t revents, void *userdata) { + _cleanup_set_free_ Set *around = NULL, *gone = NULL; + Manager *m = userdata; + const char *what; + Iterator i; + Unit *u; + int r; + + assert(m); + assert(revents & EPOLLIN); + + if (fd == mnt_monitor_get_fd(m->mount_monitor)) { + bool rescan = false; + + /* Drain all events and verify that the event is valid. + * + * Note that libmount also monitors /run/mount mkdir if the + * directory does not exist yet. The mkdir may generate event + * which is irrelevant for us. + * + * error: r < 0; valid: r == 0, false positive: rc == 1 */ + do { + r = mnt_monitor_next_change(m->mount_monitor, NULL, NULL); + if (r == 0) + rescan = true; + else if (r < 0) + return log_error_errno(r, "Failed to drain libmount events"); + } while (r == 0); + + log_debug("libmount event [rescan: %s]", yes_no(rescan)); + if (!rescan) + return 0; + } + + r = mount_load_proc_self_mountinfo(m, true); + if (r < 0) { + /* Reset flags, just in case, for later calls */ + LIST_FOREACH(units_by_type, u, m->units_by_type[UNIT_MOUNT]) { + Mount *mount = MOUNT(u); + + mount->is_mounted = mount->just_mounted = mount->just_changed = false; + } + + return 0; + } + + manager_dispatch_load_queue(m); + + LIST_FOREACH(units_by_type, u, m->units_by_type[UNIT_MOUNT]) { + Mount *mount = MOUNT(u); + + if (!mount->is_mounted) { + + /* A mount point is not around right now. It + * might be gone, or might never have + * existed. */ + + if (mount->from_proc_self_mountinfo && + mount->parameters_proc_self_mountinfo.what) { + + /* Remember that this device might just have disappeared */ + if (set_ensure_allocated(&gone, &string_hash_ops) < 0 || + set_put(gone, mount->parameters_proc_self_mountinfo.what) < 0) + log_oom(); /* we don't care too much about OOM here... */ + } + + mount->from_proc_self_mountinfo = false; + + switch (mount->state) { + + case MOUNT_MOUNTED: + /* This has just been unmounted by + * somebody else, follow the state + * change. */ + mount->result = MOUNT_SUCCESS; /* make sure we forget any earlier umount failures */ + mount_enter_dead(mount, MOUNT_SUCCESS); + break; + + default: + break; + } + + } else if (mount->just_mounted || mount->just_changed) { + + /* A mount point was added or changed */ + + switch (mount->state) { + + case MOUNT_DEAD: + case MOUNT_FAILED: + /* This has just been mounted by + * somebody else, follow the state + * change. */ + mount_enter_mounted(mount, MOUNT_SUCCESS); + break; + + case MOUNT_MOUNTING: + mount_set_state(mount, MOUNT_MOUNTING_DONE); + break; + + default: + /* Nothing really changed, but let's + * issue an notification call + * nonetheless, in case somebody is + * waiting for this. (e.g. file system + * ro/rw remounts.) */ + mount_set_state(mount, mount->state); + break; + } + } + + if (mount->is_mounted && + mount->from_proc_self_mountinfo && + mount->parameters_proc_self_mountinfo.what) { + + if (set_ensure_allocated(&around, &string_hash_ops) < 0 || + set_put(around, mount->parameters_proc_self_mountinfo.what) < 0) + log_oom(); + } + + /* Reset the flags for later calls */ + mount->is_mounted = mount->just_mounted = mount->just_changed = false; + } + + SET_FOREACH(what, gone, i) { + if (set_contains(around, what)) + continue; + + /* Let the device units know that the device is no longer mounted */ + (void) device_found_node(m, what, false, DEVICE_FOUND_MOUNT, true); + } + + return 0; +} + +static void mount_reset_failed(Unit *u) { + Mount *m = MOUNT(u); + + assert(m); + + if (m->state == MOUNT_FAILED) + mount_set_state(m, MOUNT_DEAD); + + m->result = MOUNT_SUCCESS; + m->reload_result = MOUNT_SUCCESS; +} + +static int mount_kill(Unit *u, KillWho who, int signo, sd_bus_error *error) { + return unit_kill_common(u, who, signo, -1, MOUNT(u)->control_pid, error); +} + +static int mount_control_pid(Unit *u) { + Mount *m = MOUNT(u); + + assert(m); + + return m->control_pid; +} + +static const char* const mount_exec_command_table[_MOUNT_EXEC_COMMAND_MAX] = { + [MOUNT_EXEC_MOUNT] = "ExecMount", + [MOUNT_EXEC_UNMOUNT] = "ExecUnmount", + [MOUNT_EXEC_REMOUNT] = "ExecRemount", +}; + +DEFINE_STRING_TABLE_LOOKUP(mount_exec_command, MountExecCommand); + +static const char* const mount_result_table[_MOUNT_RESULT_MAX] = { + [MOUNT_SUCCESS] = "success", + [MOUNT_FAILURE_RESOURCES] = "resources", + [MOUNT_FAILURE_TIMEOUT] = "timeout", + [MOUNT_FAILURE_EXIT_CODE] = "exit-code", + [MOUNT_FAILURE_SIGNAL] = "signal", + [MOUNT_FAILURE_CORE_DUMP] = "core-dump", + [MOUNT_FAILURE_START_LIMIT_HIT] = "start-limit-hit", +}; + +DEFINE_STRING_TABLE_LOOKUP(mount_result, MountResult); + +const UnitVTable mount_vtable = { + .object_size = sizeof(Mount), + .exec_context_offset = offsetof(Mount, exec_context), + .cgroup_context_offset = offsetof(Mount, cgroup_context), + .kill_context_offset = offsetof(Mount, kill_context), + .exec_runtime_offset = offsetof(Mount, exec_runtime), + + .sections = + "Unit\0" + "Mount\0" + "Install\0", + .private_section = "Mount", + + .init = mount_init, + .load = mount_load, + .done = mount_done, + + .coldplug = mount_coldplug, + + .dump = mount_dump, + + .start = mount_start, + .stop = mount_stop, + .reload = mount_reload, + + .kill = mount_kill, + + .serialize = mount_serialize, + .deserialize_item = mount_deserialize_item, + + .active_state = mount_active_state, + .sub_state_to_string = mount_sub_state_to_string, + + .check_gc = mount_check_gc, + + .sigchld_event = mount_sigchld_event, + + .reset_failed = mount_reset_failed, + + .control_pid = mount_control_pid, + + .bus_vtable = bus_mount_vtable, + .bus_set_property = bus_mount_set_property, + .bus_commit_properties = bus_mount_commit_properties, + + .get_timeout = mount_get_timeout, + + .can_transient = true, + + .enumerate = mount_enumerate, + .shutdown = mount_shutdown, + + .status_message_formats = { + .starting_stopping = { + [0] = "Mounting %s...", + [1] = "Unmounting %s...", + }, + .finished_start_job = { + [JOB_DONE] = "Mounted %s.", + [JOB_FAILED] = "Failed to mount %s.", + [JOB_TIMEOUT] = "Timed out mounting %s.", + }, + .finished_stop_job = { + [JOB_DONE] = "Unmounted %s.", + [JOB_FAILED] = "Failed unmounting %s.", + [JOB_TIMEOUT] = "Timed out unmounting %s.", + }, + }, +}; diff --git a/src/grp-system/libcore/src/namespace.c b/src/grp-system/libcore/src/namespace.c new file mode 100644 index 0000000000..f67b0136bf --- /dev/null +++ b/src/grp-system/libcore/src/namespace.c @@ -0,0 +1,664 @@ +/*** + This file is part of systemd. + + Copyright 2010 Lennart Poettering + + systemd is free software; you can redistribute it and/or modify it + under the terms of the GNU Lesser General Public License as published by + the Free Software Foundation; either version 2.1 of the License, or + (at your option) any later version. + + systemd is distributed in the hope that it will be useful, but + WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + Lesser General Public License for more details. + + You should have received a copy of the GNU Lesser General Public License + along with systemd; If not, see . +***/ + +#include +#include +#include +#include +#include +#include +#include + +#include + +#include "core/loopback-setup.h" +#include "core/namespace.h" +#include "systemd-basic/alloc-util.h" +#include "systemd-basic/fd-util.h" +#include "systemd-basic/missing.h" +#include "systemd-basic/mkdir.h" +#include "systemd-basic/mount-util.h" +#include "systemd-basic/path-util.h" +#include "systemd-basic/selinux-util.h" +#include "systemd-basic/socket-util.h" +#include "systemd-basic/string-table.h" +#include "systemd-basic/string-util.h" +#include "systemd-basic/strv.h" +#include "systemd-basic/umask-util.h" +#include "systemd-basic/user-util.h" +#include "systemd-basic/util.h" +#include "systemd-shared/dev-setup.h" + +#define DEV_MOUNT_OPTIONS (MS_NOSUID|MS_STRICTATIME|MS_NOEXEC) + +typedef enum MountMode { + /* This is ordered by priority! */ + INACCESSIBLE, + READONLY, + PRIVATE_TMP, + PRIVATE_VAR_TMP, + PRIVATE_DEV, + READWRITE +} MountMode; + +typedef struct BindMount { + const char *path; + MountMode mode; + bool done; + bool ignore; +} BindMount; + +static int append_mounts(BindMount **p, char **strv, MountMode mode) { + char **i; + + assert(p); + + STRV_FOREACH(i, strv) { + + (*p)->ignore = false; + (*p)->done = false; + + if ((mode == INACCESSIBLE || mode == READONLY || mode == READWRITE) && (*i)[0] == '-') { + (*p)->ignore = true; + (*i)++; + } + + if (!path_is_absolute(*i)) + return -EINVAL; + + (*p)->path = *i; + (*p)->mode = mode; + (*p)++; + } + + return 0; +} + +static int mount_path_compare(const void *a, const void *b) { + const BindMount *p = a, *q = b; + int d; + + d = path_compare(p->path, q->path); + + if (d == 0) { + /* If the paths are equal, check the mode */ + if (p->mode < q->mode) + return -1; + + if (p->mode > q->mode) + return 1; + + return 0; + } + + /* If the paths are not equal, then order prefixes first */ + return d; +} + +static void drop_duplicates(BindMount *m, unsigned *n) { + BindMount *f, *t, *previous; + + assert(m); + assert(n); + + for (f = m, t = m, previous = NULL; f < m+*n; f++) { + + /* The first one wins */ + if (previous && path_equal(f->path, previous->path)) + continue; + + *t = *f; + + previous = t; + + t++; + } + + *n = t - m; +} + +static int mount_dev(BindMount *m) { + static const char devnodes[] = + "/dev/null\0" + "/dev/zero\0" + "/dev/full\0" + "/dev/random\0" + "/dev/urandom\0" + "/dev/tty\0"; + + char temporary_mount[] = "/tmp/namespace-dev-XXXXXX"; + const char *d, *dev = NULL, *devpts = NULL, *devshm = NULL, *devhugepages = NULL, *devmqueue = NULL, *devlog = NULL, *devptmx = NULL; + _cleanup_umask_ mode_t u; + int r; + + assert(m); + + u = umask(0000); + + if (!mkdtemp(temporary_mount)) + return -errno; + + dev = strjoina(temporary_mount, "/dev"); + (void) mkdir(dev, 0755); + if (mount("tmpfs", dev, "tmpfs", DEV_MOUNT_OPTIONS, "mode=755") < 0) { + r = -errno; + goto fail; + } + + devpts = strjoina(temporary_mount, "/dev/pts"); + (void) mkdir(devpts, 0755); + if (mount("/dev/pts", devpts, NULL, MS_BIND, NULL) < 0) { + r = -errno; + goto fail; + } + + devptmx = strjoina(temporary_mount, "/dev/ptmx"); + if (symlink("pts/ptmx", devptmx) < 0) { + r = -errno; + goto fail; + } + + devshm = strjoina(temporary_mount, "/dev/shm"); + (void) mkdir(devshm, 01777); + r = mount("/dev/shm", devshm, NULL, MS_BIND, NULL); + if (r < 0) { + r = -errno; + goto fail; + } + + devmqueue = strjoina(temporary_mount, "/dev/mqueue"); + (void) mkdir(devmqueue, 0755); + (void) mount("/dev/mqueue", devmqueue, NULL, MS_BIND, NULL); + + devhugepages = strjoina(temporary_mount, "/dev/hugepages"); + (void) mkdir(devhugepages, 0755); + (void) mount("/dev/hugepages", devhugepages, NULL, MS_BIND, NULL); + + devlog = strjoina(temporary_mount, "/dev/log"); + (void) symlink("/run/systemd/journal/dev-log", devlog); + + NULSTR_FOREACH(d, devnodes) { + _cleanup_free_ char *dn = NULL; + struct stat st; + + r = stat(d, &st); + if (r < 0) { + + if (errno == ENOENT) + continue; + + r = -errno; + goto fail; + } + + if (!S_ISBLK(st.st_mode) && + !S_ISCHR(st.st_mode)) { + r = -EINVAL; + goto fail; + } + + if (st.st_rdev == 0) + continue; + + dn = strappend(temporary_mount, d); + if (!dn) { + r = -ENOMEM; + goto fail; + } + + mac_selinux_create_file_prepare(d, st.st_mode); + r = mknod(dn, st.st_mode, st.st_rdev); + mac_selinux_create_file_clear(); + + if (r < 0) { + r = -errno; + goto fail; + } + } + + dev_setup(temporary_mount, UID_INVALID, GID_INVALID); + + /* Create the /dev directory if missing. It is more likely to be + * missing when the service is started with RootDirectory. This is + * consistent with mount units creating the mount points when missing. + */ + (void) mkdir_p_label(m->path, 0755); + + /* Unmount everything in old /dev */ + umount_recursive(m->path, 0); + if (mount(dev, m->path, NULL, MS_MOVE, NULL) < 0) { + r = -errno; + goto fail; + } + + rmdir(dev); + rmdir(temporary_mount); + + return 0; + +fail: + if (devpts) + umount(devpts); + + if (devshm) + umount(devshm); + + if (devhugepages) + umount(devhugepages); + + if (devmqueue) + umount(devmqueue); + + umount(dev); + rmdir(dev); + rmdir(temporary_mount); + + return r; +} + +static int apply_mount( + BindMount *m, + const char *tmp_dir, + const char *var_tmp_dir) { + + const char *what; + int r; + struct stat target; + + assert(m); + + switch (m->mode) { + + case INACCESSIBLE: + + /* First, get rid of everything that is below if there + * is anything... Then, overmount it with an + * inaccessible path. */ + umount_recursive(m->path, 0); + + if (lstat(m->path, &target) < 0) { + if (m->ignore && errno == ENOENT) + return 0; + return -errno; + } + + what = mode_to_inaccessible_node(target.st_mode); + if (!what) { + log_debug("File type not supported for inaccessible mounts. Note that symlinks are not allowed"); + return -ELOOP; + } + break; + case READONLY: + case READWRITE: + /* Nothing to mount here, we just later toggle the + * MS_RDONLY bit for the mount point */ + return 0; + + case PRIVATE_TMP: + what = tmp_dir; + break; + + case PRIVATE_VAR_TMP: + what = var_tmp_dir; + break; + + case PRIVATE_DEV: + return mount_dev(m); + + default: + assert_not_reached("Unknown mode"); + } + + assert(what); + + r = mount(what, m->path, NULL, MS_BIND|MS_REC, NULL); + if (r >= 0) { + log_debug("Successfully mounted %s to %s", what, m->path); + return r; + } else { + if (m->ignore && errno == ENOENT) + return 0; + return log_debug_errno(errno, "Failed to mount %s to %s: %m", what, m->path); + } +} + +static int make_read_only(BindMount *m) { + int r; + + assert(m); + + if (IN_SET(m->mode, INACCESSIBLE, READONLY)) + r = bind_remount_recursive(m->path, true); + else if (IN_SET(m->mode, READWRITE, PRIVATE_TMP, PRIVATE_VAR_TMP, PRIVATE_DEV)) { + r = bind_remount_recursive(m->path, false); + if (r == 0 && m->mode == PRIVATE_DEV) /* can be readonly but the submounts can't*/ + if (mount(NULL, m->path, NULL, MS_REMOUNT|DEV_MOUNT_OPTIONS|MS_RDONLY, NULL) < 0) + r = -errno; + } else + r = 0; + + if (m->ignore && r == -ENOENT) + return 0; + + return r; +} + +int setup_namespace( + const char* root_directory, + char** read_write_paths, + char** read_only_paths, + char** inaccessible_paths, + const char* tmp_dir, + const char* var_tmp_dir, + bool private_dev, + ProtectHome protect_home, + ProtectSystem protect_system, + unsigned long mount_flags) { + + BindMount *m, *mounts = NULL; + unsigned n; + int r = 0; + + if (mount_flags == 0) + mount_flags = MS_SHARED; + + if (unshare(CLONE_NEWNS) < 0) + return -errno; + + n = !!tmp_dir + !!var_tmp_dir + + strv_length(read_write_paths) + + strv_length(read_only_paths) + + strv_length(inaccessible_paths) + + private_dev + + (protect_home != PROTECT_HOME_NO ? 3 : 0) + + (protect_system != PROTECT_SYSTEM_NO ? 2 : 0) + + (protect_system == PROTECT_SYSTEM_FULL ? 1 : 0); + + if (n > 0) { + m = mounts = (BindMount *) alloca0(n * sizeof(BindMount)); + r = append_mounts(&m, read_write_paths, READWRITE); + if (r < 0) + return r; + + r = append_mounts(&m, read_only_paths, READONLY); + if (r < 0) + return r; + + r = append_mounts(&m, inaccessible_paths, INACCESSIBLE); + if (r < 0) + return r; + + if (tmp_dir) { + m->path = prefix_roota(root_directory, "/tmp"); + m->mode = PRIVATE_TMP; + m++; + } + + if (var_tmp_dir) { + m->path = prefix_roota(root_directory, "/var/tmp"); + m->mode = PRIVATE_VAR_TMP; + m++; + } + + if (private_dev) { + m->path = prefix_roota(root_directory, "/dev"); + m->mode = PRIVATE_DEV; + m++; + } + + if (protect_home != PROTECT_HOME_NO) { + const char *home_dir, *run_user_dir, *root_dir; + + home_dir = prefix_roota(root_directory, "/home"); + home_dir = strjoina("-", home_dir); + run_user_dir = prefix_roota(root_directory, "/run/user"); + run_user_dir = strjoina("-", run_user_dir); + root_dir = prefix_roota(root_directory, "/root"); + root_dir = strjoina("-", root_dir); + + r = append_mounts(&m, STRV_MAKE(home_dir, run_user_dir, root_dir), + protect_home == PROTECT_HOME_READ_ONLY ? READONLY : INACCESSIBLE); + if (r < 0) + return r; + } + + if (protect_system != PROTECT_SYSTEM_NO) { + const char *usr_dir, *boot_dir, *etc_dir; + + usr_dir = prefix_roota(root_directory, "/usr"); + boot_dir = prefix_roota(root_directory, "/boot"); + boot_dir = strjoina("-", boot_dir); + etc_dir = prefix_roota(root_directory, "/etc"); + + r = append_mounts(&m, protect_system == PROTECT_SYSTEM_FULL + ? STRV_MAKE(usr_dir, boot_dir, etc_dir) + : STRV_MAKE(usr_dir, boot_dir), READONLY); + if (r < 0) + return r; + } + + assert(mounts + n == m); + + qsort(mounts, n, sizeof(BindMount), mount_path_compare); + drop_duplicates(mounts, &n); + } + + if (n > 0 || root_directory) { + /* Remount / as SLAVE so that nothing now mounted in the namespace + shows up in the parent */ + if (mount(NULL, "/", NULL, MS_SLAVE|MS_REC, NULL) < 0) + return -errno; + } + + if (root_directory) { + /* Turn directory into bind mount */ + if (mount(root_directory, root_directory, NULL, MS_BIND|MS_REC, NULL) < 0) + return -errno; + } + + if (n > 0) { + for (m = mounts; m < mounts + n; ++m) { + r = apply_mount(m, tmp_dir, var_tmp_dir); + if (r < 0) + goto fail; + } + + for (m = mounts; m < mounts + n; ++m) { + r = make_read_only(m); + if (r < 0) + goto fail; + } + } + + if (root_directory) { + /* MS_MOVE does not work on MS_SHARED so the remount MS_SHARED will be done later */ + r = mount_move_root(root_directory); + + /* at this point, we cannot rollback */ + if (r < 0) + return r; + } + + /* Remount / as the desired mode. Not that this will not + * reestablish propagation from our side to the host, since + * what's disconnected is disconnected. */ + if (mount(NULL, "/", NULL, mount_flags | MS_REC, NULL) < 0) + /* at this point, we cannot rollback */ + return -errno; + + return 0; + +fail: + if (n > 0) { + for (m = mounts; m < mounts + n; ++m) + if (m->done) + (void) umount2(m->path, MNT_DETACH); + } + + return r; +} + +static int setup_one_tmp_dir(const char *id, const char *prefix, char **path) { + _cleanup_free_ char *x = NULL; + char bid[SD_ID128_STRING_MAX]; + sd_id128_t boot_id; + int r; + + assert(id); + assert(prefix); + assert(path); + + /* We include the boot id in the directory so that after a + * reboot we can easily identify obsolete directories. */ + + r = sd_id128_get_boot(&boot_id); + if (r < 0) + return r; + + x = strjoin(prefix, "/systemd-private-", sd_id128_to_string(boot_id, bid), "-", id, "-XXXXXX", NULL); + if (!x) + return -ENOMEM; + + RUN_WITH_UMASK(0077) + if (!mkdtemp(x)) + return -errno; + + RUN_WITH_UMASK(0000) { + char *y; + + y = strjoina(x, "/tmp"); + + if (mkdir(y, 0777 | S_ISVTX) < 0) + return -errno; + } + + *path = x; + x = NULL; + + return 0; +} + +int setup_tmp_dirs(const char *id, char **tmp_dir, char **var_tmp_dir) { + char *a, *b; + int r; + + assert(id); + assert(tmp_dir); + assert(var_tmp_dir); + + r = setup_one_tmp_dir(id, "/tmp", &a); + if (r < 0) + return r; + + r = setup_one_tmp_dir(id, "/var/tmp", &b); + if (r < 0) { + char *t; + + t = strjoina(a, "/tmp"); + rmdir(t); + rmdir(a); + + free(a); + return r; + } + + *tmp_dir = a; + *var_tmp_dir = b; + + return 0; +} + +int setup_netns(int netns_storage_socket[2]) { + _cleanup_close_ int netns = -1; + int r, q; + + assert(netns_storage_socket); + assert(netns_storage_socket[0] >= 0); + assert(netns_storage_socket[1] >= 0); + + /* We use the passed socketpair as a storage buffer for our + * namespace reference fd. Whatever process runs this first + * shall create a new namespace, all others should just join + * it. To serialize that we use a file lock on the socket + * pair. + * + * It's a bit crazy, but hey, works great! */ + + if (lockf(netns_storage_socket[0], F_LOCK, 0) < 0) + return -errno; + + netns = receive_one_fd(netns_storage_socket[0], MSG_DONTWAIT); + if (netns == -EAGAIN) { + /* Nothing stored yet, so let's create a new namespace */ + + if (unshare(CLONE_NEWNET) < 0) { + r = -errno; + goto fail; + } + + loopback_setup(); + + netns = open("/proc/self/ns/net", O_RDONLY|O_CLOEXEC|O_NOCTTY); + if (netns < 0) { + r = -errno; + goto fail; + } + + r = 1; + + } else if (netns < 0) { + r = netns; + goto fail; + + } else { + /* Yay, found something, so let's join the namespace */ + if (setns(netns, CLONE_NEWNET) < 0) { + r = -errno; + goto fail; + } + + r = 0; + } + + q = send_one_fd(netns_storage_socket[1], netns, MSG_DONTWAIT); + if (q < 0) { + r = q; + goto fail; + } + +fail: + (void) lockf(netns_storage_socket[0], F_ULOCK, 0); + return r; +} + +static const char *const protect_home_table[_PROTECT_HOME_MAX] = { + [PROTECT_HOME_NO] = "no", + [PROTECT_HOME_YES] = "yes", + [PROTECT_HOME_READ_ONLY] = "read-only", +}; + +DEFINE_STRING_TABLE_LOOKUP(protect_home, ProtectHome); + +static const char *const protect_system_table[_PROTECT_SYSTEM_MAX] = { + [PROTECT_SYSTEM_NO] = "no", + [PROTECT_SYSTEM_YES] = "yes", + [PROTECT_SYSTEM_FULL] = "full", +}; + +DEFINE_STRING_TABLE_LOOKUP(protect_system, ProtectSystem); diff --git a/src/grp-system/libcore/src/path.c b/src/grp-system/libcore/src/path.c new file mode 100644 index 0000000000..49c81dffec --- /dev/null +++ b/src/grp-system/libcore/src/path.c @@ -0,0 +1,785 @@ +/*** + This file is part of systemd. + + Copyright 2010 Lennart Poettering + + systemd is free software; you can redistribute it and/or modify it + under the terms of the GNU Lesser General Public License as published by + the Free Software Foundation; either version 2.1 of the License, or + (at your option) any later version. + + systemd is distributed in the hope that it will be useful, but + WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + Lesser General Public License for more details. + + You should have received a copy of the GNU Lesser General Public License + along with systemd; If not, see . +***/ + +#include +#include +#include +#include + +#include "core/path.h" +#include "core/unit.h" +#include "sd-bus/bus-error.h" +#include "sd-bus/bus-util.h" +#include "systemd-basic/fd-util.h" +#include "systemd-basic/fs-util.h" +#include "systemd-basic/glob-util.h" +#include "systemd-basic/macro.h" +#include "systemd-basic/mkdir.h" +#include "systemd-basic/special.h" +#include "systemd-basic/stat-util.h" +#include "systemd-basic/string-table.h" +#include "systemd-basic/string-util.h" +#include "systemd-basic/unit-name.h" + +#include "dbus-path.h" + +static const UnitActiveState state_translation_table[_PATH_STATE_MAX] = { + [PATH_DEAD] = UNIT_INACTIVE, + [PATH_WAITING] = UNIT_ACTIVE, + [PATH_RUNNING] = UNIT_ACTIVE, + [PATH_FAILED] = UNIT_FAILED +}; + +static int path_dispatch_io(sd_event_source *source, int fd, uint32_t revents, void *userdata); + +int path_spec_watch(PathSpec *s, sd_event_io_handler_t handler) { + + static const int flags_table[_PATH_TYPE_MAX] = { + [PATH_EXISTS] = IN_DELETE_SELF|IN_MOVE_SELF|IN_ATTRIB, + [PATH_EXISTS_GLOB] = IN_DELETE_SELF|IN_MOVE_SELF|IN_ATTRIB, + [PATH_CHANGED] = IN_DELETE_SELF|IN_MOVE_SELF|IN_ATTRIB|IN_CLOSE_WRITE|IN_CREATE|IN_DELETE|IN_MOVED_FROM|IN_MOVED_TO, + [PATH_MODIFIED] = IN_DELETE_SELF|IN_MOVE_SELF|IN_ATTRIB|IN_CLOSE_WRITE|IN_CREATE|IN_DELETE|IN_MOVED_FROM|IN_MOVED_TO|IN_MODIFY, + [PATH_DIRECTORY_NOT_EMPTY] = IN_DELETE_SELF|IN_MOVE_SELF|IN_ATTRIB|IN_CREATE|IN_MOVED_TO + }; + + bool exists = false; + char *slash, *oldslash = NULL; + int r; + + assert(s); + assert(s->unit); + assert(handler); + + path_spec_unwatch(s); + + s->inotify_fd = inotify_init1(IN_NONBLOCK|IN_CLOEXEC); + if (s->inotify_fd < 0) { + r = -errno; + goto fail; + } + + r = sd_event_add_io(s->unit->manager->event, &s->event_source, s->inotify_fd, EPOLLIN, handler, s); + if (r < 0) + goto fail; + + (void) sd_event_source_set_description(s->event_source, "path"); + + /* This assumes the path was passed through path_kill_slashes()! */ + + for (slash = strchr(s->path, '/'); ; slash = strchr(slash+1, '/')) { + char *cut = NULL; + int flags; + char tmp; + + if (slash) { + cut = slash + (slash == s->path); + tmp = *cut; + *cut = '\0'; + + flags = IN_MOVE_SELF | IN_DELETE_SELF | IN_ATTRIB | IN_CREATE | IN_MOVED_TO; + } else + flags = flags_table[s->type]; + + r = inotify_add_watch(s->inotify_fd, s->path, flags); + if (r < 0) { + if (errno == EACCES || errno == ENOENT) { + if (cut) + *cut = tmp; + break; + } + + r = log_warning_errno(errno, "Failed to add watch on %s: %s", s->path, errno == ENOSPC ? "too many watches" : strerror(-r)); + if (cut) + *cut = tmp; + goto fail; + } else { + exists = true; + + /* Path exists, we don't need to watch parent too closely. */ + if (oldslash) { + char *cut2 = oldslash + (oldslash == s->path); + char tmp2 = *cut2; + *cut2 = '\0'; + + (void) inotify_add_watch(s->inotify_fd, s->path, IN_MOVE_SELF); + /* Error is ignored, the worst can happen is we get spurious events. */ + + *cut2 = tmp2; + } + } + + if (cut) + *cut = tmp; + + if (slash) + oldslash = slash; + else { + /* whole path has been iterated over */ + s->primary_wd = r; + break; + } + } + + if (!exists) { + r = log_error_errno(errno, "Failed to add watch on any of the components of %s: %m", s->path); + /* either EACCESS or ENOENT */ + goto fail; + } + + return 0; + +fail: + path_spec_unwatch(s); + return r; +} + +void path_spec_unwatch(PathSpec *s) { + assert(s); + + s->event_source = sd_event_source_unref(s->event_source); + s->inotify_fd = safe_close(s->inotify_fd); +} + +int path_spec_fd_event(PathSpec *s, uint32_t revents) { + union inotify_event_buffer buffer; + struct inotify_event *e; + ssize_t l; + int r = 0; + + if (revents != EPOLLIN) { + log_error("Got invalid poll event on inotify."); + return -EINVAL; + } + + l = read(s->inotify_fd, &buffer, sizeof(buffer)); + if (l < 0) { + if (errno == EAGAIN || errno == EINTR) + return 0; + + return log_error_errno(errno, "Failed to read inotify event: %m"); + } + + FOREACH_INOTIFY_EVENT(e, buffer, l) { + if ((s->type == PATH_CHANGED || s->type == PATH_MODIFIED) && + s->primary_wd == e->wd) + r = 1; + } + + return r; +} + +static bool path_spec_check_good(PathSpec *s, bool initial) { + bool good = false; + + switch (s->type) { + + case PATH_EXISTS: + good = access(s->path, F_OK) >= 0; + break; + + case PATH_EXISTS_GLOB: + good = glob_exists(s->path) > 0; + break; + + case PATH_DIRECTORY_NOT_EMPTY: { + int k; + + k = dir_is_empty(s->path); + good = !(k == -ENOENT || k > 0); + break; + } + + case PATH_CHANGED: + case PATH_MODIFIED: { + bool b; + + b = access(s->path, F_OK) >= 0; + good = !initial && b != s->previous_exists; + s->previous_exists = b; + break; + } + + default: + ; + } + + return good; +} + +static void path_spec_mkdir(PathSpec *s, mode_t mode) { + int r; + + if (s->type == PATH_EXISTS || s->type == PATH_EXISTS_GLOB) + return; + + r = mkdir_p_label(s->path, mode); + if (r < 0) + log_warning_errno(r, "mkdir(%s) failed: %m", s->path); +} + +static void path_spec_dump(PathSpec *s, FILE *f, const char *prefix) { + fprintf(f, + "%s%s: %s\n", + prefix, + path_type_to_string(s->type), + s->path); +} + +void path_spec_done(PathSpec *s) { + assert(s); + assert(s->inotify_fd == -1); + + free(s->path); +} + +static void path_init(Unit *u) { + Path *p = PATH(u); + + assert(u); + assert(u->load_state == UNIT_STUB); + + p->directory_mode = 0755; +} + +void path_free_specs(Path *p) { + PathSpec *s; + + assert(p); + + while ((s = p->specs)) { + path_spec_unwatch(s); + LIST_REMOVE(spec, p->specs, s); + path_spec_done(s); + free(s); + } +} + +static void path_done(Unit *u) { + Path *p = PATH(u); + + assert(p); + + path_free_specs(p); +} + +static int path_add_mount_links(Path *p) { + PathSpec *s; + int r; + + assert(p); + + LIST_FOREACH(spec, s, p->specs) { + r = unit_require_mounts_for(UNIT(p), s->path); + if (r < 0) + return r; + } + + return 0; +} + +static int path_verify(Path *p) { + assert(p); + + if (UNIT(p)->load_state != UNIT_LOADED) + return 0; + + if (!p->specs) { + log_unit_error(UNIT(p), "Path unit lacks path setting. Refusing."); + return -EINVAL; + } + + return 0; +} + +static int path_add_default_dependencies(Path *p) { + int r; + + assert(p); + + if (!UNIT(p)->default_dependencies) + return 0; + + r = unit_add_dependency_by_name(UNIT(p), UNIT_BEFORE, SPECIAL_PATHS_TARGET, NULL, true); + if (r < 0) + return r; + + if (MANAGER_IS_SYSTEM(UNIT(p)->manager)) { + r = unit_add_two_dependencies_by_name(UNIT(p), UNIT_AFTER, UNIT_REQUIRES, SPECIAL_SYSINIT_TARGET, NULL, true); + if (r < 0) + return r; + } + + return unit_add_two_dependencies_by_name(UNIT(p), UNIT_BEFORE, UNIT_CONFLICTS, SPECIAL_SHUTDOWN_TARGET, NULL, true); +} + +static int path_load(Unit *u) { + Path *p = PATH(u); + int r; + + assert(u); + assert(u->load_state == UNIT_STUB); + + r = unit_load_fragment_and_dropin(u); + if (r < 0) + return r; + + if (u->load_state == UNIT_LOADED) { + + if (set_isempty(u->dependencies[UNIT_TRIGGERS])) { + Unit *x; + + r = unit_load_related_unit(u, ".service", &x); + if (r < 0) + return r; + + r = unit_add_two_dependencies(u, UNIT_BEFORE, UNIT_TRIGGERS, x, true); + if (r < 0) + return r; + } + + r = path_add_mount_links(p); + if (r < 0) + return r; + + r = path_add_default_dependencies(p); + if (r < 0) + return r; + } + + return path_verify(p); +} + +static void path_dump(Unit *u, FILE *f, const char *prefix) { + Path *p = PATH(u); + Unit *trigger; + PathSpec *s; + + assert(p); + assert(f); + + trigger = UNIT_TRIGGER(u); + + fprintf(f, + "%sPath State: %s\n" + "%sResult: %s\n" + "%sUnit: %s\n" + "%sMakeDirectory: %s\n" + "%sDirectoryMode: %04o\n", + prefix, path_state_to_string(p->state), + prefix, path_result_to_string(p->result), + prefix, trigger ? trigger->id : "n/a", + prefix, yes_no(p->make_directory), + prefix, p->directory_mode); + + LIST_FOREACH(spec, s, p->specs) + path_spec_dump(s, f, prefix); +} + +static void path_unwatch(Path *p) { + PathSpec *s; + + assert(p); + + LIST_FOREACH(spec, s, p->specs) + path_spec_unwatch(s); +} + +static int path_watch(Path *p) { + int r; + PathSpec *s; + + assert(p); + + LIST_FOREACH(spec, s, p->specs) { + r = path_spec_watch(s, path_dispatch_io); + if (r < 0) + return r; + } + + return 0; +} + +static void path_set_state(Path *p, PathState state) { + PathState old_state; + assert(p); + + old_state = p->state; + p->state = state; + + if (state != PATH_WAITING && + (state != PATH_RUNNING || p->inotify_triggered)) + path_unwatch(p); + + if (state != old_state) + log_unit_debug(UNIT(p), "Changed %s -> %s", path_state_to_string(old_state), path_state_to_string(state)); + + unit_notify(UNIT(p), state_translation_table[old_state], state_translation_table[state], true); +} + +static void path_enter_waiting(Path *p, bool initial, bool recheck); + +static int path_coldplug(Unit *u) { + Path *p = PATH(u); + + assert(p); + assert(p->state == PATH_DEAD); + + if (p->deserialized_state != p->state) { + + if (p->deserialized_state == PATH_WAITING || + p->deserialized_state == PATH_RUNNING) + path_enter_waiting(p, true, true); + else + path_set_state(p, p->deserialized_state); + } + + return 0; +} + +static void path_enter_dead(Path *p, PathResult f) { + assert(p); + + if (f != PATH_SUCCESS) + p->result = f; + + path_set_state(p, p->result != PATH_SUCCESS ? PATH_FAILED : PATH_DEAD); +} + +static void path_enter_running(Path *p) { + _cleanup_(sd_bus_error_free) sd_bus_error error = SD_BUS_ERROR_NULL; + Unit *trigger; + int r; + + assert(p); + + /* Don't start job if we are supposed to go down */ + if (unit_stop_pending(UNIT(p))) + return; + + trigger = UNIT_TRIGGER(UNIT(p)); + if (!trigger) { + log_unit_error(UNIT(p), "Unit to trigger vanished."); + path_enter_dead(p, PATH_FAILURE_RESOURCES); + return; + } + + r = manager_add_job(UNIT(p)->manager, JOB_START, trigger, JOB_REPLACE, &error, NULL); + if (r < 0) + goto fail; + + p->inotify_triggered = false; + + r = path_watch(p); + if (r < 0) + goto fail; + + path_set_state(p, PATH_RUNNING); + return; + +fail: + log_unit_warning(UNIT(p), "Failed to queue unit startup job: %s", bus_error_message(&error, r)); + path_enter_dead(p, PATH_FAILURE_RESOURCES); +} + +static bool path_check_good(Path *p, bool initial) { + PathSpec *s; + bool good = false; + + assert(p); + + LIST_FOREACH(spec, s, p->specs) { + good = path_spec_check_good(s, initial); + + if (good) + break; + } + + return good; +} + +static void path_enter_waiting(Path *p, bool initial, bool recheck) { + int r; + + if (recheck) + if (path_check_good(p, initial)) { + log_unit_debug(UNIT(p), "Got triggered."); + path_enter_running(p); + return; + } + + r = path_watch(p); + if (r < 0) + goto fail; + + /* Hmm, so now we have created inotify watches, but the file + * might have appeared/been removed by now, so we must + * recheck */ + + if (recheck) + if (path_check_good(p, false)) { + log_unit_debug(UNIT(p), "Got triggered."); + path_enter_running(p); + return; + } + + path_set_state(p, PATH_WAITING); + return; + +fail: + log_unit_warning_errno(UNIT(p), r, "Failed to enter waiting state: %m"); + path_enter_dead(p, PATH_FAILURE_RESOURCES); +} + +static void path_mkdir(Path *p) { + PathSpec *s; + + assert(p); + + if (!p->make_directory) + return; + + LIST_FOREACH(spec, s, p->specs) + path_spec_mkdir(s, p->directory_mode); +} + +static int path_start(Unit *u) { + Path *p = PATH(u); + Unit *trigger; + int r; + + assert(p); + assert(p->state == PATH_DEAD || p->state == PATH_FAILED); + + trigger = UNIT_TRIGGER(u); + if (!trigger || trigger->load_state != UNIT_LOADED) { + log_unit_error(u, "Refusing to start, unit to trigger not loaded."); + return -ENOENT; + } + + r = unit_start_limit_test(u); + if (r < 0) { + path_enter_dead(p, PATH_FAILURE_START_LIMIT_HIT); + return r; + } + + path_mkdir(p); + + p->result = PATH_SUCCESS; + path_enter_waiting(p, true, true); + + return 1; +} + +static int path_stop(Unit *u) { + Path *p = PATH(u); + + assert(p); + assert(p->state == PATH_WAITING || p->state == PATH_RUNNING); + + path_enter_dead(p, PATH_SUCCESS); + return 1; +} + +static int path_serialize(Unit *u, FILE *f, FDSet *fds) { + Path *p = PATH(u); + + assert(u); + assert(f); + assert(fds); + + unit_serialize_item(u, f, "state", path_state_to_string(p->state)); + unit_serialize_item(u, f, "result", path_result_to_string(p->result)); + + return 0; +} + +static int path_deserialize_item(Unit *u, const char *key, const char *value, FDSet *fds) { + Path *p = PATH(u); + + assert(u); + assert(key); + assert(value); + assert(fds); + + if (streq(key, "state")) { + PathState state; + + state = path_state_from_string(value); + if (state < 0) + log_unit_debug(u, "Failed to parse state value: %s", value); + else + p->deserialized_state = state; + + } else if (streq(key, "result")) { + PathResult f; + + f = path_result_from_string(value); + if (f < 0) + log_unit_debug(u, "Failed to parse result value: %s", value); + else if (f != PATH_SUCCESS) + p->result = f; + + } else + log_unit_debug(u, "Unknown serialization key: %s", key); + + return 0; +} + +_pure_ static UnitActiveState path_active_state(Unit *u) { + assert(u); + + return state_translation_table[PATH(u)->state]; +} + +_pure_ static const char *path_sub_state_to_string(Unit *u) { + assert(u); + + return path_state_to_string(PATH(u)->state); +} + +static int path_dispatch_io(sd_event_source *source, int fd, uint32_t revents, void *userdata) { + PathSpec *s = userdata; + Path *p; + int changed; + + assert(s); + assert(s->unit); + assert(fd >= 0); + + p = PATH(s->unit); + + if (p->state != PATH_WAITING && + p->state != PATH_RUNNING) + return 0; + + /* log_debug("inotify wakeup on %s.", u->id); */ + + LIST_FOREACH(spec, s, p->specs) + if (path_spec_owns_inotify_fd(s, fd)) + break; + + if (!s) { + log_error("Got event on unknown fd."); + goto fail; + } + + changed = path_spec_fd_event(s, revents); + if (changed < 0) + goto fail; + + /* If we are already running, then remember that one event was + * dispatched so that we restart the service only if something + * actually changed on disk */ + p->inotify_triggered = true; + + if (changed) + path_enter_running(p); + else + path_enter_waiting(p, false, true); + + return 0; + +fail: + path_enter_dead(p, PATH_FAILURE_RESOURCES); + return 0; +} + +static void path_trigger_notify(Unit *u, Unit *other) { + Path *p = PATH(u); + + assert(u); + assert(other); + + /* Invoked whenever the unit we trigger changes state or gains + * or loses a job */ + + if (other->load_state != UNIT_LOADED) + return; + + if (p->state == PATH_RUNNING && + UNIT_IS_INACTIVE_OR_FAILED(unit_active_state(other))) { + log_unit_debug(UNIT(p), "Got notified about unit deactivation."); + + /* Hmm, so inotify was triggered since the + * last activation, so I guess we need to + * recheck what is going on. */ + path_enter_waiting(p, false, p->inotify_triggered); + } +} + +static void path_reset_failed(Unit *u) { + Path *p = PATH(u); + + assert(p); + + if (p->state == PATH_FAILED) + path_set_state(p, PATH_DEAD); + + p->result = PATH_SUCCESS; +} + +static const char* const path_type_table[_PATH_TYPE_MAX] = { + [PATH_EXISTS] = "PathExists", + [PATH_EXISTS_GLOB] = "PathExistsGlob", + [PATH_DIRECTORY_NOT_EMPTY] = "DirectoryNotEmpty", + [PATH_CHANGED] = "PathChanged", + [PATH_MODIFIED] = "PathModified", +}; + +DEFINE_STRING_TABLE_LOOKUP(path_type, PathType); + +static const char* const path_result_table[_PATH_RESULT_MAX] = { + [PATH_SUCCESS] = "success", + [PATH_FAILURE_RESOURCES] = "resources", + [PATH_FAILURE_START_LIMIT_HIT] = "start-limit-hit", +}; + +DEFINE_STRING_TABLE_LOOKUP(path_result, PathResult); + +const UnitVTable path_vtable = { + .object_size = sizeof(Path), + + .sections = + "Unit\0" + "Path\0" + "Install\0", + + .init = path_init, + .done = path_done, + .load = path_load, + + .coldplug = path_coldplug, + + .dump = path_dump, + + .start = path_start, + .stop = path_stop, + + .serialize = path_serialize, + .deserialize_item = path_deserialize_item, + + .active_state = path_active_state, + .sub_state_to_string = path_sub_state_to_string, + + .trigger_notify = path_trigger_notify, + + .reset_failed = path_reset_failed, + + .bus_vtable = bus_path_vtable +}; diff --git a/src/grp-system/libcore/src/scope.c b/src/grp-system/libcore/src/scope.c new file mode 100644 index 0000000000..11816469a6 --- /dev/null +++ b/src/grp-system/libcore/src/scope.c @@ -0,0 +1,622 @@ +/*** + This file is part of systemd. + + Copyright 2013 Lennart Poettering + + systemd is free software; you can redistribute it and/or modify it + under the terms of the GNU Lesser General Public License as published by + the Free Software Foundation; either version 2.1 of the License, or + (at your option) any later version. + + systemd is distributed in the hope that it will be useful, but + WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + Lesser General Public License for more details. + + You should have received a copy of the GNU Lesser General Public License + along with systemd; If not, see . +***/ + +#include +#include + +#include "core/scope.h" +#include "core/unit.h" +#include "systemd-basic/alloc-util.h" +#include "systemd-basic/log.h" +#include "systemd-basic/special.h" +#include "systemd-basic/string-table.h" +#include "systemd-basic/string-util.h" +#include "systemd-basic/strv.h" +#include "systemd-basic/unit-name.h" + +#include "dbus-scope.h" +#include "load-dropin.h" + +static const UnitActiveState state_translation_table[_SCOPE_STATE_MAX] = { + [SCOPE_DEAD] = UNIT_INACTIVE, + [SCOPE_RUNNING] = UNIT_ACTIVE, + [SCOPE_ABANDONED] = UNIT_ACTIVE, + [SCOPE_STOP_SIGTERM] = UNIT_DEACTIVATING, + [SCOPE_STOP_SIGKILL] = UNIT_DEACTIVATING, + [SCOPE_FAILED] = UNIT_FAILED +}; + +static int scope_dispatch_timer(sd_event_source *source, usec_t usec, void *userdata); + +static void scope_init(Unit *u) { + Scope *s = SCOPE(u); + + assert(u); + assert(u->load_state == UNIT_STUB); + + s->timeout_stop_usec = u->manager->default_timeout_stop_usec; + u->ignore_on_isolate = true; +} + +static void scope_done(Unit *u) { + Scope *s = SCOPE(u); + + assert(u); + + free(s->controller); + + s->timer_event_source = sd_event_source_unref(s->timer_event_source); +} + +static int scope_arm_timer(Scope *s, usec_t usec) { + int r; + + assert(s); + + if (s->timer_event_source) { + r = sd_event_source_set_time(s->timer_event_source, usec); + if (r < 0) + return r; + + return sd_event_source_set_enabled(s->timer_event_source, SD_EVENT_ONESHOT); + } + + if (usec == USEC_INFINITY) + return 0; + + r = sd_event_add_time( + UNIT(s)->manager->event, + &s->timer_event_source, + CLOCK_MONOTONIC, + usec, 0, + scope_dispatch_timer, s); + if (r < 0) + return r; + + (void) sd_event_source_set_description(s->timer_event_source, "scope-timer"); + + return 0; +} + +static void scope_set_state(Scope *s, ScopeState state) { + ScopeState old_state; + assert(s); + + old_state = s->state; + s->state = state; + + if (!IN_SET(state, SCOPE_STOP_SIGTERM, SCOPE_STOP_SIGKILL)) + s->timer_event_source = sd_event_source_unref(s->timer_event_source); + + if (IN_SET(state, SCOPE_DEAD, SCOPE_FAILED)) + unit_unwatch_all_pids(UNIT(s)); + + if (state != old_state) + log_debug("%s changed %s -> %s", UNIT(s)->id, scope_state_to_string(old_state), scope_state_to_string(state)); + + unit_notify(UNIT(s), state_translation_table[old_state], state_translation_table[state], true); +} + +static int scope_add_default_dependencies(Scope *s) { + int r; + + assert(s); + + if (!UNIT(s)->default_dependencies) + return 0; + + /* Make sure scopes are unloaded on shutdown */ + r = unit_add_two_dependencies_by_name( + UNIT(s), + UNIT_BEFORE, UNIT_CONFLICTS, + SPECIAL_SHUTDOWN_TARGET, NULL, true); + if (r < 0) + return r; + + return 0; +} + +static int scope_verify(Scope *s) { + assert(s); + + if (UNIT(s)->load_state != UNIT_LOADED) + return 0; + + if (set_isempty(UNIT(s)->pids) && + !MANAGER_IS_RELOADING(UNIT(s)->manager) && + !unit_has_name(UNIT(s), SPECIAL_INIT_SCOPE)) { + log_unit_error(UNIT(s), "Scope has no PIDs. Refusing."); + return -EINVAL; + } + + return 0; +} + +static int scope_load(Unit *u) { + Scope *s = SCOPE(u); + int r; + + assert(s); + assert(u->load_state == UNIT_STUB); + + if (!u->transient && !MANAGER_IS_RELOADING(u->manager)) + /* Refuse to load non-transient scope units, but allow them while reloading. */ + return -ENOENT; + + r = unit_load_fragment_and_dropin_optional(u); + if (r < 0) + return r; + + if (u->load_state == UNIT_LOADED) { + r = unit_patch_contexts(u); + if (r < 0) + return r; + + r = unit_set_default_slice(u); + if (r < 0) + return r; + + r = scope_add_default_dependencies(s); + if (r < 0) + return r; + } + + return scope_verify(s); +} + +static int scope_coldplug(Unit *u) { + Scope *s = SCOPE(u); + int r; + + assert(s); + assert(s->state == SCOPE_DEAD); + + if (s->deserialized_state == s->state) + return 0; + + if (IN_SET(s->deserialized_state, SCOPE_STOP_SIGKILL, SCOPE_STOP_SIGTERM)) { + r = scope_arm_timer(s, usec_add(u->state_change_timestamp.monotonic, s->timeout_stop_usec)); + if (r < 0) + return r; + } + + if (!IN_SET(s->deserialized_state, SCOPE_DEAD, SCOPE_FAILED)) + unit_watch_all_pids(UNIT(s)); + + scope_set_state(s, s->deserialized_state); + return 0; +} + +static void scope_dump(Unit *u, FILE *f, const char *prefix) { + Scope *s = SCOPE(u); + + assert(s); + assert(f); + + fprintf(f, + "%sScope State: %s\n" + "%sResult: %s\n", + prefix, scope_state_to_string(s->state), + prefix, scope_result_to_string(s->result)); + + cgroup_context_dump(&s->cgroup_context, f, prefix); + kill_context_dump(&s->kill_context, f, prefix); +} + +static void scope_enter_dead(Scope *s, ScopeResult f) { + assert(s); + + if (f != SCOPE_SUCCESS) + s->result = f; + + scope_set_state(s, s->result != SCOPE_SUCCESS ? SCOPE_FAILED : SCOPE_DEAD); +} + +static void scope_enter_signal(Scope *s, ScopeState state, ScopeResult f) { + bool skip_signal = false; + int r; + + assert(s); + + if (f != SCOPE_SUCCESS) + s->result = f; + + unit_watch_all_pids(UNIT(s)); + + /* If we have a controller set let's ask the controller nicely + * to terminate the scope, instead of us going directly into + * SIGTERM berserk mode */ + if (state == SCOPE_STOP_SIGTERM) + skip_signal = bus_scope_send_request_stop(s) > 0; + + if (!skip_signal) { + r = unit_kill_context( + UNIT(s), + &s->kill_context, + state != SCOPE_STOP_SIGTERM ? KILL_KILL : + s->was_abandoned ? KILL_TERMINATE_AND_LOG : + KILL_TERMINATE, + -1, -1, false); + if (r < 0) + goto fail; + } else + r = 1; + + if (r > 0) { + r = scope_arm_timer(s, usec_add(now(CLOCK_MONOTONIC), s->timeout_stop_usec)); + if (r < 0) + goto fail; + + scope_set_state(s, state); + } else if (state == SCOPE_STOP_SIGTERM) + scope_enter_signal(s, SCOPE_STOP_SIGKILL, SCOPE_SUCCESS); + else + scope_enter_dead(s, SCOPE_SUCCESS); + + return; + +fail: + log_unit_warning_errno(UNIT(s), r, "Failed to kill processes: %m"); + + scope_enter_dead(s, SCOPE_FAILURE_RESOURCES); +} + +static int scope_start(Unit *u) { + Scope *s = SCOPE(u); + int r; + + assert(s); + + if (unit_has_name(u, SPECIAL_INIT_SCOPE)) + return -EPERM; + + if (s->state == SCOPE_FAILED) + return -EPERM; + + /* We can't fulfill this right now, please try again later */ + if (s->state == SCOPE_STOP_SIGTERM || + s->state == SCOPE_STOP_SIGKILL) + return -EAGAIN; + + assert(s->state == SCOPE_DEAD); + + if (!u->transient && !MANAGER_IS_RELOADING(u->manager)) + return -ENOENT; + + (void) unit_realize_cgroup(u); + (void) unit_reset_cpu_usage(u); + + r = unit_attach_pids_to_cgroup(u); + if (r < 0) { + log_unit_warning_errno(UNIT(s), r, "Failed to add PIDs to scope's control group: %m"); + scope_enter_dead(s, SCOPE_FAILURE_RESOURCES); + return r; + } + + s->result = SCOPE_SUCCESS; + + scope_set_state(s, SCOPE_RUNNING); + return 1; +} + +static int scope_stop(Unit *u) { + Scope *s = SCOPE(u); + + assert(s); + + if (s->state == SCOPE_STOP_SIGTERM || + s->state == SCOPE_STOP_SIGKILL) + return 0; + + assert(s->state == SCOPE_RUNNING || + s->state == SCOPE_ABANDONED); + + scope_enter_signal(s, SCOPE_STOP_SIGTERM, SCOPE_SUCCESS); + return 1; +} + +static void scope_reset_failed(Unit *u) { + Scope *s = SCOPE(u); + + assert(s); + + if (s->state == SCOPE_FAILED) + scope_set_state(s, SCOPE_DEAD); + + s->result = SCOPE_SUCCESS; +} + +static int scope_kill(Unit *u, KillWho who, int signo, sd_bus_error *error) { + return unit_kill_common(u, who, signo, -1, -1, error); +} + +static int scope_get_timeout(Unit *u, usec_t *timeout) { + Scope *s = SCOPE(u); + usec_t t; + int r; + + if (!s->timer_event_source) + return 0; + + r = sd_event_source_get_time(s->timer_event_source, &t); + if (r < 0) + return r; + if (t == USEC_INFINITY) + return 0; + + *timeout = t; + return 1; +} + +static int scope_serialize(Unit *u, FILE *f, FDSet *fds) { + Scope *s = SCOPE(u); + + assert(s); + assert(f); + assert(fds); + + unit_serialize_item(u, f, "state", scope_state_to_string(s->state)); + unit_serialize_item(u, f, "was-abandoned", yes_no(s->was_abandoned)); + return 0; +} + +static int scope_deserialize_item(Unit *u, const char *key, const char *value, FDSet *fds) { + Scope *s = SCOPE(u); + + assert(u); + assert(key); + assert(value); + assert(fds); + + if (streq(key, "state")) { + ScopeState state; + + state = scope_state_from_string(value); + if (state < 0) + log_unit_debug(u, "Failed to parse state value: %s", value); + else + s->deserialized_state = state; + + } else if (streq(key, "was-abandoned")) { + int k; + + k = parse_boolean(value); + if (k < 0) + log_unit_debug(u, "Failed to parse boolean value: %s", value); + else + s->was_abandoned = k; + } else + log_unit_debug(u, "Unknown serialization key: %s", key); + + return 0; +} + +static bool scope_check_gc(Unit *u) { + assert(u); + + /* Never clean up scopes that still have a process around, + * even if the scope is formally dead. */ + + if (!u->cgroup_path) + return false; + + return cg_is_empty_recursive(SYSTEMD_CGROUP_CONTROLLER, u->cgroup_path) <= 0; +} + +static void scope_notify_cgroup_empty_event(Unit *u) { + Scope *s = SCOPE(u); + assert(u); + + log_unit_debug(u, "cgroup is empty"); + + if (IN_SET(s->state, SCOPE_RUNNING, SCOPE_ABANDONED, SCOPE_STOP_SIGTERM, SCOPE_STOP_SIGKILL)) + scope_enter_dead(s, SCOPE_SUCCESS); +} + +static void scope_sigchld_event(Unit *u, pid_t pid, int code, int status) { + + /* If we get a SIGCHLD event for one of the processes we were + interested in, then we look for others to watch, under the + assumption that we'll sooner or later get a SIGCHLD for + them, as the original process we watched was probably the + parent of them, and they are hence now our children. */ + + unit_tidy_watch_pids(u, 0, 0); + unit_watch_all_pids(u); + + /* If the PID set is empty now, then let's finish this off + (On unified we use proper notifications) */ + if (cg_unified() <= 0 && set_isempty(u->pids)) + scope_notify_cgroup_empty_event(u); +} + +static int scope_dispatch_timer(sd_event_source *source, usec_t usec, void *userdata) { + Scope *s = SCOPE(userdata); + + assert(s); + assert(s->timer_event_source == source); + + switch (s->state) { + + case SCOPE_STOP_SIGTERM: + if (s->kill_context.send_sigkill) { + log_unit_warning(UNIT(s), "Stopping timed out. Killing."); + scope_enter_signal(s, SCOPE_STOP_SIGKILL, SCOPE_FAILURE_TIMEOUT); + } else { + log_unit_warning(UNIT(s), "Stopping timed out. Skipping SIGKILL."); + scope_enter_dead(s, SCOPE_FAILURE_TIMEOUT); + } + + break; + + case SCOPE_STOP_SIGKILL: + log_unit_warning(UNIT(s), "Still around after SIGKILL. Ignoring."); + scope_enter_dead(s, SCOPE_FAILURE_TIMEOUT); + break; + + default: + assert_not_reached("Timeout at wrong time."); + } + + return 0; +} + +int scope_abandon(Scope *s) { + assert(s); + + if (unit_has_name(UNIT(s), SPECIAL_INIT_SCOPE)) + return -EPERM; + + if (!IN_SET(s->state, SCOPE_RUNNING, SCOPE_ABANDONED)) + return -ESTALE; + + s->was_abandoned = true; + s->controller = mfree(s->controller); + + /* The client is no longer watching the remaining processes, + * so let's step in here, under the assumption that the + * remaining processes will be sooner or later reassigned to + * us as parent. */ + + unit_tidy_watch_pids(UNIT(s), 0, 0); + unit_watch_all_pids(UNIT(s)); + + /* If the PID set is empty now, then let's finish this off */ + if (set_isempty(UNIT(s)->pids)) + scope_notify_cgroup_empty_event(UNIT(s)); + else + scope_set_state(s, SCOPE_ABANDONED); + + return 0; +} + +_pure_ static UnitActiveState scope_active_state(Unit *u) { + assert(u); + + return state_translation_table[SCOPE(u)->state]; +} + +_pure_ static const char *scope_sub_state_to_string(Unit *u) { + assert(u); + + return scope_state_to_string(SCOPE(u)->state); +} + +static void scope_enumerate(Manager *m) { + Unit *u; + int r; + + assert(m); + + /* Let's unconditionally add the "init.scope" special unit + * that encapsulates PID 1. Note that PID 1 already is in the + * cgroup for this, we hence just need to allocate the object + * for it and that's it. */ + + u = manager_get_unit(m, SPECIAL_INIT_SCOPE); + if (!u) { + u = unit_new(m, sizeof(Scope)); + if (!u) { + log_oom(); + return; + } + + r = unit_add_name(u, SPECIAL_INIT_SCOPE); + if (r < 0) { + unit_free(u); + log_error_errno(r, "Failed to add init.scope name"); + return; + } + } + + u->transient = true; + u->default_dependencies = false; + u->no_gc = true; + u->ignore_on_isolate = true; + u->refuse_manual_start = true; + u->refuse_manual_stop = true; + SCOPE(u)->deserialized_state = SCOPE_RUNNING; + SCOPE(u)->kill_context.kill_signal = SIGRTMIN+14; + + /* Prettify things, if we can. */ + if (!u->description) + u->description = strdup("System and Service Manager"); + if (!u->documentation) + (void) strv_extend(&u->documentation, "man:systemd(1)"); + + unit_add_to_load_queue(u); + unit_add_to_dbus_queue(u); +} + +static const char* const scope_result_table[_SCOPE_RESULT_MAX] = { + [SCOPE_SUCCESS] = "success", + [SCOPE_FAILURE_RESOURCES] = "resources", + [SCOPE_FAILURE_TIMEOUT] = "timeout", +}; + +DEFINE_STRING_TABLE_LOOKUP(scope_result, ScopeResult); + +const UnitVTable scope_vtable = { + .object_size = sizeof(Scope), + .cgroup_context_offset = offsetof(Scope, cgroup_context), + .kill_context_offset = offsetof(Scope, kill_context), + + .sections = + "Unit\0" + "Scope\0" + "Install\0", + .private_section = "Scope", + + .can_transient = true, + + .init = scope_init, + .load = scope_load, + .done = scope_done, + + .coldplug = scope_coldplug, + + .dump = scope_dump, + + .start = scope_start, + .stop = scope_stop, + + .kill = scope_kill, + + .get_timeout = scope_get_timeout, + + .serialize = scope_serialize, + .deserialize_item = scope_deserialize_item, + + .active_state = scope_active_state, + .sub_state_to_string = scope_sub_state_to_string, + + .check_gc = scope_check_gc, + + .sigchld_event = scope_sigchld_event, + + .reset_failed = scope_reset_failed, + + .notify_cgroup_empty = scope_notify_cgroup_empty_event, + + .bus_vtable = bus_scope_vtable, + .bus_set_property = bus_scope_set_property, + .bus_commit_properties = bus_scope_commit_properties, + + .enumerate = scope_enumerate, +}; diff --git a/src/grp-system/libcore/src/selinux-access.c b/src/grp-system/libcore/src/selinux-access.c new file mode 100644 index 0000000000..6807af86c1 --- /dev/null +++ b/src/grp-system/libcore/src/selinux-access.c @@ -0,0 +1,284 @@ +/*** + This file is part of systemd. + + Copyright 2012 Dan Walsh + + systemd is free software; you can redistribute it and/or modify it + under the terms of the GNU Lesser General Public License as published by + the Free Software Foundation; either version 2.1 of the License, or + (at your option) any later version. + + systemd is distributed in the hope that it will be useful, but + WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + Lesser General Public License for more details. + + You should have received a copy of the GNU Lesser General Public License + along with systemd; If not, see . +***/ + +#include "selinux-access.h" + +#ifdef HAVE_SELINUX + +#include +#include +#include +#include +#ifdef HAVE_AUDIT +#include +#endif + +#include + +#include "sd-bus/bus-util.h" +#include "systemd-basic/alloc-util.h" +#include "systemd-basic/log.h" +#include "systemd-basic/path-util.h" +#include "systemd-basic/selinux-util.h" +#include "systemd-basic/stdio-util.h" +#include "systemd-basic/strv.h" +#include "systemd-basic/util.h" + +#include "audit-fd.h" + +static bool initialized = false; + +struct audit_info { + sd_bus_creds *creds; + const char *path; + const char *cmdline; +}; + +/* + Any time an access gets denied this callback will be called + with the audit data. We then need to just copy the audit data into the msgbuf. +*/ +static int audit_callback( + void *auditdata, + security_class_t cls, + char *msgbuf, + size_t msgbufsize) { + + const struct audit_info *audit = auditdata; + uid_t uid = 0, login_uid = 0; + gid_t gid = 0; + char login_uid_buf[DECIMAL_STR_MAX(uid_t) + 1] = "n/a"; + char uid_buf[DECIMAL_STR_MAX(uid_t) + 1] = "n/a"; + char gid_buf[DECIMAL_STR_MAX(gid_t) + 1] = "n/a"; + + if (sd_bus_creds_get_audit_login_uid(audit->creds, &login_uid) >= 0) + xsprintf(login_uid_buf, UID_FMT, login_uid); + if (sd_bus_creds_get_euid(audit->creds, &uid) >= 0) + xsprintf(uid_buf, UID_FMT, uid); + if (sd_bus_creds_get_egid(audit->creds, &gid) >= 0) + xsprintf(gid_buf, GID_FMT, gid); + + snprintf(msgbuf, msgbufsize, + "auid=%s uid=%s gid=%s%s%s%s%s%s%s", + login_uid_buf, uid_buf, gid_buf, + audit->path ? " path=\"" : "", strempty(audit->path), audit->path ? "\"" : "", + audit->cmdline ? " cmdline=\"" : "", strempty(audit->cmdline), audit->cmdline ? "\"" : ""); + + return 0; +} + +static int callback_type_to_priority(int type) { + switch(type) { + + case SELINUX_ERROR: + return LOG_ERR; + + case SELINUX_WARNING: + return LOG_WARNING; + + case SELINUX_INFO: + return LOG_INFO; + + case SELINUX_AVC: + default: + return LOG_NOTICE; + } +} + +/* + libselinux uses this callback when access gets denied or other + events happen. If audit is turned on, messages will be reported + using audit netlink, otherwise they will be logged using the usual + channels. + + Code copied from dbus and modified. +*/ +_printf_(2, 3) static int log_callback(int type, const char *fmt, ...) { + va_list ap; + const char *fmt2; + +#ifdef HAVE_AUDIT + int fd; + + fd = get_audit_fd(); + + if (fd >= 0) { + _cleanup_free_ char *buf = NULL; + int r; + + va_start(ap, fmt); + r = vasprintf(&buf, fmt, ap); + va_end(ap); + + if (r >= 0) { + audit_log_user_avc_message(fd, AUDIT_USER_AVC, buf, NULL, NULL, NULL, 0); + return 0; + } + } +#endif + + fmt2 = strjoina("selinux: ", fmt); + + va_start(ap, fmt); + log_internalv(LOG_AUTH | callback_type_to_priority(type), 0, __FILE__, __LINE__, __FUNCTION__, fmt2, ap); + va_end(ap); + + return 0; +} + +static int access_init(sd_bus_error *error) { + + if (!mac_selinux_use()) + return 0; + + if (initialized) + return 1; + + if (avc_open(NULL, 0) != 0) { + int enforce, saved_errno = errno; + + enforce = security_getenforce(); + log_full_errno(enforce != 0 ? LOG_ERR : LOG_WARNING, saved_errno, "Failed to open the SELinux AVC: %m"); + + /* If enforcement isn't on, then let's suppress this + * error, and just don't do any AVC checks. The + * warning we printed is hence all the admin will + * see. */ + if (enforce == 0) + return 0; + + /* Return an access denied error, if we couldn't load + * the AVC but enforcing mode was on, or we couldn't + * determine whether it is one. */ + return sd_bus_error_setf(error, SD_BUS_ERROR_ACCESS_DENIED, "Failed to open the SELinux AVC: %s", strerror(saved_errno)); + } + + selinux_set_callback(SELINUX_CB_AUDIT, (union selinux_callback) audit_callback); + selinux_set_callback(SELINUX_CB_LOG, (union selinux_callback) log_callback); + + initialized = true; + return 1; +} + +/* + This function communicates with the kernel to check whether or not it should + allow the access. + If the machine is in permissive mode it will return ok. Audit messages will + still be generated if the access would be denied in enforcing mode. +*/ +int mac_selinux_generic_access_check( + sd_bus_message *message, + const char *path, + const char *permission, + sd_bus_error *error) { + + _cleanup_(sd_bus_creds_unrefp) sd_bus_creds *creds = NULL; + const char *tclass = NULL, *scon = NULL; + struct audit_info audit_info = {}; + _cleanup_free_ char *cl = NULL; + char *fcon = NULL; + char **cmdline = NULL; + int r = 0; + + assert(message); + assert(permission); + assert(error); + + r = access_init(error); + if (r <= 0) + return r; + + r = sd_bus_query_sender_creds( + message, + SD_BUS_CREDS_PID|SD_BUS_CREDS_EUID|SD_BUS_CREDS_EGID| + SD_BUS_CREDS_CMDLINE|SD_BUS_CREDS_AUDIT_LOGIN_UID| + SD_BUS_CREDS_SELINUX_CONTEXT| + SD_BUS_CREDS_AUGMENT /* get more bits from /proc */, + &creds); + if (r < 0) + goto finish; + + /* The SELinux context is something we really should have + * gotten directly from the message or sender, and not be an + * augmented field. If it was augmented we cannot use it for + * authorization, since this is racy and vulnerable. Let's add + * an extra check, just in case, even though this really + * shouldn't be possible. */ + assert_return((sd_bus_creds_get_augmented_mask(creds) & SD_BUS_CREDS_SELINUX_CONTEXT) == 0, -EPERM); + + r = sd_bus_creds_get_selinux_context(creds, &scon); + if (r < 0) + goto finish; + + if (path) { + /* Get the file context of the unit file */ + + r = getfilecon_raw(path, &fcon); + if (r < 0) { + r = sd_bus_error_setf(error, SD_BUS_ERROR_ACCESS_DENIED, "Failed to get file context on %s.", path); + goto finish; + } + + tclass = "service"; + } else { + r = getcon_raw(&fcon); + if (r < 0) { + r = sd_bus_error_setf(error, SD_BUS_ERROR_ACCESS_DENIED, "Failed to get current context."); + goto finish; + } + + tclass = "system"; + } + + sd_bus_creds_get_cmdline(creds, &cmdline); + cl = strv_join(cmdline, " "); + + audit_info.creds = creds; + audit_info.path = path; + audit_info.cmdline = cl; + + r = selinux_check_access(scon, fcon, tclass, permission, &audit_info); + if (r < 0) + r = sd_bus_error_setf(error, SD_BUS_ERROR_ACCESS_DENIED, "SELinux policy denies access."); + + log_debug("SELinux access check scon=%s tcon=%s tclass=%s perm=%s path=%s cmdline=%s: %i", scon, fcon, tclass, permission, path, cl, r); + +finish: + freecon(fcon); + + if (r < 0 && security_getenforce() != 1) { + sd_bus_error_free(error); + r = 0; + } + + return r; +} + +#else + +int mac_selinux_generic_access_check( + sd_bus_message *message, + const char *path, + const char *permission, + sd_bus_error *error) { + + return 0; +} + +#endif diff --git a/src/grp-system/libcore/src/selinux-access.h b/src/grp-system/libcore/src/selinux-access.h new file mode 100644 index 0000000000..19adabdee3 --- /dev/null +++ b/src/grp-system/libcore/src/selinux-access.h @@ -0,0 +1,45 @@ +#pragma once + +/*** + This file is part of systemd. + + Copyright 2012 Dan Walsh + + systemd is free software; you can redistribute it and/or modify it + under the terms of the GNU Lesser General Public License as published by + the Free Software Foundation; either version 2.1 of the License, or + (at your option) any later version. + + systemd is distributed in the hope that it will be useful, but + WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + Lesser General Public License for more details. + + You should have received a copy of the GNU Lesser General Public License + along with systemd; If not, see . +***/ + +#include + +#include "core/manager.h" +#include "sd-bus/bus-util.h" + +int mac_selinux_generic_access_check(sd_bus_message *message, const char *path, const char *permission, sd_bus_error *error); + +#ifdef HAVE_SELINUX + +#define mac_selinux_access_check(message, permission, error) \ + mac_selinux_generic_access_check((message), NULL, (permission), (error)) + +#define mac_selinux_unit_access_check(unit, message, permission, error) \ + ({ \ + Unit *_unit = (unit); \ + mac_selinux_generic_access_check((message), _unit->source_path ?: _unit->fragment_path, (permission), (error)); \ + }) + +#else + +#define mac_selinux_access_check(message, permission, error) 0 +#define mac_selinux_unit_access_check(unit, message, permission, error) 0 + +#endif diff --git a/src/grp-system/libcore/src/selinux-setup.c b/src/grp-system/libcore/src/selinux-setup.c new file mode 100644 index 0000000000..d81a8e7fa1 --- /dev/null +++ b/src/grp-system/libcore/src/selinux-setup.c @@ -0,0 +1,121 @@ +/*** + This file is part of systemd. + + Copyright 2010 Lennart Poettering + + systemd is free software; you can redistribute it and/or modify it + under the terms of the GNU Lesser General Public License as published by + the Free Software Foundation; either version 2.1 of the License, or + (at your option) any later version. + + systemd is distributed in the hope that it will be useful, but + WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + Lesser General Public License for more details. + + You should have received a copy of the GNU Lesser General Public License + along with systemd; If not, see . +***/ + +#include +#include +#include + +#ifdef HAVE_SELINUX +#include +#endif + +#include "core/selinux-setup.h" +#include "systemd-basic/log.h" +#include "systemd-basic/macro.h" +#include "systemd-basic/selinux-util.h" +#include "systemd-basic/string-util.h" +#include "systemd-basic/util.h" + +#ifdef HAVE_SELINUX +_printf_(2,3) +static int null_log(int type, const char *fmt, ...) { + return 0; +} +#endif + +int mac_selinux_setup(bool *loaded_policy) { + +#ifdef HAVE_SELINUX + int enforce = 0; + usec_t before_load, after_load; + char *con; + int r; + union selinux_callback cb; + bool initialized = false; + + assert(loaded_policy); + + /* Turn off all of SELinux' own logging, we want to do that */ + cb.func_log = null_log; + selinux_set_callback(SELINUX_CB_LOG, cb); + + /* Don't load policy in the initrd if we don't appear to have + * it. For the real root, we check below if we've already + * loaded policy, and return gracefully. + */ + if (in_initrd() && access(selinux_path(), F_OK) < 0) + return 0; + + /* Already initialized by somebody else? */ + r = getcon_raw(&con); + if (r == 0) { + initialized = !streq(con, "kernel"); + freecon(con); + } + + /* Make sure we have no fds open while loading the policy and + * transitioning */ + log_close(); + + /* Now load the policy */ + before_load = now(CLOCK_MONOTONIC); + r = selinux_init_load_policy(&enforce); + if (r == 0) { + _cleanup_(mac_selinux_freep) char *label = NULL; + char timespan[FORMAT_TIMESPAN_MAX]; + + mac_selinux_retest(); + + /* Transition to the new context */ + r = mac_selinux_get_create_label_from_exe(SYSTEMD_BINARY_PATH, &label); + if (r < 0 || !label) { + log_open(); + log_error("Failed to compute init label, ignoring."); + } else { + r = setcon_raw(label); + + log_open(); + if (r < 0) + log_error("Failed to transition into init label '%s', ignoring.", label); + } + + after_load = now(CLOCK_MONOTONIC); + + log_info("Successfully loaded SELinux policy in %s.", + format_timespan(timespan, sizeof(timespan), after_load - before_load, 0)); + + *loaded_policy = true; + + } else { + log_open(); + + if (enforce > 0) { + if (!initialized) { + log_emergency("Failed to load SELinux policy."); + return -EIO; + } + + log_warning("Failed to load new SELinux policy. Continuing with old policy."); + } else + log_debug("Unable to load SELinux policy. Ignoring."); + } +#endif + + return 0; +} diff --git a/src/grp-system/libcore/src/service.c b/src/grp-system/libcore/src/service.c new file mode 100644 index 0000000000..df114edf09 --- /dev/null +++ b/src/grp-system/libcore/src/service.c @@ -0,0 +1,3392 @@ +/*** + This file is part of systemd. + + Copyright 2010 Lennart Poettering + + systemd is free software; you can redistribute it and/or modify it + under the terms of the GNU Lesser General Public License as published by + the Free Software Foundation; either version 2.1 of the License, or + (at your option) any later version. + + systemd is distributed in the hope that it will be useful, but + WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + Lesser General Public License for more details. + + You should have received a copy of the GNU Lesser General Public License + along with systemd; If not, see . +***/ + +#include +#include +#include + +#include "core/load-fragment.h" +#include "core/manager.h" +#include "core/service.h" +#include "core/unit.h" +#include "sd-bus/bus-error.h" +#include "sd-bus/bus-kernel.h" +#include "sd-bus/bus-util.h" +#include "systemd-basic/alloc-util.h" +#include "systemd-basic/async.h" +#include "systemd-basic/def.h" +#include "systemd-basic/env-util.h" +#include "systemd-basic/escape.h" +#include "systemd-basic/exit-status.h" +#include "systemd-basic/fd-util.h" +#include "systemd-basic/fileio.h" +#include "systemd-basic/formats-util.h" +#include "systemd-basic/fs-util.h" +#include "systemd-basic/log.h" +#include "systemd-basic/parse-util.h" +#include "systemd-basic/path-util.h" +#include "systemd-basic/process-util.h" +#include "systemd-basic/signal-util.h" +#include "systemd-basic/special.h" +#include "systemd-basic/string-table.h" +#include "systemd-basic/string-util.h" +#include "systemd-basic/strv.h" +#include "systemd-basic/unit-name.h" +#include "systemd-basic/utf8.h" +#include "systemd-basic/util.h" + +#include "dbus-service.h" +#include "load-dropin.h" +#include "unit-printf.h" + +static const UnitActiveState state_translation_table[_SERVICE_STATE_MAX] = { + [SERVICE_DEAD] = UNIT_INACTIVE, + [SERVICE_START_PRE] = UNIT_ACTIVATING, + [SERVICE_START] = UNIT_ACTIVATING, + [SERVICE_START_POST] = UNIT_ACTIVATING, + [SERVICE_RUNNING] = UNIT_ACTIVE, + [SERVICE_EXITED] = UNIT_ACTIVE, + [SERVICE_RELOAD] = UNIT_RELOADING, + [SERVICE_STOP] = UNIT_DEACTIVATING, + [SERVICE_STOP_SIGABRT] = UNIT_DEACTIVATING, + [SERVICE_STOP_SIGTERM] = UNIT_DEACTIVATING, + [SERVICE_STOP_SIGKILL] = UNIT_DEACTIVATING, + [SERVICE_STOP_POST] = UNIT_DEACTIVATING, + [SERVICE_FINAL_SIGTERM] = UNIT_DEACTIVATING, + [SERVICE_FINAL_SIGKILL] = UNIT_DEACTIVATING, + [SERVICE_FAILED] = UNIT_FAILED, + [SERVICE_AUTO_RESTART] = UNIT_ACTIVATING +}; + +/* For Type=idle we never want to delay any other jobs, hence we + * consider idle jobs active as soon as we start working on them */ +static const UnitActiveState state_translation_table_idle[_SERVICE_STATE_MAX] = { + [SERVICE_DEAD] = UNIT_INACTIVE, + [SERVICE_START_PRE] = UNIT_ACTIVE, + [SERVICE_START] = UNIT_ACTIVE, + [SERVICE_START_POST] = UNIT_ACTIVE, + [SERVICE_RUNNING] = UNIT_ACTIVE, + [SERVICE_EXITED] = UNIT_ACTIVE, + [SERVICE_RELOAD] = UNIT_RELOADING, + [SERVICE_STOP] = UNIT_DEACTIVATING, + [SERVICE_STOP_SIGABRT] = UNIT_DEACTIVATING, + [SERVICE_STOP_SIGTERM] = UNIT_DEACTIVATING, + [SERVICE_STOP_SIGKILL] = UNIT_DEACTIVATING, + [SERVICE_STOP_POST] = UNIT_DEACTIVATING, + [SERVICE_FINAL_SIGTERM] = UNIT_DEACTIVATING, + [SERVICE_FINAL_SIGKILL] = UNIT_DEACTIVATING, + [SERVICE_FAILED] = UNIT_FAILED, + [SERVICE_AUTO_RESTART] = UNIT_ACTIVATING +}; + +static int service_dispatch_io(sd_event_source *source, int fd, uint32_t events, void *userdata); +static int service_dispatch_timer(sd_event_source *source, usec_t usec, void *userdata); +static int service_dispatch_watchdog(sd_event_source *source, usec_t usec, void *userdata); + +static void service_enter_signal(Service *s, ServiceState state, ServiceResult f); +static void service_enter_reload_by_notify(Service *s); + +static void service_init(Unit *u) { + Service *s = SERVICE(u); + + assert(u); + assert(u->load_state == UNIT_STUB); + + s->timeout_start_usec = u->manager->default_timeout_start_usec; + s->timeout_stop_usec = u->manager->default_timeout_stop_usec; + s->restart_usec = u->manager->default_restart_usec; + s->runtime_max_usec = USEC_INFINITY; + s->type = _SERVICE_TYPE_INVALID; + s->socket_fd = -1; + s->stdin_fd = s->stdout_fd = s->stderr_fd = -1; + s->guess_main_pid = true; + + s->control_command_id = _SERVICE_EXEC_COMMAND_INVALID; +} + +static void service_unwatch_control_pid(Service *s) { + assert(s); + + if (s->control_pid <= 0) + return; + + unit_unwatch_pid(UNIT(s), s->control_pid); + s->control_pid = 0; +} + +static void service_unwatch_main_pid(Service *s) { + assert(s); + + if (s->main_pid <= 0) + return; + + unit_unwatch_pid(UNIT(s), s->main_pid); + s->main_pid = 0; +} + +static void service_unwatch_pid_file(Service *s) { + if (!s->pid_file_pathspec) + return; + + log_unit_debug(UNIT(s), "Stopping watch for PID file %s", s->pid_file_pathspec->path); + path_spec_unwatch(s->pid_file_pathspec); + path_spec_done(s->pid_file_pathspec); + s->pid_file_pathspec = mfree(s->pid_file_pathspec); +} + +static int service_set_main_pid(Service *s, pid_t pid) { + pid_t ppid; + + assert(s); + + if (pid <= 1) + return -EINVAL; + + if (pid == getpid()) + return -EINVAL; + + if (s->main_pid == pid && s->main_pid_known) + return 0; + + if (s->main_pid != pid) { + service_unwatch_main_pid(s); + exec_status_start(&s->main_exec_status, pid); + } + + s->main_pid = pid; + s->main_pid_known = true; + + if (get_process_ppid(pid, &ppid) >= 0 && ppid != getpid()) { + log_unit_warning(UNIT(s), "Supervising process "PID_FMT" which is not our child. We'll most likely not notice when it exits.", pid); + s->main_pid_alien = true; + } else + s->main_pid_alien = false; + + return 0; +} + +void service_close_socket_fd(Service *s) { + assert(s); + + /* Undo the effect of service_set_socket_fd(). */ + + s->socket_fd = asynchronous_close(s->socket_fd); + + if (UNIT_ISSET(s->accept_socket)) { + socket_connection_unref(SOCKET(UNIT_DEREF(s->accept_socket))); + unit_ref_unset(&s->accept_socket); + } +} + +static void service_stop_watchdog(Service *s) { + assert(s); + + s->watchdog_event_source = sd_event_source_unref(s->watchdog_event_source); + s->watchdog_timestamp = DUAL_TIMESTAMP_NULL; +} + +static usec_t service_get_watchdog_usec(Service *s) { + assert(s); + + if (s->watchdog_override_enable) + return s->watchdog_override_usec; + else + return s->watchdog_usec; +} + +static void service_start_watchdog(Service *s) { + int r; + usec_t watchdog_usec; + + assert(s); + + watchdog_usec = service_get_watchdog_usec(s); + if (watchdog_usec == 0 || watchdog_usec == USEC_INFINITY) + return; + + if (s->watchdog_event_source) { + r = sd_event_source_set_time(s->watchdog_event_source, usec_add(s->watchdog_timestamp.monotonic, watchdog_usec)); + if (r < 0) { + log_unit_warning_errno(UNIT(s), r, "Failed to reset watchdog timer: %m"); + return; + } + + r = sd_event_source_set_enabled(s->watchdog_event_source, SD_EVENT_ONESHOT); + } else { + r = sd_event_add_time( + UNIT(s)->manager->event, + &s->watchdog_event_source, + CLOCK_MONOTONIC, + usec_add(s->watchdog_timestamp.monotonic, watchdog_usec), 0, + service_dispatch_watchdog, s); + if (r < 0) { + log_unit_warning_errno(UNIT(s), r, "Failed to add watchdog timer: %m"); + return; + } + + (void) sd_event_source_set_description(s->watchdog_event_source, "service-watchdog"); + + /* Let's process everything else which might be a sign + * of living before we consider a service died. */ + r = sd_event_source_set_priority(s->watchdog_event_source, SD_EVENT_PRIORITY_IDLE); + } + + if (r < 0) + log_unit_warning_errno(UNIT(s), r, "Failed to install watchdog timer: %m"); +} + +static void service_reset_watchdog(Service *s) { + assert(s); + + dual_timestamp_get(&s->watchdog_timestamp); + service_start_watchdog(s); +} + +static void service_reset_watchdog_timeout(Service *s, usec_t watchdog_override_usec) { + assert(s); + + s->watchdog_override_enable = true; + s->watchdog_override_usec = watchdog_override_usec; + service_reset_watchdog(s); + + log_unit_debug(UNIT(s), "watchdog_usec="USEC_FMT, s->watchdog_usec); + log_unit_debug(UNIT(s), "watchdog_override_usec="USEC_FMT, s->watchdog_override_usec); +} + +static void service_fd_store_unlink(ServiceFDStore *fs) { + + if (!fs) + return; + + if (fs->service) { + assert(fs->service->n_fd_store > 0); + LIST_REMOVE(fd_store, fs->service->fd_store, fs); + fs->service->n_fd_store--; + } + + if (fs->event_source) { + sd_event_source_set_enabled(fs->event_source, SD_EVENT_OFF); + sd_event_source_unref(fs->event_source); + } + + free(fs->fdname); + safe_close(fs->fd); + free(fs); +} + +static void service_release_resources(Unit *u) { + Service *s = SERVICE(u); + + assert(s); + + if (!s->fd_store && s->stdin_fd < 0 && s->stdout_fd < 0 && s->stderr_fd < 0) + return; + + log_unit_debug(u, "Releasing all resources."); + + s->stdin_fd = safe_close(s->stdin_fd); + s->stdout_fd = safe_close(s->stdout_fd); + s->stderr_fd = safe_close(s->stderr_fd); + + while (s->fd_store) + service_fd_store_unlink(s->fd_store); + + assert(s->n_fd_store == 0); +} + +static void service_done(Unit *u) { + Service *s = SERVICE(u); + + assert(s); + + s->pid_file = mfree(s->pid_file); + s->status_text = mfree(s->status_text); + + s->exec_runtime = exec_runtime_unref(s->exec_runtime); + exec_command_free_array(s->exec_command, _SERVICE_EXEC_COMMAND_MAX); + s->control_command = NULL; + s->main_command = NULL; + + exit_status_set_free(&s->restart_prevent_status); + exit_status_set_free(&s->restart_force_status); + exit_status_set_free(&s->success_status); + + /* This will leak a process, but at least no memory or any of + * our resources */ + service_unwatch_main_pid(s); + service_unwatch_control_pid(s); + service_unwatch_pid_file(s); + + if (s->bus_name) { + unit_unwatch_bus_name(u, s->bus_name); + s->bus_name = mfree(s->bus_name); + } + + s->bus_name_owner = mfree(s->bus_name_owner); + + service_close_socket_fd(s); + + unit_ref_unset(&s->accept_socket); + + service_stop_watchdog(s); + + s->timer_event_source = sd_event_source_unref(s->timer_event_source); + + service_release_resources(u); +} + +static int on_fd_store_io(sd_event_source *e, int fd, uint32_t revents, void *userdata) { + ServiceFDStore *fs = userdata; + + assert(e); + assert(fs); + + /* If we get either EPOLLHUP or EPOLLERR, it's time to remove this entry from the fd store */ + service_fd_store_unlink(fs); + return 0; +} + +static int service_add_fd_store(Service *s, int fd, const char *name) { + ServiceFDStore *fs; + int r; + + assert(s); + assert(fd >= 0); + + if (s->n_fd_store >= s->n_fd_store_max) + return 0; + + LIST_FOREACH(fd_store, fs, s->fd_store) { + r = same_fd(fs->fd, fd); + if (r < 0) + return r; + if (r > 0) { + /* Already included */ + safe_close(fd); + return 1; + } + } + + fs = new0(ServiceFDStore, 1); + if (!fs) + return -ENOMEM; + + fs->fd = fd; + fs->service = s; + fs->fdname = strdup(name ?: "stored"); + if (!fs->fdname) { + free(fs); + return -ENOMEM; + } + + r = sd_event_add_io(UNIT(s)->manager->event, &fs->event_source, fd, 0, on_fd_store_io, fs); + if (r < 0) { + free(fs->fdname); + free(fs); + return r; + } + + (void) sd_event_source_set_description(fs->event_source, "service-fd-store"); + + LIST_PREPEND(fd_store, s->fd_store, fs); + s->n_fd_store++; + + return 1; +} + +static int service_add_fd_store_set(Service *s, FDSet *fds, const char *name) { + int r; + + assert(s); + + if (fdset_size(fds) <= 0) + return 0; + + while (s->n_fd_store < s->n_fd_store_max) { + _cleanup_close_ int fd = -1; + + fd = fdset_steal_first(fds); + if (fd < 0) + break; + + r = service_add_fd_store(s, fd, name); + if (r < 0) + return log_unit_error_errno(UNIT(s), r, "Couldn't add fd to fd store: %m"); + if (r > 0) { + log_unit_debug(UNIT(s), "Added fd to fd store."); + fd = -1; + } + } + + if (fdset_size(fds) > 0) + log_unit_warning(UNIT(s), "Tried to store more fds than FileDescriptorStoreMax=%u allows, closing remaining.", s->n_fd_store_max); + + return 0; +} + +static int service_arm_timer(Service *s, usec_t usec) { + int r; + + assert(s); + + if (s->timer_event_source) { + r = sd_event_source_set_time(s->timer_event_source, usec); + if (r < 0) + return r; + + return sd_event_source_set_enabled(s->timer_event_source, SD_EVENT_ONESHOT); + } + + if (usec == USEC_INFINITY) + return 0; + + r = sd_event_add_time( + UNIT(s)->manager->event, + &s->timer_event_source, + CLOCK_MONOTONIC, + usec, 0, + service_dispatch_timer, s); + if (r < 0) + return r; + + (void) sd_event_source_set_description(s->timer_event_source, "service-timer"); + + return 0; +} + +static int service_verify(Service *s) { + assert(s); + + if (UNIT(s)->load_state != UNIT_LOADED) + return 0; + + if (!s->exec_command[SERVICE_EXEC_START] && !s->exec_command[SERVICE_EXEC_STOP]) { + log_unit_error(UNIT(s), "Service lacks both ExecStart= and ExecStop= setting. Refusing."); + return -EINVAL; + } + + if (s->type != SERVICE_ONESHOT && !s->exec_command[SERVICE_EXEC_START]) { + log_unit_error(UNIT(s), "Service has no ExecStart= setting, which is only allowed for Type=oneshot services. Refusing."); + return -EINVAL; + } + + if (!s->remain_after_exit && !s->exec_command[SERVICE_EXEC_START]) { + log_unit_error(UNIT(s), "Service has no ExecStart= setting, which is only allowed for RemainAfterExit=yes services. Refusing."); + return -EINVAL; + } + + if (s->type != SERVICE_ONESHOT && s->exec_command[SERVICE_EXEC_START]->command_next) { + log_unit_error(UNIT(s), "Service has more than one ExecStart= setting, which is only allowed for Type=oneshot services. Refusing."); + return -EINVAL; + } + + if (s->type == SERVICE_ONESHOT && s->restart != SERVICE_RESTART_NO) { + log_unit_error(UNIT(s), "Service has Restart= setting other than no, which isn't allowed for Type=oneshot services. Refusing."); + return -EINVAL; + } + + if (s->type == SERVICE_ONESHOT && !exit_status_set_is_empty(&s->restart_force_status)) { + log_unit_error(UNIT(s), "Service has RestartForceStatus= set, which isn't allowed for Type=oneshot services. Refusing."); + return -EINVAL; + } + + if (s->type == SERVICE_DBUS && !s->bus_name) { + log_unit_error(UNIT(s), "Service is of type D-Bus but no D-Bus service name has been specified. Refusing."); + return -EINVAL; + } + + if (s->bus_name && s->type != SERVICE_DBUS) + log_unit_warning(UNIT(s), "Service has a D-Bus service name specified, but is not of type dbus. Ignoring."); + + if (s->exec_context.pam_name && !(s->kill_context.kill_mode == KILL_CONTROL_GROUP || s->kill_context.kill_mode == KILL_MIXED)) { + log_unit_error(UNIT(s), "Service has PAM enabled. Kill mode must be set to 'control-group' or 'mixed'. Refusing."); + return -EINVAL; + } + + if (s->usb_function_descriptors && !s->usb_function_strings) + log_unit_warning(UNIT(s), "Service has USBFunctionDescriptors= setting, but no USBFunctionStrings=. Ignoring."); + + if (!s->usb_function_descriptors && s->usb_function_strings) + log_unit_warning(UNIT(s), "Service has USBFunctionStrings= setting, but no USBFunctionDescriptors=. Ignoring."); + + if (s->runtime_max_usec != USEC_INFINITY && s->type == SERVICE_ONESHOT) + log_unit_warning(UNIT(s), "MaxRuntimeSec= has no effect in combination with Type=oneshot. Ignoring."); + + return 0; +} + +static int service_add_default_dependencies(Service *s) { + int r; + + assert(s); + + if (!UNIT(s)->default_dependencies) + return 0; + + /* Add a number of automatic dependencies useful for the + * majority of services. */ + + if (MANAGER_IS_SYSTEM(UNIT(s)->manager)) { + /* First, pull in the really early boot stuff, and + * require it, so that we fail if we can't acquire + * it. */ + + r = unit_add_two_dependencies_by_name(UNIT(s), UNIT_AFTER, UNIT_REQUIRES, SPECIAL_SYSINIT_TARGET, NULL, true); + if (r < 0) + return r; + } else { + + /* In the --user instance there's no sysinit.target, + * in that case require basic.target instead. */ + + r = unit_add_dependency_by_name(UNIT(s), UNIT_REQUIRES, SPECIAL_BASIC_TARGET, NULL, true); + if (r < 0) + return r; + } + + /* Second, if the rest of the base system is in the same + * transaction, order us after it, but do not pull it in or + * even require it. */ + r = unit_add_dependency_by_name(UNIT(s), UNIT_AFTER, SPECIAL_BASIC_TARGET, NULL, true); + if (r < 0) + return r; + + /* Third, add us in for normal shutdown. */ + return unit_add_two_dependencies_by_name(UNIT(s), UNIT_BEFORE, UNIT_CONFLICTS, SPECIAL_SHUTDOWN_TARGET, NULL, true); +} + +static void service_fix_output(Service *s) { + assert(s); + + /* If nothing has been explicitly configured, patch default + * output in. If input is socket/tty we avoid this however, + * since in that case we want output to default to the same + * place as we read input from. */ + + if (s->exec_context.std_error == EXEC_OUTPUT_INHERIT && + s->exec_context.std_output == EXEC_OUTPUT_INHERIT && + s->exec_context.std_input == EXEC_INPUT_NULL) + s->exec_context.std_error = UNIT(s)->manager->default_std_error; + + if (s->exec_context.std_output == EXEC_OUTPUT_INHERIT && + s->exec_context.std_input == EXEC_INPUT_NULL) + s->exec_context.std_output = UNIT(s)->manager->default_std_output; +} + +static int service_setup_bus_name(Service *s) { + int r; + + assert(s); + + if (!s->bus_name) + return 0; + + r = unit_add_dependency_by_name(UNIT(s), UNIT_REQUIRES, SPECIAL_DBUS_SOCKET, NULL, true); + if (r < 0) + return log_unit_error_errno(UNIT(s), r, "Failed to add dependency on " SPECIAL_DBUS_SOCKET ": %m"); + + /* Regardless if kdbus is used or not, we always want to be ordered against dbus.socket if both are in the transaction. */ + r = unit_add_dependency_by_name(UNIT(s), UNIT_AFTER, SPECIAL_DBUS_SOCKET, NULL, true); + if (r < 0) + return log_unit_error_errno(UNIT(s), r, "Failed to add dependency on " SPECIAL_DBUS_SOCKET ": %m"); + + r = unit_watch_bus_name(UNIT(s), s->bus_name); + if (r == -EEXIST) + return log_unit_error_errno(UNIT(s), r, "Two services allocated for the same bus name %s, refusing operation.", s->bus_name); + if (r < 0) + return log_unit_error_errno(UNIT(s), r, "Cannot watch bus name %s: %m", s->bus_name); + + return 0; +} + +static int service_add_extras(Service *s) { + int r; + + assert(s); + + if (s->type == _SERVICE_TYPE_INVALID) { + /* Figure out a type automatically */ + if (s->bus_name) + s->type = SERVICE_DBUS; + else if (s->exec_command[SERVICE_EXEC_START]) + s->type = SERVICE_SIMPLE; + else + s->type = SERVICE_ONESHOT; + } + + /* Oneshot services have disabled start timeout by default */ + if (s->type == SERVICE_ONESHOT && !s->start_timeout_defined) + s->timeout_start_usec = USEC_INFINITY; + + service_fix_output(s); + + r = unit_patch_contexts(UNIT(s)); + if (r < 0) + return r; + + r = unit_add_exec_dependencies(UNIT(s), &s->exec_context); + if (r < 0) + return r; + + r = unit_set_default_slice(UNIT(s)); + if (r < 0) + return r; + + if (s->type == SERVICE_NOTIFY && s->notify_access == NOTIFY_NONE) + s->notify_access = NOTIFY_MAIN; + + if (s->watchdog_usec > 0 && s->notify_access == NOTIFY_NONE) + s->notify_access = NOTIFY_MAIN; + + r = service_add_default_dependencies(s); + if (r < 0) + return r; + + r = service_setup_bus_name(s); + if (r < 0) + return r; + + return 0; +} + +static int service_load(Unit *u) { + Service *s = SERVICE(u); + int r; + + assert(s); + + /* Load a .service file */ + r = unit_load_fragment(u); + if (r < 0) + return r; + + /* Still nothing found? Then let's give up */ + if (u->load_state == UNIT_STUB) + return -ENOENT; + + /* This is a new unit? Then let's add in some extras */ + if (u->load_state == UNIT_LOADED) { + + /* We were able to load something, then let's add in + * the dropin directories. */ + r = unit_load_dropin(u); + if (r < 0) + return r; + + /* This is a new unit? Then let's add in some + * extras */ + r = service_add_extras(s); + if (r < 0) + return r; + } + + return service_verify(s); +} + +static void service_dump(Unit *u, FILE *f, const char *prefix) { + ServiceExecCommand c; + Service *s = SERVICE(u); + const char *prefix2; + + assert(s); + + prefix = strempty(prefix); + prefix2 = strjoina(prefix, "\t"); + + fprintf(f, + "%sService State: %s\n" + "%sResult: %s\n" + "%sReload Result: %s\n" + "%sPermissionsStartOnly: %s\n" + "%sRootDirectoryStartOnly: %s\n" + "%sRemainAfterExit: %s\n" + "%sGuessMainPID: %s\n" + "%sType: %s\n" + "%sRestart: %s\n" + "%sNotifyAccess: %s\n" + "%sNotifyState: %s\n", + prefix, service_state_to_string(s->state), + prefix, service_result_to_string(s->result), + prefix, service_result_to_string(s->reload_result), + prefix, yes_no(s->permissions_start_only), + prefix, yes_no(s->root_directory_start_only), + prefix, yes_no(s->remain_after_exit), + prefix, yes_no(s->guess_main_pid), + prefix, service_type_to_string(s->type), + prefix, service_restart_to_string(s->restart), + prefix, notify_access_to_string(s->notify_access), + prefix, notify_state_to_string(s->notify_state)); + + if (s->control_pid > 0) + fprintf(f, + "%sControl PID: "PID_FMT"\n", + prefix, s->control_pid); + + if (s->main_pid > 0) + fprintf(f, + "%sMain PID: "PID_FMT"\n" + "%sMain PID Known: %s\n" + "%sMain PID Alien: %s\n", + prefix, s->main_pid, + prefix, yes_no(s->main_pid_known), + prefix, yes_no(s->main_pid_alien)); + + if (s->pid_file) + fprintf(f, + "%sPIDFile: %s\n", + prefix, s->pid_file); + + if (s->bus_name) + fprintf(f, + "%sBusName: %s\n" + "%sBus Name Good: %s\n", + prefix, s->bus_name, + prefix, yes_no(s->bus_name_good)); + + kill_context_dump(&s->kill_context, f, prefix); + exec_context_dump(&s->exec_context, f, prefix); + + for (c = 0; c < _SERVICE_EXEC_COMMAND_MAX; c++) { + + if (!s->exec_command[c]) + continue; + + fprintf(f, "%s-> %s:\n", + prefix, service_exec_command_to_string(c)); + + exec_command_dump_list(s->exec_command[c], f, prefix2); + } + + if (s->status_text) + fprintf(f, "%sStatus Text: %s\n", + prefix, s->status_text); + + if (s->n_fd_store_max > 0) + fprintf(f, + "%sFile Descriptor Store Max: %u\n" + "%sFile Descriptor Store Current: %u\n", + prefix, s->n_fd_store_max, + prefix, s->n_fd_store); +} + +static int service_load_pid_file(Service *s, bool may_warn) { + _cleanup_free_ char *k = NULL; + int r; + pid_t pid; + + assert(s); + + if (!s->pid_file) + return -ENOENT; + + r = read_one_line_file(s->pid_file, &k); + if (r < 0) { + if (may_warn) + log_unit_info_errno(UNIT(s), r, "PID file %s not readable (yet?) after %s: %m", s->pid_file, service_state_to_string(s->state)); + return r; + } + + r = parse_pid(k, &pid); + if (r < 0) { + if (may_warn) + log_unit_info_errno(UNIT(s), r, "Failed to read PID from file %s: %m", s->pid_file); + return r; + } + + if (!pid_is_alive(pid)) { + if (may_warn) + log_unit_info(UNIT(s), "PID "PID_FMT" read from file %s does not exist or is a zombie.", pid, s->pid_file); + return -ESRCH; + } + + if (s->main_pid_known) { + if (pid == s->main_pid) + return 0; + + log_unit_debug(UNIT(s), "Main PID changing: "PID_FMT" -> "PID_FMT, s->main_pid, pid); + + service_unwatch_main_pid(s); + s->main_pid_known = false; + } else + log_unit_debug(UNIT(s), "Main PID loaded: "PID_FMT, pid); + + r = service_set_main_pid(s, pid); + if (r < 0) + return r; + + r = unit_watch_pid(UNIT(s), pid); + if (r < 0) { + /* FIXME: we need to do something here */ + log_unit_warning_errno(UNIT(s), r, "Failed to watch PID "PID_FMT" for service: %m", pid); + return r; + } + + return 0; +} + +static void service_search_main_pid(Service *s) { + pid_t pid = 0; + int r; + + assert(s); + + /* If we know it anyway, don't ever fallback to unreliable + * heuristics */ + if (s->main_pid_known) + return; + + if (!s->guess_main_pid) + return; + + assert(s->main_pid <= 0); + + if (unit_search_main_pid(UNIT(s), &pid) < 0) + return; + + log_unit_debug(UNIT(s), "Main PID guessed: "PID_FMT, pid); + if (service_set_main_pid(s, pid) < 0) + return; + + r = unit_watch_pid(UNIT(s), pid); + if (r < 0) + /* FIXME: we need to do something here */ + log_unit_warning_errno(UNIT(s), r, "Failed to watch PID "PID_FMT" from: %m", pid); +} + +static void service_set_state(Service *s, ServiceState state) { + ServiceState old_state; + const UnitActiveState *table; + + assert(s); + + table = s->type == SERVICE_IDLE ? state_translation_table_idle : state_translation_table; + + old_state = s->state; + s->state = state; + + service_unwatch_pid_file(s); + + if (!IN_SET(state, + SERVICE_START_PRE, SERVICE_START, SERVICE_START_POST, + SERVICE_RUNNING, + SERVICE_RELOAD, + SERVICE_STOP, SERVICE_STOP_SIGABRT, SERVICE_STOP_SIGTERM, SERVICE_STOP_SIGKILL, SERVICE_STOP_POST, + SERVICE_FINAL_SIGTERM, SERVICE_FINAL_SIGKILL, + SERVICE_AUTO_RESTART)) + s->timer_event_source = sd_event_source_unref(s->timer_event_source); + + if (!IN_SET(state, + SERVICE_START, SERVICE_START_POST, + SERVICE_RUNNING, SERVICE_RELOAD, + SERVICE_STOP, SERVICE_STOP_SIGABRT, SERVICE_STOP_SIGTERM, SERVICE_STOP_SIGKILL, SERVICE_STOP_POST, + SERVICE_FINAL_SIGTERM, SERVICE_FINAL_SIGKILL)) { + service_unwatch_main_pid(s); + s->main_command = NULL; + } + + if (!IN_SET(state, + SERVICE_START_PRE, SERVICE_START, SERVICE_START_POST, + SERVICE_RELOAD, + SERVICE_STOP, SERVICE_STOP_SIGABRT, SERVICE_STOP_SIGTERM, SERVICE_STOP_SIGKILL, SERVICE_STOP_POST, + SERVICE_FINAL_SIGTERM, SERVICE_FINAL_SIGKILL)) { + service_unwatch_control_pid(s); + s->control_command = NULL; + s->control_command_id = _SERVICE_EXEC_COMMAND_INVALID; + } + + if (IN_SET(state, SERVICE_DEAD, SERVICE_FAILED, SERVICE_AUTO_RESTART)) + unit_unwatch_all_pids(UNIT(s)); + + if (!IN_SET(state, + SERVICE_START_PRE, SERVICE_START, SERVICE_START_POST, + SERVICE_RUNNING, SERVICE_RELOAD, + SERVICE_STOP, SERVICE_STOP_SIGABRT, SERVICE_STOP_SIGTERM, SERVICE_STOP_SIGKILL, SERVICE_STOP_POST, + SERVICE_FINAL_SIGTERM, SERVICE_FINAL_SIGKILL) && + !(state == SERVICE_DEAD && UNIT(s)->job)) + service_close_socket_fd(s); + + if (!IN_SET(state, SERVICE_START_POST, SERVICE_RUNNING, SERVICE_RELOAD)) + service_stop_watchdog(s); + + /* For the inactive states unit_notify() will trim the cgroup, + * but for exit we have to do that ourselves... */ + if (state == SERVICE_EXITED && !MANAGER_IS_RELOADING(UNIT(s)->manager)) + unit_prune_cgroup(UNIT(s)); + + /* For remain_after_exit services, let's see if we can "release" the + * hold on the console, since unit_notify() only does that in case of + * change of state */ + if (state == SERVICE_EXITED && + s->remain_after_exit && + UNIT(s)->manager->n_on_console > 0) { + + ExecContext *ec; + + ec = unit_get_exec_context(UNIT(s)); + if (ec && exec_context_may_touch_console(ec)) { + Manager *m = UNIT(s)->manager; + + m->n_on_console--; + if (m->n_on_console == 0) + /* unset no_console_output flag, since the console is free */ + m->no_console_output = false; + } + } + + if (old_state != state) + log_unit_debug(UNIT(s), "Changed %s -> %s", service_state_to_string(old_state), service_state_to_string(state)); + + unit_notify(UNIT(s), table[old_state], table[state], s->reload_result == SERVICE_SUCCESS); +} + +static usec_t service_coldplug_timeout(Service *s) { + assert(s); + + switch (s->deserialized_state) { + + case SERVICE_START_PRE: + case SERVICE_START: + case SERVICE_START_POST: + case SERVICE_RELOAD: + return usec_add(UNIT(s)->state_change_timestamp.monotonic, s->timeout_start_usec); + + case SERVICE_RUNNING: + return usec_add(UNIT(s)->active_enter_timestamp.monotonic, s->runtime_max_usec); + + case SERVICE_STOP: + case SERVICE_STOP_SIGABRT: + case SERVICE_STOP_SIGTERM: + case SERVICE_STOP_SIGKILL: + case SERVICE_STOP_POST: + case SERVICE_FINAL_SIGTERM: + case SERVICE_FINAL_SIGKILL: + return usec_add(UNIT(s)->state_change_timestamp.monotonic, s->timeout_stop_usec); + + case SERVICE_AUTO_RESTART: + return usec_add(UNIT(s)->inactive_enter_timestamp.monotonic, s->restart_usec); + + default: + return USEC_INFINITY; + } +} + +static int service_coldplug(Unit *u) { + Service *s = SERVICE(u); + int r; + + assert(s); + assert(s->state == SERVICE_DEAD); + + if (s->deserialized_state == s->state) + return 0; + + r = service_arm_timer(s, service_coldplug_timeout(s)); + if (r < 0) + return r; + + if (s->main_pid > 0 && + pid_is_unwaited(s->main_pid) && + ((s->deserialized_state == SERVICE_START && IN_SET(s->type, SERVICE_FORKING, SERVICE_DBUS, SERVICE_ONESHOT, SERVICE_NOTIFY)) || + IN_SET(s->deserialized_state, + SERVICE_START, SERVICE_START_POST, + SERVICE_RUNNING, SERVICE_RELOAD, + SERVICE_STOP, SERVICE_STOP_SIGABRT, SERVICE_STOP_SIGTERM, SERVICE_STOP_SIGKILL, SERVICE_STOP_POST, + SERVICE_FINAL_SIGTERM, SERVICE_FINAL_SIGKILL))) { + r = unit_watch_pid(UNIT(s), s->main_pid); + if (r < 0) + return r; + } + + if (s->control_pid > 0 && + pid_is_unwaited(s->control_pid) && + IN_SET(s->deserialized_state, + SERVICE_START_PRE, SERVICE_START, SERVICE_START_POST, + SERVICE_RELOAD, + SERVICE_STOP, SERVICE_STOP_SIGABRT, SERVICE_STOP_SIGTERM, SERVICE_STOP_SIGKILL, SERVICE_STOP_POST, + SERVICE_FINAL_SIGTERM, SERVICE_FINAL_SIGKILL)) { + r = unit_watch_pid(UNIT(s), s->control_pid); + if (r < 0) + return r; + } + + if (!IN_SET(s->deserialized_state, SERVICE_DEAD, SERVICE_FAILED, SERVICE_AUTO_RESTART)) + unit_watch_all_pids(UNIT(s)); + + if (IN_SET(s->deserialized_state, SERVICE_START_POST, SERVICE_RUNNING, SERVICE_RELOAD)) + service_start_watchdog(s); + + service_set_state(s, s->deserialized_state); + return 0; +} + +static int service_collect_fds(Service *s, int **fds, char ***fd_names) { + _cleanup_strv_free_ char **rfd_names = NULL; + _cleanup_free_ int *rfds = NULL; + int rn_fds = 0, r; + + assert(s); + assert(fds); + assert(fd_names); + + if (s->socket_fd >= 0) { + + /* Pass the per-connection socket */ + + rfds = new(int, 1); + if (!rfds) + return -ENOMEM; + rfds[0] = s->socket_fd; + + rfd_names = strv_new("connection", NULL); + if (!rfd_names) + return -ENOMEM; + + rn_fds = 1; + } else { + Iterator i; + Unit *u; + + /* Pass all our configured sockets for singleton services */ + + SET_FOREACH(u, UNIT(s)->dependencies[UNIT_TRIGGERED_BY], i) { + _cleanup_free_ int *cfds = NULL; + Socket *sock; + int cn_fds; + + if (u->type != UNIT_SOCKET) + continue; + + sock = SOCKET(u); + + cn_fds = socket_collect_fds(sock, &cfds); + if (cn_fds < 0) + return cn_fds; + + if (cn_fds <= 0) + continue; + + if (!rfds) { + rfds = cfds; + rn_fds = cn_fds; + + cfds = NULL; + } else { + int *t; + + t = realloc(rfds, (rn_fds + cn_fds) * sizeof(int)); + if (!t) + return -ENOMEM; + + memcpy(t + rn_fds, cfds, cn_fds * sizeof(int)); + + rfds = t; + rn_fds += cn_fds; + } + + r = strv_extend_n(&rfd_names, socket_fdname(sock), cn_fds); + if (r < 0) + return r; + } + } + + if (s->n_fd_store > 0) { + ServiceFDStore *fs; + char **nl; + int *t; + + t = realloc(rfds, (rn_fds + s->n_fd_store) * sizeof(int)); + if (!t) + return -ENOMEM; + + rfds = t; + + nl = realloc(rfd_names, (rn_fds + s->n_fd_store + 1) * sizeof(char*)); + if (!nl) + return -ENOMEM; + + rfd_names = nl; + + LIST_FOREACH(fd_store, fs, s->fd_store) { + rfds[rn_fds] = fs->fd; + rfd_names[rn_fds] = strdup(strempty(fs->fdname)); + if (!rfd_names[rn_fds]) + return -ENOMEM; + + rn_fds++; + } + + rfd_names[rn_fds] = NULL; + } + + *fds = rfds; + *fd_names = rfd_names; + + rfds = NULL; + rfd_names = NULL; + + return rn_fds; +} + +static int service_spawn( + Service *s, + ExecCommand *c, + usec_t timeout, + bool pass_fds, + bool apply_permissions, + bool apply_chroot, + bool apply_tty_stdin, + bool is_control, + pid_t *_pid) { + + _cleanup_strv_free_ char **argv = NULL, **final_env = NULL, **our_env = NULL, **fd_names = NULL; + _cleanup_free_ int *fds = NULL; + unsigned n_fds = 0, n_env = 0; + const char *path; + pid_t pid; + + ExecParameters exec_params = { + .apply_permissions = apply_permissions, + .apply_chroot = apply_chroot, + .apply_tty_stdin = apply_tty_stdin, + .stdin_fd = -1, + .stdout_fd = -1, + .stderr_fd = -1, + }; + + int r; + + assert(s); + assert(c); + assert(_pid); + + (void) unit_realize_cgroup(UNIT(s)); + if (s->reset_cpu_usage) { + (void) unit_reset_cpu_usage(UNIT(s)); + s->reset_cpu_usage = false; + } + + r = unit_setup_exec_runtime(UNIT(s)); + if (r < 0) + return r; + + if (pass_fds || + s->exec_context.std_input == EXEC_INPUT_SOCKET || + s->exec_context.std_output == EXEC_OUTPUT_SOCKET || + s->exec_context.std_error == EXEC_OUTPUT_SOCKET) { + + r = service_collect_fds(s, &fds, &fd_names); + if (r < 0) + return r; + + n_fds = r; + } + + r = service_arm_timer(s, usec_add(now(CLOCK_MONOTONIC), timeout)); + if (r < 0) + return r; + + r = unit_full_printf_strv(UNIT(s), c->argv, &argv); + if (r < 0) + return r; + + our_env = new0(char*, 6); + if (!our_env) + return -ENOMEM; + + if (is_control ? s->notify_access == NOTIFY_ALL : s->notify_access != NOTIFY_NONE) + if (asprintf(our_env + n_env++, "NOTIFY_SOCKET=%s", UNIT(s)->manager->notify_socket) < 0) + return -ENOMEM; + + if (s->main_pid > 0) + if (asprintf(our_env + n_env++, "MAINPID="PID_FMT, s->main_pid) < 0) + return -ENOMEM; + + if (!MANAGER_IS_SYSTEM(UNIT(s)->manager)) + if (asprintf(our_env + n_env++, "MANAGERPID="PID_FMT, getpid()) < 0) + return -ENOMEM; + + if (s->socket_fd >= 0) { + union sockaddr_union sa; + socklen_t salen = sizeof(sa); + + r = getpeername(s->socket_fd, &sa.sa, &salen); + if (r < 0) + return -errno; + + if (IN_SET(sa.sa.sa_family, AF_INET, AF_INET6)) { + _cleanup_free_ char *addr = NULL; + char *t; + int port; + + r = sockaddr_pretty(&sa.sa, salen, true, false, &addr); + if (r < 0) + return r; + + t = strappend("REMOTE_ADDR=", addr); + if (!t) + return -ENOMEM; + our_env[n_env++] = t; + + port = sockaddr_port(&sa.sa); + if (port < 0) + return port; + + if (asprintf(&t, "REMOTE_PORT=%u", port) < 0) + return -ENOMEM; + our_env[n_env++] = t; + } + } + + final_env = strv_env_merge(2, UNIT(s)->manager->environment, our_env, NULL); + if (!final_env) + return -ENOMEM; + + if (is_control && UNIT(s)->cgroup_path) { + path = strjoina(UNIT(s)->cgroup_path, "/control"); + (void) cg_create(SYSTEMD_CGROUP_CONTROLLER, path); + } else + path = UNIT(s)->cgroup_path; + + exec_params.argv = argv; + exec_params.fds = fds; + exec_params.fd_names = fd_names; + exec_params.n_fds = n_fds; + exec_params.environment = final_env; + exec_params.confirm_spawn = UNIT(s)->manager->confirm_spawn; + exec_params.cgroup_supported = UNIT(s)->manager->cgroup_supported; + exec_params.cgroup_path = path; + exec_params.cgroup_delegate = s->cgroup_context.delegate; + exec_params.runtime_prefix = manager_get_runtime_prefix(UNIT(s)->manager); + exec_params.watchdog_usec = s->watchdog_usec; + exec_params.selinux_context_net = s->socket_fd_selinux_context_net; + if (s->type == SERVICE_IDLE) + exec_params.idle_pipe = UNIT(s)->manager->idle_pipe; + exec_params.stdin_fd = s->stdin_fd; + exec_params.stdout_fd = s->stdout_fd; + exec_params.stderr_fd = s->stderr_fd; + + r = exec_spawn(UNIT(s), + c, + &s->exec_context, + &exec_params, + s->exec_runtime, + &pid); + if (r < 0) + return r; + + r = unit_watch_pid(UNIT(s), pid); + if (r < 0) + /* FIXME: we need to do something here */ + return r; + + *_pid = pid; + + return 0; +} + +static int main_pid_good(Service *s) { + assert(s); + + /* Returns 0 if the pid is dead, 1 if it is good, -1 if we + * don't know */ + + /* If we know the pid file, then let's just check if it is + * still valid */ + if (s->main_pid_known) { + + /* If it's an alien child let's check if it is still + * alive ... */ + if (s->main_pid_alien && s->main_pid > 0) + return pid_is_alive(s->main_pid); + + /* .. otherwise assume we'll get a SIGCHLD for it, + * which we really should wait for to collect exit + * status and code */ + return s->main_pid > 0; + } + + /* We don't know the pid */ + return -EAGAIN; +} + +_pure_ static int control_pid_good(Service *s) { + assert(s); + + return s->control_pid > 0; +} + +static int cgroup_good(Service *s) { + int r; + + assert(s); + + if (!UNIT(s)->cgroup_path) + return 0; + + r = cg_is_empty_recursive(SYSTEMD_CGROUP_CONTROLLER, UNIT(s)->cgroup_path); + if (r < 0) + return r; + + return !r; +} + +static bool service_shall_restart(Service *s) { + assert(s); + + /* Don't restart after manual stops */ + if (s->forbid_restart) + return false; + + /* Never restart if this is configured as special exception */ + if (exit_status_set_test(&s->restart_prevent_status, s->main_exec_status.code, s->main_exec_status.status)) + return false; + + /* Restart if the exit code/status are configured as restart triggers */ + if (exit_status_set_test(&s->restart_force_status, s->main_exec_status.code, s->main_exec_status.status)) + return true; + + switch (s->restart) { + + case SERVICE_RESTART_NO: + return false; + + case SERVICE_RESTART_ALWAYS: + return true; + + case SERVICE_RESTART_ON_SUCCESS: + return s->result == SERVICE_SUCCESS; + + case SERVICE_RESTART_ON_FAILURE: + return s->result != SERVICE_SUCCESS; + + case SERVICE_RESTART_ON_ABNORMAL: + return !IN_SET(s->result, SERVICE_SUCCESS, SERVICE_FAILURE_EXIT_CODE); + + case SERVICE_RESTART_ON_WATCHDOG: + return s->result == SERVICE_FAILURE_WATCHDOG; + + case SERVICE_RESTART_ON_ABORT: + return IN_SET(s->result, SERVICE_FAILURE_SIGNAL, SERVICE_FAILURE_CORE_DUMP); + + default: + assert_not_reached("unknown restart setting"); + } +} + +static void service_enter_dead(Service *s, ServiceResult f, bool allow_restart) { + int r; + assert(s); + + if (f != SERVICE_SUCCESS) + s->result = f; + + service_set_state(s, s->result != SERVICE_SUCCESS ? SERVICE_FAILED : SERVICE_DEAD); + + if (s->result != SERVICE_SUCCESS) { + log_unit_warning(UNIT(s), "Failed with result '%s'.", service_result_to_string(s->result)); + failure_action(UNIT(s)->manager, s->failure_action, UNIT(s)->reboot_arg); + } + + if (allow_restart && service_shall_restart(s)) { + + r = service_arm_timer(s, usec_add(now(CLOCK_MONOTONIC), s->restart_usec)); + if (r < 0) + goto fail; + + service_set_state(s, SERVICE_AUTO_RESTART); + } + + /* The next restart might not be a manual stop, hence reset the flag indicating manual stops */ + s->forbid_restart = false; + + /* We want fresh tmpdirs in case service is started again immediately */ + exec_runtime_destroy(s->exec_runtime); + s->exec_runtime = exec_runtime_unref(s->exec_runtime); + + /* Also, remove the runtime directory in */ + exec_context_destroy_runtime_directory(&s->exec_context, manager_get_runtime_prefix(UNIT(s)->manager)); + + /* Try to delete the pid file. At this point it will be + * out-of-date, and some software might be confused by it, so + * let's remove it. */ + if (s->pid_file) + (void) unlink(s->pid_file); + + return; + +fail: + log_unit_warning_errno(UNIT(s), r, "Failed to run install restart timer: %m"); + service_enter_dead(s, SERVICE_FAILURE_RESOURCES, false); +} + +static void service_enter_stop_post(Service *s, ServiceResult f) { + int r; + assert(s); + + if (f != SERVICE_SUCCESS) + s->result = f; + + service_unwatch_control_pid(s); + unit_watch_all_pids(UNIT(s)); + + s->control_command = s->exec_command[SERVICE_EXEC_STOP_POST]; + if (s->control_command) { + s->control_command_id = SERVICE_EXEC_STOP_POST; + + r = service_spawn(s, + s->control_command, + s->timeout_stop_usec, + false, + !s->permissions_start_only, + !s->root_directory_start_only, + true, + true, + &s->control_pid); + if (r < 0) + goto fail; + + service_set_state(s, SERVICE_STOP_POST); + } else + service_enter_signal(s, SERVICE_FINAL_SIGTERM, SERVICE_SUCCESS); + + return; + +fail: + log_unit_warning_errno(UNIT(s), r, "Failed to run 'stop-post' task: %m"); + service_enter_signal(s, SERVICE_FINAL_SIGTERM, SERVICE_FAILURE_RESOURCES); +} + +static int state_to_kill_operation(ServiceState state) { + switch (state) { + + case SERVICE_STOP_SIGABRT: + return KILL_ABORT; + + case SERVICE_STOP_SIGTERM: + case SERVICE_FINAL_SIGTERM: + return KILL_TERMINATE; + + case SERVICE_STOP_SIGKILL: + case SERVICE_FINAL_SIGKILL: + return KILL_KILL; + + default: + return _KILL_OPERATION_INVALID; + } +} + +static void service_enter_signal(Service *s, ServiceState state, ServiceResult f) { + int r; + + assert(s); + + if (f != SERVICE_SUCCESS) + s->result = f; + + unit_watch_all_pids(UNIT(s)); + + r = unit_kill_context( + UNIT(s), + &s->kill_context, + state_to_kill_operation(state), + s->main_pid, + s->control_pid, + s->main_pid_alien); + + if (r < 0) + goto fail; + + if (r > 0) { + r = service_arm_timer(s, usec_add(now(CLOCK_MONOTONIC), s->timeout_stop_usec)); + if (r < 0) + goto fail; + + service_set_state(s, state); + } else if (IN_SET(state, SERVICE_STOP_SIGABRT, SERVICE_STOP_SIGTERM) && s->kill_context.send_sigkill) + service_enter_signal(s, SERVICE_STOP_SIGKILL, SERVICE_SUCCESS); + else if (IN_SET(state, SERVICE_STOP_SIGABRT, SERVICE_STOP_SIGTERM, SERVICE_STOP_SIGKILL)) + service_enter_stop_post(s, SERVICE_SUCCESS); + else if (state == SERVICE_FINAL_SIGTERM && s->kill_context.send_sigkill) + service_enter_signal(s, SERVICE_FINAL_SIGKILL, SERVICE_SUCCESS); + else + service_enter_dead(s, SERVICE_SUCCESS, true); + + return; + +fail: + log_unit_warning_errno(UNIT(s), r, "Failed to kill processes: %m"); + + if (IN_SET(state, SERVICE_STOP_SIGABRT, SERVICE_STOP_SIGTERM, SERVICE_STOP_SIGKILL)) + service_enter_stop_post(s, SERVICE_FAILURE_RESOURCES); + else + service_enter_dead(s, SERVICE_FAILURE_RESOURCES, true); +} + +static void service_enter_stop_by_notify(Service *s) { + assert(s); + + unit_watch_all_pids(UNIT(s)); + + service_arm_timer(s, usec_add(now(CLOCK_MONOTONIC), s->timeout_stop_usec)); + + /* The service told us it's stopping, so it's as if we SIGTERM'd it. */ + service_set_state(s, SERVICE_STOP_SIGTERM); +} + +static void service_enter_stop(Service *s, ServiceResult f) { + int r; + + assert(s); + + if (f != SERVICE_SUCCESS) + s->result = f; + + service_unwatch_control_pid(s); + unit_watch_all_pids(UNIT(s)); + + s->control_command = s->exec_command[SERVICE_EXEC_STOP]; + if (s->control_command) { + s->control_command_id = SERVICE_EXEC_STOP; + + r = service_spawn(s, + s->control_command, + s->timeout_stop_usec, + false, + !s->permissions_start_only, + !s->root_directory_start_only, + false, + true, + &s->control_pid); + if (r < 0) + goto fail; + + service_set_state(s, SERVICE_STOP); + } else + service_enter_signal(s, SERVICE_STOP_SIGTERM, SERVICE_SUCCESS); + + return; + +fail: + log_unit_warning_errno(UNIT(s), r, "Failed to run 'stop' task: %m"); + service_enter_signal(s, SERVICE_STOP_SIGTERM, SERVICE_FAILURE_RESOURCES); +} + +static bool service_good(Service *s) { + int main_pid_ok; + assert(s); + + if (s->type == SERVICE_DBUS && !s->bus_name_good) + return false; + + main_pid_ok = main_pid_good(s); + if (main_pid_ok > 0) /* It's alive */ + return true; + if (main_pid_ok == 0) /* It's dead */ + return false; + + /* OK, we don't know anything about the main PID, maybe + * because there is none. Let's check the control group + * instead. */ + + return cgroup_good(s) != 0; +} + +static void service_enter_running(Service *s, ServiceResult f) { + assert(s); + + if (f != SERVICE_SUCCESS) + s->result = f; + + service_unwatch_control_pid(s); + + if (service_good(s)) { + + /* If there are any queued up sd_notify() + * notifications, process them now */ + if (s->notify_state == NOTIFY_RELOADING) + service_enter_reload_by_notify(s); + else if (s->notify_state == NOTIFY_STOPPING) + service_enter_stop_by_notify(s); + else { + service_set_state(s, SERVICE_RUNNING); + service_arm_timer(s, usec_add(UNIT(s)->active_enter_timestamp.monotonic, s->runtime_max_usec)); + } + + } else if (s->remain_after_exit) + service_set_state(s, SERVICE_EXITED); + else + service_enter_stop(s, SERVICE_SUCCESS); +} + +static void service_enter_start_post(Service *s) { + int r; + assert(s); + + service_unwatch_control_pid(s); + service_reset_watchdog(s); + + s->control_command = s->exec_command[SERVICE_EXEC_START_POST]; + if (s->control_command) { + s->control_command_id = SERVICE_EXEC_START_POST; + + r = service_spawn(s, + s->control_command, + s->timeout_start_usec, + false, + !s->permissions_start_only, + !s->root_directory_start_only, + false, + true, + &s->control_pid); + if (r < 0) + goto fail; + + service_set_state(s, SERVICE_START_POST); + } else + service_enter_running(s, SERVICE_SUCCESS); + + return; + +fail: + log_unit_warning_errno(UNIT(s), r, "Failed to run 'start-post' task: %m"); + service_enter_stop(s, SERVICE_FAILURE_RESOURCES); +} + +static void service_kill_control_processes(Service *s) { + char *p; + + if (!UNIT(s)->cgroup_path) + return; + + p = strjoina(UNIT(s)->cgroup_path, "/control"); + cg_kill_recursive(SYSTEMD_CGROUP_CONTROLLER, p, SIGKILL, CGROUP_SIGCONT|CGROUP_IGNORE_SELF|CGROUP_REMOVE, NULL, NULL, NULL); +} + +static void service_enter_start(Service *s) { + ExecCommand *c; + usec_t timeout; + pid_t pid; + int r; + + assert(s); + + service_unwatch_control_pid(s); + service_unwatch_main_pid(s); + + /* We want to ensure that nobody leaks processes from + * START_PRE here, so let's go on a killing spree, People + * should not spawn long running processes from START_PRE. */ + service_kill_control_processes(s); + + if (s->type == SERVICE_FORKING) { + s->control_command_id = SERVICE_EXEC_START; + c = s->control_command = s->exec_command[SERVICE_EXEC_START]; + + s->main_command = NULL; + } else { + s->control_command_id = _SERVICE_EXEC_COMMAND_INVALID; + s->control_command = NULL; + + c = s->main_command = s->exec_command[SERVICE_EXEC_START]; + } + + if (!c) { + assert(s->type == SERVICE_ONESHOT); + service_enter_start_post(s); + return; + } + + if (IN_SET(s->type, SERVICE_SIMPLE, SERVICE_IDLE)) + /* For simple + idle this is the main process. We don't apply any timeout here, but + * service_enter_running() will later apply the .runtime_max_usec timeout. */ + timeout = USEC_INFINITY; + else + timeout = s->timeout_start_usec; + + r = service_spawn(s, + c, + timeout, + true, + true, + true, + true, + false, + &pid); + if (r < 0) + goto fail; + + if (IN_SET(s->type, SERVICE_SIMPLE, SERVICE_IDLE)) { + /* For simple services we immediately start + * the START_POST binaries. */ + + service_set_main_pid(s, pid); + service_enter_start_post(s); + + } else if (s->type == SERVICE_FORKING) { + + /* For forking services we wait until the start + * process exited. */ + + s->control_pid = pid; + service_set_state(s, SERVICE_START); + + } else if (IN_SET(s->type, SERVICE_ONESHOT, SERVICE_DBUS, SERVICE_NOTIFY)) { + + /* For oneshot services we wait until the start + * process exited, too, but it is our main process. */ + + /* For D-Bus services we know the main pid right away, + * but wait for the bus name to appear on the + * bus. Notify services are similar. */ + + service_set_main_pid(s, pid); + service_set_state(s, SERVICE_START); + } else + assert_not_reached("Unknown service type"); + + return; + +fail: + log_unit_warning_errno(UNIT(s), r, "Failed to run 'start' task: %m"); + service_enter_signal(s, SERVICE_FINAL_SIGTERM, SERVICE_FAILURE_RESOURCES); +} + +static void service_enter_start_pre(Service *s) { + int r; + + assert(s); + + service_unwatch_control_pid(s); + + s->control_command = s->exec_command[SERVICE_EXEC_START_PRE]; + if (s->control_command) { + /* Before we start anything, let's clear up what might + * be left from previous runs. */ + service_kill_control_processes(s); + + s->control_command_id = SERVICE_EXEC_START_PRE; + + r = service_spawn(s, + s->control_command, + s->timeout_start_usec, + false, + !s->permissions_start_only, + !s->root_directory_start_only, + true, + true, + &s->control_pid); + if (r < 0) + goto fail; + + service_set_state(s, SERVICE_START_PRE); + } else + service_enter_start(s); + + return; + +fail: + log_unit_warning_errno(UNIT(s), r, "Failed to run 'start-pre' task: %m"); + service_enter_dead(s, SERVICE_FAILURE_RESOURCES, true); +} + +static void service_enter_restart(Service *s) { + _cleanup_(sd_bus_error_free) sd_bus_error error = SD_BUS_ERROR_NULL; + int r; + + assert(s); + + if (UNIT(s)->job && UNIT(s)->job->type == JOB_STOP) { + /* Don't restart things if we are going down anyway */ + log_unit_info(UNIT(s), "Stop job pending for unit, delaying automatic restart."); + + r = service_arm_timer(s, usec_add(now(CLOCK_MONOTONIC), s->restart_usec)); + if (r < 0) + goto fail; + + return; + } + + /* Any units that are bound to this service must also be + * restarted. We use JOB_RESTART (instead of the more obvious + * JOB_START) here so that those dependency jobs will be added + * as well. */ + r = manager_add_job(UNIT(s)->manager, JOB_RESTART, UNIT(s), JOB_FAIL, &error, NULL); + if (r < 0) + goto fail; + + /* Note that we stay in the SERVICE_AUTO_RESTART state here, + * it will be canceled as part of the service_stop() call that + * is executed as part of JOB_RESTART. */ + + log_unit_debug(UNIT(s), "Scheduled restart job."); + return; + +fail: + log_unit_warning(UNIT(s), "Failed to schedule restart job: %s", bus_error_message(&error, -r)); + service_enter_dead(s, SERVICE_FAILURE_RESOURCES, false); +} + +static void service_enter_reload_by_notify(Service *s) { + assert(s); + + service_arm_timer(s, usec_add(now(CLOCK_MONOTONIC), s->timeout_start_usec)); + service_set_state(s, SERVICE_RELOAD); +} + +static void service_enter_reload(Service *s) { + int r; + + assert(s); + + service_unwatch_control_pid(s); + s->reload_result = SERVICE_SUCCESS; + + s->control_command = s->exec_command[SERVICE_EXEC_RELOAD]; + if (s->control_command) { + s->control_command_id = SERVICE_EXEC_RELOAD; + + r = service_spawn(s, + s->control_command, + s->timeout_start_usec, + false, + !s->permissions_start_only, + !s->root_directory_start_only, + false, + true, + &s->control_pid); + if (r < 0) + goto fail; + + service_set_state(s, SERVICE_RELOAD); + } else + service_enter_running(s, SERVICE_SUCCESS); + + return; + +fail: + log_unit_warning_errno(UNIT(s), r, "Failed to run 'reload' task: %m"); + s->reload_result = SERVICE_FAILURE_RESOURCES; + service_enter_running(s, SERVICE_SUCCESS); +} + +static void service_run_next_control(Service *s) { + usec_t timeout; + int r; + + assert(s); + assert(s->control_command); + assert(s->control_command->command_next); + + assert(s->control_command_id != SERVICE_EXEC_START); + + s->control_command = s->control_command->command_next; + service_unwatch_control_pid(s); + + if (IN_SET(s->state, SERVICE_START_PRE, SERVICE_START, SERVICE_START_POST, SERVICE_RUNNING, SERVICE_RELOAD)) + timeout = s->timeout_start_usec; + else + timeout = s->timeout_stop_usec; + + r = service_spawn(s, + s->control_command, + timeout, + false, + !s->permissions_start_only, + !s->root_directory_start_only, + s->control_command_id == SERVICE_EXEC_START_PRE || + s->control_command_id == SERVICE_EXEC_STOP_POST, + true, + &s->control_pid); + if (r < 0) + goto fail; + + return; + +fail: + log_unit_warning_errno(UNIT(s), r, "Failed to run next control task: %m"); + + if (s->state == SERVICE_START_PRE) + service_enter_signal(s, SERVICE_FINAL_SIGTERM, SERVICE_FAILURE_RESOURCES); + else if (s->state == SERVICE_STOP) + service_enter_signal(s, SERVICE_STOP_SIGTERM, SERVICE_FAILURE_RESOURCES); + else if (s->state == SERVICE_STOP_POST) + service_enter_dead(s, SERVICE_FAILURE_RESOURCES, true); + else if (s->state == SERVICE_RELOAD) { + s->reload_result = SERVICE_FAILURE_RESOURCES; + service_enter_running(s, SERVICE_SUCCESS); + } else + service_enter_stop(s, SERVICE_FAILURE_RESOURCES); +} + +static void service_run_next_main(Service *s) { + pid_t pid; + int r; + + assert(s); + assert(s->main_command); + assert(s->main_command->command_next); + assert(s->type == SERVICE_ONESHOT); + + s->main_command = s->main_command->command_next; + service_unwatch_main_pid(s); + + r = service_spawn(s, + s->main_command, + s->timeout_start_usec, + true, + true, + true, + true, + false, + &pid); + if (r < 0) + goto fail; + + service_set_main_pid(s, pid); + + return; + +fail: + log_unit_warning_errno(UNIT(s), r, "Failed to run next main task: %m"); + service_enter_stop(s, SERVICE_FAILURE_RESOURCES); +} + +static int service_start(Unit *u) { + Service *s = SERVICE(u); + int r; + + assert(s); + + /* We cannot fulfill this request right now, try again later + * please! */ + if (IN_SET(s->state, + SERVICE_STOP, SERVICE_STOP_SIGABRT, SERVICE_STOP_SIGTERM, SERVICE_STOP_SIGKILL, SERVICE_STOP_POST, + SERVICE_FINAL_SIGTERM, SERVICE_FINAL_SIGKILL)) + return -EAGAIN; + + /* Already on it! */ + if (IN_SET(s->state, SERVICE_START_PRE, SERVICE_START, SERVICE_START_POST)) + return 0; + + /* A service that will be restarted must be stopped first to + * trigger BindsTo and/or OnFailure dependencies. If a user + * does not want to wait for the holdoff time to elapse, the + * service should be manually restarted, not started. We + * simply return EAGAIN here, so that any start jobs stay + * queued, and assume that the auto restart timer will + * eventually trigger the restart. */ + if (s->state == SERVICE_AUTO_RESTART) + return -EAGAIN; + + assert(IN_SET(s->state, SERVICE_DEAD, SERVICE_FAILED)); + + /* Make sure we don't enter a busy loop of some kind. */ + r = unit_start_limit_test(u); + if (r < 0) { + service_enter_dead(s, SERVICE_FAILURE_START_LIMIT_HIT, false); + return r; + } + + s->result = SERVICE_SUCCESS; + s->reload_result = SERVICE_SUCCESS; + s->main_pid_known = false; + s->main_pid_alien = false; + s->forbid_restart = false; + s->reset_cpu_usage = true; + + s->status_text = mfree(s->status_text); + s->status_errno = 0; + + s->notify_state = NOTIFY_UNKNOWN; + + s->watchdog_override_enable = false; + s->watchdog_override_usec = 0; + + service_enter_start_pre(s); + return 1; +} + +static int service_stop(Unit *u) { + Service *s = SERVICE(u); + + assert(s); + + /* Don't create restart jobs from manual stops. */ + s->forbid_restart = true; + + /* Already on it */ + if (IN_SET(s->state, + SERVICE_STOP, SERVICE_STOP_SIGABRT, SERVICE_STOP_SIGTERM, SERVICE_STOP_SIGKILL, SERVICE_STOP_POST, + SERVICE_FINAL_SIGTERM, SERVICE_FINAL_SIGKILL)) + return 0; + + /* A restart will be scheduled or is in progress. */ + if (s->state == SERVICE_AUTO_RESTART) { + service_set_state(s, SERVICE_DEAD); + return 0; + } + + /* If there's already something running we go directly into + * kill mode. */ + if (IN_SET(s->state, SERVICE_START_PRE, SERVICE_START, SERVICE_START_POST, SERVICE_RELOAD)) { + service_enter_signal(s, SERVICE_STOP_SIGTERM, SERVICE_SUCCESS); + return 0; + } + + assert(IN_SET(s->state, SERVICE_RUNNING, SERVICE_EXITED)); + + service_enter_stop(s, SERVICE_SUCCESS); + return 1; +} + +static int service_reload(Unit *u) { + Service *s = SERVICE(u); + + assert(s); + + assert(s->state == SERVICE_RUNNING || s->state == SERVICE_EXITED); + + service_enter_reload(s); + return 1; +} + +_pure_ static bool service_can_reload(Unit *u) { + Service *s = SERVICE(u); + + assert(s); + + return !!s->exec_command[SERVICE_EXEC_RELOAD]; +} + +static int service_serialize(Unit *u, FILE *f, FDSet *fds) { + Service *s = SERVICE(u); + ServiceFDStore *fs; + int r; + + assert(u); + assert(f); + assert(fds); + + unit_serialize_item(u, f, "state", service_state_to_string(s->state)); + unit_serialize_item(u, f, "result", service_result_to_string(s->result)); + unit_serialize_item(u, f, "reload-result", service_result_to_string(s->reload_result)); + + if (s->control_pid > 0) + unit_serialize_item_format(u, f, "control-pid", PID_FMT, s->control_pid); + + if (s->main_pid_known && s->main_pid > 0) + unit_serialize_item_format(u, f, "main-pid", PID_FMT, s->main_pid); + + unit_serialize_item(u, f, "main-pid-known", yes_no(s->main_pid_known)); + unit_serialize_item(u, f, "bus-name-good", yes_no(s->bus_name_good)); + unit_serialize_item(u, f, "bus-name-owner", s->bus_name_owner); + + r = unit_serialize_item_escaped(u, f, "status-text", s->status_text); + if (r < 0) + return r; + + /* FIXME: There's a minor uncleanliness here: if there are + * multiple commands attached here, we will start from the + * first one again */ + if (s->control_command_id >= 0) + unit_serialize_item(u, f, "control-command", service_exec_command_to_string(s->control_command_id)); + + r = unit_serialize_item_fd(u, f, fds, "stdin-fd", s->stdin_fd); + if (r < 0) + return r; + r = unit_serialize_item_fd(u, f, fds, "stdout-fd", s->stdout_fd); + if (r < 0) + return r; + r = unit_serialize_item_fd(u, f, fds, "stderr-fd", s->stderr_fd); + if (r < 0) + return r; + + r = unit_serialize_item_fd(u, f, fds, "socket-fd", s->socket_fd); + if (r < 0) + return r; + + LIST_FOREACH(fd_store, fs, s->fd_store) { + _cleanup_free_ char *c = NULL; + int copy; + + copy = fdset_put_dup(fds, fs->fd); + if (copy < 0) + return copy; + + c = cescape(fs->fdname); + + unit_serialize_item_format(u, f, "fd-store-fd", "%i %s", copy, strempty(c)); + } + + if (s->main_exec_status.pid > 0) { + unit_serialize_item_format(u, f, "main-exec-status-pid", PID_FMT, s->main_exec_status.pid); + dual_timestamp_serialize(f, "main-exec-status-start", &s->main_exec_status.start_timestamp); + dual_timestamp_serialize(f, "main-exec-status-exit", &s->main_exec_status.exit_timestamp); + + if (dual_timestamp_is_set(&s->main_exec_status.exit_timestamp)) { + unit_serialize_item_format(u, f, "main-exec-status-code", "%i", s->main_exec_status.code); + unit_serialize_item_format(u, f, "main-exec-status-status", "%i", s->main_exec_status.status); + } + } + + dual_timestamp_serialize(f, "watchdog-timestamp", &s->watchdog_timestamp); + + unit_serialize_item(u, f, "forbid-restart", yes_no(s->forbid_restart)); + + if (s->watchdog_override_enable) + unit_serialize_item_format(u, f, "watchdog-override-usec", USEC_FMT, s->watchdog_override_usec); + + return 0; +} + +static int service_deserialize_item(Unit *u, const char *key, const char *value, FDSet *fds) { + Service *s = SERVICE(u); + int r; + + assert(u); + assert(key); + assert(value); + assert(fds); + + if (streq(key, "state")) { + ServiceState state; + + state = service_state_from_string(value); + if (state < 0) + log_unit_debug(u, "Failed to parse state value: %s", value); + else + s->deserialized_state = state; + } else if (streq(key, "result")) { + ServiceResult f; + + f = service_result_from_string(value); + if (f < 0) + log_unit_debug(u, "Failed to parse result value: %s", value); + else if (f != SERVICE_SUCCESS) + s->result = f; + + } else if (streq(key, "reload-result")) { + ServiceResult f; + + f = service_result_from_string(value); + if (f < 0) + log_unit_debug(u, "Failed to parse reload result value: %s", value); + else if (f != SERVICE_SUCCESS) + s->reload_result = f; + + } else if (streq(key, "control-pid")) { + pid_t pid; + + if (parse_pid(value, &pid) < 0) + log_unit_debug(u, "Failed to parse control-pid value: %s", value); + else + s->control_pid = pid; + } else if (streq(key, "main-pid")) { + pid_t pid; + + if (parse_pid(value, &pid) < 0) + log_unit_debug(u, "Failed to parse main-pid value: %s", value); + else { + service_set_main_pid(s, pid); + unit_watch_pid(UNIT(s), pid); + } + } else if (streq(key, "main-pid-known")) { + int b; + + b = parse_boolean(value); + if (b < 0) + log_unit_debug(u, "Failed to parse main-pid-known value: %s", value); + else + s->main_pid_known = b; + } else if (streq(key, "bus-name-good")) { + int b; + + b = parse_boolean(value); + if (b < 0) + log_unit_debug(u, "Failed to parse bus-name-good value: %s", value); + else + s->bus_name_good = b; + } else if (streq(key, "bus-name-owner")) { + r = free_and_strdup(&s->bus_name_owner, value); + if (r < 0) + log_unit_error_errno(u, r, "Unable to deserialize current bus owner %s: %m", value); + } else if (streq(key, "status-text")) { + char *t; + + r = cunescape(value, 0, &t); + if (r < 0) + log_unit_debug_errno(u, r, "Failed to unescape status text: %s", value); + else { + free(s->status_text); + s->status_text = t; + } + + } else if (streq(key, "control-command")) { + ServiceExecCommand id; + + id = service_exec_command_from_string(value); + if (id < 0) + log_unit_debug(u, "Failed to parse exec-command value: %s", value); + else { + s->control_command_id = id; + s->control_command = s->exec_command[id]; + } + } else if (streq(key, "socket-fd")) { + int fd; + + if (safe_atoi(value, &fd) < 0 || fd < 0 || !fdset_contains(fds, fd)) + log_unit_debug(u, "Failed to parse socket-fd value: %s", value); + else { + asynchronous_close(s->socket_fd); + s->socket_fd = fdset_remove(fds, fd); + } + } else if (streq(key, "fd-store-fd")) { + const char *fdv; + size_t pf; + int fd; + + pf = strcspn(value, WHITESPACE); + fdv = strndupa(value, pf); + + if (safe_atoi(fdv, &fd) < 0 || fd < 0 || !fdset_contains(fds, fd)) + log_unit_debug(u, "Failed to parse fd-store-fd value: %s", value); + else { + _cleanup_free_ char *t = NULL; + const char *fdn; + + fdn = value + pf; + fdn += strspn(fdn, WHITESPACE); + (void) cunescape(fdn, 0, &t); + + r = service_add_fd_store(s, fd, t); + if (r < 0) + log_unit_error_errno(u, r, "Failed to add fd to store: %m"); + else if (r > 0) + fdset_remove(fds, fd); + } + + } else if (streq(key, "main-exec-status-pid")) { + pid_t pid; + + if (parse_pid(value, &pid) < 0) + log_unit_debug(u, "Failed to parse main-exec-status-pid value: %s", value); + else + s->main_exec_status.pid = pid; + } else if (streq(key, "main-exec-status-code")) { + int i; + + if (safe_atoi(value, &i) < 0) + log_unit_debug(u, "Failed to parse main-exec-status-code value: %s", value); + else + s->main_exec_status.code = i; + } else if (streq(key, "main-exec-status-status")) { + int i; + + if (safe_atoi(value, &i) < 0) + log_unit_debug(u, "Failed to parse main-exec-status-status value: %s", value); + else + s->main_exec_status.status = i; + } else if (streq(key, "main-exec-status-start")) + dual_timestamp_deserialize(value, &s->main_exec_status.start_timestamp); + else if (streq(key, "main-exec-status-exit")) + dual_timestamp_deserialize(value, &s->main_exec_status.exit_timestamp); + else if (streq(key, "watchdog-timestamp")) + dual_timestamp_deserialize(value, &s->watchdog_timestamp); + else if (streq(key, "forbid-restart")) { + int b; + + b = parse_boolean(value); + if (b < 0) + log_unit_debug(u, "Failed to parse forbid-restart value: %s", value); + else + s->forbid_restart = b; + } else if (streq(key, "stdin-fd")) { + int fd; + + if (safe_atoi(value, &fd) < 0 || fd < 0 || !fdset_contains(fds, fd)) + log_unit_debug(u, "Failed to parse stdin-fd value: %s", value); + else { + asynchronous_close(s->stdin_fd); + s->stdin_fd = fdset_remove(fds, fd); + s->exec_context.stdio_as_fds = true; + } + } else if (streq(key, "stdout-fd")) { + int fd; + + if (safe_atoi(value, &fd) < 0 || fd < 0 || !fdset_contains(fds, fd)) + log_unit_debug(u, "Failed to parse stdout-fd value: %s", value); + else { + asynchronous_close(s->stdout_fd); + s->stdout_fd = fdset_remove(fds, fd); + s->exec_context.stdio_as_fds = true; + } + } else if (streq(key, "stderr-fd")) { + int fd; + + if (safe_atoi(value, &fd) < 0 || fd < 0 || !fdset_contains(fds, fd)) + log_unit_debug(u, "Failed to parse stderr-fd value: %s", value); + else { + asynchronous_close(s->stderr_fd); + s->stderr_fd = fdset_remove(fds, fd); + s->exec_context.stdio_as_fds = true; + } + } else if (streq(key, "watchdog-override-usec")) { + usec_t watchdog_override_usec; + if (timestamp_deserialize(value, &watchdog_override_usec) < 0) + log_unit_debug(u, "Failed to parse watchdog_override_usec value: %s", value); + else { + s->watchdog_override_enable = true; + s->watchdog_override_usec = watchdog_override_usec; + } + } else + log_unit_debug(u, "Unknown serialization key: %s", key); + + return 0; +} + +_pure_ static UnitActiveState service_active_state(Unit *u) { + const UnitActiveState *table; + + assert(u); + + table = SERVICE(u)->type == SERVICE_IDLE ? state_translation_table_idle : state_translation_table; + + return table[SERVICE(u)->state]; +} + +static const char *service_sub_state_to_string(Unit *u) { + assert(u); + + return service_state_to_string(SERVICE(u)->state); +} + +static bool service_check_gc(Unit *u) { + Service *s = SERVICE(u); + + assert(s); + + /* Never clean up services that still have a process around, + * even if the service is formally dead. */ + if (cgroup_good(s) > 0 || + main_pid_good(s) > 0 || + control_pid_good(s) > 0) + return true; + + return false; +} + +static int service_retry_pid_file(Service *s) { + int r; + + assert(s->pid_file); + assert(s->state == SERVICE_START || s->state == SERVICE_START_POST); + + r = service_load_pid_file(s, false); + if (r < 0) + return r; + + service_unwatch_pid_file(s); + + service_enter_running(s, SERVICE_SUCCESS); + return 0; +} + +static int service_watch_pid_file(Service *s) { + int r; + + log_unit_debug(UNIT(s), "Setting watch for PID file %s", s->pid_file_pathspec->path); + + r = path_spec_watch(s->pid_file_pathspec, service_dispatch_io); + if (r < 0) + goto fail; + + /* the pidfile might have appeared just before we set the watch */ + log_unit_debug(UNIT(s), "Trying to read PID file %s in case it changed", s->pid_file_pathspec->path); + service_retry_pid_file(s); + + return 0; +fail: + log_unit_error_errno(UNIT(s), r, "Failed to set a watch for PID file %s: %m", s->pid_file_pathspec->path); + service_unwatch_pid_file(s); + return r; +} + +static int service_demand_pid_file(Service *s) { + PathSpec *ps; + + assert(s->pid_file); + assert(!s->pid_file_pathspec); + + ps = new0(PathSpec, 1); + if (!ps) + return -ENOMEM; + + ps->unit = UNIT(s); + ps->path = strdup(s->pid_file); + if (!ps->path) { + free(ps); + return -ENOMEM; + } + + path_kill_slashes(ps->path); + + /* PATH_CHANGED would not be enough. There are daemons (sendmail) that + * keep their PID file open all the time. */ + ps->type = PATH_MODIFIED; + ps->inotify_fd = -1; + + s->pid_file_pathspec = ps; + + return service_watch_pid_file(s); +} + +static int service_dispatch_io(sd_event_source *source, int fd, uint32_t events, void *userdata) { + PathSpec *p = userdata; + Service *s; + + assert(p); + + s = SERVICE(p->unit); + + assert(s); + assert(fd >= 0); + assert(s->state == SERVICE_START || s->state == SERVICE_START_POST); + assert(s->pid_file_pathspec); + assert(path_spec_owns_inotify_fd(s->pid_file_pathspec, fd)); + + log_unit_debug(UNIT(s), "inotify event"); + + if (path_spec_fd_event(p, events) < 0) + goto fail; + + if (service_retry_pid_file(s) == 0) + return 0; + + if (service_watch_pid_file(s) < 0) + goto fail; + + return 0; + +fail: + service_unwatch_pid_file(s); + service_enter_signal(s, SERVICE_STOP_SIGTERM, SERVICE_FAILURE_RESOURCES); + return 0; +} + +static void service_notify_cgroup_empty_event(Unit *u) { + Service *s = SERVICE(u); + + assert(u); + + log_unit_debug(u, "cgroup is empty"); + + switch (s->state) { + + /* Waiting for SIGCHLD is usually more interesting, + * because it includes return codes/signals. Which is + * why we ignore the cgroup events for most cases, + * except when we don't know pid which to expect the + * SIGCHLD for. */ + + case SERVICE_START: + case SERVICE_START_POST: + /* If we were hoping for the daemon to write its PID file, + * we can give up now. */ + if (s->pid_file_pathspec) { + log_unit_warning(u, "Daemon never wrote its PID file. Failing."); + + service_unwatch_pid_file(s); + if (s->state == SERVICE_START) + service_enter_signal(s, SERVICE_FINAL_SIGTERM, SERVICE_FAILURE_RESOURCES); + else + service_enter_stop(s, SERVICE_FAILURE_RESOURCES); + } + break; + + case SERVICE_RUNNING: + /* service_enter_running() will figure out what to do */ + service_enter_running(s, SERVICE_SUCCESS); + break; + + case SERVICE_STOP_SIGABRT: + case SERVICE_STOP_SIGTERM: + case SERVICE_STOP_SIGKILL: + + if (main_pid_good(s) <= 0 && !control_pid_good(s)) + service_enter_stop_post(s, SERVICE_SUCCESS); + + break; + + case SERVICE_STOP_POST: + case SERVICE_FINAL_SIGTERM: + case SERVICE_FINAL_SIGKILL: + if (main_pid_good(s) <= 0 && !control_pid_good(s)) + service_enter_dead(s, SERVICE_SUCCESS, true); + + break; + + default: + ; + } +} + +static void service_sigchld_event(Unit *u, pid_t pid, int code, int status) { + Service *s = SERVICE(u); + ServiceResult f; + + assert(s); + assert(pid >= 0); + + if (UNIT(s)->fragment_path ? is_clean_exit(code, status, &s->success_status) : + is_clean_exit_lsb(code, status, &s->success_status)) + f = SERVICE_SUCCESS; + else if (code == CLD_EXITED) + f = SERVICE_FAILURE_EXIT_CODE; + else if (code == CLD_KILLED) + f = SERVICE_FAILURE_SIGNAL; + else if (code == CLD_DUMPED) + f = SERVICE_FAILURE_CORE_DUMP; + else + assert_not_reached("Unknown code"); + + if (s->main_pid == pid) { + /* Forking services may occasionally move to a new PID. + * As long as they update the PID file before exiting the old + * PID, they're fine. */ + if (service_load_pid_file(s, false) == 0) + return; + + s->main_pid = 0; + exec_status_exit(&s->main_exec_status, &s->exec_context, pid, code, status); + + if (s->main_command) { + /* If this is not a forking service than the + * main process got started and hence we copy + * the exit status so that it is recorded both + * as main and as control process exit + * status */ + + s->main_command->exec_status = s->main_exec_status; + + if (s->main_command->ignore) + f = SERVICE_SUCCESS; + } else if (s->exec_command[SERVICE_EXEC_START]) { + + /* If this is a forked process, then we should + * ignore the return value if this was + * configured for the starter process */ + + if (s->exec_command[SERVICE_EXEC_START]->ignore) + f = SERVICE_SUCCESS; + } + + log_struct(f == SERVICE_SUCCESS ? LOG_DEBUG : LOG_NOTICE, + LOG_UNIT_ID(u), + LOG_UNIT_MESSAGE(u, "Main process exited, code=%s, status=%i/%s", + sigchld_code_to_string(code), status, + strna(code == CLD_EXITED + ? exit_status_to_string(status, EXIT_STATUS_FULL) + : signal_to_string(status))), + "EXIT_CODE=%s", sigchld_code_to_string(code), + "EXIT_STATUS=%i", status, + NULL); + + if (f != SERVICE_SUCCESS) + s->result = f; + + if (s->main_command && + s->main_command->command_next && + f == SERVICE_SUCCESS) { + + /* There is another command to * + * execute, so let's do that. */ + + log_unit_debug(u, "Running next main command for state %s.", service_state_to_string(s->state)); + service_run_next_main(s); + + } else { + + /* The service exited, so the service is officially + * gone. */ + s->main_command = NULL; + + switch (s->state) { + + case SERVICE_START_POST: + case SERVICE_RELOAD: + case SERVICE_STOP: + /* Need to wait until the operation is + * done */ + break; + + case SERVICE_START: + if (s->type == SERVICE_ONESHOT) { + /* This was our main goal, so let's go on */ + if (f == SERVICE_SUCCESS) + service_enter_start_post(s); + else + service_enter_signal(s, SERVICE_FINAL_SIGTERM, f); + break; + } + + /* Fall through */ + + case SERVICE_RUNNING: + service_enter_running(s, f); + break; + + case SERVICE_STOP_SIGABRT: + case SERVICE_STOP_SIGTERM: + case SERVICE_STOP_SIGKILL: + + if (!control_pid_good(s)) + service_enter_stop_post(s, f); + + /* If there is still a control process, wait for that first */ + break; + + case SERVICE_STOP_POST: + case SERVICE_FINAL_SIGTERM: + case SERVICE_FINAL_SIGKILL: + + if (!control_pid_good(s)) + service_enter_dead(s, f, true); + break; + + default: + assert_not_reached("Uh, main process died at wrong time."); + } + } + + } else if (s->control_pid == pid) { + s->control_pid = 0; + + if (s->control_command) { + exec_status_exit(&s->control_command->exec_status, &s->exec_context, pid, code, status); + + if (s->control_command->ignore) + f = SERVICE_SUCCESS; + } + + log_unit_full(u, f == SERVICE_SUCCESS ? LOG_DEBUG : LOG_NOTICE, 0, + "Control process exited, code=%s status=%i", + sigchld_code_to_string(code), status); + + if (f != SERVICE_SUCCESS) + s->result = f; + + /* Immediately get rid of the cgroup, so that the + * kernel doesn't delay the cgroup empty messages for + * the service cgroup any longer than necessary */ + service_kill_control_processes(s); + + if (s->control_command && + s->control_command->command_next && + f == SERVICE_SUCCESS) { + + /* There is another command to * + * execute, so let's do that. */ + + log_unit_debug(u, "Running next control command for state %s.", service_state_to_string(s->state)); + service_run_next_control(s); + + } else { + /* No further commands for this step, so let's + * figure out what to do next */ + + s->control_command = NULL; + s->control_command_id = _SERVICE_EXEC_COMMAND_INVALID; + + log_unit_debug(u, "Got final SIGCHLD for state %s.", service_state_to_string(s->state)); + + switch (s->state) { + + case SERVICE_START_PRE: + if (f == SERVICE_SUCCESS) + service_enter_start(s); + else + service_enter_signal(s, SERVICE_FINAL_SIGTERM, f); + break; + + case SERVICE_START: + if (s->type != SERVICE_FORKING) + /* Maybe spurious event due to a reload that changed the type? */ + break; + + if (f != SERVICE_SUCCESS) { + service_enter_signal(s, SERVICE_FINAL_SIGTERM, f); + break; + } + + if (s->pid_file) { + bool has_start_post; + int r; + + /* Let's try to load the pid file here if we can. + * The PID file might actually be created by a START_POST + * script. In that case don't worry if the loading fails. */ + + has_start_post = !!s->exec_command[SERVICE_EXEC_START_POST]; + r = service_load_pid_file(s, !has_start_post); + if (!has_start_post && r < 0) { + r = service_demand_pid_file(s); + if (r < 0 || !cgroup_good(s)) + service_enter_signal(s, SERVICE_FINAL_SIGTERM, SERVICE_FAILURE_RESOURCES); + break; + } + } else + service_search_main_pid(s); + + service_enter_start_post(s); + break; + + case SERVICE_START_POST: + if (f != SERVICE_SUCCESS) { + service_enter_signal(s, SERVICE_STOP_SIGTERM, f); + break; + } + + if (s->pid_file) { + int r; + + r = service_load_pid_file(s, true); + if (r < 0) { + r = service_demand_pid_file(s); + if (r < 0 || !cgroup_good(s)) + service_enter_stop(s, SERVICE_FAILURE_RESOURCES); + break; + } + } else + service_search_main_pid(s); + + service_enter_running(s, SERVICE_SUCCESS); + break; + + case SERVICE_RELOAD: + if (f == SERVICE_SUCCESS) + if (service_load_pid_file(s, true) < 0) + service_search_main_pid(s); + + s->reload_result = f; + service_enter_running(s, SERVICE_SUCCESS); + break; + + case SERVICE_STOP: + service_enter_signal(s, SERVICE_STOP_SIGTERM, f); + break; + + case SERVICE_STOP_SIGABRT: + case SERVICE_STOP_SIGTERM: + case SERVICE_STOP_SIGKILL: + if (main_pid_good(s) <= 0) + service_enter_stop_post(s, f); + + /* If there is still a service + * process around, wait until + * that one quit, too */ + break; + + case SERVICE_STOP_POST: + case SERVICE_FINAL_SIGTERM: + case SERVICE_FINAL_SIGKILL: + if (main_pid_good(s) <= 0) + service_enter_dead(s, f, true); + break; + + default: + assert_not_reached("Uh, control process died at wrong time."); + } + } + } + + /* Notify clients about changed exit status */ + unit_add_to_dbus_queue(u); + + /* We got one SIGCHLD for the service, let's watch all + * processes that are now running of the service, and watch + * that. Among the PIDs we then watch will be children + * reassigned to us, which hopefully allows us to identify + * when all children are gone */ + unit_tidy_watch_pids(u, s->main_pid, s->control_pid); + unit_watch_all_pids(u); + + /* If the PID set is empty now, then let's finish this off + (On unified we use proper notifications) */ + if (cg_unified() <= 0 && set_isempty(u->pids)) + service_notify_cgroup_empty_event(u); +} + +static int service_dispatch_timer(sd_event_source *source, usec_t usec, void *userdata) { + Service *s = SERVICE(userdata); + + assert(s); + assert(source == s->timer_event_source); + + switch (s->state) { + + case SERVICE_START_PRE: + case SERVICE_START: + log_unit_warning(UNIT(s), "%s operation timed out. Terminating.", s->state == SERVICE_START ? "Start" : "Start-pre"); + service_enter_signal(s, SERVICE_FINAL_SIGTERM, SERVICE_FAILURE_TIMEOUT); + break; + + case SERVICE_START_POST: + log_unit_warning(UNIT(s), "Start-post operation timed out. Stopping."); + service_enter_signal(s, SERVICE_STOP_SIGTERM, SERVICE_FAILURE_TIMEOUT); + break; + + case SERVICE_RUNNING: + log_unit_warning(UNIT(s), "Service reached runtime time limit. Stopping."); + service_enter_stop(s, SERVICE_FAILURE_TIMEOUT); + break; + + case SERVICE_RELOAD: + log_unit_warning(UNIT(s), "Reload operation timed out. Killing reload process."); + service_kill_control_processes(s); + s->reload_result = SERVICE_FAILURE_TIMEOUT; + service_enter_running(s, SERVICE_SUCCESS); + break; + + case SERVICE_STOP: + log_unit_warning(UNIT(s), "Stopping timed out. Terminating."); + service_enter_signal(s, SERVICE_STOP_SIGTERM, SERVICE_FAILURE_TIMEOUT); + break; + + case SERVICE_STOP_SIGABRT: + log_unit_warning(UNIT(s), "State 'stop-sigabrt' timed out. Terminating."); + service_enter_signal(s, SERVICE_STOP_SIGTERM, SERVICE_FAILURE_TIMEOUT); + break; + + case SERVICE_STOP_SIGTERM: + if (s->kill_context.send_sigkill) { + log_unit_warning(UNIT(s), "State 'stop-sigterm' timed out. Killing."); + service_enter_signal(s, SERVICE_STOP_SIGKILL, SERVICE_FAILURE_TIMEOUT); + } else { + log_unit_warning(UNIT(s), "State 'stop-sigterm' timed out. Skipping SIGKILL."); + service_enter_stop_post(s, SERVICE_FAILURE_TIMEOUT); + } + + break; + + case SERVICE_STOP_SIGKILL: + /* Uh, we sent a SIGKILL and it is still not gone? + * Must be something we cannot kill, so let's just be + * weirded out and continue */ + + log_unit_warning(UNIT(s), "Processes still around after SIGKILL. Ignoring."); + service_enter_stop_post(s, SERVICE_FAILURE_TIMEOUT); + break; + + case SERVICE_STOP_POST: + log_unit_warning(UNIT(s), "State 'stop-post' timed out. Terminating."); + service_enter_signal(s, SERVICE_FINAL_SIGTERM, SERVICE_FAILURE_TIMEOUT); + break; + + case SERVICE_FINAL_SIGTERM: + if (s->kill_context.send_sigkill) { + log_unit_warning(UNIT(s), "State 'stop-final-sigterm' timed out. Killing."); + service_enter_signal(s, SERVICE_FINAL_SIGKILL, SERVICE_FAILURE_TIMEOUT); + } else { + log_unit_warning(UNIT(s), "State 'stop-final-sigterm' timed out. Skipping SIGKILL. Entering failed mode."); + service_enter_dead(s, SERVICE_FAILURE_TIMEOUT, false); + } + + break; + + case SERVICE_FINAL_SIGKILL: + log_unit_warning(UNIT(s), "Processes still around after final SIGKILL. Entering failed mode."); + service_enter_dead(s, SERVICE_FAILURE_TIMEOUT, true); + break; + + case SERVICE_AUTO_RESTART: + log_unit_info(UNIT(s), + s->restart_usec > 0 ? + "Service hold-off time over, scheduling restart." : + "Service has no hold-off time, scheduling restart."); + service_enter_restart(s); + break; + + default: + assert_not_reached("Timeout at wrong time."); + } + + return 0; +} + +static int service_dispatch_watchdog(sd_event_source *source, usec_t usec, void *userdata) { + Service *s = SERVICE(userdata); + char t[FORMAT_TIMESPAN_MAX]; + usec_t watchdog_usec; + + assert(s); + assert(source == s->watchdog_event_source); + + watchdog_usec = service_get_watchdog_usec(s); + + log_unit_error(UNIT(s), "Watchdog timeout (limit %s)!", + format_timespan(t, sizeof(t), watchdog_usec, 1)); + + service_enter_signal(s, SERVICE_STOP_SIGABRT, SERVICE_FAILURE_WATCHDOG); + + return 0; +} + +static void service_notify_message(Unit *u, pid_t pid, char **tags, FDSet *fds) { + Service *s = SERVICE(u); + _cleanup_free_ char *cc = NULL; + bool notify_dbus = false; + const char *e; + + assert(u); + + cc = strv_join(tags, ", "); + + if (s->notify_access == NOTIFY_NONE) { + log_unit_warning(u, "Got notification message from PID "PID_FMT", but reception is disabled.", pid); + return; + } else if (s->notify_access == NOTIFY_MAIN && pid != s->main_pid) { + if (s->main_pid != 0) + log_unit_warning(u, "Got notification message from PID "PID_FMT", but reception only permitted for main PID "PID_FMT, pid, s->main_pid); + else + log_unit_debug(u, "Got notification message from PID "PID_FMT", but reception only permitted for main PID which is currently not known", pid); + return; + } else + log_unit_debug(u, "Got notification message from PID "PID_FMT" (%s)", pid, isempty(cc) ? "n/a" : cc); + + /* Interpret MAINPID= */ + e = strv_find_startswith(tags, "MAINPID="); + if (e && IN_SET(s->state, SERVICE_START, SERVICE_START_POST, SERVICE_RUNNING, SERVICE_RELOAD)) { + if (parse_pid(e, &pid) < 0) + log_unit_warning(u, "Failed to parse MAINPID= field in notification message: %s", e); + else { + service_set_main_pid(s, pid); + unit_watch_pid(UNIT(s), pid); + notify_dbus = true; + } + } + + /* Interpret RELOADING= */ + if (strv_find(tags, "RELOADING=1")) { + + s->notify_state = NOTIFY_RELOADING; + + if (s->state == SERVICE_RUNNING) + service_enter_reload_by_notify(s); + + notify_dbus = true; + } + + /* Interpret READY= */ + if (strv_find(tags, "READY=1")) { + + s->notify_state = NOTIFY_READY; + + /* Type=notify services inform us about completed + * initialization with READY=1 */ + if (s->type == SERVICE_NOTIFY && s->state == SERVICE_START) + service_enter_start_post(s); + + /* Sending READY=1 while we are reloading informs us + * that the reloading is complete */ + if (s->state == SERVICE_RELOAD && s->control_pid == 0) + service_enter_running(s, SERVICE_SUCCESS); + + notify_dbus = true; + } + + /* Interpret STOPPING= */ + if (strv_find(tags, "STOPPING=1")) { + + s->notify_state = NOTIFY_STOPPING; + + if (s->state == SERVICE_RUNNING) + service_enter_stop_by_notify(s); + + notify_dbus = true; + } + + /* Interpret STATUS= */ + e = strv_find_startswith(tags, "STATUS="); + if (e) { + _cleanup_free_ char *t = NULL; + + if (!isempty(e)) { + if (!utf8_is_valid(e)) + log_unit_warning(u, "Status message in notification message is not UTF-8 clean."); + else { + t = strdup(e); + if (!t) + log_oom(); + } + } + + if (!streq_ptr(s->status_text, t)) { + + free(s->status_text); + s->status_text = t; + t = NULL; + + notify_dbus = true; + } + } + + /* Interpret ERRNO= */ + e = strv_find_startswith(tags, "ERRNO="); + if (e) { + int status_errno; + + if (safe_atoi(e, &status_errno) < 0 || status_errno < 0) + log_unit_warning(u, "Failed to parse ERRNO= field in notification message: %s", e); + else { + if (s->status_errno != status_errno) { + s->status_errno = status_errno; + notify_dbus = true; + } + } + } + + /* Interpret WATCHDOG= */ + if (strv_find(tags, "WATCHDOG=1")) + service_reset_watchdog(s); + + if (strv_find(tags, "FDSTORE=1")) { + const char *name; + + name = strv_find_startswith(tags, "FDNAME="); + if (name && !fdname_is_valid(name)) { + log_unit_warning(u, "Passed FDNAME= name is invalid, ignoring."); + name = NULL; + } + + service_add_fd_store_set(s, fds, name); + } + + e = strv_find_startswith(tags, "WATCHDOG_USEC="); + if (e) { + usec_t watchdog_override_usec; + if (safe_atou64(e, &watchdog_override_usec) < 0) + log_unit_warning(u, "Failed to parse WATCHDOG_USEC=%s", e); + else + service_reset_watchdog_timeout(s, watchdog_override_usec); + } + + /* Notify clients about changed status or main pid */ + if (notify_dbus) + unit_add_to_dbus_queue(u); +} + +static int service_get_timeout(Unit *u, usec_t *timeout) { + Service *s = SERVICE(u); + uint64_t t; + int r; + + if (!s->timer_event_source) + return 0; + + r = sd_event_source_get_time(s->timer_event_source, &t); + if (r < 0) + return r; + if (t == USEC_INFINITY) + return 0; + + *timeout = t; + return 1; +} + +static void service_bus_name_owner_change( + Unit *u, + const char *name, + const char *old_owner, + const char *new_owner) { + + Service *s = SERVICE(u); + int r; + + assert(s); + assert(name); + + assert(streq(s->bus_name, name)); + assert(old_owner || new_owner); + + if (old_owner && new_owner) + log_unit_debug(u, "D-Bus name %s changed owner from %s to %s", name, old_owner, new_owner); + else if (old_owner) + log_unit_debug(u, "D-Bus name %s no longer registered by %s", name, old_owner); + else + log_unit_debug(u, "D-Bus name %s now registered by %s", name, new_owner); + + s->bus_name_good = !!new_owner; + + /* Track the current owner, so we can reconstruct changes after a daemon reload */ + r = free_and_strdup(&s->bus_name_owner, new_owner); + if (r < 0) { + log_unit_error_errno(u, r, "Unable to set new bus name owner %s: %m", new_owner); + return; + } + + if (s->type == SERVICE_DBUS) { + + /* service_enter_running() will figure out what to + * do */ + if (s->state == SERVICE_RUNNING) + service_enter_running(s, SERVICE_SUCCESS); + else if (s->state == SERVICE_START && new_owner) + service_enter_start_post(s); + + } else if (new_owner && + s->main_pid <= 0 && + (s->state == SERVICE_START || + s->state == SERVICE_START_POST || + s->state == SERVICE_RUNNING || + s->state == SERVICE_RELOAD)) { + + _cleanup_(sd_bus_creds_unrefp) sd_bus_creds *creds = NULL; + pid_t pid; + + /* Try to acquire PID from bus service */ + + r = sd_bus_get_name_creds(u->manager->api_bus, name, SD_BUS_CREDS_PID, &creds); + if (r >= 0) + r = sd_bus_creds_get_pid(creds, &pid); + if (r >= 0) { + log_unit_debug(u, "D-Bus name %s is now owned by process %u", name, (unsigned) pid); + + service_set_main_pid(s, pid); + unit_watch_pid(UNIT(s), pid); + } + } +} + +int service_set_socket_fd(Service *s, int fd, Socket *sock, bool selinux_context_net) { + _cleanup_free_ char *peer = NULL; + int r; + + assert(s); + assert(fd >= 0); + + /* This is called by the socket code when instantiating a new service for a stream socket and the socket needs + * to be configured. We take ownership of the passed fd on success. */ + + if (UNIT(s)->load_state != UNIT_LOADED) + return -EINVAL; + + if (s->socket_fd >= 0) + return -EBUSY; + + if (s->state != SERVICE_DEAD) + return -EAGAIN; + + if (getpeername_pretty(fd, true, &peer) >= 0) { + + if (UNIT(s)->description) { + _cleanup_free_ char *a; + + a = strjoin(UNIT(s)->description, " (", peer, ")", NULL); + if (!a) + return -ENOMEM; + + r = unit_set_description(UNIT(s), a); + } else + r = unit_set_description(UNIT(s), peer); + + if (r < 0) + return r; + } + + r = unit_add_two_dependencies(UNIT(sock), UNIT_BEFORE, UNIT_TRIGGERS, UNIT(s), false); + if (r < 0) + return r; + + s->socket_fd = fd; + s->socket_fd_selinux_context_net = selinux_context_net; + + unit_ref_set(&s->accept_socket, UNIT(sock)); + return 0; +} + +static void service_reset_failed(Unit *u) { + Service *s = SERVICE(u); + + assert(s); + + if (s->state == SERVICE_FAILED) + service_set_state(s, SERVICE_DEAD); + + s->result = SERVICE_SUCCESS; + s->reload_result = SERVICE_SUCCESS; +} + +static int service_kill(Unit *u, KillWho who, int signo, sd_bus_error *error) { + Service *s = SERVICE(u); + + return unit_kill_common(u, who, signo, s->main_pid, s->control_pid, error); +} + +static int service_main_pid(Unit *u) { + Service *s = SERVICE(u); + + assert(s); + + return s->main_pid; +} + +static int service_control_pid(Unit *u) { + Service *s = SERVICE(u); + + assert(s); + + return s->control_pid; +} + +static const char* const service_restart_table[_SERVICE_RESTART_MAX] = { + [SERVICE_RESTART_NO] = "no", + [SERVICE_RESTART_ON_SUCCESS] = "on-success", + [SERVICE_RESTART_ON_FAILURE] = "on-failure", + [SERVICE_RESTART_ON_ABNORMAL] = "on-abnormal", + [SERVICE_RESTART_ON_WATCHDOG] = "on-watchdog", + [SERVICE_RESTART_ON_ABORT] = "on-abort", + [SERVICE_RESTART_ALWAYS] = "always", +}; + +DEFINE_STRING_TABLE_LOOKUP(service_restart, ServiceRestart); + +static const char* const service_type_table[_SERVICE_TYPE_MAX] = { + [SERVICE_SIMPLE] = "simple", + [SERVICE_FORKING] = "forking", + [SERVICE_ONESHOT] = "oneshot", + [SERVICE_DBUS] = "dbus", + [SERVICE_NOTIFY] = "notify", + [SERVICE_IDLE] = "idle" +}; + +DEFINE_STRING_TABLE_LOOKUP(service_type, ServiceType); + +static const char* const service_exec_command_table[_SERVICE_EXEC_COMMAND_MAX] = { + [SERVICE_EXEC_START_PRE] = "ExecStartPre", + [SERVICE_EXEC_START] = "ExecStart", + [SERVICE_EXEC_START_POST] = "ExecStartPost", + [SERVICE_EXEC_RELOAD] = "ExecReload", + [SERVICE_EXEC_STOP] = "ExecStop", + [SERVICE_EXEC_STOP_POST] = "ExecStopPost", +}; + +DEFINE_STRING_TABLE_LOOKUP(service_exec_command, ServiceExecCommand); + +static const char* const notify_access_table[_NOTIFY_ACCESS_MAX] = { + [NOTIFY_NONE] = "none", + [NOTIFY_MAIN] = "main", + [NOTIFY_ALL] = "all" +}; + +DEFINE_STRING_TABLE_LOOKUP(notify_access, NotifyAccess); + +static const char* const notify_state_table[_NOTIFY_STATE_MAX] = { + [NOTIFY_UNKNOWN] = "unknown", + [NOTIFY_READY] = "ready", + [NOTIFY_RELOADING] = "reloading", + [NOTIFY_STOPPING] = "stopping", +}; + +DEFINE_STRING_TABLE_LOOKUP(notify_state, NotifyState); + +static const char* const service_result_table[_SERVICE_RESULT_MAX] = { + [SERVICE_SUCCESS] = "success", + [SERVICE_FAILURE_RESOURCES] = "resources", + [SERVICE_FAILURE_TIMEOUT] = "timeout", + [SERVICE_FAILURE_EXIT_CODE] = "exit-code", + [SERVICE_FAILURE_SIGNAL] = "signal", + [SERVICE_FAILURE_CORE_DUMP] = "core-dump", + [SERVICE_FAILURE_WATCHDOG] = "watchdog", + [SERVICE_FAILURE_START_LIMIT_HIT] = "start-limit-hit", +}; + +DEFINE_STRING_TABLE_LOOKUP(service_result, ServiceResult); + +const UnitVTable service_vtable = { + .object_size = sizeof(Service), + .exec_context_offset = offsetof(Service, exec_context), + .cgroup_context_offset = offsetof(Service, cgroup_context), + .kill_context_offset = offsetof(Service, kill_context), + .exec_runtime_offset = offsetof(Service, exec_runtime), + + .sections = + "Unit\0" + "Service\0" + "Install\0", + .private_section = "Service", + + .init = service_init, + .done = service_done, + .load = service_load, + .release_resources = service_release_resources, + + .coldplug = service_coldplug, + + .dump = service_dump, + + .start = service_start, + .stop = service_stop, + .reload = service_reload, + + .can_reload = service_can_reload, + + .kill = service_kill, + + .serialize = service_serialize, + .deserialize_item = service_deserialize_item, + + .active_state = service_active_state, + .sub_state_to_string = service_sub_state_to_string, + + .check_gc = service_check_gc, + + .sigchld_event = service_sigchld_event, + + .reset_failed = service_reset_failed, + + .notify_cgroup_empty = service_notify_cgroup_empty_event, + .notify_message = service_notify_message, + + .main_pid = service_main_pid, + .control_pid = service_control_pid, + + .bus_name_owner_change = service_bus_name_owner_change, + + .bus_vtable = bus_service_vtable, + .bus_set_property = bus_service_set_property, + .bus_commit_properties = bus_service_commit_properties, + + .get_timeout = service_get_timeout, + .can_transient = true, + + .status_message_formats = { + .starting_stopping = { + [0] = "Starting %s...", + [1] = "Stopping %s...", + }, + .finished_start_job = { + [JOB_DONE] = "Started %s.", + [JOB_FAILED] = "Failed to start %s.", + }, + .finished_stop_job = { + [JOB_DONE] = "Stopped %s.", + [JOB_FAILED] = "Stopped (with error) %s.", + }, + }, +}; diff --git a/src/grp-system/libcore/src/show-status.c b/src/grp-system/libcore/src/show-status.c new file mode 100644 index 0000000000..dec0eb00be --- /dev/null +++ b/src/grp-system/libcore/src/show-status.c @@ -0,0 +1,124 @@ +/*** + This file is part of systemd. + + Copyright 2014 Lennart Poettering + + systemd is free software; you can redistribute it and/or modify it + under the terms of the GNU Lesser General Public License as published by + the Free Software Foundation; either version 2.1 of the License, or + (at your option) any later version. + + systemd is distributed in the hope that it will be useful, but + WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + Lesser General Public License for more details. + + You should have received a copy of the GNU Lesser General Public License + along with systemd; If not, see . +***/ + +#include "core/show-status.h" +#include "systemd-basic/alloc-util.h" +#include "systemd-basic/fd-util.h" +#include "systemd-basic/io-util.h" +#include "systemd-basic/parse-util.h" +#include "systemd-basic/string-util.h" +#include "systemd-basic/terminal-util.h" +#include "systemd-basic/util.h" + +int parse_show_status(const char *v, ShowStatus *ret) { + int r; + + assert(v); + assert(ret); + + if (streq(v, "auto")) { + *ret = SHOW_STATUS_AUTO; + return 0; + } + + r = parse_boolean(v); + if (r < 0) + return r; + + *ret = r ? SHOW_STATUS_YES : SHOW_STATUS_NO; + return 0; +} + +int status_vprintf(const char *status, bool ellipse, bool ephemeral, const char *format, va_list ap) { + static const char status_indent[] = " "; /* "[" STATUS "] " */ + _cleanup_free_ char *s = NULL; + _cleanup_close_ int fd = -1; + struct iovec iovec[6] = {}; + int n = 0; + static bool prev_ephemeral; + + assert(format); + + /* This is independent of logging, as status messages are + * optional and go exclusively to the console. */ + + if (vasprintf(&s, format, ap) < 0) + return log_oom(); + + fd = open_terminal("/dev/console", O_WRONLY|O_NOCTTY|O_CLOEXEC); + if (fd < 0) + return fd; + + if (ellipse) { + char *e; + size_t emax, sl; + int c; + + c = fd_columns(fd); + if (c <= 0) + c = 80; + + sl = status ? sizeof(status_indent)-1 : 0; + + emax = c - sl - 1; + if (emax < 3) + emax = 3; + + e = ellipsize(s, emax, 50); + if (e) { + free(s); + s = e; + } + } + + if (prev_ephemeral) + IOVEC_SET_STRING(iovec[n++], "\r" ANSI_ERASE_TO_END_OF_LINE); + prev_ephemeral = ephemeral; + + if (status) { + if (!isempty(status)) { + IOVEC_SET_STRING(iovec[n++], "["); + IOVEC_SET_STRING(iovec[n++], status); + IOVEC_SET_STRING(iovec[n++], "] "); + } else + IOVEC_SET_STRING(iovec[n++], status_indent); + } + + IOVEC_SET_STRING(iovec[n++], s); + if (!ephemeral) + IOVEC_SET_STRING(iovec[n++], "\n"); + + if (writev(fd, iovec, n) < 0) + return -errno; + + return 0; +} + +int status_printf(const char *status, bool ellipse, bool ephemeral, const char *format, ...) { + va_list ap; + int r; + + assert(format); + + va_start(ap, format); + r = status_vprintf(status, ellipse, ephemeral, format, ap); + va_end(ap); + + return r; +} diff --git a/src/grp-system/libcore/src/slice.c b/src/grp-system/libcore/src/slice.c new file mode 100644 index 0000000000..4ab5548572 --- /dev/null +++ b/src/grp-system/libcore/src/slice.c @@ -0,0 +1,347 @@ +/*** + This file is part of systemd. + + Copyright 2013 Lennart Poettering + + systemd is free software; you can redistribute it and/or modify it + under the terms of the GNU Lesser General Public License as published by + the Free Software Foundation; either version 2.1 of the License, or + (at your option) any later version. + + systemd is distributed in the hope that it will be useful, but + WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + Lesser General Public License for more details. + + You should have received a copy of the GNU Lesser General Public License + along with systemd; If not, see . +***/ + +#include + +#include "core/slice.h" +#include "core/unit.h" +#include "systemd-basic/alloc-util.h" +#include "systemd-basic/log.h" +#include "systemd-basic/special.h" +#include "systemd-basic/string-util.h" +#include "systemd-basic/strv.h" +#include "systemd-basic/unit-name.h" + +#include "dbus-slice.h" + +static const UnitActiveState state_translation_table[_SLICE_STATE_MAX] = { + [SLICE_DEAD] = UNIT_INACTIVE, + [SLICE_ACTIVE] = UNIT_ACTIVE +}; + +static void slice_init(Unit *u) { + assert(u); + assert(u->load_state == UNIT_STUB); + + u->ignore_on_isolate = true; +} + +static void slice_set_state(Slice *t, SliceState state) { + SliceState old_state; + assert(t); + + old_state = t->state; + t->state = state; + + if (state != old_state) + log_debug("%s changed %s -> %s", + UNIT(t)->id, + slice_state_to_string(old_state), + slice_state_to_string(state)); + + unit_notify(UNIT(t), state_translation_table[old_state], state_translation_table[state], true); +} + +static int slice_add_parent_slice(Slice *s) { + char *a, *dash; + Unit *parent; + int r; + + assert(s); + + if (UNIT_ISSET(UNIT(s)->slice)) + return 0; + + if (unit_has_name(UNIT(s), SPECIAL_ROOT_SLICE)) + return 0; + + a = strdupa(UNIT(s)->id); + dash = strrchr(a, '-'); + if (dash) + strcpy(dash, ".slice"); + else + a = (char*) SPECIAL_ROOT_SLICE; + + r = manager_load_unit(UNIT(s)->manager, a, NULL, NULL, &parent); + if (r < 0) + return r; + + unit_ref_set(&UNIT(s)->slice, parent); + return 0; +} + +static int slice_add_default_dependencies(Slice *s) { + int r; + + assert(s); + + if (!UNIT(s)->default_dependencies) + return 0; + + /* Make sure slices are unloaded on shutdown */ + r = unit_add_two_dependencies_by_name( + UNIT(s), + UNIT_BEFORE, UNIT_CONFLICTS, + SPECIAL_SHUTDOWN_TARGET, NULL, true); + if (r < 0) + return r; + + return 0; +} + +static int slice_verify(Slice *s) { + _cleanup_free_ char *parent = NULL; + int r; + + assert(s); + + if (UNIT(s)->load_state != UNIT_LOADED) + return 0; + + if (!slice_name_is_valid(UNIT(s)->id)) { + log_unit_error(UNIT(s), "Slice name %s is not valid. Refusing.", UNIT(s)->id); + return -EINVAL; + } + + r = slice_build_parent_slice(UNIT(s)->id, &parent); + if (r < 0) + return log_unit_error_errno(UNIT(s), r, "Failed to determine parent slice: %m"); + + if (parent ? !unit_has_name(UNIT_DEREF(UNIT(s)->slice), parent) : UNIT_ISSET(UNIT(s)->slice)) { + log_unit_error(UNIT(s), "Located outside of parent slice. Refusing."); + return -EINVAL; + } + + return 0; +} + +static int slice_load(Unit *u) { + Slice *s = SLICE(u); + int r; + + assert(s); + assert(u->load_state == UNIT_STUB); + + r = unit_load_fragment_and_dropin_optional(u); + if (r < 0) + return r; + + /* This is a new unit? Then let's add in some extras */ + if (u->load_state == UNIT_LOADED) { + + r = unit_patch_contexts(u); + if (r < 0) + return r; + + r = slice_add_parent_slice(s); + if (r < 0) + return r; + + r = slice_add_default_dependencies(s); + if (r < 0) + return r; + } + + return slice_verify(s); +} + +static int slice_coldplug(Unit *u) { + Slice *t = SLICE(u); + + assert(t); + assert(t->state == SLICE_DEAD); + + if (t->deserialized_state != t->state) + slice_set_state(t, t->deserialized_state); + + return 0; +} + +static void slice_dump(Unit *u, FILE *f, const char *prefix) { + Slice *t = SLICE(u); + + assert(t); + assert(f); + + fprintf(f, + "%sSlice State: %s\n", + prefix, slice_state_to_string(t->state)); + + cgroup_context_dump(&t->cgroup_context, f, prefix); +} + +static int slice_start(Unit *u) { + Slice *t = SLICE(u); + + assert(t); + assert(t->state == SLICE_DEAD); + + (void) unit_realize_cgroup(u); + (void) unit_reset_cpu_usage(u); + + slice_set_state(t, SLICE_ACTIVE); + return 1; +} + +static int slice_stop(Unit *u) { + Slice *t = SLICE(u); + + assert(t); + assert(t->state == SLICE_ACTIVE); + + /* We do not need to destroy the cgroup explicitly, + * unit_notify() will do that for us anyway. */ + + slice_set_state(t, SLICE_DEAD); + return 1; +} + +static int slice_kill(Unit *u, KillWho who, int signo, sd_bus_error *error) { + return unit_kill_common(u, who, signo, -1, -1, error); +} + +static int slice_serialize(Unit *u, FILE *f, FDSet *fds) { + Slice *s = SLICE(u); + + assert(s); + assert(f); + assert(fds); + + unit_serialize_item(u, f, "state", slice_state_to_string(s->state)); + return 0; +} + +static int slice_deserialize_item(Unit *u, const char *key, const char *value, FDSet *fds) { + Slice *s = SLICE(u); + + assert(u); + assert(key); + assert(value); + assert(fds); + + if (streq(key, "state")) { + SliceState state; + + state = slice_state_from_string(value); + if (state < 0) + log_debug("Failed to parse state value %s", value); + else + s->deserialized_state = state; + + } else + log_debug("Unknown serialization key '%s'", key); + + return 0; +} + +_pure_ static UnitActiveState slice_active_state(Unit *u) { + assert(u); + + return state_translation_table[SLICE(u)->state]; +} + +_pure_ static const char *slice_sub_state_to_string(Unit *u) { + assert(u); + + return slice_state_to_string(SLICE(u)->state); +} + +static void slice_enumerate(Manager *m) { + Unit *u; + int r; + + assert(m); + + u = manager_get_unit(m, SPECIAL_ROOT_SLICE); + if (!u) { + u = unit_new(m, sizeof(Slice)); + if (!u) { + log_oom(); + return; + } + + r = unit_add_name(u, SPECIAL_ROOT_SLICE); + if (r < 0) { + unit_free(u); + log_error_errno(r, "Failed to add -.slice name"); + return; + } + } + + u->default_dependencies = false; + u->no_gc = true; + u->ignore_on_isolate = true; + u->refuse_manual_start = true; + u->refuse_manual_stop = true; + SLICE(u)->deserialized_state = SLICE_ACTIVE; + + if (!u->description) + u->description = strdup("Root Slice"); + if (!u->documentation) + (void) strv_extend(&u->documentation, "man:systemd.special(7)"); + + unit_add_to_load_queue(u); + unit_add_to_dbus_queue(u); +} + +const UnitVTable slice_vtable = { + .object_size = sizeof(Slice), + .cgroup_context_offset = offsetof(Slice, cgroup_context), + + .sections = + "Unit\0" + "Slice\0" + "Install\0", + .private_section = "Slice", + + .can_transient = true, + + .init = slice_init, + .load = slice_load, + + .coldplug = slice_coldplug, + + .dump = slice_dump, + + .start = slice_start, + .stop = slice_stop, + + .kill = slice_kill, + + .serialize = slice_serialize, + .deserialize_item = slice_deserialize_item, + + .active_state = slice_active_state, + .sub_state_to_string = slice_sub_state_to_string, + + .bus_vtable = bus_slice_vtable, + .bus_set_property = bus_slice_set_property, + .bus_commit_properties = bus_slice_commit_properties, + + .enumerate = slice_enumerate, + + .status_message_formats = { + .finished_start_job = { + [JOB_DONE] = "Created slice %s.", + }, + .finished_stop_job = { + [JOB_DONE] = "Removed slice %s.", + }, + }, +}; diff --git a/src/grp-system/libcore/src/smack-setup.c b/src/grp-system/libcore/src/smack-setup.c new file mode 100644 index 0000000000..5d94873419 --- /dev/null +++ b/src/grp-system/libcore/src/smack-setup.c @@ -0,0 +1,346 @@ +/*** + This file is part of systemd. + + Copyright (C) 2013 Intel Corporation + Authors: + Nathaniel Chen + + systemd is free software; you can redistribute it and/or modify it + under the terms of the GNU Lesser General Public License as published + by the Free Software Foundation; either version 2.1 of the License, + or (at your option) any later version. + + systemd is distributed in the hope that it will be useful, but + WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + Lesser General Public License for more details. + + You should have received a copy of the GNU Lesser General Public License + along with systemd; If not, see . +***/ + +#include +#include +#include +#include +#include +#include + +#include "core/smack-setup.h" +#include "systemd-basic/alloc-util.h" +#include "systemd-basic/dirent-util.h" +#include "systemd-basic/fd-util.h" +#include "systemd-basic/fileio.h" +#include "systemd-basic/log.h" +#include "systemd-basic/macro.h" +#include "systemd-basic/string-util.h" +#include "systemd-basic/util.h" + +#ifdef HAVE_SMACK + +static int write_access2_rules(const char* srcdir) { + _cleanup_close_ int load2_fd = -1, change_fd = -1; + _cleanup_closedir_ DIR *dir = NULL; + struct dirent *entry; + char buf[NAME_MAX]; + int dfd = -1; + int r = 0; + + load2_fd = open("/sys/fs/smackfs/load2", O_RDWR|O_CLOEXEC|O_NONBLOCK|O_NOCTTY); + if (load2_fd < 0) { + if (errno != ENOENT) + log_warning_errno(errno, "Failed to open '/sys/fs/smackfs/load2': %m"); + return -errno; /* negative error */ + } + + change_fd = open("/sys/fs/smackfs/change-rule", O_RDWR|O_CLOEXEC|O_NONBLOCK|O_NOCTTY); + if (change_fd < 0) { + if (errno != ENOENT) + log_warning_errno(errno, "Failed to open '/sys/fs/smackfs/change-rule': %m"); + return -errno; /* negative error */ + } + + /* write rules to load2 or change-rule from every file in the directory */ + dir = opendir(srcdir); + if (!dir) { + if (errno != ENOENT) + log_warning_errno(errno, "Failed to opendir '%s': %m", srcdir); + return errno; /* positive on purpose */ + } + + dfd = dirfd(dir); + assert(dfd >= 0); + + FOREACH_DIRENT(entry, dir, return 0) { + int fd; + _cleanup_fclose_ FILE *policy = NULL; + + if (!dirent_is_file(entry)) + continue; + + fd = openat(dfd, entry->d_name, O_RDONLY|O_CLOEXEC); + if (fd < 0) { + if (r == 0) + r = -errno; + log_warning_errno(errno, "Failed to open '%s': %m", entry->d_name); + continue; + } + + policy = fdopen(fd, "re"); + if (!policy) { + if (r == 0) + r = -errno; + safe_close(fd); + log_error_errno(errno, "Failed to open '%s': %m", entry->d_name); + continue; + } + + /* load2 write rules in the kernel require a line buffered stream */ + FOREACH_LINE(buf, policy, + log_error_errno(errno, "Failed to read line from '%s': %m", + entry->d_name)) { + + _cleanup_free_ char *sbj = NULL, *obj = NULL, *acc1 = NULL, *acc2 = NULL; + + if (isempty(truncate_nl(buf))) + continue; + + /* if 3 args -> load rule : subject object access1 */ + /* if 4 args -> change rule : subject object access1 access2 */ + if (sscanf(buf, "%ms %ms %ms %ms", &sbj, &obj, &acc1, &acc2) < 3) { + log_error_errno(errno, "Failed to parse rule '%s' in '%s', ignoring.", buf, entry->d_name); + continue; + } + + if (write(isempty(acc2) ? load2_fd : change_fd, buf, strlen(buf)) < 0) { + if (r == 0) + r = -errno; + log_error_errno(errno, "Failed to write '%s' to '%s' in '%s'", + buf, isempty(acc2) ? "/sys/fs/smackfs/load2" : "/sys/fs/smackfs/change-rule", entry->d_name); + } + } + } + + return r; +} + +static int write_cipso2_rules(const char* srcdir) { + _cleanup_close_ int cipso2_fd = -1; + _cleanup_closedir_ DIR *dir = NULL; + struct dirent *entry; + char buf[NAME_MAX]; + int dfd = -1; + int r = 0; + + cipso2_fd = open("/sys/fs/smackfs/cipso2", O_RDWR|O_CLOEXEC|O_NONBLOCK|O_NOCTTY); + if (cipso2_fd < 0) { + if (errno != ENOENT) + log_warning_errno(errno, "Failed to open '/sys/fs/smackfs/cipso2': %m"); + return -errno; /* negative error */ + } + + /* write rules to cipso2 from every file in the directory */ + dir = opendir(srcdir); + if (!dir) { + if (errno != ENOENT) + log_warning_errno(errno, "Failed to opendir '%s': %m", srcdir); + return errno; /* positive on purpose */ + } + + dfd = dirfd(dir); + assert(dfd >= 0); + + FOREACH_DIRENT(entry, dir, return 0) { + int fd; + _cleanup_fclose_ FILE *policy = NULL; + + if (!dirent_is_file(entry)) + continue; + + fd = openat(dfd, entry->d_name, O_RDONLY|O_CLOEXEC); + if (fd < 0) { + if (r == 0) + r = -errno; + log_error_errno(errno, "Failed to open '%s': %m", entry->d_name); + continue; + } + + policy = fdopen(fd, "re"); + if (!policy) { + if (r == 0) + r = -errno; + safe_close(fd); + log_error_errno(errno, "Failed to open '%s': %m", entry->d_name); + continue; + } + + /* cipso2 write rules in the kernel require a line buffered stream */ + FOREACH_LINE(buf, policy, + log_error_errno(errno, "Failed to read line from '%s': %m", + entry->d_name)) { + + if (isempty(truncate_nl(buf))) + continue; + + if (write(cipso2_fd, buf, strlen(buf)) < 0) { + if (r == 0) + r = -errno; + log_error_errno(errno, "Failed to write '%s' to '/sys/fs/smackfs/cipso2' in '%s'", + buf, entry->d_name); + break; + } + } + } + + return r; +} + +static int write_netlabel_rules(const char* srcdir) { + _cleanup_fclose_ FILE *dst = NULL; + _cleanup_closedir_ DIR *dir = NULL; + struct dirent *entry; + char buf[NAME_MAX]; + int dfd = -1; + int r = 0; + + dst = fopen("/sys/fs/smackfs/netlabel", "we"); + if (!dst) { + if (errno != ENOENT) + log_warning_errno(errno, "Failed to open /sys/fs/smackfs/netlabel: %m"); + return -errno; /* negative error */ + } + + /* write rules to dst from every file in the directory */ + dir = opendir(srcdir); + if (!dir) { + if (errno != ENOENT) + log_warning_errno(errno, "Failed to opendir %s: %m", srcdir); + return errno; /* positive on purpose */ + } + + dfd = dirfd(dir); + assert(dfd >= 0); + + FOREACH_DIRENT(entry, dir, return 0) { + int fd; + _cleanup_fclose_ FILE *policy = NULL; + + fd = openat(dfd, entry->d_name, O_RDONLY|O_CLOEXEC); + if (fd < 0) { + if (r == 0) + r = -errno; + log_warning_errno(errno, "Failed to open %s: %m", entry->d_name); + continue; + } + + policy = fdopen(fd, "re"); + if (!policy) { + if (r == 0) + r = -errno; + safe_close(fd); + log_error_errno(errno, "Failed to open %s: %m", entry->d_name); + continue; + } + + /* load2 write rules in the kernel require a line buffered stream */ + FOREACH_LINE(buf, policy, + log_error_errno(errno, "Failed to read line from %s: %m", + entry->d_name)) { + if (!fputs(buf, dst)) { + if (r == 0) + r = -EINVAL; + log_error_errno(errno, "Failed to write line to /sys/fs/smackfs/netlabel"); + break; + } + if (fflush(dst)) { + if (r == 0) + r = -errno; + log_error_errno(errno, "Failed to flush writes to /sys/fs/smackfs/netlabel: %m"); + break; + } + } + } + + return r; +} + +#endif + +int mac_smack_setup(bool *loaded_policy) { + +#ifdef HAVE_SMACK + + int r; + + assert(loaded_policy); + + r = write_access2_rules("/etc/smack/accesses.d/"); + switch(r) { + case -ENOENT: + log_debug("Smack is not enabled in the kernel."); + return 0; + case ENOENT: + log_debug("Smack access rules directory '/etc/smack/accesses.d/' not found"); + return 0; + case 0: + log_info("Successfully loaded Smack policies."); + break; + default: + log_warning_errno(r, "Failed to load Smack access rules, ignoring: %m"); + return 0; + } + +#ifdef SMACK_RUN_LABEL + r = write_string_file("/proc/self/attr/current", SMACK_RUN_LABEL, 0); + if (r < 0) + log_warning_errno(r, "Failed to set SMACK label \"" SMACK_RUN_LABEL "\" on self: %m"); + r = write_string_file("/sys/fs/smackfs/ambient", SMACK_RUN_LABEL, 0); + if (r < 0) + log_warning_errno(r, "Failed to set SMACK ambient label \"" SMACK_RUN_LABEL "\": %m"); + r = write_string_file("/sys/fs/smackfs/netlabel", + "0.0.0.0/0 " SMACK_RUN_LABEL, 0); + if (r < 0) + log_warning_errno(r, "Failed to set SMACK netlabel rule \"0.0.0.0/0 " SMACK_RUN_LABEL "\": %m"); + r = write_string_file("/sys/fs/smackfs/netlabel", "127.0.0.1 -CIPSO", 0); + if (r < 0) + log_warning_errno(r, "Failed to set SMACK netlabel rule \"127.0.0.1 -CIPSO\": %m"); +#endif + + r = write_cipso2_rules("/etc/smack/cipso.d/"); + switch(r) { + case -ENOENT: + log_debug("Smack/CIPSO is not enabled in the kernel."); + return 0; + case ENOENT: + log_debug("Smack/CIPSO access rules directory '/etc/smack/cipso.d/' not found"); + break; + case 0: + log_info("Successfully loaded Smack/CIPSO policies."); + break; + default: + log_warning_errno(r, "Failed to load Smack/CIPSO access rules, ignoring: %m"); + break; + } + + r = write_netlabel_rules("/etc/smack/netlabel.d/"); + switch(r) { + case -ENOENT: + log_debug("Smack/CIPSO is not enabled in the kernel."); + return 0; + case ENOENT: + log_debug("Smack network host rules directory '/etc/smack/netlabel.d/' not found"); + break; + case 0: + log_info("Successfully loaded Smack network host rules."); + break; + default: + log_warning_errno(r, "Failed to load Smack network host rules: %m, ignoring."); + break; + } + + *loaded_policy = true; + +#endif + + return 0; +} diff --git a/src/grp-system/libcore/src/socket.c b/src/grp-system/libcore/src/socket.c new file mode 100644 index 0000000000..7e57cc7e79 --- /dev/null +++ b/src/grp-system/libcore/src/socket.c @@ -0,0 +1,2994 @@ +/*** + This file is part of systemd. + + Copyright 2010 Lennart Poettering + + systemd is free software; you can redistribute it and/or modify it + under the terms of the GNU Lesser General Public License as published by + the Free Software Foundation; either version 2.1 of the License, or + (at your option) any later version. + + systemd is distributed in the hope that it will be useful, but + WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + Lesser General Public License for more details. + + You should have received a copy of the GNU Lesser General Public License + along with systemd; If not, see . +***/ + +#include +#include +#include +#include +#include +#include +#include +#include +#include + +#include + +#include "core/socket.h" +#include "core/unit.h" +#include "sd-bus/bus-error.h" +#include "sd-bus/bus-util.h" +#include "systemd-basic/alloc-util.h" +#include "systemd-basic/copy.h" +#include "systemd-basic/def.h" +#include "systemd-basic/exit-status.h" +#include "systemd-basic/fd-util.h" +#include "systemd-basic/formats-util.h" +#include "systemd-basic/io-util.h" +#include "systemd-basic/label.h" +#include "systemd-basic/log.h" +#include "systemd-basic/missing.h" +#include "systemd-basic/mkdir.h" +#include "systemd-basic/parse-util.h" +#include "systemd-basic/path-util.h" +#include "systemd-basic/process-util.h" +#include "systemd-basic/selinux-util.h" +#include "systemd-basic/signal-util.h" +#include "systemd-basic/smack-util.h" +#include "systemd-basic/special.h" +#include "systemd-basic/string-table.h" +#include "systemd-basic/string-util.h" +#include "systemd-basic/strv.h" +#include "systemd-basic/unit-name.h" +#include "systemd-basic/user-util.h" + +#include "dbus-socket.h" +#include "unit-printf.h" + +static const UnitActiveState state_translation_table[_SOCKET_STATE_MAX] = { + [SOCKET_DEAD] = UNIT_INACTIVE, + [SOCKET_START_PRE] = UNIT_ACTIVATING, + [SOCKET_START_CHOWN] = UNIT_ACTIVATING, + [SOCKET_START_POST] = UNIT_ACTIVATING, + [SOCKET_LISTENING] = UNIT_ACTIVE, + [SOCKET_RUNNING] = UNIT_ACTIVE, + [SOCKET_STOP_PRE] = UNIT_DEACTIVATING, + [SOCKET_STOP_PRE_SIGTERM] = UNIT_DEACTIVATING, + [SOCKET_STOP_PRE_SIGKILL] = UNIT_DEACTIVATING, + [SOCKET_STOP_POST] = UNIT_DEACTIVATING, + [SOCKET_FINAL_SIGTERM] = UNIT_DEACTIVATING, + [SOCKET_FINAL_SIGKILL] = UNIT_DEACTIVATING, + [SOCKET_FAILED] = UNIT_FAILED +}; + +static int socket_dispatch_io(sd_event_source *source, int fd, uint32_t revents, void *userdata); +static int socket_dispatch_timer(sd_event_source *source, usec_t usec, void *userdata); + +static void socket_init(Unit *u) { + Socket *s = SOCKET(u); + + assert(u); + assert(u->load_state == UNIT_STUB); + + s->backlog = SOMAXCONN; + s->timeout_usec = u->manager->default_timeout_start_usec; + s->directory_mode = 0755; + s->socket_mode = 0666; + + s->max_connections = 64; + + s->priority = -1; + s->ip_tos = -1; + s->ip_ttl = -1; + s->mark = -1; + + s->exec_context.std_output = u->manager->default_std_output; + s->exec_context.std_error = u->manager->default_std_error; + + s->control_command_id = _SOCKET_EXEC_COMMAND_INVALID; + + s->trigger_limit.interval = USEC_INFINITY; + s->trigger_limit.burst = (unsigned) -1; +} + +static void socket_unwatch_control_pid(Socket *s) { + assert(s); + + if (s->control_pid <= 0) + return; + + unit_unwatch_pid(UNIT(s), s->control_pid); + s->control_pid = 0; +} + +static void socket_cleanup_fd_list(SocketPort *p) { + assert(p); + + close_many(p->auxiliary_fds, p->n_auxiliary_fds); + p->auxiliary_fds = mfree(p->auxiliary_fds); + p->n_auxiliary_fds = 0; +} + +void socket_free_ports(Socket *s) { + SocketPort *p; + + assert(s); + + while ((p = s->ports)) { + LIST_REMOVE(port, s->ports, p); + + sd_event_source_unref(p->event_source); + + socket_cleanup_fd_list(p); + safe_close(p->fd); + free(p->path); + free(p); + } +} + +static void socket_done(Unit *u) { + Socket *s = SOCKET(u); + + assert(s); + + socket_free_ports(s); + + s->exec_runtime = exec_runtime_unref(s->exec_runtime); + exec_command_free_array(s->exec_command, _SOCKET_EXEC_COMMAND_MAX); + s->control_command = NULL; + + socket_unwatch_control_pid(s); + + unit_ref_unset(&s->service); + + s->tcp_congestion = mfree(s->tcp_congestion); + s->bind_to_device = mfree(s->bind_to_device); + + s->smack = mfree(s->smack); + s->smack_ip_in = mfree(s->smack_ip_in); + s->smack_ip_out = mfree(s->smack_ip_out); + + strv_free(s->symlinks); + + s->user = mfree(s->user); + s->group = mfree(s->group); + + s->fdname = mfree(s->fdname); + + s->timer_event_source = sd_event_source_unref(s->timer_event_source); +} + +static int socket_arm_timer(Socket *s, usec_t usec) { + int r; + + assert(s); + + if (s->timer_event_source) { + r = sd_event_source_set_time(s->timer_event_source, usec); + if (r < 0) + return r; + + return sd_event_source_set_enabled(s->timer_event_source, SD_EVENT_ONESHOT); + } + + if (usec == USEC_INFINITY) + return 0; + + r = sd_event_add_time( + UNIT(s)->manager->event, + &s->timer_event_source, + CLOCK_MONOTONIC, + usec, 0, + socket_dispatch_timer, s); + if (r < 0) + return r; + + (void) sd_event_source_set_description(s->timer_event_source, "socket-timer"); + + return 0; +} + +int socket_instantiate_service(Socket *s) { + _cleanup_free_ char *prefix = NULL, *name = NULL; + int r; + Unit *u; + + assert(s); + + /* This fills in s->service if it isn't filled in yet. For + * Accept=yes sockets we create the next connection service + * here. For Accept=no this is mostly a NOP since the service + * is figured out at load time anyway. */ + + if (UNIT_DEREF(s->service)) + return 0; + + if (!s->accept) + return 0; + + r = unit_name_to_prefix(UNIT(s)->id, &prefix); + if (r < 0) + return r; + + if (asprintf(&name, "%s@%u.service", prefix, s->n_accepted) < 0) + return -ENOMEM; + + r = manager_load_unit(UNIT(s)->manager, name, NULL, NULL, &u); + if (r < 0) + return r; + + unit_ref_set(&s->service, u); + + return unit_add_two_dependencies(UNIT(s), UNIT_BEFORE, UNIT_TRIGGERS, u, false); +} + +static bool have_non_accept_socket(Socket *s) { + SocketPort *p; + + assert(s); + + if (!s->accept) + return true; + + LIST_FOREACH(port, p, s->ports) { + + if (p->type != SOCKET_SOCKET) + return true; + + if (!socket_address_can_accept(&p->address)) + return true; + } + + return false; +} + +static int socket_add_mount_links(Socket *s) { + SocketPort *p; + int r; + + assert(s); + + LIST_FOREACH(port, p, s->ports) { + const char *path = NULL; + + if (p->type == SOCKET_SOCKET) + path = socket_address_get_path(&p->address); + else if (IN_SET(p->type, SOCKET_FIFO, SOCKET_SPECIAL, SOCKET_USB_FUNCTION)) + path = p->path; + + if (!path) + continue; + + r = unit_require_mounts_for(UNIT(s), path); + if (r < 0) + return r; + } + + return 0; +} + +static int socket_add_device_link(Socket *s) { + char *t; + + assert(s); + + if (!s->bind_to_device || streq(s->bind_to_device, "lo")) + return 0; + + t = strjoina("/sys/subsystem/net/devices/", s->bind_to_device); + return unit_add_node_link(UNIT(s), t, false, UNIT_BINDS_TO); +} + +static int socket_add_default_dependencies(Socket *s) { + int r; + assert(s); + + if (!UNIT(s)->default_dependencies) + return 0; + + r = unit_add_dependency_by_name(UNIT(s), UNIT_BEFORE, SPECIAL_SOCKETS_TARGET, NULL, true); + if (r < 0) + return r; + + if (MANAGER_IS_SYSTEM(UNIT(s)->manager)) { + r = unit_add_two_dependencies_by_name(UNIT(s), UNIT_AFTER, UNIT_REQUIRES, SPECIAL_SYSINIT_TARGET, NULL, true); + if (r < 0) + return r; + } + + return unit_add_two_dependencies_by_name(UNIT(s), UNIT_BEFORE, UNIT_CONFLICTS, SPECIAL_SHUTDOWN_TARGET, NULL, true); +} + +_pure_ static bool socket_has_exec(Socket *s) { + unsigned i; + assert(s); + + for (i = 0; i < _SOCKET_EXEC_COMMAND_MAX; i++) + if (s->exec_command[i]) + return true; + + return false; +} + +static int socket_add_extras(Socket *s) { + Unit *u = UNIT(s); + int r; + + assert(s); + + /* Pick defaults for the trigger limit, if nothing was explicitly configured. We pick a relatively high limit + * in Accept=yes mode, and a lower limit for Accept=no. Reason: in Accept=yes mode we are invoking accept() + * ourselves before the trigger limit can hit, thus incoming connections are taken off the socket queue quickly + * and reliably. This is different for Accept=no, where the spawned service has to take the incoming traffic + * off the queues, which it might not necessarily do. Moreover, while Accept=no services are supposed to + * process whatever is queued in one go, and thus should normally never have to be started frequently. This is + * different for Accept=yes where each connection is processed by a new service instance, and thus frequent + * service starts are typical. */ + + if (s->trigger_limit.interval == USEC_INFINITY) + s->trigger_limit.interval = 2 * USEC_PER_SEC; + + if (s->trigger_limit.burst == (unsigned) -1) { + if (s->accept) + s->trigger_limit.burst = 200; + else + s->trigger_limit.burst = 20; + } + + if (have_non_accept_socket(s)) { + + if (!UNIT_DEREF(s->service)) { + Unit *x; + + r = unit_load_related_unit(u, ".service", &x); + if (r < 0) + return r; + + unit_ref_set(&s->service, x); + } + + r = unit_add_two_dependencies(u, UNIT_BEFORE, UNIT_TRIGGERS, UNIT_DEREF(s->service), true); + if (r < 0) + return r; + } + + r = socket_add_mount_links(s); + if (r < 0) + return r; + + r = socket_add_device_link(s); + if (r < 0) + return r; + + r = unit_patch_contexts(u); + if (r < 0) + return r; + + if (socket_has_exec(s)) { + r = unit_add_exec_dependencies(u, &s->exec_context); + if (r < 0) + return r; + + r = unit_set_default_slice(u); + if (r < 0) + return r; + } + + r = socket_add_default_dependencies(s); + if (r < 0) + return r; + + return 0; +} + +static const char *socket_find_symlink_target(Socket *s) { + const char *found = NULL; + SocketPort *p; + + LIST_FOREACH(port, p, s->ports) { + const char *f = NULL; + + switch (p->type) { + + case SOCKET_FIFO: + f = p->path; + break; + + case SOCKET_SOCKET: + if (p->address.sockaddr.un.sun_path[0] != 0) + f = p->address.sockaddr.un.sun_path; + break; + + default: + break; + } + + if (f) { + if (found) + return NULL; + + found = f; + } + } + + return found; +} + +static int socket_verify(Socket *s) { + assert(s); + + if (UNIT(s)->load_state != UNIT_LOADED) + return 0; + + if (!s->ports) { + log_unit_error(UNIT(s), "Unit lacks Listen setting. Refusing."); + return -EINVAL; + } + + if (s->accept && have_non_accept_socket(s)) { + log_unit_error(UNIT(s), "Unit configured for accepting sockets, but sockets are non-accepting. Refusing."); + return -EINVAL; + } + + if (s->accept && s->max_connections <= 0) { + log_unit_error(UNIT(s), "MaxConnection= setting too small. Refusing."); + return -EINVAL; + } + + if (s->accept && UNIT_DEREF(s->service)) { + log_unit_error(UNIT(s), "Explicit service configuration for accepting socket units not supported. Refusing."); + return -EINVAL; + } + + if (s->exec_context.pam_name && s->kill_context.kill_mode != KILL_CONTROL_GROUP) { + log_unit_error(UNIT(s), "Unit has PAM enabled. Kill mode must be set to 'control-group'. Refusing."); + return -EINVAL; + } + + if (!strv_isempty(s->symlinks) && !socket_find_symlink_target(s)) { + log_unit_error(UNIT(s), "Unit has symlinks set but none or more than one node in the file system. Refusing."); + return -EINVAL; + } + + return 0; +} + +static int socket_load(Unit *u) { + Socket *s = SOCKET(u); + int r; + + assert(u); + assert(u->load_state == UNIT_STUB); + + r = unit_load_fragment_and_dropin(u); + if (r < 0) + return r; + + if (u->load_state == UNIT_LOADED) { + /* This is a new unit? Then let's add in some extras */ + r = socket_add_extras(s); + if (r < 0) + return r; + } + + return socket_verify(s); +} + +_const_ static const char* listen_lookup(int family, int type) { + + if (family == AF_NETLINK) + return "ListenNetlink"; + + if (type == SOCK_STREAM) + return "ListenStream"; + else if (type == SOCK_DGRAM) + return "ListenDatagram"; + else if (type == SOCK_SEQPACKET) + return "ListenSequentialPacket"; + + assert_not_reached("Unknown socket type"); + return NULL; +} + +static void socket_dump(Unit *u, FILE *f, const char *prefix) { + char time_string[FORMAT_TIMESPAN_MAX]; + SocketExecCommand c; + Socket *s = SOCKET(u); + SocketPort *p; + const char *prefix2; + + assert(s); + assert(f); + + prefix = strempty(prefix); + prefix2 = strjoina(prefix, "\t"); + + fprintf(f, + "%sSocket State: %s\n" + "%sResult: %s\n" + "%sBindIPv6Only: %s\n" + "%sBacklog: %u\n" + "%sSocketMode: %04o\n" + "%sDirectoryMode: %04o\n" + "%sKeepAlive: %s\n" + "%sNoDelay: %s\n" + "%sFreeBind: %s\n" + "%sTransparent: %s\n" + "%sBroadcast: %s\n" + "%sPassCredentials: %s\n" + "%sPassSecurity: %s\n" + "%sTCPCongestion: %s\n" + "%sRemoveOnStop: %s\n" + "%sWritable: %s\n" + "%sFDName: %s\n" + "%sSELinuxContextFromNet: %s\n", + prefix, socket_state_to_string(s->state), + prefix, socket_result_to_string(s->result), + prefix, socket_address_bind_ipv6_only_to_string(s->bind_ipv6_only), + prefix, s->backlog, + prefix, s->socket_mode, + prefix, s->directory_mode, + prefix, yes_no(s->keep_alive), + prefix, yes_no(s->no_delay), + prefix, yes_no(s->free_bind), + prefix, yes_no(s->transparent), + prefix, yes_no(s->broadcast), + prefix, yes_no(s->pass_cred), + prefix, yes_no(s->pass_sec), + prefix, strna(s->tcp_congestion), + prefix, yes_no(s->remove_on_stop), + prefix, yes_no(s->writable), + prefix, socket_fdname(s), + prefix, yes_no(s->selinux_context_from_net)); + + if (s->control_pid > 0) + fprintf(f, + "%sControl PID: "PID_FMT"\n", + prefix, s->control_pid); + + if (s->bind_to_device) + fprintf(f, + "%sBindToDevice: %s\n", + prefix, s->bind_to_device); + + if (s->accept) + fprintf(f, + "%sAccepted: %u\n" + "%sNConnections: %u\n" + "%sMaxConnections: %u\n", + prefix, s->n_accepted, + prefix, s->n_connections, + prefix, s->max_connections); + + if (s->priority >= 0) + fprintf(f, + "%sPriority: %i\n", + prefix, s->priority); + + if (s->receive_buffer > 0) + fprintf(f, + "%sReceiveBuffer: %zu\n", + prefix, s->receive_buffer); + + if (s->send_buffer > 0) + fprintf(f, + "%sSendBuffer: %zu\n", + prefix, s->send_buffer); + + if (s->ip_tos >= 0) + fprintf(f, + "%sIPTOS: %i\n", + prefix, s->ip_tos); + + if (s->ip_ttl >= 0) + fprintf(f, + "%sIPTTL: %i\n", + prefix, s->ip_ttl); + + if (s->pipe_size > 0) + fprintf(f, + "%sPipeSize: %zu\n", + prefix, s->pipe_size); + + if (s->mark >= 0) + fprintf(f, + "%sMark: %i\n", + prefix, s->mark); + + if (s->mq_maxmsg > 0) + fprintf(f, + "%sMessageQueueMaxMessages: %li\n", + prefix, s->mq_maxmsg); + + if (s->mq_msgsize > 0) + fprintf(f, + "%sMessageQueueMessageSize: %li\n", + prefix, s->mq_msgsize); + + if (s->reuse_port) + fprintf(f, + "%sReusePort: %s\n", + prefix, yes_no(s->reuse_port)); + + if (s->smack) + fprintf(f, + "%sSmackLabel: %s\n", + prefix, s->smack); + + if (s->smack_ip_in) + fprintf(f, + "%sSmackLabelIPIn: %s\n", + prefix, s->smack_ip_in); + + if (s->smack_ip_out) + fprintf(f, + "%sSmackLabelIPOut: %s\n", + prefix, s->smack_ip_out); + + if (!isempty(s->user) || !isempty(s->group)) + fprintf(f, + "%sSocketUser: %s\n" + "%sSocketGroup: %s\n", + prefix, strna(s->user), + prefix, strna(s->group)); + + if (s->keep_alive_time > 0) + fprintf(f, + "%sKeepAliveTimeSec: %s\n", + prefix, format_timespan(time_string, FORMAT_TIMESPAN_MAX, s->keep_alive_time, USEC_PER_SEC)); + + if (s->keep_alive_interval) + fprintf(f, + "%sKeepAliveIntervalSec: %s\n", + prefix, format_timespan(time_string, FORMAT_TIMESPAN_MAX, s->keep_alive_interval, USEC_PER_SEC)); + + if (s->keep_alive_cnt) + fprintf(f, + "%sKeepAliveProbes: %u\n", + prefix, s->keep_alive_cnt); + + if (s->defer_accept) + fprintf(f, + "%sDeferAcceptSec: %s\n", + prefix, format_timespan(time_string, FORMAT_TIMESPAN_MAX, s->defer_accept, USEC_PER_SEC)); + + LIST_FOREACH(port, p, s->ports) { + + if (p->type == SOCKET_SOCKET) { + const char *t; + int r; + char *k = NULL; + + r = socket_address_print(&p->address, &k); + if (r < 0) + t = strerror(-r); + else + t = k; + + fprintf(f, "%s%s: %s\n", prefix, listen_lookup(socket_address_family(&p->address), p->address.type), t); + free(k); + } else if (p->type == SOCKET_SPECIAL) + fprintf(f, "%sListenSpecial: %s\n", prefix, p->path); + else if (p->type == SOCKET_USB_FUNCTION) + fprintf(f, "%sListenUSBFunction: %s\n", prefix, p->path); + else if (p->type == SOCKET_MQUEUE) + fprintf(f, "%sListenMessageQueue: %s\n", prefix, p->path); + else + fprintf(f, "%sListenFIFO: %s\n", prefix, p->path); + } + + fprintf(f, + "%sTriggerLimitIntervalSec: %s\n" + "%sTriggerLimitBurst: %u\n", + prefix, format_timespan(time_string, FORMAT_TIMESPAN_MAX, s->trigger_limit.interval, USEC_PER_SEC), + prefix, s->trigger_limit.burst); + + exec_context_dump(&s->exec_context, f, prefix); + kill_context_dump(&s->kill_context, f, prefix); + + for (c = 0; c < _SOCKET_EXEC_COMMAND_MAX; c++) { + if (!s->exec_command[c]) + continue; + + fprintf(f, "%s-> %s:\n", + prefix, socket_exec_command_to_string(c)); + + exec_command_dump_list(s->exec_command[c], f, prefix2); + } +} + +static int instance_from_socket(int fd, unsigned nr, char **instance) { + socklen_t l; + char *r; + union sockaddr_union local, remote; + + assert(fd >= 0); + assert(instance); + + l = sizeof(local); + if (getsockname(fd, &local.sa, &l) < 0) + return -errno; + + l = sizeof(remote); + if (getpeername(fd, &remote.sa, &l) < 0) + return -errno; + + switch (local.sa.sa_family) { + + case AF_INET: { + uint32_t + a = be32toh(local.in.sin_addr.s_addr), + b = be32toh(remote.in.sin_addr.s_addr); + + if (asprintf(&r, + "%u-%u.%u.%u.%u:%u-%u.%u.%u.%u:%u", + nr, + a >> 24, (a >> 16) & 0xFF, (a >> 8) & 0xFF, a & 0xFF, + be16toh(local.in.sin_port), + b >> 24, (b >> 16) & 0xFF, (b >> 8) & 0xFF, b & 0xFF, + be16toh(remote.in.sin_port)) < 0) + return -ENOMEM; + + break; + } + + case AF_INET6: { + static const unsigned char ipv4_prefix[] = { + 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0xFF, 0xFF + }; + + if (memcmp(&local.in6.sin6_addr, ipv4_prefix, sizeof(ipv4_prefix)) == 0 && + memcmp(&remote.in6.sin6_addr, ipv4_prefix, sizeof(ipv4_prefix)) == 0) { + const uint8_t + *a = local.in6.sin6_addr.s6_addr+12, + *b = remote.in6.sin6_addr.s6_addr+12; + + if (asprintf(&r, + "%u-%u.%u.%u.%u:%u-%u.%u.%u.%u:%u", + nr, + a[0], a[1], a[2], a[3], + be16toh(local.in6.sin6_port), + b[0], b[1], b[2], b[3], + be16toh(remote.in6.sin6_port)) < 0) + return -ENOMEM; + } else { + char a[INET6_ADDRSTRLEN], b[INET6_ADDRSTRLEN]; + + if (asprintf(&r, + "%u-%s:%u-%s:%u", + nr, + inet_ntop(AF_INET6, &local.in6.sin6_addr, a, sizeof(a)), + be16toh(local.in6.sin6_port), + inet_ntop(AF_INET6, &remote.in6.sin6_addr, b, sizeof(b)), + be16toh(remote.in6.sin6_port)) < 0) + return -ENOMEM; + } + + break; + } + + case AF_UNIX: { + struct ucred ucred; + int k; + + k = getpeercred(fd, &ucred); + if (k >= 0) { + if (asprintf(&r, + "%u-"PID_FMT"-"UID_FMT, + nr, ucred.pid, ucred.uid) < 0) + return -ENOMEM; + } else if (k == -ENODATA) { + /* This handles the case where somebody is + * connecting from another pid/uid namespace + * (e.g. from outside of our container). */ + if (asprintf(&r, + "%u-unknown", + nr) < 0) + return -ENOMEM; + } else + return k; + + break; + } + + default: + assert_not_reached("Unhandled socket type."); + } + + *instance = r; + return 0; +} + +static void socket_close_fds(Socket *s) { + SocketPort *p; + char **i; + + assert(s); + + LIST_FOREACH(port, p, s->ports) { + bool was_open; + + was_open = p->fd >= 0; + + p->event_source = sd_event_source_unref(p->event_source); + p->fd = safe_close(p->fd); + socket_cleanup_fd_list(p); + + /* One little note: we should normally not delete any sockets in the file system here! After all some + * other process we spawned might still have a reference of this fd and wants to continue to use + * it. Therefore we normally delete sockets in the file system before we create a new one, not after we + * stopped using one! That all said, if the user explicitly requested this, we'll delete them here + * anyway, but only then. */ + + if (!was_open || !s->remove_on_stop) + continue; + + switch (p->type) { + + case SOCKET_FIFO: + (void) unlink(p->path); + break; + + case SOCKET_MQUEUE: + (void) mq_unlink(p->path); + break; + + case SOCKET_SOCKET: + (void) socket_address_unlink(&p->address); + break; + + default: + break; + } + } + + if (s->remove_on_stop) + STRV_FOREACH(i, s->symlinks) + (void) unlink(*i); +} + +static void socket_apply_socket_options(Socket *s, int fd) { + int r; + + assert(s); + assert(fd >= 0); + + if (s->keep_alive) { + int b = s->keep_alive; + if (setsockopt(fd, SOL_SOCKET, SO_KEEPALIVE, &b, sizeof(b)) < 0) + log_unit_warning_errno(UNIT(s), errno, "SO_KEEPALIVE failed: %m"); + } + + if (s->keep_alive_time) { + int value = s->keep_alive_time / USEC_PER_SEC; + if (setsockopt(fd, SOL_TCP, TCP_KEEPIDLE, &value, sizeof(value)) < 0) + log_unit_warning_errno(UNIT(s), errno, "TCP_KEEPIDLE failed: %m"); + } + + if (s->keep_alive_interval) { + int value = s->keep_alive_interval / USEC_PER_SEC; + if (setsockopt(fd, SOL_TCP, TCP_KEEPINTVL, &value, sizeof(value)) < 0) + log_unit_warning_errno(UNIT(s), errno, "TCP_KEEPINTVL failed: %m"); + } + + if (s->keep_alive_cnt) { + int value = s->keep_alive_cnt; + if (setsockopt(fd, SOL_TCP, TCP_KEEPCNT, &value, sizeof(value)) < 0) + log_unit_warning_errno(UNIT(s), errno, "TCP_KEEPCNT failed: %m"); + } + + if (s->defer_accept) { + int value = s->defer_accept / USEC_PER_SEC; + if (setsockopt(fd, SOL_TCP, TCP_DEFER_ACCEPT, &value, sizeof(value)) < 0) + log_unit_warning_errno(UNIT(s), errno, "TCP_DEFER_ACCEPT failed: %m"); + } + + if (s->no_delay) { + int b = s->no_delay; + + if (s->socket_protocol == IPPROTO_SCTP) { + if (setsockopt(fd, SOL_SCTP, SCTP_NODELAY, &b, sizeof(b)) < 0) + log_unit_warning_errno(UNIT(s), errno, "SCTP_NODELAY failed: %m"); + } else { + if (setsockopt(fd, SOL_TCP, TCP_NODELAY, &b, sizeof(b)) < 0) + log_unit_warning_errno(UNIT(s), errno, "TCP_NODELAY failed: %m"); + } + } + + if (s->broadcast) { + int one = 1; + if (setsockopt(fd, SOL_SOCKET, SO_BROADCAST, &one, sizeof(one)) < 0) + log_unit_warning_errno(UNIT(s), errno, "SO_BROADCAST failed: %m"); + } + + if (s->pass_cred) { + int one = 1; + if (setsockopt(fd, SOL_SOCKET, SO_PASSCRED, &one, sizeof(one)) < 0) + log_unit_warning_errno(UNIT(s), errno, "SO_PASSCRED failed: %m"); + } + + if (s->pass_sec) { + int one = 1; + if (setsockopt(fd, SOL_SOCKET, SO_PASSSEC, &one, sizeof(one)) < 0) + log_unit_warning_errno(UNIT(s), errno, "SO_PASSSEC failed: %m"); + } + + if (s->priority >= 0) + if (setsockopt(fd, SOL_SOCKET, SO_PRIORITY, &s->priority, sizeof(s->priority)) < 0) + log_unit_warning_errno(UNIT(s), errno, "SO_PRIORITY failed: %m"); + + if (s->receive_buffer > 0) { + int value = (int) s->receive_buffer; + + /* We first try with SO_RCVBUFFORCE, in case we have the perms for that */ + + if (setsockopt(fd, SOL_SOCKET, SO_RCVBUFFORCE, &value, sizeof(value)) < 0) + if (setsockopt(fd, SOL_SOCKET, SO_RCVBUF, &value, sizeof(value)) < 0) + log_unit_warning_errno(UNIT(s), errno, "SO_RCVBUF failed: %m"); + } + + if (s->send_buffer > 0) { + int value = (int) s->send_buffer; + if (setsockopt(fd, SOL_SOCKET, SO_SNDBUFFORCE, &value, sizeof(value)) < 0) + if (setsockopt(fd, SOL_SOCKET, SO_SNDBUF, &value, sizeof(value)) < 0) + log_unit_warning_errno(UNIT(s), errno, "SO_SNDBUF failed: %m"); + } + + if (s->mark >= 0) + if (setsockopt(fd, SOL_SOCKET, SO_MARK, &s->mark, sizeof(s->mark)) < 0) + log_unit_warning_errno(UNIT(s), errno, "SO_MARK failed: %m"); + + if (s->ip_tos >= 0) + if (setsockopt(fd, IPPROTO_IP, IP_TOS, &s->ip_tos, sizeof(s->ip_tos)) < 0) + log_unit_warning_errno(UNIT(s), errno, "IP_TOS failed: %m"); + + if (s->ip_ttl >= 0) { + int x; + + r = setsockopt(fd, IPPROTO_IP, IP_TTL, &s->ip_ttl, sizeof(s->ip_ttl)); + + if (socket_ipv6_is_supported()) + x = setsockopt(fd, IPPROTO_IPV6, IPV6_UNICAST_HOPS, &s->ip_ttl, sizeof(s->ip_ttl)); + else { + x = -1; + errno = EAFNOSUPPORT; + } + + if (r < 0 && x < 0) + log_unit_warning_errno(UNIT(s), errno, "IP_TTL/IPV6_UNICAST_HOPS failed: %m"); + } + + if (s->tcp_congestion) + if (setsockopt(fd, SOL_TCP, TCP_CONGESTION, s->tcp_congestion, strlen(s->tcp_congestion)+1) < 0) + log_unit_warning_errno(UNIT(s), errno, "TCP_CONGESTION failed: %m"); + + if (s->smack_ip_in) { + r = mac_smack_apply_fd(fd, SMACK_ATTR_IPIN, s->smack_ip_in); + if (r < 0) + log_unit_error_errno(UNIT(s), r, "mac_smack_apply_ip_in_fd: %m"); + } + + if (s->smack_ip_out) { + r = mac_smack_apply_fd(fd, SMACK_ATTR_IPOUT, s->smack_ip_out); + if (r < 0) + log_unit_error_errno(UNIT(s), r, "mac_smack_apply_ip_out_fd: %m"); + } +} + +static void socket_apply_fifo_options(Socket *s, int fd) { + int r; + + assert(s); + assert(fd >= 0); + + if (s->pipe_size > 0) + if (fcntl(fd, F_SETPIPE_SZ, s->pipe_size) < 0) + log_unit_warning_errno(UNIT(s), errno, "Setting pipe size failed, ignoring: %m"); + + if (s->smack) { + r = mac_smack_apply_fd(fd, SMACK_ATTR_ACCESS, s->smack); + if (r < 0) + log_unit_error_errno(UNIT(s), r, "SMACK relabelling failed, ignoring: %m"); + } +} + +static int fifo_address_create( + const char *path, + mode_t directory_mode, + mode_t socket_mode) { + + _cleanup_close_ int fd = -1; + mode_t old_mask; + struct stat st; + int r; + + assert(path); + + mkdir_parents_label(path, directory_mode); + + r = mac_selinux_create_file_prepare(path, S_IFIFO); + if (r < 0) + return r; + + /* Enforce the right access mode for the fifo */ + old_mask = umask(~ socket_mode); + + /* Include the original umask in our mask */ + (void) umask(~socket_mode | old_mask); + + r = mkfifo(path, socket_mode); + (void) umask(old_mask); + + if (r < 0 && errno != EEXIST) { + r = -errno; + goto fail; + } + + fd = open(path, O_RDWR | O_CLOEXEC | O_NOCTTY | O_NONBLOCK | O_NOFOLLOW); + if (fd < 0) { + r = -errno; + goto fail; + } + + mac_selinux_create_file_clear(); + + if (fstat(fd, &st) < 0) { + r = -errno; + goto fail; + } + + if (!S_ISFIFO(st.st_mode) || + (st.st_mode & 0777) != (socket_mode & ~old_mask) || + st.st_uid != getuid() || + st.st_gid != getgid()) { + r = -EEXIST; + goto fail; + } + + r = fd; + fd = -1; + + return r; + +fail: + mac_selinux_create_file_clear(); + return r; +} + +static int special_address_create(const char *path, bool writable) { + _cleanup_close_ int fd = -1; + struct stat st; + int r; + + assert(path); + + fd = open(path, (writable ? O_RDWR : O_RDONLY)|O_CLOEXEC|O_NOCTTY|O_NONBLOCK|O_NOFOLLOW); + if (fd < 0) + return -errno; + + if (fstat(fd, &st) < 0) + return -errno; + + /* Check whether this is a /proc, /sys or /dev file or char device */ + if (!S_ISREG(st.st_mode) && !S_ISCHR(st.st_mode)) + return -EEXIST; + + r = fd; + fd = -1; + + return r; +} + +static int usbffs_address_create(const char *path) { + _cleanup_close_ int fd = -1; + struct stat st; + int r; + + assert(path); + + fd = open(path, O_RDWR|O_CLOEXEC|O_NOCTTY|O_NONBLOCK|O_NOFOLLOW); + if (fd < 0) + return -errno; + + if (fstat(fd, &st) < 0) + return -errno; + + /* Check whether this is a regular file (ffs endpoint)*/ + if (!S_ISREG(st.st_mode)) + return -EEXIST; + + r = fd; + fd = -1; + + return r; +} + +static int mq_address_create( + const char *path, + mode_t mq_mode, + long maxmsg, + long msgsize) { + + _cleanup_close_ int fd = -1; + struct stat st; + mode_t old_mask; + struct mq_attr _attr, *attr = NULL; + int r; + + assert(path); + + if (maxmsg > 0 && msgsize > 0) { + _attr = (struct mq_attr) { + .mq_flags = O_NONBLOCK, + .mq_maxmsg = maxmsg, + .mq_msgsize = msgsize, + }; + attr = &_attr; + } + + /* Enforce the right access mode for the mq */ + old_mask = umask(~ mq_mode); + + /* Include the original umask in our mask */ + (void) umask(~mq_mode | old_mask); + fd = mq_open(path, O_RDONLY|O_CLOEXEC|O_NONBLOCK|O_CREAT, mq_mode, attr); + (void) umask(old_mask); + + if (fd < 0) + return -errno; + + if (fstat(fd, &st) < 0) + return -errno; + + if ((st.st_mode & 0777) != (mq_mode & ~old_mask) || + st.st_uid != getuid() || + st.st_gid != getgid()) + return -EEXIST; + + r = fd; + fd = -1; + + return r; +} + +static int socket_symlink(Socket *s) { + const char *p; + char **i; + + assert(s); + + p = socket_find_symlink_target(s); + if (!p) + return 0; + + STRV_FOREACH(i, s->symlinks) + symlink_label(p, *i); + + return 0; +} + +static int usbffs_write_descs(int fd, Service *s) { + int r; + + if (!s->usb_function_descriptors || !s->usb_function_strings) + return -EINVAL; + + r = copy_file_fd(s->usb_function_descriptors, fd, false); + if (r < 0) + return r; + + return copy_file_fd(s->usb_function_strings, fd, false); +} + +static int usbffs_select_ep(const struct dirent *d) { + return d->d_name[0] != '.' && !streq(d->d_name, "ep0"); +} + +static int usbffs_dispatch_eps(SocketPort *p) { + _cleanup_free_ struct dirent **ent = NULL; + _cleanup_free_ char *path = NULL; + int r, i, n, k; + + path = dirname_malloc(p->path); + if (!path) + return -ENOMEM; + + r = scandir(path, &ent, usbffs_select_ep, alphasort); + if (r < 0) + return -errno; + + n = r; + p->auxiliary_fds = new(int, n); + if (!p->auxiliary_fds) + return -ENOMEM; + + p->n_auxiliary_fds = n; + + k = 0; + for (i = 0; i < n; ++i) { + _cleanup_free_ char *ep = NULL; + + ep = path_make_absolute(ent[i]->d_name, path); + if (!ep) + return -ENOMEM; + + path_kill_slashes(ep); + + r = usbffs_address_create(ep); + if (r < 0) + goto fail; + + p->auxiliary_fds[k] = r; + + ++k; + free(ent[i]); + } + + return r; + +fail: + close_many(p->auxiliary_fds, k); + p->auxiliary_fds = mfree(p->auxiliary_fds); + p->n_auxiliary_fds = 0; + + return r; +} + +static int socket_determine_selinux_label(Socket *s, char **ret) { + ExecCommand *c; + int r; + + assert(s); + assert(ret); + + if (s->selinux_context_from_net) { + /* If this is requested, get label from the network label */ + + r = mac_selinux_get_our_label(ret); + if (r == -EOPNOTSUPP) + goto no_label; + + } else { + /* Otherwise, get it from the executable we are about to start */ + r = socket_instantiate_service(s); + if (r < 0) + return r; + + if (!UNIT_ISSET(s->service)) + goto no_label; + + c = SERVICE(UNIT_DEREF(s->service))->exec_command[SERVICE_EXEC_START]; + if (!c) + goto no_label; + + r = mac_selinux_get_create_label_from_exe(c->path, ret); + if (r == -EPERM || r == -EOPNOTSUPP) + goto no_label; + } + + return r; + +no_label: + *ret = NULL; + return 0; +} + +static int socket_open_fds(Socket *s) { + _cleanup_(mac_selinux_freep) char *label = NULL; + bool know_label = false; + SocketPort *p; + int r; + + assert(s); + + LIST_FOREACH(port, p, s->ports) { + + if (p->fd >= 0) + continue; + + switch (p->type) { + + case SOCKET_SOCKET: + + if (!know_label) { + /* Figure out label, if we don't it know yet. We do it once, for the first socket where + * we need this and remember it for the rest. */ + + r = socket_determine_selinux_label(s, &label); + if (r < 0) + goto rollback; + + know_label = true; + } + + /* Apply the socket protocol */ + switch (p->address.type) { + + case SOCK_STREAM: + case SOCK_SEQPACKET: + if (s->socket_protocol == IPPROTO_SCTP) + p->address.protocol = s->socket_protocol; + break; + + case SOCK_DGRAM: + if (s->socket_protocol == IPPROTO_UDPLITE) + p->address.protocol = s->socket_protocol; + break; + } + + r = socket_address_listen( + &p->address, + SOCK_CLOEXEC|SOCK_NONBLOCK, + s->backlog, + s->bind_ipv6_only, + s->bind_to_device, + s->reuse_port, + s->free_bind, + s->transparent, + s->directory_mode, + s->socket_mode, + label); + if (r < 0) + goto rollback; + + p->fd = r; + socket_apply_socket_options(s, p->fd); + socket_symlink(s); + break; + + case SOCKET_SPECIAL: + + p->fd = special_address_create(p->path, s->writable); + if (p->fd < 0) { + r = p->fd; + goto rollback; + } + break; + + case SOCKET_FIFO: + + p->fd = fifo_address_create( + p->path, + s->directory_mode, + s->socket_mode); + if (p->fd < 0) { + r = p->fd; + goto rollback; + } + + socket_apply_fifo_options(s, p->fd); + socket_symlink(s); + break; + + case SOCKET_MQUEUE: + + p->fd = mq_address_create( + p->path, + s->socket_mode, + s->mq_maxmsg, + s->mq_msgsize); + if (p->fd < 0) { + r = p->fd; + goto rollback; + } + break; + + case SOCKET_USB_FUNCTION: { + _cleanup_free_ char *ep = NULL; + + ep = path_make_absolute("ep0", p->path); + + p->fd = usbffs_address_create(ep); + if (p->fd < 0) { + r = p->fd; + goto rollback; + } + + r = usbffs_write_descs(p->fd, SERVICE(UNIT_DEREF(s->service))); + if (r < 0) + goto rollback; + + r = usbffs_dispatch_eps(p); + if (r < 0) + goto rollback; + + break; + } + default: + assert_not_reached("Unknown port type"); + } + } + + return 0; + +rollback: + socket_close_fds(s); + return r; +} + +static void socket_unwatch_fds(Socket *s) { + SocketPort *p; + int r; + + assert(s); + + LIST_FOREACH(port, p, s->ports) { + if (p->fd < 0) + continue; + + if (!p->event_source) + continue; + + r = sd_event_source_set_enabled(p->event_source, SD_EVENT_OFF); + if (r < 0) + log_unit_debug_errno(UNIT(s), r, "Failed to disable event source: %m"); + } +} + +static int socket_watch_fds(Socket *s) { + SocketPort *p; + int r; + + assert(s); + + LIST_FOREACH(port, p, s->ports) { + if (p->fd < 0) + continue; + + if (p->event_source) { + r = sd_event_source_set_enabled(p->event_source, SD_EVENT_ON); + if (r < 0) + goto fail; + } else { + r = sd_event_add_io(UNIT(s)->manager->event, &p->event_source, p->fd, EPOLLIN, socket_dispatch_io, p); + if (r < 0) + goto fail; + + (void) sd_event_source_set_description(p->event_source, "socket-port-io"); + } + } + + return 0; + +fail: + log_unit_warning_errno(UNIT(s), r, "Failed to watch listening fds: %m"); + socket_unwatch_fds(s); + return r; +} + +enum { + SOCKET_OPEN_NONE, + SOCKET_OPEN_SOME, + SOCKET_OPEN_ALL, +}; + +static int socket_check_open(Socket *s) { + bool have_open = false, have_closed = false; + SocketPort *p; + + assert(s); + + LIST_FOREACH(port, p, s->ports) { + if (p->fd < 0) + have_closed = true; + else + have_open = true; + + if (have_open && have_closed) + return SOCKET_OPEN_SOME; + } + + if (have_open) + return SOCKET_OPEN_ALL; + + return SOCKET_OPEN_NONE; +} + +static void socket_set_state(Socket *s, SocketState state) { + SocketState old_state; + assert(s); + + old_state = s->state; + s->state = state; + + if (!IN_SET(state, + SOCKET_START_PRE, + SOCKET_START_CHOWN, + SOCKET_START_POST, + SOCKET_STOP_PRE, + SOCKET_STOP_PRE_SIGTERM, + SOCKET_STOP_PRE_SIGKILL, + SOCKET_STOP_POST, + SOCKET_FINAL_SIGTERM, + SOCKET_FINAL_SIGKILL)) { + + s->timer_event_source = sd_event_source_unref(s->timer_event_source); + socket_unwatch_control_pid(s); + s->control_command = NULL; + s->control_command_id = _SOCKET_EXEC_COMMAND_INVALID; + } + + if (state != SOCKET_LISTENING) + socket_unwatch_fds(s); + + if (!IN_SET(state, + SOCKET_START_CHOWN, + SOCKET_START_POST, + SOCKET_LISTENING, + SOCKET_RUNNING, + SOCKET_STOP_PRE, + SOCKET_STOP_PRE_SIGTERM, + SOCKET_STOP_PRE_SIGKILL)) + socket_close_fds(s); + + if (state != old_state) + log_unit_debug(UNIT(s), "Changed %s -> %s", socket_state_to_string(old_state), socket_state_to_string(state)); + + unit_notify(UNIT(s), state_translation_table[old_state], state_translation_table[state], true); +} + +static int socket_coldplug(Unit *u) { + Socket *s = SOCKET(u); + int r; + + assert(s); + assert(s->state == SOCKET_DEAD); + + if (s->deserialized_state == s->state) + return 0; + + if (s->control_pid > 0 && + pid_is_unwaited(s->control_pid) && + IN_SET(s->deserialized_state, + SOCKET_START_PRE, + SOCKET_START_CHOWN, + SOCKET_START_POST, + SOCKET_STOP_PRE, + SOCKET_STOP_PRE_SIGTERM, + SOCKET_STOP_PRE_SIGKILL, + SOCKET_STOP_POST, + SOCKET_FINAL_SIGTERM, + SOCKET_FINAL_SIGKILL)) { + + r = unit_watch_pid(UNIT(s), s->control_pid); + if (r < 0) + return r; + + r = socket_arm_timer(s, usec_add(u->state_change_timestamp.monotonic, s->timeout_usec)); + if (r < 0) + return r; + } + + if (IN_SET(s->deserialized_state, + SOCKET_START_CHOWN, + SOCKET_START_POST, + SOCKET_LISTENING, + SOCKET_RUNNING)) { + + /* Originally, we used to simply reopen all sockets here that we didn't have file descriptors + * for. However, this is problematic, as we won't traverse throught the SOCKET_START_CHOWN state for + * them, and thus the UID/GID wouldn't be right. Hence, instead simply check if we have all fds open, + * and if there's a mismatch, warn loudly. */ + + r = socket_check_open(s); + if (r == SOCKET_OPEN_NONE) + log_unit_warning(UNIT(s), + "Socket unit configuration has changed while unit has been running, " + "no open socket file descriptor left. " + "The socket unit is not functional until restarted."); + else if (r == SOCKET_OPEN_SOME) + log_unit_warning(UNIT(s), + "Socket unit configuration has changed while unit has been running, " + "and some socket file descriptors have not been opened yet. " + "The socket unit is not fully functional until restarted."); + } + + if (s->deserialized_state == SOCKET_LISTENING) { + r = socket_watch_fds(s); + if (r < 0) + return r; + } + + socket_set_state(s, s->deserialized_state); + return 0; +} + +static int socket_spawn(Socket *s, ExecCommand *c, pid_t *_pid) { + _cleanup_free_ char **argv = NULL; + pid_t pid; + int r; + ExecParameters exec_params = { + .apply_permissions = true, + .apply_chroot = true, + .apply_tty_stdin = true, + .stdin_fd = -1, + .stdout_fd = -1, + .stderr_fd = -1, + }; + + assert(s); + assert(c); + assert(_pid); + + (void) unit_realize_cgroup(UNIT(s)); + if (s->reset_cpu_usage) { + (void) unit_reset_cpu_usage(UNIT(s)); + s->reset_cpu_usage = false; + } + + r = unit_setup_exec_runtime(UNIT(s)); + if (r < 0) + return r; + + r = socket_arm_timer(s, usec_add(now(CLOCK_MONOTONIC), s->timeout_usec)); + if (r < 0) + return r; + + r = unit_full_printf_strv(UNIT(s), c->argv, &argv); + if (r < 0) + return r; + + exec_params.argv = argv; + exec_params.environment = UNIT(s)->manager->environment; + exec_params.confirm_spawn = UNIT(s)->manager->confirm_spawn; + exec_params.cgroup_supported = UNIT(s)->manager->cgroup_supported; + exec_params.cgroup_path = UNIT(s)->cgroup_path; + exec_params.cgroup_delegate = s->cgroup_context.delegate; + exec_params.runtime_prefix = manager_get_runtime_prefix(UNIT(s)->manager); + + r = exec_spawn(UNIT(s), + c, + &s->exec_context, + &exec_params, + s->exec_runtime, + &pid); + if (r < 0) + return r; + + r = unit_watch_pid(UNIT(s), pid); + if (r < 0) + /* FIXME: we need to do something here */ + return r; + + *_pid = pid; + return 0; +} + +static int socket_chown(Socket *s, pid_t *_pid) { + pid_t pid; + int r; + + r = socket_arm_timer(s, usec_add(now(CLOCK_MONOTONIC), s->timeout_usec)); + if (r < 0) + goto fail; + + /* We have to resolve the user names out-of-process, hence + * let's fork here. It's messy, but well, what can we do? */ + + pid = fork(); + if (pid < 0) + return -errno; + + if (pid == 0) { + SocketPort *p; + uid_t uid = UID_INVALID; + gid_t gid = GID_INVALID; + int ret; + + (void) default_signals(SIGNALS_CRASH_HANDLER, SIGNALS_IGNORE, -1); + (void) ignore_signals(SIGPIPE, -1); + log_forget_fds(); + + if (!isempty(s->user)) { + const char *user = s->user; + + r = get_user_creds(&user, &uid, &gid, NULL, NULL); + if (r < 0) { + ret = EXIT_USER; + goto fail_child; + } + } + + if (!isempty(s->group)) { + const char *group = s->group; + + r = get_group_creds(&group, &gid); + if (r < 0) { + ret = EXIT_GROUP; + goto fail_child; + } + } + + LIST_FOREACH(port, p, s->ports) { + const char *path = NULL; + + if (p->type == SOCKET_SOCKET) + path = socket_address_get_path(&p->address); + else if (p->type == SOCKET_FIFO) + path = p->path; + + if (!path) + continue; + + if (chown(path, uid, gid) < 0) { + r = -errno; + ret = EXIT_CHOWN; + goto fail_child; + } + } + + _exit(0); + + fail_child: + log_open(); + log_error_errno(r, "Failed to chown socket at step %s: %m", exit_status_to_string(ret, EXIT_STATUS_SYSTEMD)); + + _exit(ret); + } + + r = unit_watch_pid(UNIT(s), pid); + if (r < 0) + goto fail; + + *_pid = pid; + return 0; + +fail: + s->timer_event_source = sd_event_source_unref(s->timer_event_source); + return r; +} + +static void socket_enter_dead(Socket *s, SocketResult f) { + assert(s); + + if (f != SOCKET_SUCCESS) + s->result = f; + + exec_runtime_destroy(s->exec_runtime); + s->exec_runtime = exec_runtime_unref(s->exec_runtime); + + exec_context_destroy_runtime_directory(&s->exec_context, manager_get_runtime_prefix(UNIT(s)->manager)); + + socket_set_state(s, s->result != SOCKET_SUCCESS ? SOCKET_FAILED : SOCKET_DEAD); +} + +static void socket_enter_signal(Socket *s, SocketState state, SocketResult f); + +static void socket_enter_stop_post(Socket *s, SocketResult f) { + int r; + assert(s); + + if (f != SOCKET_SUCCESS) + s->result = f; + + socket_unwatch_control_pid(s); + s->control_command_id = SOCKET_EXEC_STOP_POST; + s->control_command = s->exec_command[SOCKET_EXEC_STOP_POST]; + + if (s->control_command) { + r = socket_spawn(s, s->control_command, &s->control_pid); + if (r < 0) + goto fail; + + socket_set_state(s, SOCKET_STOP_POST); + } else + socket_enter_signal(s, SOCKET_FINAL_SIGTERM, SOCKET_SUCCESS); + + return; + +fail: + log_unit_warning_errno(UNIT(s), r, "Failed to run 'stop-post' task: %m"); + socket_enter_signal(s, SOCKET_FINAL_SIGTERM, SOCKET_FAILURE_RESOURCES); +} + +static void socket_enter_signal(Socket *s, SocketState state, SocketResult f) { + int r; + + assert(s); + + if (f != SOCKET_SUCCESS) + s->result = f; + + r = unit_kill_context( + UNIT(s), + &s->kill_context, + (state != SOCKET_STOP_PRE_SIGTERM && state != SOCKET_FINAL_SIGTERM) ? + KILL_KILL : KILL_TERMINATE, + -1, + s->control_pid, + false); + if (r < 0) + goto fail; + + if (r > 0) { + r = socket_arm_timer(s, usec_add(now(CLOCK_MONOTONIC), s->timeout_usec)); + if (r < 0) + goto fail; + + socket_set_state(s, state); + } else if (state == SOCKET_STOP_PRE_SIGTERM) + socket_enter_signal(s, SOCKET_STOP_PRE_SIGKILL, SOCKET_SUCCESS); + else if (state == SOCKET_STOP_PRE_SIGKILL) + socket_enter_stop_post(s, SOCKET_SUCCESS); + else if (state == SOCKET_FINAL_SIGTERM) + socket_enter_signal(s, SOCKET_FINAL_SIGKILL, SOCKET_SUCCESS); + else + socket_enter_dead(s, SOCKET_SUCCESS); + + return; + +fail: + log_unit_warning_errno(UNIT(s), r, "Failed to kill processes: %m"); + + if (state == SOCKET_STOP_PRE_SIGTERM || state == SOCKET_STOP_PRE_SIGKILL) + socket_enter_stop_post(s, SOCKET_FAILURE_RESOURCES); + else + socket_enter_dead(s, SOCKET_FAILURE_RESOURCES); +} + +static void socket_enter_stop_pre(Socket *s, SocketResult f) { + int r; + assert(s); + + if (f != SOCKET_SUCCESS) + s->result = f; + + socket_unwatch_control_pid(s); + s->control_command_id = SOCKET_EXEC_STOP_PRE; + s->control_command = s->exec_command[SOCKET_EXEC_STOP_PRE]; + + if (s->control_command) { + r = socket_spawn(s, s->control_command, &s->control_pid); + if (r < 0) + goto fail; + + socket_set_state(s, SOCKET_STOP_PRE); + } else + socket_enter_stop_post(s, SOCKET_SUCCESS); + + return; + +fail: + log_unit_warning_errno(UNIT(s), r, "Failed to run 'stop-pre' task: %m"); + socket_enter_stop_post(s, SOCKET_FAILURE_RESOURCES); +} + +static void socket_enter_listening(Socket *s) { + int r; + assert(s); + + r = socket_watch_fds(s); + if (r < 0) { + log_unit_warning_errno(UNIT(s), r, "Failed to watch sockets: %m"); + goto fail; + } + + socket_set_state(s, SOCKET_LISTENING); + return; + +fail: + socket_enter_stop_pre(s, SOCKET_FAILURE_RESOURCES); +} + +static void socket_enter_start_post(Socket *s) { + int r; + assert(s); + + socket_unwatch_control_pid(s); + s->control_command_id = SOCKET_EXEC_START_POST; + s->control_command = s->exec_command[SOCKET_EXEC_START_POST]; + + if (s->control_command) { + r = socket_spawn(s, s->control_command, &s->control_pid); + if (r < 0) { + log_unit_warning_errno(UNIT(s), r, "Failed to run 'start-post' task: %m"); + goto fail; + } + + socket_set_state(s, SOCKET_START_POST); + } else + socket_enter_listening(s); + + return; + +fail: + socket_enter_stop_pre(s, SOCKET_FAILURE_RESOURCES); +} + +static void socket_enter_start_chown(Socket *s) { + int r; + + assert(s); + + r = socket_open_fds(s); + if (r < 0) { + log_unit_warning_errno(UNIT(s), r, "Failed to listen on sockets: %m"); + goto fail; + } + + if (!isempty(s->user) || !isempty(s->group)) { + + socket_unwatch_control_pid(s); + s->control_command_id = SOCKET_EXEC_START_CHOWN; + s->control_command = NULL; + + r = socket_chown(s, &s->control_pid); + if (r < 0) { + log_unit_warning_errno(UNIT(s), r, "Failed to fork 'start-chown' task: %m"); + goto fail; + } + + socket_set_state(s, SOCKET_START_CHOWN); + } else + socket_enter_start_post(s); + + return; + +fail: + socket_enter_stop_pre(s, SOCKET_FAILURE_RESOURCES); +} + +static void socket_enter_start_pre(Socket *s) { + int r; + assert(s); + + socket_unwatch_control_pid(s); + s->control_command_id = SOCKET_EXEC_START_PRE; + s->control_command = s->exec_command[SOCKET_EXEC_START_PRE]; + + if (s->control_command) { + r = socket_spawn(s, s->control_command, &s->control_pid); + if (r < 0) { + log_unit_warning_errno(UNIT(s), r, "Failed to run 'start-pre' task: %m"); + goto fail; + } + + socket_set_state(s, SOCKET_START_PRE); + } else + socket_enter_start_chown(s); + + return; + +fail: + socket_enter_dead(s, SOCKET_FAILURE_RESOURCES); +} + +static void flush_ports(Socket *s) { + SocketPort *p; + + /* Flush all incoming traffic, regardless if actual bytes or new connections, so that this socket isn't busy + * anymore */ + + LIST_FOREACH(port, p, s->ports) { + if (p->fd < 0) + continue; + + (void) flush_accept(p->fd); + (void) flush_fd(p->fd); + } +} + +static void socket_enter_running(Socket *s, int cfd) { + _cleanup_(sd_bus_error_free) sd_bus_error error = SD_BUS_ERROR_NULL; + int r; + + /* Note that this call takes possession of the connection fd passed. It either has to assign it somewhere or + * close it. */ + + assert(s); + + /* We don't take connections anymore if we are supposed to shut down anyway */ + if (unit_stop_pending(UNIT(s))) { + + log_unit_debug(UNIT(s), "Suppressing connection request since unit stop is scheduled."); + + if (cfd >= 0) + cfd = safe_close(cfd); + else + flush_ports(s); + + return; + } + + if (!ratelimit_test(&s->trigger_limit)) { + safe_close(cfd); + log_unit_warning(UNIT(s), "Trigger limit hit, refusing further activation."); + socket_enter_stop_pre(s, SOCKET_FAILURE_TRIGGER_LIMIT_HIT); + return; + } + + if (cfd < 0) { + Iterator i; + Unit *other; + bool pending = false; + + /* If there's already a start pending don't bother to + * do anything */ + SET_FOREACH(other, UNIT(s)->dependencies[UNIT_TRIGGERS], i) + if (unit_active_or_pending(other)) { + pending = true; + break; + } + + if (!pending) { + if (!UNIT_ISSET(s->service)) { + log_unit_error(UNIT(s), "Service to activate vanished, refusing activation."); + r = -ENOENT; + goto fail; + } + + r = manager_add_job(UNIT(s)->manager, JOB_START, UNIT_DEREF(s->service), JOB_REPLACE, &error, NULL); + if (r < 0) + goto fail; + } + + socket_set_state(s, SOCKET_RUNNING); + } else { + _cleanup_free_ char *prefix = NULL, *instance = NULL, *name = NULL; + Service *service; + + if (s->n_connections >= s->max_connections) { + log_unit_warning(UNIT(s), "Too many incoming connections (%u), refusing connection attempt.", s->n_connections); + safe_close(cfd); + return; + } + + r = socket_instantiate_service(s); + if (r < 0) + goto fail; + + r = instance_from_socket(cfd, s->n_accepted, &instance); + if (r < 0) { + if (r != -ENOTCONN) + goto fail; + + /* ENOTCONN is legitimate if TCP RST was received. + * This connection is over, but the socket unit lives on. */ + log_unit_debug(UNIT(s), "Got ENOTCONN on incoming socket, assuming aborted connection attempt, ignoring."); + safe_close(cfd); + return; + } + + r = unit_name_to_prefix(UNIT(s)->id, &prefix); + if (r < 0) + goto fail; + + r = unit_name_build(prefix, instance, ".service", &name); + if (r < 0) + goto fail; + + r = unit_add_name(UNIT_DEREF(s->service), name); + if (r < 0) + goto fail; + + service = SERVICE(UNIT_DEREF(s->service)); + unit_ref_unset(&s->service); + + s->n_accepted++; + unit_choose_id(UNIT(service), name); + + r = service_set_socket_fd(service, cfd, s, s->selinux_context_from_net); + if (r < 0) + goto fail; + + cfd = -1; /* We passed ownership of the fd to the service now. Forget it here. */ + s->n_connections++; + + r = manager_add_job(UNIT(s)->manager, JOB_START, UNIT(service), JOB_REPLACE, &error, NULL); + if (r < 0) { + /* We failed to activate the new service, but it still exists. Let's make sure the service + * closes and forgets the connection fd again, immediately. */ + service_close_socket_fd(service); + goto fail; + } + + /* Notify clients about changed counters */ + unit_add_to_dbus_queue(UNIT(s)); + } + + return; + +fail: + log_unit_warning(UNIT(s), "Failed to queue service startup job (Maybe the service file is missing or not a %s unit?): %s", + cfd >= 0 ? "template" : "non-template", + bus_error_message(&error, r)); + + socket_enter_stop_pre(s, SOCKET_FAILURE_RESOURCES); + safe_close(cfd); +} + +static void socket_run_next(Socket *s) { + int r; + + assert(s); + assert(s->control_command); + assert(s->control_command->command_next); + + socket_unwatch_control_pid(s); + + s->control_command = s->control_command->command_next; + + r = socket_spawn(s, s->control_command, &s->control_pid); + if (r < 0) + goto fail; + + return; + +fail: + log_unit_warning_errno(UNIT(s), r, "Failed to run next task: %m"); + + if (s->state == SOCKET_START_POST) + socket_enter_stop_pre(s, SOCKET_FAILURE_RESOURCES); + else if (s->state == SOCKET_STOP_POST) + socket_enter_dead(s, SOCKET_FAILURE_RESOURCES); + else + socket_enter_signal(s, SOCKET_FINAL_SIGTERM, SOCKET_FAILURE_RESOURCES); +} + +static int socket_start(Unit *u) { + Socket *s = SOCKET(u); + int r; + + assert(s); + + /* We cannot fulfill this request right now, try again later + * please! */ + if (IN_SET(s->state, + SOCKET_STOP_PRE, + SOCKET_STOP_PRE_SIGKILL, + SOCKET_STOP_PRE_SIGTERM, + SOCKET_STOP_POST, + SOCKET_FINAL_SIGTERM, + SOCKET_FINAL_SIGKILL)) + return -EAGAIN; + + /* Already on it! */ + if (IN_SET(s->state, + SOCKET_START_PRE, + SOCKET_START_CHOWN, + SOCKET_START_POST)) + return 0; + + /* Cannot run this without the service being around */ + if (UNIT_ISSET(s->service)) { + Service *service; + + service = SERVICE(UNIT_DEREF(s->service)); + + if (UNIT(service)->load_state != UNIT_LOADED) { + log_unit_error(u, "Socket service %s not loaded, refusing.", UNIT(service)->id); + return -ENOENT; + } + + /* If the service is already active we cannot start the + * socket */ + if (service->state != SERVICE_DEAD && + service->state != SERVICE_FAILED && + service->state != SERVICE_AUTO_RESTART) { + log_unit_error(u, "Socket service %s already active, refusing.", UNIT(service)->id); + return -EBUSY; + } + } + + assert(s->state == SOCKET_DEAD || s->state == SOCKET_FAILED); + + r = unit_start_limit_test(u); + if (r < 0) { + socket_enter_dead(s, SOCKET_FAILURE_START_LIMIT_HIT); + return r; + } + + s->result = SOCKET_SUCCESS; + s->reset_cpu_usage = true; + + socket_enter_start_pre(s); + + return 1; +} + +static int socket_stop(Unit *u) { + Socket *s = SOCKET(u); + + assert(s); + + /* Already on it */ + if (IN_SET(s->state, + SOCKET_STOP_PRE, + SOCKET_STOP_PRE_SIGTERM, + SOCKET_STOP_PRE_SIGKILL, + SOCKET_STOP_POST, + SOCKET_FINAL_SIGTERM, + SOCKET_FINAL_SIGKILL)) + return 0; + + /* If there's already something running we go directly into + * kill mode. */ + if (IN_SET(s->state, + SOCKET_START_PRE, + SOCKET_START_CHOWN, + SOCKET_START_POST)) { + socket_enter_signal(s, SOCKET_STOP_PRE_SIGTERM, SOCKET_SUCCESS); + return -EAGAIN; + } + + assert(s->state == SOCKET_LISTENING || s->state == SOCKET_RUNNING); + + socket_enter_stop_pre(s, SOCKET_SUCCESS); + return 1; +} + +static int socket_serialize(Unit *u, FILE *f, FDSet *fds) { + Socket *s = SOCKET(u); + SocketPort *p; + int r; + + assert(u); + assert(f); + assert(fds); + + unit_serialize_item(u, f, "state", socket_state_to_string(s->state)); + unit_serialize_item(u, f, "result", socket_result_to_string(s->result)); + unit_serialize_item_format(u, f, "n-accepted", "%u", s->n_accepted); + + if (s->control_pid > 0) + unit_serialize_item_format(u, f, "control-pid", PID_FMT, s->control_pid); + + if (s->control_command_id >= 0) + unit_serialize_item(u, f, "control-command", socket_exec_command_to_string(s->control_command_id)); + + LIST_FOREACH(port, p, s->ports) { + int copy; + + if (p->fd < 0) + continue; + + copy = fdset_put_dup(fds, p->fd); + if (copy < 0) + return copy; + + if (p->type == SOCKET_SOCKET) { + _cleanup_free_ char *t = NULL; + + r = socket_address_print(&p->address, &t); + if (r < 0) + return r; + + if (socket_address_family(&p->address) == AF_NETLINK) + unit_serialize_item_format(u, f, "netlink", "%i %s", copy, t); + else + unit_serialize_item_format(u, f, "socket", "%i %i %s", copy, p->address.type, t); + + } else if (p->type == SOCKET_SPECIAL) + unit_serialize_item_format(u, f, "special", "%i %s", copy, p->path); + else if (p->type == SOCKET_MQUEUE) + unit_serialize_item_format(u, f, "mqueue", "%i %s", copy, p->path); + else if (p->type == SOCKET_USB_FUNCTION) + unit_serialize_item_format(u, f, "ffs", "%i %s", copy, p->path); + else { + assert(p->type == SOCKET_FIFO); + unit_serialize_item_format(u, f, "fifo", "%i %s", copy, p->path); + } + } + + return 0; +} + +static int socket_deserialize_item(Unit *u, const char *key, const char *value, FDSet *fds) { + Socket *s = SOCKET(u); + + assert(u); + assert(key); + assert(value); + + if (streq(key, "state")) { + SocketState state; + + state = socket_state_from_string(value); + if (state < 0) + log_unit_debug(u, "Failed to parse state value: %s", value); + else + s->deserialized_state = state; + } else if (streq(key, "result")) { + SocketResult f; + + f = socket_result_from_string(value); + if (f < 0) + log_unit_debug(u, "Failed to parse result value: %s", value); + else if (f != SOCKET_SUCCESS) + s->result = f; + + } else if (streq(key, "n-accepted")) { + unsigned k; + + if (safe_atou(value, &k) < 0) + log_unit_debug(u, "Failed to parse n-accepted value: %s", value); + else + s->n_accepted += k; + } else if (streq(key, "control-pid")) { + pid_t pid; + + if (parse_pid(value, &pid) < 0) + log_unit_debug(u, "Failed to parse control-pid value: %s", value); + else + s->control_pid = pid; + } else if (streq(key, "control-command")) { + SocketExecCommand id; + + id = socket_exec_command_from_string(value); + if (id < 0) + log_unit_debug(u, "Failed to parse exec-command value: %s", value); + else { + s->control_command_id = id; + s->control_command = s->exec_command[id]; + } + } else if (streq(key, "fifo")) { + int fd, skip = 0; + SocketPort *p; + + if (sscanf(value, "%i %n", &fd, &skip) < 1 || fd < 0 || !fdset_contains(fds, fd)) + log_unit_debug(u, "Failed to parse fifo value: %s", value); + else { + + LIST_FOREACH(port, p, s->ports) + if (p->type == SOCKET_FIFO && + path_equal_or_files_same(p->path, value+skip)) + break; + + if (p) { + safe_close(p->fd); + p->fd = fdset_remove(fds, fd); + } + } + + } else if (streq(key, "special")) { + int fd, skip = 0; + SocketPort *p; + + if (sscanf(value, "%i %n", &fd, &skip) < 1 || fd < 0 || !fdset_contains(fds, fd)) + log_unit_debug(u, "Failed to parse special value: %s", value); + else { + + LIST_FOREACH(port, p, s->ports) + if (p->type == SOCKET_SPECIAL && + path_equal_or_files_same(p->path, value+skip)) + break; + + if (p) { + safe_close(p->fd); + p->fd = fdset_remove(fds, fd); + } + } + + } else if (streq(key, "mqueue")) { + int fd, skip = 0; + SocketPort *p; + + if (sscanf(value, "%i %n", &fd, &skip) < 1 || fd < 0 || !fdset_contains(fds, fd)) + log_unit_debug(u, "Failed to parse mqueue value: %s", value); + else { + + LIST_FOREACH(port, p, s->ports) + if (p->type == SOCKET_MQUEUE && + streq(p->path, value+skip)) + break; + + if (p) { + safe_close(p->fd); + p->fd = fdset_remove(fds, fd); + } + } + + } else if (streq(key, "socket")) { + int fd, type, skip = 0; + SocketPort *p; + + if (sscanf(value, "%i %i %n", &fd, &type, &skip) < 2 || fd < 0 || type < 0 || !fdset_contains(fds, fd)) + log_unit_debug(u, "Failed to parse socket value: %s", value); + else { + + LIST_FOREACH(port, p, s->ports) + if (socket_address_is(&p->address, value+skip, type)) + break; + + if (p) { + safe_close(p->fd); + p->fd = fdset_remove(fds, fd); + } + } + + } else if (streq(key, "netlink")) { + int fd, skip = 0; + SocketPort *p; + + if (sscanf(value, "%i %n", &fd, &skip) < 1 || fd < 0 || !fdset_contains(fds, fd)) + log_unit_debug(u, "Failed to parse socket value: %s", value); + else { + + LIST_FOREACH(port, p, s->ports) + if (socket_address_is_netlink(&p->address, value+skip)) + break; + + if (p) { + safe_close(p->fd); + p->fd = fdset_remove(fds, fd); + } + } + + } else if (streq(key, "ffs")) { + int fd, skip = 0; + SocketPort *p; + + if (sscanf(value, "%i %n", &fd, &skip) < 1 || fd < 0 || !fdset_contains(fds, fd)) + log_unit_debug(u, "Failed to parse ffs value: %s", value); + else { + + LIST_FOREACH(port, p, s->ports) + if (p->type == SOCKET_USB_FUNCTION && + path_equal_or_files_same(p->path, value+skip)) + break; + + if (p) { + safe_close(p->fd); + p->fd = fdset_remove(fds, fd); + } + } + + } else + log_unit_debug(UNIT(s), "Unknown serialization key: %s", key); + + return 0; +} + +static void socket_distribute_fds(Unit *u, FDSet *fds) { + Socket *s = SOCKET(u); + SocketPort *p; + + assert(u); + + LIST_FOREACH(port, p, s->ports) { + Iterator i; + int fd; + + if (p->type != SOCKET_SOCKET) + continue; + + if (p->fd >= 0) + continue; + + FDSET_FOREACH(fd, fds, i) { + if (socket_address_matches_fd(&p->address, fd)) { + p->fd = fdset_remove(fds, fd); + s->deserialized_state = SOCKET_LISTENING; + break; + } + } + } +} + +_pure_ static UnitActiveState socket_active_state(Unit *u) { + assert(u); + + return state_translation_table[SOCKET(u)->state]; +} + +_pure_ static const char *socket_sub_state_to_string(Unit *u) { + assert(u); + + return socket_state_to_string(SOCKET(u)->state); +} + +const char* socket_port_type_to_string(SocketPort *p) { + + assert(p); + + switch (p->type) { + + case SOCKET_SOCKET: + + switch (p->address.type) { + + case SOCK_STREAM: + return "Stream"; + + case SOCK_DGRAM: + return "Datagram"; + + case SOCK_SEQPACKET: + return "SequentialPacket"; + + case SOCK_RAW: + if (socket_address_family(&p->address) == AF_NETLINK) + return "Netlink"; + + default: + return NULL; + } + + case SOCKET_SPECIAL: + return "Special"; + + case SOCKET_MQUEUE: + return "MessageQueue"; + + case SOCKET_FIFO: + return "FIFO"; + + case SOCKET_USB_FUNCTION: + return "USBFunction"; + + default: + return NULL; + } +} + +_pure_ static bool socket_check_gc(Unit *u) { + Socket *s = SOCKET(u); + + assert(u); + + return s->n_connections > 0; +} + +static int socket_dispatch_io(sd_event_source *source, int fd, uint32_t revents, void *userdata) { + SocketPort *p = userdata; + int cfd = -1; + + assert(p); + assert(fd >= 0); + + if (p->socket->state != SOCKET_LISTENING) + return 0; + + log_unit_debug(UNIT(p->socket), "Incoming traffic"); + + if (revents != EPOLLIN) { + + if (revents & EPOLLHUP) + log_unit_error(UNIT(p->socket), "Got POLLHUP on a listening socket. The service probably invoked shutdown() on it, and should better not do that."); + else + log_unit_error(UNIT(p->socket), "Got unexpected poll event (0x%x) on socket.", revents); + goto fail; + } + + if (p->socket->accept && + p->type == SOCKET_SOCKET && + socket_address_can_accept(&p->address)) { + + for (;;) { + + cfd = accept4(fd, NULL, NULL, SOCK_NONBLOCK); + if (cfd < 0) { + + if (errno == EINTR) + continue; + + log_unit_error_errno(UNIT(p->socket), errno, "Failed to accept socket: %m"); + goto fail; + } + + break; + } + + socket_apply_socket_options(p->socket, cfd); + } + + socket_enter_running(p->socket, cfd); + return 0; + +fail: + socket_enter_stop_pre(p->socket, SOCKET_FAILURE_RESOURCES); + return 0; +} + +static void socket_sigchld_event(Unit *u, pid_t pid, int code, int status) { + Socket *s = SOCKET(u); + SocketResult f; + + assert(s); + assert(pid >= 0); + + if (pid != s->control_pid) + return; + + s->control_pid = 0; + + if (is_clean_exit(code, status, NULL)) + f = SOCKET_SUCCESS; + else if (code == CLD_EXITED) + f = SOCKET_FAILURE_EXIT_CODE; + else if (code == CLD_KILLED) + f = SOCKET_FAILURE_SIGNAL; + else if (code == CLD_DUMPED) + f = SOCKET_FAILURE_CORE_DUMP; + else + assert_not_reached("Unknown sigchld code"); + + if (s->control_command) { + exec_status_exit(&s->control_command->exec_status, &s->exec_context, pid, code, status); + + if (s->control_command->ignore) + f = SOCKET_SUCCESS; + } + + log_unit_full(u, f == SOCKET_SUCCESS ? LOG_DEBUG : LOG_NOTICE, 0, + "Control process exited, code=%s status=%i", + sigchld_code_to_string(code), status); + + if (f != SOCKET_SUCCESS) + s->result = f; + + if (s->control_command && + s->control_command->command_next && + f == SOCKET_SUCCESS) { + + log_unit_debug(u, "Running next command for state %s", socket_state_to_string(s->state)); + socket_run_next(s); + } else { + s->control_command = NULL; + s->control_command_id = _SOCKET_EXEC_COMMAND_INVALID; + + /* No further commands for this step, so let's figure + * out what to do next */ + + log_unit_debug(u, "Got final SIGCHLD for state %s", socket_state_to_string(s->state)); + + switch (s->state) { + + case SOCKET_START_PRE: + if (f == SOCKET_SUCCESS) + socket_enter_start_chown(s); + else + socket_enter_signal(s, SOCKET_FINAL_SIGTERM, f); + break; + + case SOCKET_START_CHOWN: + if (f == SOCKET_SUCCESS) + socket_enter_start_post(s); + else + socket_enter_stop_pre(s, f); + break; + + case SOCKET_START_POST: + if (f == SOCKET_SUCCESS) + socket_enter_listening(s); + else + socket_enter_stop_pre(s, f); + break; + + case SOCKET_STOP_PRE: + case SOCKET_STOP_PRE_SIGTERM: + case SOCKET_STOP_PRE_SIGKILL: + socket_enter_stop_post(s, f); + break; + + case SOCKET_STOP_POST: + case SOCKET_FINAL_SIGTERM: + case SOCKET_FINAL_SIGKILL: + socket_enter_dead(s, f); + break; + + default: + assert_not_reached("Uh, control process died at wrong time."); + } + } + + /* Notify clients about changed exit status */ + unit_add_to_dbus_queue(u); +} + +static int socket_dispatch_timer(sd_event_source *source, usec_t usec, void *userdata) { + Socket *s = SOCKET(userdata); + + assert(s); + assert(s->timer_event_source == source); + + switch (s->state) { + + case SOCKET_START_PRE: + log_unit_warning(UNIT(s), "Starting timed out. Terminating."); + socket_enter_signal(s, SOCKET_FINAL_SIGTERM, SOCKET_FAILURE_TIMEOUT); + break; + + case SOCKET_START_CHOWN: + case SOCKET_START_POST: + log_unit_warning(UNIT(s), "Starting timed out. Stopping."); + socket_enter_stop_pre(s, SOCKET_FAILURE_TIMEOUT); + break; + + case SOCKET_STOP_PRE: + log_unit_warning(UNIT(s), "Stopping timed out. Terminating."); + socket_enter_signal(s, SOCKET_STOP_PRE_SIGTERM, SOCKET_FAILURE_TIMEOUT); + break; + + case SOCKET_STOP_PRE_SIGTERM: + if (s->kill_context.send_sigkill) { + log_unit_warning(UNIT(s), "Stopping timed out. Killing."); + socket_enter_signal(s, SOCKET_STOP_PRE_SIGKILL, SOCKET_FAILURE_TIMEOUT); + } else { + log_unit_warning(UNIT(s), "Stopping timed out. Skipping SIGKILL. Ignoring."); + socket_enter_stop_post(s, SOCKET_FAILURE_TIMEOUT); + } + break; + + case SOCKET_STOP_PRE_SIGKILL: + log_unit_warning(UNIT(s), "Processes still around after SIGKILL. Ignoring."); + socket_enter_stop_post(s, SOCKET_FAILURE_TIMEOUT); + break; + + case SOCKET_STOP_POST: + log_unit_warning(UNIT(s), "Stopping timed out (2). Terminating."); + socket_enter_signal(s, SOCKET_FINAL_SIGTERM, SOCKET_FAILURE_TIMEOUT); + break; + + case SOCKET_FINAL_SIGTERM: + if (s->kill_context.send_sigkill) { + log_unit_warning(UNIT(s), "Stopping timed out (2). Killing."); + socket_enter_signal(s, SOCKET_FINAL_SIGKILL, SOCKET_FAILURE_TIMEOUT); + } else { + log_unit_warning(UNIT(s), "Stopping timed out (2). Skipping SIGKILL. Ignoring."); + socket_enter_dead(s, SOCKET_FAILURE_TIMEOUT); + } + break; + + case SOCKET_FINAL_SIGKILL: + log_unit_warning(UNIT(s), "Still around after SIGKILL (2). Entering failed mode."); + socket_enter_dead(s, SOCKET_FAILURE_TIMEOUT); + break; + + default: + assert_not_reached("Timeout at wrong time."); + } + + return 0; +} + +int socket_collect_fds(Socket *s, int **fds) { + int *rfds, k = 0, n = 0; + SocketPort *p; + + assert(s); + assert(fds); + + /* Called from the service code for requesting our fds */ + + LIST_FOREACH(port, p, s->ports) { + if (p->fd >= 0) + n++; + n += p->n_auxiliary_fds; + } + + if (n <= 0) { + *fds = NULL; + return 0; + } + + rfds = new(int, n); + if (!rfds) + return -ENOMEM; + + LIST_FOREACH(port, p, s->ports) { + int i; + + if (p->fd >= 0) + rfds[k++] = p->fd; + for (i = 0; i < p->n_auxiliary_fds; ++i) + rfds[k++] = p->auxiliary_fds[i]; + } + + assert(k == n); + + *fds = rfds; + return n; +} + +static void socket_reset_failed(Unit *u) { + Socket *s = SOCKET(u); + + assert(s); + + if (s->state == SOCKET_FAILED) + socket_set_state(s, SOCKET_DEAD); + + s->result = SOCKET_SUCCESS; +} + +void socket_connection_unref(Socket *s) { + assert(s); + + /* The service is dead. Yay! + * + * This is strictly for one-instance-per-connection + * services. */ + + assert(s->n_connections > 0); + s->n_connections--; + + log_unit_debug(UNIT(s), "One connection closed, %u left.", s->n_connections); +} + +static void socket_trigger_notify(Unit *u, Unit *other) { + Socket *s = SOCKET(u); + + assert(u); + assert(other); + + /* Filter out invocations with bogus state */ + if (other->load_state != UNIT_LOADED || other->type != UNIT_SERVICE) + return; + + /* Don't propagate state changes from the service if we are already down */ + if (!IN_SET(s->state, SOCKET_RUNNING, SOCKET_LISTENING)) + return; + + /* We don't care for the service state if we are in Accept=yes mode */ + if (s->accept) + return; + + /* Propagate start limit hit state */ + if (other->start_limit_hit) { + socket_enter_stop_pre(s, SOCKET_FAILURE_SERVICE_START_LIMIT_HIT); + return; + } + + /* Don't propagate anything if there's still a job queued */ + if (other->job) + return; + + if (IN_SET(SERVICE(other)->state, + SERVICE_DEAD, SERVICE_FAILED, + SERVICE_FINAL_SIGTERM, SERVICE_FINAL_SIGKILL, + SERVICE_AUTO_RESTART)) + socket_enter_listening(s); + + if (SERVICE(other)->state == SERVICE_RUNNING) + socket_set_state(s, SOCKET_RUNNING); +} + +static int socket_kill(Unit *u, KillWho who, int signo, sd_bus_error *error) { + return unit_kill_common(u, who, signo, -1, SOCKET(u)->control_pid, error); +} + +static int socket_get_timeout(Unit *u, usec_t *timeout) { + Socket *s = SOCKET(u); + usec_t t; + int r; + + if (!s->timer_event_source) + return 0; + + r = sd_event_source_get_time(s->timer_event_source, &t); + if (r < 0) + return r; + if (t == USEC_INFINITY) + return 0; + + *timeout = t; + return 1; +} + +char *socket_fdname(Socket *s) { + assert(s); + + /* Returns the name to use for $LISTEN_NAMES. If the user + * didn't specify anything specifically, use the socket unit's + * name as fallback. */ + + if (s->fdname) + return s->fdname; + + return UNIT(s)->id; +} + +static int socket_control_pid(Unit *u) { + Socket *s = SOCKET(u); + + assert(s); + + return s->control_pid; +} + +static const char* const socket_exec_command_table[_SOCKET_EXEC_COMMAND_MAX] = { + [SOCKET_EXEC_START_PRE] = "StartPre", + [SOCKET_EXEC_START_CHOWN] = "StartChown", + [SOCKET_EXEC_START_POST] = "StartPost", + [SOCKET_EXEC_STOP_PRE] = "StopPre", + [SOCKET_EXEC_STOP_POST] = "StopPost" +}; + +DEFINE_STRING_TABLE_LOOKUP(socket_exec_command, SocketExecCommand); + +static const char* const socket_result_table[_SOCKET_RESULT_MAX] = { + [SOCKET_SUCCESS] = "success", + [SOCKET_FAILURE_RESOURCES] = "resources", + [SOCKET_FAILURE_TIMEOUT] = "timeout", + [SOCKET_FAILURE_EXIT_CODE] = "exit-code", + [SOCKET_FAILURE_SIGNAL] = "signal", + [SOCKET_FAILURE_CORE_DUMP] = "core-dump", + [SOCKET_FAILURE_START_LIMIT_HIT] = "start-limit-hit", + [SOCKET_FAILURE_TRIGGER_LIMIT_HIT] = "trigger-limit-hit", + [SOCKET_FAILURE_SERVICE_START_LIMIT_HIT] = "service-start-limit-hit" +}; + +DEFINE_STRING_TABLE_LOOKUP(socket_result, SocketResult); + +const UnitVTable socket_vtable = { + .object_size = sizeof(Socket), + .exec_context_offset = offsetof(Socket, exec_context), + .cgroup_context_offset = offsetof(Socket, cgroup_context), + .kill_context_offset = offsetof(Socket, kill_context), + .exec_runtime_offset = offsetof(Socket, exec_runtime), + + .sections = + "Unit\0" + "Socket\0" + "Install\0", + .private_section = "Socket", + + .init = socket_init, + .done = socket_done, + .load = socket_load, + + .coldplug = socket_coldplug, + + .dump = socket_dump, + + .start = socket_start, + .stop = socket_stop, + + .kill = socket_kill, + + .get_timeout = socket_get_timeout, + + .serialize = socket_serialize, + .deserialize_item = socket_deserialize_item, + .distribute_fds = socket_distribute_fds, + + .active_state = socket_active_state, + .sub_state_to_string = socket_sub_state_to_string, + + .check_gc = socket_check_gc, + + .sigchld_event = socket_sigchld_event, + + .trigger_notify = socket_trigger_notify, + + .reset_failed = socket_reset_failed, + + .control_pid = socket_control_pid, + + .bus_vtable = bus_socket_vtable, + .bus_set_property = bus_socket_set_property, + .bus_commit_properties = bus_socket_commit_properties, + + .status_message_formats = { + /*.starting_stopping = { + [0] = "Starting socket %s...", + [1] = "Stopping socket %s...", + },*/ + .finished_start_job = { + [JOB_DONE] = "Listening on %s.", + [JOB_FAILED] = "Failed to listen on %s.", + [JOB_TIMEOUT] = "Timed out starting %s.", + }, + .finished_stop_job = { + [JOB_DONE] = "Closed %s.", + [JOB_FAILED] = "Failed stopping %s.", + [JOB_TIMEOUT] = "Timed out stopping %s.", + }, + }, +}; diff --git a/src/grp-system/libcore/src/swap.c b/src/grp-system/libcore/src/swap.c new file mode 100644 index 0000000000..29f3971720 --- /dev/null +++ b/src/grp-system/libcore/src/swap.c @@ -0,0 +1,1533 @@ +/*** + This file is part of systemd. + + Copyright 2010 Lennart Poettering + + systemd is free software; you can redistribute it and/or modify it + under the terms of the GNU Lesser General Public License as published by + the Free Software Foundation; either version 2.1 of the License, or + (at your option) any later version. + + systemd is distributed in the hope that it will be useful, but + WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + Lesser General Public License for more details. + + You should have received a copy of the GNU Lesser General Public License + along with systemd; If not, see . +***/ + +#include +#include +#include +#include + +#include + +#include "core/swap.h" +#include "core/unit.h" +#include "systemd-basic/alloc-util.h" +#include "systemd-basic/escape.h" +#include "systemd-basic/exit-status.h" +#include "systemd-basic/fd-util.h" +#include "systemd-basic/formats-util.h" +#include "systemd-basic/parse-util.h" +#include "systemd-basic/path-util.h" +#include "systemd-basic/process-util.h" +#include "systemd-basic/special.h" +#include "systemd-basic/string-table.h" +#include "systemd-basic/string-util.h" +#include "systemd-basic/unit-name.h" +#include "systemd-basic/virt.h" +#include "systemd-shared/fstab-util.h" +#include "systemd-shared/udev-util.h" + +#include "dbus-swap.h" + +static const UnitActiveState state_translation_table[_SWAP_STATE_MAX] = { + [SWAP_DEAD] = UNIT_INACTIVE, + [SWAP_ACTIVATING] = UNIT_ACTIVATING, + [SWAP_ACTIVATING_DONE] = UNIT_ACTIVE, + [SWAP_ACTIVE] = UNIT_ACTIVE, + [SWAP_DEACTIVATING] = UNIT_DEACTIVATING, + [SWAP_ACTIVATING_SIGTERM] = UNIT_DEACTIVATING, + [SWAP_ACTIVATING_SIGKILL] = UNIT_DEACTIVATING, + [SWAP_DEACTIVATING_SIGTERM] = UNIT_DEACTIVATING, + [SWAP_DEACTIVATING_SIGKILL] = UNIT_DEACTIVATING, + [SWAP_FAILED] = UNIT_FAILED +}; + +static int swap_dispatch_timer(sd_event_source *source, usec_t usec, void *userdata); +static int swap_dispatch_io(sd_event_source *source, int fd, uint32_t revents, void *userdata); + +static void swap_unset_proc_swaps(Swap *s) { + assert(s); + + if (!s->from_proc_swaps) + return; + + s->parameters_proc_swaps.what = mfree(s->parameters_proc_swaps.what); + + s->from_proc_swaps = false; +} + +static int swap_set_devnode(Swap *s, const char *devnode) { + Hashmap *swaps; + Swap *first; + int r; + + assert(s); + + r = hashmap_ensure_allocated(&UNIT(s)->manager->swaps_by_devnode, &string_hash_ops); + if (r < 0) + return r; + + swaps = UNIT(s)->manager->swaps_by_devnode; + + if (s->devnode) { + first = hashmap_get(swaps, s->devnode); + + LIST_REMOVE(same_devnode, first, s); + if (first) + hashmap_replace(swaps, first->devnode, first); + else + hashmap_remove(swaps, s->devnode); + + s->devnode = mfree(s->devnode); + } + + if (devnode) { + s->devnode = strdup(devnode); + if (!s->devnode) + return -ENOMEM; + + first = hashmap_get(swaps, s->devnode); + LIST_PREPEND(same_devnode, first, s); + + return hashmap_replace(swaps, first->devnode, first); + } + + return 0; +} + +static void swap_init(Unit *u) { + Swap *s = SWAP(u); + + assert(s); + assert(UNIT(s)->load_state == UNIT_STUB); + + s->timeout_usec = u->manager->default_timeout_start_usec; + + s->exec_context.std_output = u->manager->default_std_output; + s->exec_context.std_error = u->manager->default_std_error; + + s->parameters_proc_swaps.priority = s->parameters_fragment.priority = -1; + + s->control_command_id = _SWAP_EXEC_COMMAND_INVALID; + + u->ignore_on_isolate = true; +} + +static void swap_unwatch_control_pid(Swap *s) { + assert(s); + + if (s->control_pid <= 0) + return; + + unit_unwatch_pid(UNIT(s), s->control_pid); + s->control_pid = 0; +} + +static void swap_done(Unit *u) { + Swap *s = SWAP(u); + + assert(s); + + swap_unset_proc_swaps(s); + swap_set_devnode(s, NULL); + + s->what = mfree(s->what); + s->parameters_fragment.what = mfree(s->parameters_fragment.what); + s->parameters_fragment.options = mfree(s->parameters_fragment.options); + + s->exec_runtime = exec_runtime_unref(s->exec_runtime); + exec_command_done_array(s->exec_command, _SWAP_EXEC_COMMAND_MAX); + s->control_command = NULL; + + swap_unwatch_control_pid(s); + + s->timer_event_source = sd_event_source_unref(s->timer_event_source); +} + +static int swap_arm_timer(Swap *s, usec_t usec) { + int r; + + assert(s); + + if (s->timer_event_source) { + r = sd_event_source_set_time(s->timer_event_source, usec); + if (r < 0) + return r; + + return sd_event_source_set_enabled(s->timer_event_source, SD_EVENT_ONESHOT); + } + + if (usec == USEC_INFINITY) + return 0; + + r = sd_event_add_time( + UNIT(s)->manager->event, + &s->timer_event_source, + CLOCK_MONOTONIC, + usec, 0, + swap_dispatch_timer, s); + if (r < 0) + return r; + + (void) sd_event_source_set_description(s->timer_event_source, "swap-timer"); + + return 0; +} + +static int swap_add_device_links(Swap *s) { + assert(s); + + if (!s->what) + return 0; + + if (!s->from_fragment) + return 0; + + if (is_device_path(s->what)) + return unit_add_node_link(UNIT(s), s->what, MANAGER_IS_SYSTEM(UNIT(s)->manager), UNIT_BINDS_TO); + else + /* File based swap devices need to be ordered after + * systemd-remount-fs.service, since they might need a + * writable file system. */ + return unit_add_dependency_by_name(UNIT(s), UNIT_AFTER, SPECIAL_REMOUNT_FS_SERVICE, NULL, true); +} + +static int swap_add_default_dependencies(Swap *s) { + int r; + + assert(s); + + if (!UNIT(s)->default_dependencies) + return 0; + + if (!MANAGER_IS_SYSTEM(UNIT(s)->manager)) + return 0; + + if (detect_container() > 0) + return 0; + + /* swap units generated for the swap dev links are missing the + * ordering dep against the swap target. */ + r = unit_add_dependency_by_name(UNIT(s), UNIT_BEFORE, SPECIAL_SWAP_TARGET, NULL, true); + if (r < 0) + return r; + + return unit_add_two_dependencies_by_name(UNIT(s), UNIT_BEFORE, UNIT_CONFLICTS, SPECIAL_UMOUNT_TARGET, NULL, true); +} + +static int swap_verify(Swap *s) { + _cleanup_free_ char *e = NULL; + int r; + + if (UNIT(s)->load_state != UNIT_LOADED) + return 0; + + r = unit_name_from_path(s->what, ".swap", &e); + if (r < 0) + return log_unit_error_errno(UNIT(s), r, "Failed to generate unit name from path: %m"); + + if (!unit_has_name(UNIT(s), e)) { + log_unit_error(UNIT(s), "Value of What= and unit name do not match, not loading."); + return -EINVAL; + } + + if (s->exec_context.pam_name && s->kill_context.kill_mode != KILL_CONTROL_GROUP) { + log_unit_error(UNIT(s), "Unit has PAM enabled. Kill mode must be set to 'control-group'. Refusing to load."); + return -EINVAL; + } + + return 0; +} + +static int swap_load_devnode(Swap *s) { + _cleanup_udev_device_unref_ struct udev_device *d = NULL; + struct stat st; + const char *p; + + assert(s); + + if (stat(s->what, &st) < 0 || !S_ISBLK(st.st_mode)) + return 0; + + d = udev_device_new_from_devnum(UNIT(s)->manager->udev, 'b', st.st_rdev); + if (!d) + return 0; + + p = udev_device_get_devnode(d); + if (!p) + return 0; + + return swap_set_devnode(s, p); +} + +static int swap_load(Unit *u) { + int r; + Swap *s = SWAP(u); + + assert(s); + assert(u->load_state == UNIT_STUB); + + /* Load a .swap file */ + r = unit_load_fragment_and_dropin_optional(u); + if (r < 0) + return r; + + if (u->load_state == UNIT_LOADED) { + + if (UNIT(s)->fragment_path) + s->from_fragment = true; + + if (!s->what) { + if (s->parameters_fragment.what) + s->what = strdup(s->parameters_fragment.what); + else if (s->parameters_proc_swaps.what) + s->what = strdup(s->parameters_proc_swaps.what); + else { + r = unit_name_to_path(u->id, &s->what); + if (r < 0) + return r; + } + + if (!s->what) + return -ENOMEM; + } + + path_kill_slashes(s->what); + + if (!UNIT(s)->description) { + r = unit_set_description(u, s->what); + if (r < 0) + return r; + } + + r = unit_require_mounts_for(UNIT(s), s->what); + if (r < 0) + return r; + + r = swap_add_device_links(s); + if (r < 0) + return r; + + r = swap_load_devnode(s); + if (r < 0) + return r; + + r = unit_patch_contexts(u); + if (r < 0) + return r; + + r = unit_add_exec_dependencies(u, &s->exec_context); + if (r < 0) + return r; + + r = unit_set_default_slice(u); + if (r < 0) + return r; + + r = swap_add_default_dependencies(s); + if (r < 0) + return r; + } + + return swap_verify(s); +} + +static int swap_setup_unit( + Manager *m, + const char *what, + const char *what_proc_swaps, + int priority, + bool set_flags) { + + _cleanup_free_ char *e = NULL; + bool delete = false; + Unit *u = NULL; + int r; + SwapParameters *p; + + assert(m); + assert(what); + assert(what_proc_swaps); + + r = unit_name_from_path(what, ".swap", &e); + if (r < 0) + return log_unit_error_errno(u, r, "Failed to generate unit name from path: %m"); + + u = manager_get_unit(m, e); + + if (u && + SWAP(u)->from_proc_swaps && + !path_equal(SWAP(u)->parameters_proc_swaps.what, what_proc_swaps)) { + log_error("Swap %s appeared twice with different device paths %s and %s", e, SWAP(u)->parameters_proc_swaps.what, what_proc_swaps); + return -EEXIST; + } + + if (!u) { + delete = true; + + u = unit_new(m, sizeof(Swap)); + if (!u) + return log_oom(); + + r = unit_add_name(u, e); + if (r < 0) + goto fail; + + SWAP(u)->what = strdup(what); + if (!SWAP(u)->what) { + r = -ENOMEM; + goto fail; + } + + unit_add_to_load_queue(u); + } else + delete = false; + + p = &SWAP(u)->parameters_proc_swaps; + + if (!p->what) { + p->what = strdup(what_proc_swaps); + if (!p->what) { + r = -ENOMEM; + goto fail; + } + } + + if (set_flags) { + SWAP(u)->is_active = true; + SWAP(u)->just_activated = !SWAP(u)->from_proc_swaps; + } + + SWAP(u)->from_proc_swaps = true; + + p->priority = priority; + + unit_add_to_dbus_queue(u); + return 0; + +fail: + log_unit_warning_errno(u, r, "Failed to load swap unit: %m"); + + if (delete && u) + unit_free(u); + + return r; +} + +static int swap_process_new(Manager *m, const char *device, int prio, bool set_flags) { + _cleanup_udev_device_unref_ struct udev_device *d = NULL; + struct udev_list_entry *item = NULL, *first = NULL; + const char *dn; + struct stat st; + int r; + + assert(m); + + r = swap_setup_unit(m, device, device, prio, set_flags); + if (r < 0) + return r; + + /* If this is a block device, then let's add duplicates for + * all other names of this block device */ + if (stat(device, &st) < 0 || !S_ISBLK(st.st_mode)) + return 0; + + d = udev_device_new_from_devnum(m->udev, 'b', st.st_rdev); + if (!d) + return 0; + + /* Add the main device node */ + dn = udev_device_get_devnode(d); + if (dn && !streq(dn, device)) + swap_setup_unit(m, dn, device, prio, set_flags); + + /* Add additional units for all symlinks */ + first = udev_device_get_devlinks_list_entry(d); + udev_list_entry_foreach(item, first) { + const char *p; + + /* Don't bother with the /dev/block links */ + p = udev_list_entry_get_name(item); + + if (streq(p, device)) + continue; + + if (path_startswith(p, "/dev/block/")) + continue; + + if (stat(p, &st) >= 0) + if (!S_ISBLK(st.st_mode) || + st.st_rdev != udev_device_get_devnum(d)) + continue; + + swap_setup_unit(m, p, device, prio, set_flags); + } + + return r; +} + +static void swap_set_state(Swap *s, SwapState state) { + SwapState old_state; + Swap *other; + + assert(s); + + old_state = s->state; + s->state = state; + + if (state != SWAP_ACTIVATING && + state != SWAP_ACTIVATING_SIGTERM && + state != SWAP_ACTIVATING_SIGKILL && + state != SWAP_ACTIVATING_DONE && + state != SWAP_DEACTIVATING && + state != SWAP_DEACTIVATING_SIGTERM && + state != SWAP_DEACTIVATING_SIGKILL) { + s->timer_event_source = sd_event_source_unref(s->timer_event_source); + swap_unwatch_control_pid(s); + s->control_command = NULL; + s->control_command_id = _SWAP_EXEC_COMMAND_INVALID; + } + + if (state != old_state) + log_unit_debug(UNIT(s), "Changed %s -> %s", swap_state_to_string(old_state), swap_state_to_string(state)); + + unit_notify(UNIT(s), state_translation_table[old_state], state_translation_table[state], true); + + /* If there other units for the same device node have a job + queued it might be worth checking again if it is runnable + now. This is necessary, since swap_start() refuses + operation with EAGAIN if there's already another job for + the same device node queued. */ + LIST_FOREACH_OTHERS(same_devnode, other, s) + if (UNIT(other)->job) + job_add_to_run_queue(UNIT(other)->job); +} + +static int swap_coldplug(Unit *u) { + Swap *s = SWAP(u); + SwapState new_state = SWAP_DEAD; + int r; + + assert(s); + assert(s->state == SWAP_DEAD); + + if (s->deserialized_state != s->state) + new_state = s->deserialized_state; + else if (s->from_proc_swaps) + new_state = SWAP_ACTIVE; + + if (new_state == s->state) + return 0; + + if (s->control_pid > 0 && + pid_is_unwaited(s->control_pid) && + IN_SET(new_state, + SWAP_ACTIVATING, + SWAP_ACTIVATING_SIGTERM, + SWAP_ACTIVATING_SIGKILL, + SWAP_ACTIVATING_DONE, + SWAP_DEACTIVATING, + SWAP_DEACTIVATING_SIGTERM, + SWAP_DEACTIVATING_SIGKILL)) { + + r = unit_watch_pid(UNIT(s), s->control_pid); + if (r < 0) + return r; + + r = swap_arm_timer(s, usec_add(u->state_change_timestamp.monotonic, s->timeout_usec)); + if (r < 0) + return r; + } + + swap_set_state(s, new_state); + return 0; +} + +static void swap_dump(Unit *u, FILE *f, const char *prefix) { + Swap *s = SWAP(u); + SwapParameters *p; + + assert(s); + assert(f); + + if (s->from_proc_swaps) + p = &s->parameters_proc_swaps; + else if (s->from_fragment) + p = &s->parameters_fragment; + else + p = NULL; + + fprintf(f, + "%sSwap State: %s\n" + "%sResult: %s\n" + "%sWhat: %s\n" + "%sFrom /proc/swaps: %s\n" + "%sFrom fragment: %s\n", + prefix, swap_state_to_string(s->state), + prefix, swap_result_to_string(s->result), + prefix, s->what, + prefix, yes_no(s->from_proc_swaps), + prefix, yes_no(s->from_fragment)); + + if (s->devnode) + fprintf(f, "%sDevice Node: %s\n", prefix, s->devnode); + + if (p) + fprintf(f, + "%sPriority: %i\n" + "%sOptions: %s\n", + prefix, p->priority, + prefix, strempty(p->options)); + + if (s->control_pid > 0) + fprintf(f, + "%sControl PID: "PID_FMT"\n", + prefix, s->control_pid); + + exec_context_dump(&s->exec_context, f, prefix); + kill_context_dump(&s->kill_context, f, prefix); +} + +static int swap_spawn(Swap *s, ExecCommand *c, pid_t *_pid) { + pid_t pid; + int r; + ExecParameters exec_params = { + .apply_permissions = true, + .apply_chroot = true, + .apply_tty_stdin = true, + .stdin_fd = -1, + .stdout_fd = -1, + .stderr_fd = -1, + }; + + assert(s); + assert(c); + assert(_pid); + + (void) unit_realize_cgroup(UNIT(s)); + if (s->reset_cpu_usage) { + (void) unit_reset_cpu_usage(UNIT(s)); + s->reset_cpu_usage = false; + } + + r = unit_setup_exec_runtime(UNIT(s)); + if (r < 0) + goto fail; + + r = swap_arm_timer(s, usec_add(now(CLOCK_MONOTONIC), s->timeout_usec)); + if (r < 0) + goto fail; + + exec_params.environment = UNIT(s)->manager->environment; + exec_params.confirm_spawn = UNIT(s)->manager->confirm_spawn; + exec_params.cgroup_supported = UNIT(s)->manager->cgroup_supported; + exec_params.cgroup_path = UNIT(s)->cgroup_path; + exec_params.cgroup_delegate = s->cgroup_context.delegate; + exec_params.runtime_prefix = manager_get_runtime_prefix(UNIT(s)->manager); + + r = exec_spawn(UNIT(s), + c, + &s->exec_context, + &exec_params, + s->exec_runtime, + &pid); + if (r < 0) + goto fail; + + r = unit_watch_pid(UNIT(s), pid); + if (r < 0) + /* FIXME: we need to do something here */ + goto fail; + + *_pid = pid; + + return 0; + +fail: + s->timer_event_source = sd_event_source_unref(s->timer_event_source); + return r; +} + +static void swap_enter_dead(Swap *s, SwapResult f) { + assert(s); + + if (f != SWAP_SUCCESS) + s->result = f; + + exec_runtime_destroy(s->exec_runtime); + s->exec_runtime = exec_runtime_unref(s->exec_runtime); + + exec_context_destroy_runtime_directory(&s->exec_context, manager_get_runtime_prefix(UNIT(s)->manager)); + + swap_set_state(s, s->result != SWAP_SUCCESS ? SWAP_FAILED : SWAP_DEAD); +} + +static void swap_enter_active(Swap *s, SwapResult f) { + assert(s); + + if (f != SWAP_SUCCESS) + s->result = f; + + swap_set_state(s, SWAP_ACTIVE); +} + +static void swap_enter_signal(Swap *s, SwapState state, SwapResult f) { + int r; + + assert(s); + + if (f != SWAP_SUCCESS) + s->result = f; + + r = unit_kill_context( + UNIT(s), + &s->kill_context, + (state != SWAP_ACTIVATING_SIGTERM && state != SWAP_DEACTIVATING_SIGTERM) ? + KILL_KILL : KILL_TERMINATE, + -1, + s->control_pid, + false); + if (r < 0) + goto fail; + + if (r > 0) { + r = swap_arm_timer(s, usec_add(now(CLOCK_MONOTONIC), s->timeout_usec)); + if (r < 0) + goto fail; + + swap_set_state(s, state); + } else if (state == SWAP_ACTIVATING_SIGTERM) + swap_enter_signal(s, SWAP_ACTIVATING_SIGKILL, SWAP_SUCCESS); + else if (state == SWAP_DEACTIVATING_SIGTERM) + swap_enter_signal(s, SWAP_DEACTIVATING_SIGKILL, SWAP_SUCCESS); + else + swap_enter_dead(s, SWAP_SUCCESS); + + return; + +fail: + log_unit_warning_errno(UNIT(s), r, "Failed to kill processes: %m"); + swap_enter_dead(s, SWAP_FAILURE_RESOURCES); +} + +static void swap_enter_activating(Swap *s) { + _cleanup_free_ char *opts = NULL; + int r; + + assert(s); + + s->control_command_id = SWAP_EXEC_ACTIVATE; + s->control_command = s->exec_command + SWAP_EXEC_ACTIVATE; + + if (s->from_fragment) { + int priority = -1; + + r = fstab_find_pri(s->parameters_fragment.options, &priority); + if (r < 0) + log_warning_errno(r, "Failed to parse swap priority \"%s\", ignoring: %m", s->parameters_fragment.options); + else if (r == 1 && s->parameters_fragment.priority >= 0) + log_warning("Duplicate swap priority configuration by Priority and Options fields."); + + if (r <= 0 && s->parameters_fragment.priority >= 0) { + if (s->parameters_fragment.options) + r = asprintf(&opts, "%s,pri=%i", s->parameters_fragment.options, s->parameters_fragment.priority); + else + r = asprintf(&opts, "pri=%i", s->parameters_fragment.priority); + if (r < 0) + goto fail; + } + } + + r = exec_command_set(s->control_command, "/sbin/swapon", NULL); + if (r < 0) + goto fail; + + if (s->parameters_fragment.options || opts) { + r = exec_command_append(s->control_command, "-o", + opts ? : s->parameters_fragment.options, NULL); + if (r < 0) + goto fail; + } + + r = exec_command_append(s->control_command, s->what, NULL); + if (r < 0) + goto fail; + + swap_unwatch_control_pid(s); + + r = swap_spawn(s, s->control_command, &s->control_pid); + if (r < 0) + goto fail; + + swap_set_state(s, SWAP_ACTIVATING); + + return; + +fail: + log_unit_warning_errno(UNIT(s), r, "Failed to run 'swapon' task: %m"); + swap_enter_dead(s, SWAP_FAILURE_RESOURCES); +} + +static void swap_enter_deactivating(Swap *s) { + int r; + + assert(s); + + s->control_command_id = SWAP_EXEC_DEACTIVATE; + s->control_command = s->exec_command + SWAP_EXEC_DEACTIVATE; + + r = exec_command_set(s->control_command, + "/sbin/swapoff", + s->what, + NULL); + if (r < 0) + goto fail; + + swap_unwatch_control_pid(s); + + r = swap_spawn(s, s->control_command, &s->control_pid); + if (r < 0) + goto fail; + + swap_set_state(s, SWAP_DEACTIVATING); + + return; + +fail: + log_unit_warning_errno(UNIT(s), r, "Failed to run 'swapoff' task: %m"); + swap_enter_active(s, SWAP_FAILURE_RESOURCES); +} + +static int swap_start(Unit *u) { + Swap *s = SWAP(u), *other; + int r; + + assert(s); + + /* We cannot fulfill this request right now, try again later + * please! */ + + if (s->state == SWAP_DEACTIVATING || + s->state == SWAP_DEACTIVATING_SIGTERM || + s->state == SWAP_DEACTIVATING_SIGKILL || + s->state == SWAP_ACTIVATING_SIGTERM || + s->state == SWAP_ACTIVATING_SIGKILL) + return -EAGAIN; + + if (s->state == SWAP_ACTIVATING) + return 0; + + assert(s->state == SWAP_DEAD || s->state == SWAP_FAILED); + + if (detect_container() > 0) + return -EPERM; + + /* If there's a job for another swap unit for the same node + * running, then let's not dispatch this one for now, and wait + * until that other job has finished. */ + LIST_FOREACH_OTHERS(same_devnode, other, s) + if (UNIT(other)->job && UNIT(other)->job->state == JOB_RUNNING) + return -EAGAIN; + + r = unit_start_limit_test(u); + if (r < 0) { + swap_enter_dead(s, SWAP_FAILURE_START_LIMIT_HIT); + return r; + } + + s->result = SWAP_SUCCESS; + s->reset_cpu_usage = true; + + swap_enter_activating(s); + return 1; +} + +static int swap_stop(Unit *u) { + Swap *s = SWAP(u); + + assert(s); + + if (s->state == SWAP_DEACTIVATING || + s->state == SWAP_DEACTIVATING_SIGTERM || + s->state == SWAP_DEACTIVATING_SIGKILL || + s->state == SWAP_ACTIVATING_SIGTERM || + s->state == SWAP_ACTIVATING_SIGKILL) + return 0; + + assert(s->state == SWAP_ACTIVATING || + s->state == SWAP_ACTIVATING_DONE || + s->state == SWAP_ACTIVE); + + if (detect_container() > 0) + return -EPERM; + + swap_enter_deactivating(s); + return 1; +} + +static int swap_serialize(Unit *u, FILE *f, FDSet *fds) { + Swap *s = SWAP(u); + + assert(s); + assert(f); + assert(fds); + + unit_serialize_item(u, f, "state", swap_state_to_string(s->state)); + unit_serialize_item(u, f, "result", swap_result_to_string(s->result)); + + if (s->control_pid > 0) + unit_serialize_item_format(u, f, "control-pid", PID_FMT, s->control_pid); + + if (s->control_command_id >= 0) + unit_serialize_item(u, f, "control-command", swap_exec_command_to_string(s->control_command_id)); + + return 0; +} + +static int swap_deserialize_item(Unit *u, const char *key, const char *value, FDSet *fds) { + Swap *s = SWAP(u); + + assert(s); + assert(fds); + + if (streq(key, "state")) { + SwapState state; + + state = swap_state_from_string(value); + if (state < 0) + log_unit_debug(u, "Failed to parse state value: %s", value); + else + s->deserialized_state = state; + } else if (streq(key, "result")) { + SwapResult f; + + f = swap_result_from_string(value); + if (f < 0) + log_unit_debug(u, "Failed to parse result value: %s", value); + else if (f != SWAP_SUCCESS) + s->result = f; + } else if (streq(key, "control-pid")) { + pid_t pid; + + if (parse_pid(value, &pid) < 0) + log_unit_debug(u, "Failed to parse control-pid value: %s", value); + else + s->control_pid = pid; + + } else if (streq(key, "control-command")) { + SwapExecCommand id; + + id = swap_exec_command_from_string(value); + if (id < 0) + log_unit_debug(u, "Failed to parse exec-command value: %s", value); + else { + s->control_command_id = id; + s->control_command = s->exec_command + id; + } + } else + log_unit_debug(u, "Unknown serialization key: %s", key); + + return 0; +} + +_pure_ static UnitActiveState swap_active_state(Unit *u) { + assert(u); + + return state_translation_table[SWAP(u)->state]; +} + +_pure_ static const char *swap_sub_state_to_string(Unit *u) { + assert(u); + + return swap_state_to_string(SWAP(u)->state); +} + +_pure_ static bool swap_check_gc(Unit *u) { + Swap *s = SWAP(u); + + assert(s); + + return s->from_proc_swaps; +} + +static void swap_sigchld_event(Unit *u, pid_t pid, int code, int status) { + Swap *s = SWAP(u); + SwapResult f; + + assert(s); + assert(pid >= 0); + + if (pid != s->control_pid) + return; + + s->control_pid = 0; + + if (is_clean_exit(code, status, NULL)) + f = SWAP_SUCCESS; + else if (code == CLD_EXITED) + f = SWAP_FAILURE_EXIT_CODE; + else if (code == CLD_KILLED) + f = SWAP_FAILURE_SIGNAL; + else if (code == CLD_DUMPED) + f = SWAP_FAILURE_CORE_DUMP; + else + assert_not_reached("Unknown code"); + + if (f != SWAP_SUCCESS) + s->result = f; + + if (s->control_command) { + exec_status_exit(&s->control_command->exec_status, &s->exec_context, pid, code, status); + + s->control_command = NULL; + s->control_command_id = _SWAP_EXEC_COMMAND_INVALID; + } + + log_unit_full(u, f == SWAP_SUCCESS ? LOG_DEBUG : LOG_NOTICE, 0, + "Swap process exited, code=%s status=%i", sigchld_code_to_string(code), status); + + switch (s->state) { + + case SWAP_ACTIVATING: + case SWAP_ACTIVATING_DONE: + case SWAP_ACTIVATING_SIGTERM: + case SWAP_ACTIVATING_SIGKILL: + + if (f == SWAP_SUCCESS) + swap_enter_active(s, f); + else + swap_enter_dead(s, f); + break; + + case SWAP_DEACTIVATING: + case SWAP_DEACTIVATING_SIGKILL: + case SWAP_DEACTIVATING_SIGTERM: + + swap_enter_dead(s, f); + break; + + default: + assert_not_reached("Uh, control process died at wrong time."); + } + + /* Notify clients about changed exit status */ + unit_add_to_dbus_queue(u); +} + +static int swap_dispatch_timer(sd_event_source *source, usec_t usec, void *userdata) { + Swap *s = SWAP(userdata); + + assert(s); + assert(s->timer_event_source == source); + + switch (s->state) { + + case SWAP_ACTIVATING: + case SWAP_ACTIVATING_DONE: + log_unit_warning(UNIT(s), "Activation timed out. Stopping."); + swap_enter_signal(s, SWAP_ACTIVATING_SIGTERM, SWAP_FAILURE_TIMEOUT); + break; + + case SWAP_DEACTIVATING: + log_unit_warning(UNIT(s), "Deactivation timed out. Stopping."); + swap_enter_signal(s, SWAP_DEACTIVATING_SIGTERM, SWAP_FAILURE_TIMEOUT); + break; + + case SWAP_ACTIVATING_SIGTERM: + if (s->kill_context.send_sigkill) { + log_unit_warning(UNIT(s), "Activation timed out. Killing."); + swap_enter_signal(s, SWAP_ACTIVATING_SIGKILL, SWAP_FAILURE_TIMEOUT); + } else { + log_unit_warning(UNIT(s), "Activation timed out. Skipping SIGKILL. Ignoring."); + swap_enter_dead(s, SWAP_FAILURE_TIMEOUT); + } + break; + + case SWAP_DEACTIVATING_SIGTERM: + if (s->kill_context.send_sigkill) { + log_unit_warning(UNIT(s), "Deactivation timed out. Killing."); + swap_enter_signal(s, SWAP_DEACTIVATING_SIGKILL, SWAP_FAILURE_TIMEOUT); + } else { + log_unit_warning(UNIT(s), "Deactivation timed out. Skipping SIGKILL. Ignoring."); + swap_enter_dead(s, SWAP_FAILURE_TIMEOUT); + } + break; + + case SWAP_ACTIVATING_SIGKILL: + case SWAP_DEACTIVATING_SIGKILL: + log_unit_warning(UNIT(s), "Swap process still around after SIGKILL. Ignoring."); + swap_enter_dead(s, SWAP_FAILURE_TIMEOUT); + break; + + default: + assert_not_reached("Timeout at wrong time."); + } + + return 0; +} + +static int swap_load_proc_swaps(Manager *m, bool set_flags) { + unsigned i; + int r = 0; + + assert(m); + + rewind(m->proc_swaps); + + (void) fscanf(m->proc_swaps, "%*s %*s %*s %*s %*s\n"); + + for (i = 1;; i++) { + _cleanup_free_ char *dev = NULL, *d = NULL; + int prio = 0, k; + + k = fscanf(m->proc_swaps, + "%ms " /* device/file */ + "%*s " /* type of swap */ + "%*s " /* swap size */ + "%*s " /* used */ + "%i\n", /* priority */ + &dev, &prio); + if (k != 2) { + if (k == EOF) + break; + + log_warning("Failed to parse /proc/swaps:%u.", i); + continue; + } + + if (cunescape(dev, UNESCAPE_RELAX, &d) < 0) + return log_oom(); + + device_found_node(m, d, true, DEVICE_FOUND_SWAP, set_flags); + + k = swap_process_new(m, d, prio, set_flags); + if (k < 0) + r = k; + } + + return r; +} + +static int swap_dispatch_io(sd_event_source *source, int fd, uint32_t revents, void *userdata) { + Manager *m = userdata; + Unit *u; + int r; + + assert(m); + assert(revents & EPOLLPRI); + + r = swap_load_proc_swaps(m, true); + if (r < 0) { + log_error_errno(r, "Failed to reread /proc/swaps: %m"); + + /* Reset flags, just in case, for late calls */ + LIST_FOREACH(units_by_type, u, m->units_by_type[UNIT_SWAP]) { + Swap *swap = SWAP(u); + + swap->is_active = swap->just_activated = false; + } + + return 0; + } + + manager_dispatch_load_queue(m); + + LIST_FOREACH(units_by_type, u, m->units_by_type[UNIT_SWAP]) { + Swap *swap = SWAP(u); + + if (!swap->is_active) { + /* This has just been deactivated */ + + swap_unset_proc_swaps(swap); + + switch (swap->state) { + + case SWAP_ACTIVE: + swap_enter_dead(swap, SWAP_SUCCESS); + break; + + default: + /* Fire again */ + swap_set_state(swap, swap->state); + break; + } + + if (swap->what) + device_found_node(m, swap->what, false, DEVICE_FOUND_SWAP, true); + + } else if (swap->just_activated) { + + /* New swap entry */ + + switch (swap->state) { + + case SWAP_DEAD: + case SWAP_FAILED: + swap_enter_active(swap, SWAP_SUCCESS); + break; + + case SWAP_ACTIVATING: + swap_set_state(swap, SWAP_ACTIVATING_DONE); + break; + + default: + /* Nothing really changed, but let's + * issue an notification call + * nonetheless, in case somebody is + * waiting for this. */ + swap_set_state(swap, swap->state); + break; + } + } + + /* Reset the flags for later calls */ + swap->is_active = swap->just_activated = false; + } + + return 1; +} + +static Unit *swap_following(Unit *u) { + Swap *s = SWAP(u); + Swap *other, *first = NULL; + + assert(s); + + /* If the user configured the swap through /etc/fstab or + * a device unit, follow that. */ + + if (s->from_fragment) + return NULL; + + LIST_FOREACH_OTHERS(same_devnode, other, s) + if (other->from_fragment) + return UNIT(other); + + /* Otherwise, make everybody follow the unit that's named after + * the swap device in the kernel */ + + if (streq_ptr(s->what, s->devnode)) + return NULL; + + LIST_FOREACH_AFTER(same_devnode, other, s) + if (streq_ptr(other->what, other->devnode)) + return UNIT(other); + + LIST_FOREACH_BEFORE(same_devnode, other, s) { + if (streq_ptr(other->what, other->devnode)) + return UNIT(other); + + first = other; + } + + /* Fall back to the first on the list */ + return UNIT(first); +} + +static int swap_following_set(Unit *u, Set **_set) { + Swap *s = SWAP(u), *other; + Set *set; + int r; + + assert(s); + assert(_set); + + if (LIST_JUST_US(same_devnode, s)) { + *_set = NULL; + return 0; + } + + set = set_new(NULL); + if (!set) + return -ENOMEM; + + LIST_FOREACH_OTHERS(same_devnode, other, s) { + r = set_put(set, other); + if (r < 0) + goto fail; + } + + *_set = set; + return 1; + +fail: + set_free(set); + return r; +} + +static void swap_shutdown(Manager *m) { + assert(m); + + m->swap_event_source = sd_event_source_unref(m->swap_event_source); + + m->proc_swaps = safe_fclose(m->proc_swaps); + + m->swaps_by_devnode = hashmap_free(m->swaps_by_devnode); +} + +static void swap_enumerate(Manager *m) { + int r; + + assert(m); + + if (!m->proc_swaps) { + m->proc_swaps = fopen("/proc/swaps", "re"); + if (!m->proc_swaps) { + if (errno == ENOENT) + log_debug("Not swap enabled, skipping enumeration"); + else + log_error_errno(errno, "Failed to open /proc/swaps: %m"); + + return; + } + + r = sd_event_add_io(m->event, &m->swap_event_source, fileno(m->proc_swaps), EPOLLPRI, swap_dispatch_io, m); + if (r < 0) { + log_error_errno(r, "Failed to watch /proc/swaps: %m"); + goto fail; + } + + /* Dispatch this before we dispatch SIGCHLD, so that + * we always get the events from /proc/swaps before + * the SIGCHLD of /sbin/swapon. */ + r = sd_event_source_set_priority(m->swap_event_source, -10); + if (r < 0) { + log_error_errno(r, "Failed to change /proc/swaps priority: %m"); + goto fail; + } + + (void) sd_event_source_set_description(m->swap_event_source, "swap-proc"); + } + + r = swap_load_proc_swaps(m, false); + if (r < 0) + goto fail; + + return; + +fail: + swap_shutdown(m); +} + +int swap_process_device_new(Manager *m, struct udev_device *dev) { + struct udev_list_entry *item = NULL, *first = NULL; + _cleanup_free_ char *e = NULL; + const char *dn; + Swap *s; + int r = 0; + + assert(m); + assert(dev); + + dn = udev_device_get_devnode(dev); + if (!dn) + return 0; + + r = unit_name_from_path(dn, ".swap", &e); + if (r < 0) + return r; + + s = hashmap_get(m->units, e); + if (s) + r = swap_set_devnode(s, dn); + + first = udev_device_get_devlinks_list_entry(dev); + udev_list_entry_foreach(item, first) { + _cleanup_free_ char *n = NULL; + int q; + + q = unit_name_from_path(udev_list_entry_get_name(item), ".swap", &n); + if (q < 0) + return q; + + s = hashmap_get(m->units, n); + if (s) { + q = swap_set_devnode(s, dn); + if (q < 0) + r = q; + } + } + + return r; +} + +int swap_process_device_remove(Manager *m, struct udev_device *dev) { + const char *dn; + int r = 0; + Swap *s; + + dn = udev_device_get_devnode(dev); + if (!dn) + return 0; + + while ((s = hashmap_get(m->swaps_by_devnode, dn))) { + int q; + + q = swap_set_devnode(s, NULL); + if (q < 0) + r = q; + } + + return r; +} + +static void swap_reset_failed(Unit *u) { + Swap *s = SWAP(u); + + assert(s); + + if (s->state == SWAP_FAILED) + swap_set_state(s, SWAP_DEAD); + + s->result = SWAP_SUCCESS; +} + +static int swap_kill(Unit *u, KillWho who, int signo, sd_bus_error *error) { + return unit_kill_common(u, who, signo, -1, SWAP(u)->control_pid, error); +} + +static int swap_get_timeout(Unit *u, usec_t *timeout) { + Swap *s = SWAP(u); + usec_t t; + int r; + + if (!s->timer_event_source) + return 0; + + r = sd_event_source_get_time(s->timer_event_source, &t); + if (r < 0) + return r; + if (t == USEC_INFINITY) + return 0; + + *timeout = t; + return 1; +} + +static bool swap_supported(void) { + static int supported = -1; + + /* If swap support is not available in the kernel, or we are + * running in a container we don't support swap units, and any + * attempts to starting one should fail immediately. */ + + if (supported < 0) + supported = + access("/proc/swaps", F_OK) >= 0 && + detect_container() <= 0; + + return supported; +} + +static int swap_control_pid(Unit *u) { + Swap *s = SWAP(u); + + assert(s); + + return s->control_pid; +} + +static const char* const swap_exec_command_table[_SWAP_EXEC_COMMAND_MAX] = { + [SWAP_EXEC_ACTIVATE] = "ExecActivate", + [SWAP_EXEC_DEACTIVATE] = "ExecDeactivate", +}; + +DEFINE_STRING_TABLE_LOOKUP(swap_exec_command, SwapExecCommand); + +static const char* const swap_result_table[_SWAP_RESULT_MAX] = { + [SWAP_SUCCESS] = "success", + [SWAP_FAILURE_RESOURCES] = "resources", + [SWAP_FAILURE_TIMEOUT] = "timeout", + [SWAP_FAILURE_EXIT_CODE] = "exit-code", + [SWAP_FAILURE_SIGNAL] = "signal", + [SWAP_FAILURE_CORE_DUMP] = "core-dump", + [SWAP_FAILURE_START_LIMIT_HIT] = "start-limit-hit", +}; + +DEFINE_STRING_TABLE_LOOKUP(swap_result, SwapResult); + +const UnitVTable swap_vtable = { + .object_size = sizeof(Swap), + .exec_context_offset = offsetof(Swap, exec_context), + .cgroup_context_offset = offsetof(Swap, cgroup_context), + .kill_context_offset = offsetof(Swap, kill_context), + .exec_runtime_offset = offsetof(Swap, exec_runtime), + + .sections = + "Unit\0" + "Swap\0" + "Install\0", + .private_section = "Swap", + + .init = swap_init, + .load = swap_load, + .done = swap_done, + + .coldplug = swap_coldplug, + + .dump = swap_dump, + + .start = swap_start, + .stop = swap_stop, + + .kill = swap_kill, + + .get_timeout = swap_get_timeout, + + .serialize = swap_serialize, + .deserialize_item = swap_deserialize_item, + + .active_state = swap_active_state, + .sub_state_to_string = swap_sub_state_to_string, + + .check_gc = swap_check_gc, + + .sigchld_event = swap_sigchld_event, + + .reset_failed = swap_reset_failed, + + .control_pid = swap_control_pid, + + .bus_vtable = bus_swap_vtable, + .bus_set_property = bus_swap_set_property, + .bus_commit_properties = bus_swap_commit_properties, + + .following = swap_following, + .following_set = swap_following_set, + + .enumerate = swap_enumerate, + .shutdown = swap_shutdown, + .supported = swap_supported, + + .status_message_formats = { + .starting_stopping = { + [0] = "Activating swap %s...", + [1] = "Deactivating swap %s...", + }, + .finished_start_job = { + [JOB_DONE] = "Activated swap %s.", + [JOB_FAILED] = "Failed to activate swap %s.", + [JOB_TIMEOUT] = "Timed out activating swap %s.", + }, + .finished_stop_job = { + [JOB_DONE] = "Deactivated swap %s.", + [JOB_FAILED] = "Failed deactivating swap %s.", + [JOB_TIMEOUT] = "Timed out deactivating swap %s.", + }, + }, +}; diff --git a/src/grp-system/libcore/src/target.c b/src/grp-system/libcore/src/target.c new file mode 100644 index 0000000000..991dbccb10 --- /dev/null +++ b/src/grp-system/libcore/src/target.c @@ -0,0 +1,224 @@ +/*** + This file is part of systemd. + + Copyright 2010 Lennart Poettering + + systemd is free software; you can redistribute it and/or modify it + under the terms of the GNU Lesser General Public License as published by + the Free Software Foundation; either version 2.1 of the License, or + (at your option) any later version. + + systemd is distributed in the hope that it will be useful, but + WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + Lesser General Public License for more details. + + You should have received a copy of the GNU Lesser General Public License + along with systemd; If not, see . +***/ + +#include "core/target.h" +#include "core/unit.h" +#include "systemd-basic/log.h" +#include "systemd-basic/special.h" +#include "systemd-basic/string-util.h" +#include "systemd-basic/unit-name.h" + +#include "dbus-target.h" + +static const UnitActiveState state_translation_table[_TARGET_STATE_MAX] = { + [TARGET_DEAD] = UNIT_INACTIVE, + [TARGET_ACTIVE] = UNIT_ACTIVE +}; + +static void target_set_state(Target *t, TargetState state) { + TargetState old_state; + assert(t); + + old_state = t->state; + t->state = state; + + if (state != old_state) + log_debug("%s changed %s -> %s", + UNIT(t)->id, + target_state_to_string(old_state), + target_state_to_string(state)); + + unit_notify(UNIT(t), state_translation_table[old_state], state_translation_table[state], true); +} + +static int target_add_default_dependencies(Target *t) { + + static const UnitDependency deps[] = { + UNIT_REQUIRES, + UNIT_REQUISITE, + UNIT_WANTS, + UNIT_BINDS_TO, + UNIT_PART_OF + }; + + Iterator i; + Unit *other; + int r; + unsigned k; + + assert(t); + + /* Imply ordering for requirement dependencies on target + * units. Note that when the user created a contradicting + * ordering manually we won't add anything in here to make + * sure we don't create a loop. */ + + for (k = 0; k < ELEMENTSOF(deps); k++) + SET_FOREACH(other, UNIT(t)->dependencies[deps[k]], i) { + r = unit_add_default_target_dependency(other, UNIT(t)); + if (r < 0) + return r; + } + + /* Make sure targets are unloaded on shutdown */ + return unit_add_dependency_by_name(UNIT(t), UNIT_CONFLICTS, SPECIAL_SHUTDOWN_TARGET, NULL, true); +} + +static int target_load(Unit *u) { + Target *t = TARGET(u); + int r; + + assert(t); + + r = unit_load_fragment_and_dropin(u); + if (r < 0) + return r; + + /* This is a new unit? Then let's add in some extras */ + if (u->load_state == UNIT_LOADED && u->default_dependencies) { + r = target_add_default_dependencies(t); + if (r < 0) + return r; + } + + return 0; +} + +static int target_coldplug(Unit *u) { + Target *t = TARGET(u); + + assert(t); + assert(t->state == TARGET_DEAD); + + if (t->deserialized_state != t->state) + target_set_state(t, t->deserialized_state); + + return 0; +} + +static void target_dump(Unit *u, FILE *f, const char *prefix) { + Target *t = TARGET(u); + + assert(t); + assert(f); + + fprintf(f, + "%sTarget State: %s\n", + prefix, target_state_to_string(t->state)); +} + +static int target_start(Unit *u) { + Target *t = TARGET(u); + + assert(t); + assert(t->state == TARGET_DEAD); + + target_set_state(t, TARGET_ACTIVE); + return 1; +} + +static int target_stop(Unit *u) { + Target *t = TARGET(u); + + assert(t); + assert(t->state == TARGET_ACTIVE); + + target_set_state(t, TARGET_DEAD); + return 1; +} + +static int target_serialize(Unit *u, FILE *f, FDSet *fds) { + Target *s = TARGET(u); + + assert(s); + assert(f); + assert(fds); + + unit_serialize_item(u, f, "state", target_state_to_string(s->state)); + return 0; +} + +static int target_deserialize_item(Unit *u, const char *key, const char *value, FDSet *fds) { + Target *s = TARGET(u); + + assert(u); + assert(key); + assert(value); + assert(fds); + + if (streq(key, "state")) { + TargetState state; + + state = target_state_from_string(value); + if (state < 0) + log_debug("Failed to parse state value %s", value); + else + s->deserialized_state = state; + + } else + log_debug("Unknown serialization key '%s'", key); + + return 0; +} + +_pure_ static UnitActiveState target_active_state(Unit *u) { + assert(u); + + return state_translation_table[TARGET(u)->state]; +} + +_pure_ static const char *target_sub_state_to_string(Unit *u) { + assert(u); + + return target_state_to_string(TARGET(u)->state); +} + +const UnitVTable target_vtable = { + .object_size = sizeof(Target), + + .sections = + "Unit\0" + "Target\0" + "Install\0", + + .load = target_load, + .coldplug = target_coldplug, + + .dump = target_dump, + + .start = target_start, + .stop = target_stop, + + .serialize = target_serialize, + .deserialize_item = target_deserialize_item, + + .active_state = target_active_state, + .sub_state_to_string = target_sub_state_to_string, + + .bus_vtable = bus_target_vtable, + + .status_message_formats = { + .finished_start_job = { + [JOB_DONE] = "Reached target %s.", + }, + .finished_stop_job = { + [JOB_DONE] = "Stopped target %s.", + }, + }, +}; diff --git a/src/grp-system/libcore/src/timer.c b/src/grp-system/libcore/src/timer.c new file mode 100644 index 0000000000..882d871dc9 --- /dev/null +++ b/src/grp-system/libcore/src/timer.c @@ -0,0 +1,860 @@ +/*** + This file is part of systemd. + + Copyright 2010 Lennart Poettering + + systemd is free software; you can redistribute it and/or modify it + under the terms of the GNU Lesser General Public License as published by + the Free Software Foundation; either version 2.1 of the License, or + (at your option) any later version. + + systemd is distributed in the hope that it will be useful, but + WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + Lesser General Public License for more details. + + You should have received a copy of the GNU Lesser General Public License + along with systemd; If not, see . +***/ + +#include + +#include "core/timer.h" +#include "core/unit.h" +#include "sd-bus/bus-error.h" +#include "sd-bus/bus-util.h" +#include "systemd-basic/alloc-util.h" +#include "systemd-basic/fs-util.h" +#include "systemd-basic/parse-util.h" +#include "systemd-basic/random-util.h" +#include "systemd-basic/special.h" +#include "systemd-basic/string-table.h" +#include "systemd-basic/string-util.h" +#include "systemd-basic/unit-name.h" +#include "systemd-basic/user-util.h" +#include "systemd-basic/virt.h" + +#include "dbus-timer.h" + +static const UnitActiveState state_translation_table[_TIMER_STATE_MAX] = { + [TIMER_DEAD] = UNIT_INACTIVE, + [TIMER_WAITING] = UNIT_ACTIVE, + [TIMER_RUNNING] = UNIT_ACTIVE, + [TIMER_ELAPSED] = UNIT_ACTIVE, + [TIMER_FAILED] = UNIT_FAILED +}; + +static int timer_dispatch(sd_event_source *s, uint64_t usec, void *userdata); + +static void timer_init(Unit *u) { + Timer *t = TIMER(u); + + assert(u); + assert(u->load_state == UNIT_STUB); + + t->next_elapse_monotonic_or_boottime = USEC_INFINITY; + t->next_elapse_realtime = USEC_INFINITY; + t->accuracy_usec = u->manager->default_timer_accuracy_usec; + t->remain_after_elapse = true; +} + +void timer_free_values(Timer *t) { + TimerValue *v; + + assert(t); + + while ((v = t->values)) { + LIST_REMOVE(value, t->values, v); + calendar_spec_free(v->calendar_spec); + free(v); + } +} + +static void timer_done(Unit *u) { + Timer *t = TIMER(u); + + assert(t); + + timer_free_values(t); + + t->monotonic_event_source = sd_event_source_unref(t->monotonic_event_source); + t->realtime_event_source = sd_event_source_unref(t->realtime_event_source); + + free(t->stamp_path); +} + +static int timer_verify(Timer *t) { + assert(t); + + if (UNIT(t)->load_state != UNIT_LOADED) + return 0; + + if (!t->values) { + log_unit_error(UNIT(t), "Timer unit lacks value setting. Refusing."); + return -EINVAL; + } + + return 0; +} + +static int timer_add_default_dependencies(Timer *t) { + int r; + TimerValue *v; + + assert(t); + + if (!UNIT(t)->default_dependencies) + return 0; + + r = unit_add_dependency_by_name(UNIT(t), UNIT_BEFORE, SPECIAL_TIMERS_TARGET, NULL, true); + if (r < 0) + return r; + + if (MANAGER_IS_SYSTEM(UNIT(t)->manager)) { + r = unit_add_two_dependencies_by_name(UNIT(t), UNIT_AFTER, UNIT_REQUIRES, SPECIAL_SYSINIT_TARGET, NULL, true); + if (r < 0) + return r; + + LIST_FOREACH(value, v, t->values) { + if (v->base == TIMER_CALENDAR) { + r = unit_add_dependency_by_name(UNIT(t), UNIT_AFTER, SPECIAL_TIME_SYNC_TARGET, NULL, true); + if (r < 0) + return r; + break; + } + } + } + + return unit_add_two_dependencies_by_name(UNIT(t), UNIT_BEFORE, UNIT_CONFLICTS, SPECIAL_SHUTDOWN_TARGET, NULL, true); +} + +static int timer_setup_persistent(Timer *t) { + int r; + + assert(t); + + if (!t->persistent) + return 0; + + if (MANAGER_IS_SYSTEM(UNIT(t)->manager)) { + + r = unit_require_mounts_for(UNIT(t), "/var/lib/systemd/timers"); + if (r < 0) + return r; + + t->stamp_path = strappend("/var/lib/systemd/timers/stamp-", UNIT(t)->id); + } else { + const char *e; + + e = getenv("XDG_DATA_HOME"); + if (e) + t->stamp_path = strjoin(e, "/systemd/timers/stamp-", UNIT(t)->id, NULL); + else { + + _cleanup_free_ char *h = NULL; + + r = get_home_dir(&h); + if (r < 0) + return log_unit_error_errno(UNIT(t), r, "Failed to determine home directory: %m"); + + t->stamp_path = strjoin(h, "/.local/share/systemd/timers/stamp-", UNIT(t)->id, NULL); + } + } + + if (!t->stamp_path) + return log_oom(); + + return 0; +} + +static int timer_load(Unit *u) { + Timer *t = TIMER(u); + int r; + + assert(u); + assert(u->load_state == UNIT_STUB); + + r = unit_load_fragment_and_dropin(u); + if (r < 0) + return r; + + if (u->load_state == UNIT_LOADED) { + + if (set_isempty(u->dependencies[UNIT_TRIGGERS])) { + Unit *x; + + r = unit_load_related_unit(u, ".service", &x); + if (r < 0) + return r; + + r = unit_add_two_dependencies(u, UNIT_BEFORE, UNIT_TRIGGERS, x, true); + if (r < 0) + return r; + } + + r = timer_setup_persistent(t); + if (r < 0) + return r; + + r = timer_add_default_dependencies(t); + if (r < 0) + return r; + } + + return timer_verify(t); +} + +static void timer_dump(Unit *u, FILE *f, const char *prefix) { + char buf[FORMAT_TIMESPAN_MAX]; + Timer *t = TIMER(u); + Unit *trigger; + TimerValue *v; + + trigger = UNIT_TRIGGER(u); + + fprintf(f, + "%sTimer State: %s\n" + "%sResult: %s\n" + "%sUnit: %s\n" + "%sPersistent: %s\n" + "%sWakeSystem: %s\n" + "%sAccuracy: %s\n" + "%sRemainAfterElapse: %s\n", + prefix, timer_state_to_string(t->state), + prefix, timer_result_to_string(t->result), + prefix, trigger ? trigger->id : "n/a", + prefix, yes_no(t->persistent), + prefix, yes_no(t->wake_system), + prefix, format_timespan(buf, sizeof(buf), t->accuracy_usec, 1), + prefix, yes_no(t->remain_after_elapse)); + + LIST_FOREACH(value, v, t->values) { + + if (v->base == TIMER_CALENDAR) { + _cleanup_free_ char *p = NULL; + + calendar_spec_to_string(v->calendar_spec, &p); + + fprintf(f, + "%s%s: %s\n", + prefix, + timer_base_to_string(v->base), + strna(p)); + } else { + char timespan1[FORMAT_TIMESPAN_MAX]; + + fprintf(f, + "%s%s: %s\n", + prefix, + timer_base_to_string(v->base), + format_timespan(timespan1, sizeof(timespan1), v->value, 0)); + } + } +} + +static void timer_set_state(Timer *t, TimerState state) { + TimerState old_state; + assert(t); + + old_state = t->state; + t->state = state; + + if (state != TIMER_WAITING) { + t->monotonic_event_source = sd_event_source_unref(t->monotonic_event_source); + t->realtime_event_source = sd_event_source_unref(t->realtime_event_source); + } + + if (state != old_state) + log_unit_debug(UNIT(t), "Changed %s -> %s", timer_state_to_string(old_state), timer_state_to_string(state)); + + unit_notify(UNIT(t), state_translation_table[old_state], state_translation_table[state], true); +} + +static void timer_enter_waiting(Timer *t, bool initial); + +static int timer_coldplug(Unit *u) { + Timer *t = TIMER(u); + + assert(t); + assert(t->state == TIMER_DEAD); + + if (t->deserialized_state == t->state) + return 0; + + if (t->deserialized_state == TIMER_WAITING) + timer_enter_waiting(t, false); + else + timer_set_state(t, t->deserialized_state); + + return 0; +} + +static void timer_enter_dead(Timer *t, TimerResult f) { + assert(t); + + if (f != TIMER_SUCCESS) + t->result = f; + + timer_set_state(t, t->result != TIMER_SUCCESS ? TIMER_FAILED : TIMER_DEAD); +} + +static void timer_enter_elapsed(Timer *t, bool leave_around) { + assert(t); + + /* If a unit is marked with RemainAfterElapse=yes we leave it + * around even after it elapsed once, so that starting it + * later again does not necessarily mean immediate + * retriggering. We unconditionally leave units with + * TIMER_UNIT_ACTIVE or TIMER_UNIT_INACTIVE triggers around, + * since they might be restarted automatically at any time + * later on. */ + + if (t->remain_after_elapse || leave_around) + timer_set_state(t, TIMER_ELAPSED); + else + timer_enter_dead(t, TIMER_SUCCESS); +} + +static usec_t monotonic_to_boottime(usec_t t) { + usec_t a, b; + + if (t <= 0) + return 0; + + a = now(clock_boottime_or_monotonic()); + b = now(CLOCK_MONOTONIC); + + if (t + a > b) + return t + a - b; + else + return 0; +} + +static void add_random(Timer *t, usec_t *v) { + char s[FORMAT_TIMESPAN_MAX]; + usec_t add; + + assert(t); + assert(v); + + if (t->random_usec == 0) + return; + if (*v == USEC_INFINITY) + return; + + add = random_u64() % t->random_usec; + + if (*v + add < *v) /* overflow */ + *v = (usec_t) -2; /* Highest possible value, that is not USEC_INFINITY */ + else + *v += add; + + log_unit_info(UNIT(t), "Adding %s random time.", format_timespan(s, sizeof(s), add, 0)); +} + +static void timer_enter_waiting(Timer *t, bool initial) { + bool found_monotonic = false, found_realtime = false; + usec_t ts_realtime, ts_monotonic; + usec_t base = 0; + bool leave_around = false; + TimerValue *v; + Unit *trigger; + int r; + + assert(t); + + trigger = UNIT_TRIGGER(UNIT(t)); + if (!trigger) { + log_unit_error(UNIT(t), "Unit to trigger vanished."); + timer_enter_dead(t, TIMER_FAILURE_RESOURCES); + return; + } + + /* If we shall wake the system we use the boottime clock + * rather than the monotonic clock. */ + + ts_realtime = now(CLOCK_REALTIME); + ts_monotonic = now(t->wake_system ? clock_boottime_or_monotonic() : CLOCK_MONOTONIC); + t->next_elapse_monotonic_or_boottime = t->next_elapse_realtime = 0; + + LIST_FOREACH(value, v, t->values) { + + if (v->disabled) + continue; + + if (v->base == TIMER_CALENDAR) { + usec_t b; + + /* If we know the last time this was + * triggered, schedule the job based relative + * to that. If we don't just start from + * now. */ + + b = t->last_trigger.realtime > 0 ? t->last_trigger.realtime : ts_realtime; + + r = calendar_spec_next_usec(v->calendar_spec, b, &v->next_elapse); + if (r < 0) + continue; + + if (!found_realtime) + t->next_elapse_realtime = v->next_elapse; + else + t->next_elapse_realtime = MIN(t->next_elapse_realtime, v->next_elapse); + + found_realtime = true; + + } else { + switch (v->base) { + + case TIMER_ACTIVE: + if (state_translation_table[t->state] == UNIT_ACTIVE) + base = UNIT(t)->inactive_exit_timestamp.monotonic; + else + base = ts_monotonic; + break; + + case TIMER_BOOT: + if (detect_container() <= 0) { + /* CLOCK_MONOTONIC equals the uptime on Linux */ + base = 0; + break; + } + /* In a container we don't want to include the time the host + * was already up when the container started, so count from + * our own startup. Fall through. */ + case TIMER_STARTUP: + base = UNIT(t)->manager->userspace_timestamp.monotonic; + break; + + case TIMER_UNIT_ACTIVE: + leave_around = true; + base = trigger->inactive_exit_timestamp.monotonic; + + if (base <= 0) + base = t->last_trigger.monotonic; + + if (base <= 0) + continue; + + break; + + case TIMER_UNIT_INACTIVE: + leave_around = true; + base = trigger->inactive_enter_timestamp.monotonic; + + if (base <= 0) + base = t->last_trigger.monotonic; + + if (base <= 0) + continue; + + break; + + default: + assert_not_reached("Unknown timer base"); + } + + if (t->wake_system) + base = monotonic_to_boottime(base); + + v->next_elapse = base + v->value; + + if (!initial && v->next_elapse < ts_monotonic && IN_SET(v->base, TIMER_ACTIVE, TIMER_BOOT, TIMER_STARTUP)) { + /* This is a one time trigger, disable it now */ + v->disabled = true; + continue; + } + + if (!found_monotonic) + t->next_elapse_monotonic_or_boottime = v->next_elapse; + else + t->next_elapse_monotonic_or_boottime = MIN(t->next_elapse_monotonic_or_boottime, v->next_elapse); + + found_monotonic = true; + } + } + + if (!found_monotonic && !found_realtime) { + log_unit_debug(UNIT(t), "Timer is elapsed."); + timer_enter_elapsed(t, leave_around); + return; + } + + if (found_monotonic) { + char buf[FORMAT_TIMESPAN_MAX]; + usec_t left; + + add_random(t, &t->next_elapse_monotonic_or_boottime); + + left = t->next_elapse_monotonic_or_boottime > ts_monotonic ? t->next_elapse_monotonic_or_boottime - ts_monotonic : 0; + log_unit_debug(UNIT(t), "Monotonic timer elapses in %s.", format_timespan(buf, sizeof(buf), left, 0)); + + if (t->monotonic_event_source) { + r = sd_event_source_set_time(t->monotonic_event_source, t->next_elapse_monotonic_or_boottime); + if (r < 0) + goto fail; + + r = sd_event_source_set_enabled(t->monotonic_event_source, SD_EVENT_ONESHOT); + if (r < 0) + goto fail; + } else { + + r = sd_event_add_time( + UNIT(t)->manager->event, + &t->monotonic_event_source, + t->wake_system ? CLOCK_BOOTTIME_ALARM : CLOCK_MONOTONIC, + t->next_elapse_monotonic_or_boottime, t->accuracy_usec, + timer_dispatch, t); + if (r < 0) + goto fail; + + (void) sd_event_source_set_description(t->monotonic_event_source, "timer-monotonic"); + } + + } else if (t->monotonic_event_source) { + + r = sd_event_source_set_enabled(t->monotonic_event_source, SD_EVENT_OFF); + if (r < 0) + goto fail; + } + + if (found_realtime) { + char buf[FORMAT_TIMESTAMP_MAX]; + + add_random(t, &t->next_elapse_realtime); + + log_unit_debug(UNIT(t), "Realtime timer elapses at %s.", format_timestamp(buf, sizeof(buf), t->next_elapse_realtime)); + + if (t->realtime_event_source) { + r = sd_event_source_set_time(t->realtime_event_source, t->next_elapse_realtime); + if (r < 0) + goto fail; + + r = sd_event_source_set_enabled(t->realtime_event_source, SD_EVENT_ONESHOT); + if (r < 0) + goto fail; + } else { + r = sd_event_add_time( + UNIT(t)->manager->event, + &t->realtime_event_source, + t->wake_system ? CLOCK_REALTIME_ALARM : CLOCK_REALTIME, + t->next_elapse_realtime, t->accuracy_usec, + timer_dispatch, t); + if (r < 0) + goto fail; + + (void) sd_event_source_set_description(t->realtime_event_source, "timer-realtime"); + } + + } else if (t->realtime_event_source) { + + r = sd_event_source_set_enabled(t->realtime_event_source, SD_EVENT_OFF); + if (r < 0) + goto fail; + } + + timer_set_state(t, TIMER_WAITING); + return; + +fail: + log_unit_warning_errno(UNIT(t), r, "Failed to enter waiting state: %m"); + timer_enter_dead(t, TIMER_FAILURE_RESOURCES); +} + +static void timer_enter_running(Timer *t) { + _cleanup_(sd_bus_error_free) sd_bus_error error = SD_BUS_ERROR_NULL; + Unit *trigger; + int r; + + assert(t); + + /* Don't start job if we are supposed to go down */ + if (unit_stop_pending(UNIT(t))) + return; + + trigger = UNIT_TRIGGER(UNIT(t)); + if (!trigger) { + log_unit_error(UNIT(t), "Unit to trigger vanished."); + timer_enter_dead(t, TIMER_FAILURE_RESOURCES); + return; + } + + r = manager_add_job(UNIT(t)->manager, JOB_START, trigger, JOB_REPLACE, &error, NULL); + if (r < 0) + goto fail; + + dual_timestamp_get(&t->last_trigger); + + if (t->stamp_path) + touch_file(t->stamp_path, true, t->last_trigger.realtime, UID_INVALID, GID_INVALID, MODE_INVALID); + + timer_set_state(t, TIMER_RUNNING); + return; + +fail: + log_unit_warning(UNIT(t), "Failed to queue unit startup job: %s", bus_error_message(&error, r)); + timer_enter_dead(t, TIMER_FAILURE_RESOURCES); +} + +static int timer_start(Unit *u) { + Timer *t = TIMER(u); + TimerValue *v; + Unit *trigger; + int r; + + assert(t); + assert(t->state == TIMER_DEAD || t->state == TIMER_FAILED); + + trigger = UNIT_TRIGGER(u); + if (!trigger || trigger->load_state != UNIT_LOADED) { + log_unit_error(u, "Refusing to start, unit to trigger not loaded."); + return -ENOENT; + } + + r = unit_start_limit_test(u); + if (r < 0) { + timer_enter_dead(t, TIMER_FAILURE_START_LIMIT_HIT); + return r; + } + + t->last_trigger = DUAL_TIMESTAMP_NULL; + + /* Reenable all timers that depend on unit activation time */ + LIST_FOREACH(value, v, t->values) + if (v->base == TIMER_ACTIVE) + v->disabled = false; + + if (t->stamp_path) { + struct stat st; + + if (stat(t->stamp_path, &st) >= 0) + t->last_trigger.realtime = timespec_load(&st.st_atim); + else if (errno == ENOENT) + /* The timer has never run before, + * make sure a stamp file exists. + */ + touch_file(t->stamp_path, true, USEC_INFINITY, UID_INVALID, GID_INVALID, MODE_INVALID); + } + + t->result = TIMER_SUCCESS; + timer_enter_waiting(t, true); + return 1; +} + +static int timer_stop(Unit *u) { + Timer *t = TIMER(u); + + assert(t); + assert(t->state == TIMER_WAITING || t->state == TIMER_RUNNING || t->state == TIMER_ELAPSED); + + timer_enter_dead(t, TIMER_SUCCESS); + return 1; +} + +static int timer_serialize(Unit *u, FILE *f, FDSet *fds) { + Timer *t = TIMER(u); + + assert(u); + assert(f); + assert(fds); + + unit_serialize_item(u, f, "state", timer_state_to_string(t->state)); + unit_serialize_item(u, f, "result", timer_result_to_string(t->result)); + + if (t->last_trigger.realtime > 0) + unit_serialize_item_format(u, f, "last-trigger-realtime", "%" PRIu64, t->last_trigger.realtime); + + if (t->last_trigger.monotonic > 0) + unit_serialize_item_format(u, f, "last-trigger-monotonic", "%" PRIu64, t->last_trigger.monotonic); + + return 0; +} + +static int timer_deserialize_item(Unit *u, const char *key, const char *value, FDSet *fds) { + Timer *t = TIMER(u); + int r; + + assert(u); + assert(key); + assert(value); + assert(fds); + + if (streq(key, "state")) { + TimerState state; + + state = timer_state_from_string(value); + if (state < 0) + log_unit_debug(u, "Failed to parse state value: %s", value); + else + t->deserialized_state = state; + } else if (streq(key, "result")) { + TimerResult f; + + f = timer_result_from_string(value); + if (f < 0) + log_unit_debug(u, "Failed to parse result value: %s", value); + else if (f != TIMER_SUCCESS) + t->result = f; + } else if (streq(key, "last-trigger-realtime")) { + + r = safe_atou64(value, &t->last_trigger.realtime); + if (r < 0) + log_unit_debug(u, "Failed to parse last-trigger-realtime value: %s", value); + + } else if (streq(key, "last-trigger-monotonic")) { + + r = safe_atou64(value, &t->last_trigger.monotonic); + if (r < 0) + log_unit_debug(u, "Failed to parse last-trigger-monotonic value: %s", value); + + } else + log_unit_debug(u, "Unknown serialization key: %s", key); + + return 0; +} + +_pure_ static UnitActiveState timer_active_state(Unit *u) { + assert(u); + + return state_translation_table[TIMER(u)->state]; +} + +_pure_ static const char *timer_sub_state_to_string(Unit *u) { + assert(u); + + return timer_state_to_string(TIMER(u)->state); +} + +static int timer_dispatch(sd_event_source *s, uint64_t usec, void *userdata) { + Timer *t = TIMER(userdata); + + assert(t); + + if (t->state != TIMER_WAITING) + return 0; + + log_unit_debug(UNIT(t), "Timer elapsed."); + timer_enter_running(t); + return 0; +} + +static void timer_trigger_notify(Unit *u, Unit *other) { + Timer *t = TIMER(u); + TimerValue *v; + + assert(u); + assert(other); + + if (other->load_state != UNIT_LOADED) + return; + + /* Reenable all timers that depend on unit state */ + LIST_FOREACH(value, v, t->values) + if (v->base == TIMER_UNIT_ACTIVE || + v->base == TIMER_UNIT_INACTIVE) + v->disabled = false; + + switch (t->state) { + + case TIMER_WAITING: + case TIMER_ELAPSED: + + /* Recalculate sleep time */ + timer_enter_waiting(t, false); + break; + + case TIMER_RUNNING: + + if (UNIT_IS_INACTIVE_OR_FAILED(unit_active_state(other))) { + log_unit_debug(UNIT(t), "Got notified about unit deactivation."); + timer_enter_waiting(t, false); + } + break; + + case TIMER_DEAD: + case TIMER_FAILED: + break; + + default: + assert_not_reached("Unknown timer state"); + } +} + +static void timer_reset_failed(Unit *u) { + Timer *t = TIMER(u); + + assert(t); + + if (t->state == TIMER_FAILED) + timer_set_state(t, TIMER_DEAD); + + t->result = TIMER_SUCCESS; +} + +static void timer_time_change(Unit *u) { + Timer *t = TIMER(u); + + assert(u); + + if (t->state != TIMER_WAITING) + return; + + log_unit_debug(u, "Time change, recalculating next elapse."); + timer_enter_waiting(t, false); +} + +static const char* const timer_base_table[_TIMER_BASE_MAX] = { + [TIMER_ACTIVE] = "OnActiveSec", + [TIMER_BOOT] = "OnBootSec", + [TIMER_STARTUP] = "OnStartupSec", + [TIMER_UNIT_ACTIVE] = "OnUnitActiveSec", + [TIMER_UNIT_INACTIVE] = "OnUnitInactiveSec", + [TIMER_CALENDAR] = "OnCalendar" +}; + +DEFINE_STRING_TABLE_LOOKUP(timer_base, TimerBase); + +static const char* const timer_result_table[_TIMER_RESULT_MAX] = { + [TIMER_SUCCESS] = "success", + [TIMER_FAILURE_RESOURCES] = "resources", + [TIMER_FAILURE_START_LIMIT_HIT] = "start-limit-hit", +}; + +DEFINE_STRING_TABLE_LOOKUP(timer_result, TimerResult); + +const UnitVTable timer_vtable = { + .object_size = sizeof(Timer), + + .sections = + "Unit\0" + "Timer\0" + "Install\0", + .private_section = "Timer", + + .init = timer_init, + .done = timer_done, + .load = timer_load, + + .coldplug = timer_coldplug, + + .dump = timer_dump, + + .start = timer_start, + .stop = timer_stop, + + .serialize = timer_serialize, + .deserialize_item = timer_deserialize_item, + + .active_state = timer_active_state, + .sub_state_to_string = timer_sub_state_to_string, + + .trigger_notify = timer_trigger_notify, + + .reset_failed = timer_reset_failed, + .time_change = timer_time_change, + + .bus_vtable = bus_timer_vtable, + .bus_set_property = bus_timer_set_property, + + .can_transient = true, +}; diff --git a/src/grp-system/libcore/src/transaction.c b/src/grp-system/libcore/src/transaction.c new file mode 100644 index 0000000000..2768be914f --- /dev/null +++ b/src/grp-system/libcore/src/transaction.c @@ -0,0 +1,1103 @@ +/*** + This file is part of systemd. + + Copyright 2010 Lennart Poettering + + systemd is free software; you can redistribute it and/or modify it + under the terms of the GNU Lesser General Public License as published by + the Free Software Foundation; either version 2.1 of the License, or + (at your option) any later version. + + systemd is distributed in the hope that it will be useful, but + WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + Lesser General Public License for more details. + + You should have received a copy of the GNU Lesser General Public License + along with systemd; If not, see . +***/ + +#include +#include + +#include "sd-bus/bus-common-errors.h" +#include "sd-bus/bus-error.h" +#include "systemd-basic/alloc-util.h" +#include "systemd-basic/terminal-util.h" + +#include "dbus-unit.h" +#include "transaction.h" + +static void transaction_unlink_job(Transaction *tr, Job *j, bool delete_dependencies); + +static void transaction_delete_job(Transaction *tr, Job *j, bool delete_dependencies) { + assert(tr); + assert(j); + + /* Deletes one job from the transaction */ + + transaction_unlink_job(tr, j, delete_dependencies); + + job_free(j); +} + +static void transaction_delete_unit(Transaction *tr, Unit *u) { + Job *j; + + /* Deletes all jobs associated with a certain unit from the + * transaction */ + + while ((j = hashmap_get(tr->jobs, u))) + transaction_delete_job(tr, j, true); +} + +void transaction_abort(Transaction *tr) { + Job *j; + + assert(tr); + + while ((j = hashmap_first(tr->jobs))) + transaction_delete_job(tr, j, false); + + assert(hashmap_isempty(tr->jobs)); +} + +static void transaction_find_jobs_that_matter_to_anchor(Job *j, unsigned generation) { + JobDependency *l; + + /* A recursive sweep through the graph that marks all units + * that matter to the anchor job, i.e. are directly or + * indirectly a dependency of the anchor job via paths that + * are fully marked as mattering. */ + + j->matters_to_anchor = true; + j->generation = generation; + + LIST_FOREACH(subject, l, j->subject_list) { + + /* This link does not matter */ + if (!l->matters) + continue; + + /* This unit has already been marked */ + if (l->object->generation == generation) + continue; + + transaction_find_jobs_that_matter_to_anchor(l->object, generation); + } +} + +static void transaction_merge_and_delete_job(Transaction *tr, Job *j, Job *other, JobType t) { + JobDependency *l, *last; + + assert(j); + assert(other); + assert(j->unit == other->unit); + assert(!j->installed); + + /* Merges 'other' into 'j' and then deletes 'other'. */ + + j->type = t; + j->state = JOB_WAITING; + j->irreversible = j->irreversible || other->irreversible; + j->matters_to_anchor = j->matters_to_anchor || other->matters_to_anchor; + + /* Patch us in as new owner of the JobDependency objects */ + last = NULL; + LIST_FOREACH(subject, l, other->subject_list) { + assert(l->subject == other); + l->subject = j; + last = l; + } + + /* Merge both lists */ + if (last) { + last->subject_next = j->subject_list; + if (j->subject_list) + j->subject_list->subject_prev = last; + j->subject_list = other->subject_list; + } + + /* Patch us in as new owner of the JobDependency objects */ + last = NULL; + LIST_FOREACH(object, l, other->object_list) { + assert(l->object == other); + l->object = j; + last = l; + } + + /* Merge both lists */ + if (last) { + last->object_next = j->object_list; + if (j->object_list) + j->object_list->object_prev = last; + j->object_list = other->object_list; + } + + /* Kill the other job */ + other->subject_list = NULL; + other->object_list = NULL; + transaction_delete_job(tr, other, true); +} + +_pure_ static bool job_is_conflicted_by(Job *j) { + JobDependency *l; + + assert(j); + + /* Returns true if this job is pulled in by a least one + * ConflictedBy dependency. */ + + LIST_FOREACH(object, l, j->object_list) + if (l->conflicts) + return true; + + return false; +} + +static int delete_one_unmergeable_job(Transaction *tr, Job *j) { + Job *k; + + assert(j); + + /* Tries to delete one item in the linked list + * j->transaction_next->transaction_next->... that conflicts + * with another one, in an attempt to make an inconsistent + * transaction work. */ + + /* We rely here on the fact that if a merged with b does not + * merge with c, either a or b merge with c neither */ + LIST_FOREACH(transaction, j, j) + LIST_FOREACH(transaction, k, j->transaction_next) { + Job *d; + + /* Is this one mergeable? Then skip it */ + if (job_type_is_mergeable(j->type, k->type)) + continue; + + /* Ok, we found two that conflict, let's see if we can + * drop one of them */ + if (!j->matters_to_anchor && !k->matters_to_anchor) { + + /* Both jobs don't matter, so let's + * find the one that is smarter to + * remove. Let's think positive and + * rather remove stops then starts -- + * except if something is being + * stopped because it is conflicted by + * another unit in which case we + * rather remove the start. */ + + log_unit_debug(j->unit, + "Looking at job %s/%s conflicted_by=%s", + j->unit->id, job_type_to_string(j->type), + yes_no(j->type == JOB_STOP && job_is_conflicted_by(j))); + log_unit_debug(k->unit, + "Looking at job %s/%s conflicted_by=%s", + k->unit->id, job_type_to_string(k->type), + yes_no(k->type == JOB_STOP && job_is_conflicted_by(k))); + + if (j->type == JOB_STOP) { + + if (job_is_conflicted_by(j)) + d = k; + else + d = j; + + } else if (k->type == JOB_STOP) { + + if (job_is_conflicted_by(k)) + d = j; + else + d = k; + } else + d = j; + + } else if (!j->matters_to_anchor) + d = j; + else if (!k->matters_to_anchor) + d = k; + else + return -ENOEXEC; + + /* Ok, we can drop one, so let's do so. */ + log_unit_debug(d->unit, + "Fixing conflicting jobs %s/%s,%s/%s by deleting job %s/%s", + j->unit->id, job_type_to_string(j->type), + k->unit->id, job_type_to_string(k->type), + d->unit->id, job_type_to_string(d->type)); + transaction_delete_job(tr, d, true); + return 0; + } + + return -EINVAL; +} + +static int transaction_merge_jobs(Transaction *tr, sd_bus_error *e) { + Job *j; + Iterator i; + int r; + + assert(tr); + + /* First step, check whether any of the jobs for one specific + * task conflict. If so, try to drop one of them. */ + HASHMAP_FOREACH(j, tr->jobs, i) { + JobType t; + Job *k; + + t = j->type; + LIST_FOREACH(transaction, k, j->transaction_next) { + if (job_type_merge_and_collapse(&t, k->type, j->unit) >= 0) + continue; + + /* OK, we could not merge all jobs for this + * action. Let's see if we can get rid of one + * of them */ + + r = delete_one_unmergeable_job(tr, j); + if (r >= 0) + /* Ok, we managed to drop one, now + * let's ask our callers to call us + * again after garbage collecting */ + return -EAGAIN; + + /* We couldn't merge anything. Failure */ + return sd_bus_error_setf(e, BUS_ERROR_TRANSACTION_JOBS_CONFLICTING, + "Transaction contains conflicting jobs '%s' and '%s' for %s. " + "Probably contradicting requirement dependencies configured.", + job_type_to_string(t), + job_type_to_string(k->type), + k->unit->id); + } + } + + /* Second step, merge the jobs. */ + HASHMAP_FOREACH(j, tr->jobs, i) { + JobType t = j->type; + Job *k; + + /* Merge all transaction jobs for j->unit */ + LIST_FOREACH(transaction, k, j->transaction_next) + assert_se(job_type_merge_and_collapse(&t, k->type, j->unit) == 0); + + while ((k = j->transaction_next)) { + if (tr->anchor_job == k) { + transaction_merge_and_delete_job(tr, k, j, t); + j = k; + } else + transaction_merge_and_delete_job(tr, j, k, t); + } + + assert(!j->transaction_next); + assert(!j->transaction_prev); + } + + return 0; +} + +static void transaction_drop_redundant(Transaction *tr) { + Job *j; + Iterator i; + + /* Goes through the transaction and removes all jobs of the units + * whose jobs are all noops. If not all of a unit's jobs are + * redundant, they are kept. */ + + assert(tr); + +rescan: + HASHMAP_FOREACH(j, tr->jobs, i) { + Job *k; + + LIST_FOREACH(transaction, k, j) { + + if (tr->anchor_job == k || + !job_type_is_redundant(k->type, unit_active_state(k->unit)) || + (k->unit->job && job_type_is_conflicting(k->type, k->unit->job->type))) + goto next_unit; + } + + /* log_debug("Found redundant job %s/%s, dropping.", j->unit->id, job_type_to_string(j->type)); */ + transaction_delete_job(tr, j, false); + goto rescan; + next_unit:; + } +} + +_pure_ static bool unit_matters_to_anchor(Unit *u, Job *j) { + assert(u); + assert(!j->transaction_prev); + + /* Checks whether at least one of the jobs for this unit + * matters to the anchor. */ + + LIST_FOREACH(transaction, j, j) + if (j->matters_to_anchor) + return true; + + return false; +} + +static int transaction_verify_order_one(Transaction *tr, Job *j, Job *from, unsigned generation, sd_bus_error *e) { + Iterator i; + Unit *u; + int r; + + assert(tr); + assert(j); + assert(!j->transaction_prev); + + /* Does a recursive sweep through the ordering graph, looking + * for a cycle. If we find a cycle we try to break it. */ + + /* Have we seen this before? */ + if (j->generation == generation) { + Job *k, *delete; + + /* If the marker is NULL we have been here already and + * decided the job was loop-free from here. Hence + * shortcut things and return right-away. */ + if (!j->marker) + return 0; + + /* So, the marker is not NULL and we already have been + * here. We have a cycle. Let's try to break it. We go + * backwards in our path and try to find a suitable + * job to remove. We use the marker to find our way + * back, since smart how we are we stored our way back + * in there. */ + log_unit_warning(j->unit, + "Found ordering cycle on %s/%s", + j->unit->id, job_type_to_string(j->type)); + + delete = NULL; + for (k = from; k; k = ((k->generation == generation && k->marker != k) ? k->marker : NULL)) { + + /* logging for j not k here to provide consistent narrative */ + log_unit_warning(j->unit, + "Found dependency on %s/%s", + k->unit->id, job_type_to_string(k->type)); + + if (!delete && hashmap_get(tr->jobs, k->unit) && !unit_matters_to_anchor(k->unit, k)) + /* Ok, we can drop this one, so let's + * do so. */ + delete = k; + + /* Check if this in fact was the beginning of + * the cycle */ + if (k == j) + break; + } + + + if (delete) { + const char *status; + /* logging for j not k here to provide consistent narrative */ + log_unit_warning(j->unit, + "Breaking ordering cycle by deleting job %s/%s", + delete->unit->id, job_type_to_string(delete->type)); + log_unit_error(delete->unit, + "Job %s/%s deleted to break ordering cycle starting with %s/%s", + delete->unit->id, job_type_to_string(delete->type), + j->unit->id, job_type_to_string(j->type)); + + if (log_get_show_color()) + status = ANSI_HIGHLIGHT_RED " SKIP " ANSI_NORMAL; + else + status = " SKIP "; + + unit_status_printf(delete->unit, status, + "Ordering cycle found, skipping %s"); + transaction_delete_unit(tr, delete->unit); + return -EAGAIN; + } + + log_error("Unable to break cycle"); + + return sd_bus_error_setf(e, BUS_ERROR_TRANSACTION_ORDER_IS_CYCLIC, + "Transaction order is cyclic. See system logs for details."); + } + + /* Make the marker point to where we come from, so that we can + * find our way backwards if we want to break a cycle. We use + * a special marker for the beginning: we point to + * ourselves. */ + j->marker = from ? from : j; + j->generation = generation; + + /* We assume that the dependencies are bidirectional, and + * hence can ignore UNIT_AFTER */ + SET_FOREACH(u, j->unit->dependencies[UNIT_BEFORE], i) { + Job *o; + + /* Is there a job for this unit? */ + o = hashmap_get(tr->jobs, u); + if (!o) { + /* Ok, there is no job for this in the + * transaction, but maybe there is already one + * running? */ + o = u->job; + if (!o) + continue; + } + + r = transaction_verify_order_one(tr, o, j, generation, e); + if (r < 0) + return r; + } + + /* Ok, let's backtrack, and remember that this entry is not on + * our path anymore. */ + j->marker = NULL; + + return 0; +} + +static int transaction_verify_order(Transaction *tr, unsigned *generation, sd_bus_error *e) { + Job *j; + int r; + Iterator i; + unsigned g; + + assert(tr); + assert(generation); + + /* Check if the ordering graph is cyclic. If it is, try to fix + * that up by dropping one of the jobs. */ + + g = (*generation)++; + + HASHMAP_FOREACH(j, tr->jobs, i) { + r = transaction_verify_order_one(tr, j, NULL, g, e); + if (r < 0) + return r; + } + + return 0; +} + +static void transaction_collect_garbage(Transaction *tr) { + Iterator i; + Job *j; + + assert(tr); + + /* Drop jobs that are not required by any other job */ + +rescan: + HASHMAP_FOREACH(j, tr->jobs, i) { + if (tr->anchor_job == j || j->object_list) { + /* log_debug("Keeping job %s/%s because of %s/%s", */ + /* j->unit->id, job_type_to_string(j->type), */ + /* j->object_list->subject ? j->object_list->subject->unit->id : "root", */ + /* j->object_list->subject ? job_type_to_string(j->object_list->subject->type) : "root"); */ + continue; + } + + /* log_debug("Garbage collecting job %s/%s", j->unit->id, job_type_to_string(j->type)); */ + transaction_delete_job(tr, j, true); + goto rescan; + } +} + +static int transaction_is_destructive(Transaction *tr, JobMode mode, sd_bus_error *e) { + Iterator i; + Job *j; + + assert(tr); + + /* Checks whether applying this transaction means that + * existing jobs would be replaced */ + + HASHMAP_FOREACH(j, tr->jobs, i) { + + /* Assume merged */ + assert(!j->transaction_prev); + assert(!j->transaction_next); + + if (j->unit->job && (mode == JOB_FAIL || j->unit->job->irreversible) && + job_type_is_conflicting(j->unit->job->type, j->type)) + return sd_bus_error_setf(e, BUS_ERROR_TRANSACTION_IS_DESTRUCTIVE, + "Transaction is destructive."); + } + + return 0; +} + +static void transaction_minimize_impact(Transaction *tr) { + Job *j; + Iterator i; + + assert(tr); + + /* Drops all unnecessary jobs that reverse already active jobs + * or that stop a running service. */ + +rescan: + HASHMAP_FOREACH(j, tr->jobs, i) { + LIST_FOREACH(transaction, j, j) { + bool stops_running_service, changes_existing_job; + + /* If it matters, we shouldn't drop it */ + if (j->matters_to_anchor) + continue; + + /* Would this stop a running service? + * Would this change an existing job? + * If so, let's drop this entry */ + + stops_running_service = + j->type == JOB_STOP && UNIT_IS_ACTIVE_OR_ACTIVATING(unit_active_state(j->unit)); + + changes_existing_job = + j->unit->job && + job_type_is_conflicting(j->type, j->unit->job->type); + + if (!stops_running_service && !changes_existing_job) + continue; + + if (stops_running_service) + log_unit_debug(j->unit, + "%s/%s would stop a running service.", + j->unit->id, job_type_to_string(j->type)); + + if (changes_existing_job) + log_unit_debug(j->unit, + "%s/%s would change existing job.", + j->unit->id, job_type_to_string(j->type)); + + /* Ok, let's get rid of this */ + log_unit_debug(j->unit, + "Deleting %s/%s to minimize impact.", + j->unit->id, job_type_to_string(j->type)); + + transaction_delete_job(tr, j, true); + goto rescan; + } + } +} + +static int transaction_apply(Transaction *tr, Manager *m, JobMode mode) { + Iterator i; + Job *j; + int r; + + /* Moves the transaction jobs to the set of active jobs */ + + if (mode == JOB_ISOLATE || mode == JOB_FLUSH) { + + /* When isolating first kill all installed jobs which + * aren't part of the new transaction */ + HASHMAP_FOREACH(j, m->jobs, i) { + assert(j->installed); + + if (j->unit->ignore_on_isolate) + continue; + + if (hashmap_get(tr->jobs, j->unit)) + continue; + + /* Not invalidating recursively. Avoids triggering + * OnFailure= actions of dependent jobs. Also avoids + * invalidating our iterator. */ + job_finish_and_invalidate(j, JOB_CANCELED, false, false); + } + } + + HASHMAP_FOREACH(j, tr->jobs, i) { + /* Assume merged */ + assert(!j->transaction_prev); + assert(!j->transaction_next); + + r = hashmap_put(m->jobs, UINT32_TO_PTR(j->id), j); + if (r < 0) + goto rollback; + } + + while ((j = hashmap_steal_first(tr->jobs))) { + Job *installed_job; + + /* Clean the job dependencies */ + transaction_unlink_job(tr, j, false); + + installed_job = job_install(j); + if (installed_job != j) { + /* j has been merged into a previously installed job */ + if (tr->anchor_job == j) + tr->anchor_job = installed_job; + hashmap_remove(m->jobs, UINT32_TO_PTR(j->id)); + job_free(j); + j = installed_job; + } + + job_add_to_run_queue(j); + job_add_to_dbus_queue(j); + job_start_timer(j); + job_shutdown_magic(j); + } + + return 0; + +rollback: + + HASHMAP_FOREACH(j, tr->jobs, i) + hashmap_remove(m->jobs, UINT32_TO_PTR(j->id)); + + return r; +} + +int transaction_activate(Transaction *tr, Manager *m, JobMode mode, sd_bus_error *e) { + Iterator i; + Job *j; + int r; + unsigned generation = 1; + + assert(tr); + + /* This applies the changes recorded in tr->jobs to + * the actual list of jobs, if possible. */ + + /* Reset the generation counter of all installed jobs. The detection of cycles + * looks at installed jobs. If they had a non-zero generation from some previous + * walk of the graph, the algorithm would break. */ + HASHMAP_FOREACH(j, m->jobs, i) + j->generation = 0; + + /* First step: figure out which jobs matter */ + transaction_find_jobs_that_matter_to_anchor(tr->anchor_job, generation++); + + /* Second step: Try not to stop any running services if + * we don't have to. Don't try to reverse running + * jobs if we don't have to. */ + if (mode == JOB_FAIL) + transaction_minimize_impact(tr); + + /* Third step: Drop redundant jobs */ + transaction_drop_redundant(tr); + + for (;;) { + /* Fourth step: Let's remove unneeded jobs that might + * be lurking. */ + if (mode != JOB_ISOLATE) + transaction_collect_garbage(tr); + + /* Fifth step: verify order makes sense and correct + * cycles if necessary and possible */ + r = transaction_verify_order(tr, &generation, e); + if (r >= 0) + break; + + if (r != -EAGAIN) { + log_warning("Requested transaction contains an unfixable cyclic ordering dependency: %s", bus_error_message(e, r)); + return r; + } + + /* Let's see if the resulting transaction ordering + * graph is still cyclic... */ + } + + for (;;) { + /* Sixth step: let's drop unmergeable entries if + * necessary and possible, merge entries we can + * merge */ + r = transaction_merge_jobs(tr, e); + if (r >= 0) + break; + + if (r != -EAGAIN) { + log_warning("Requested transaction contains unmergeable jobs: %s", bus_error_message(e, r)); + return r; + } + + /* Seventh step: an entry got dropped, let's garbage + * collect its dependencies. */ + if (mode != JOB_ISOLATE) + transaction_collect_garbage(tr); + + /* Let's see if the resulting transaction still has + * unmergeable entries ... */ + } + + /* Eights step: Drop redundant jobs again, if the merging now allows us to drop more. */ + transaction_drop_redundant(tr); + + /* Ninth step: check whether we can actually apply this */ + r = transaction_is_destructive(tr, mode, e); + if (r < 0) { + log_notice("Requested transaction contradicts existing jobs: %s", bus_error_message(e, r)); + return r; + } + + /* Tenth step: apply changes */ + r = transaction_apply(tr, m, mode); + if (r < 0) + return log_warning_errno(r, "Failed to apply transaction: %m"); + + assert(hashmap_isempty(tr->jobs)); + + if (!hashmap_isempty(m->jobs)) { + /* Are there any jobs now? Then make sure we have the + * idle pipe around. We don't really care too much + * whether this works or not, as the idle pipe is a + * feature for cosmetics, not actually useful for + * anything beyond that. */ + + if (m->idle_pipe[0] < 0 && m->idle_pipe[1] < 0 && + m->idle_pipe[2] < 0 && m->idle_pipe[3] < 0) { + (void) pipe2(m->idle_pipe, O_NONBLOCK|O_CLOEXEC); + (void) pipe2(m->idle_pipe + 2, O_NONBLOCK|O_CLOEXEC); + } + } + + return 0; +} + +static Job* transaction_add_one_job(Transaction *tr, JobType type, Unit *unit, bool *is_new) { + Job *j, *f; + + assert(tr); + assert(unit); + + /* Looks for an existing prospective job and returns that. If + * it doesn't exist it is created and added to the prospective + * jobs list. */ + + f = hashmap_get(tr->jobs, unit); + + LIST_FOREACH(transaction, j, f) { + assert(j->unit == unit); + + if (j->type == type) { + if (is_new) + *is_new = false; + return j; + } + } + + j = job_new(unit, type); + if (!j) + return NULL; + + j->generation = 0; + j->marker = NULL; + j->matters_to_anchor = false; + j->irreversible = tr->irreversible; + + LIST_PREPEND(transaction, f, j); + + if (hashmap_replace(tr->jobs, unit, f) < 0) { + LIST_REMOVE(transaction, f, j); + job_free(j); + return NULL; + } + + if (is_new) + *is_new = true; + + /* log_debug("Added job %s/%s to transaction.", unit->id, job_type_to_string(type)); */ + + return j; +} + +static void transaction_unlink_job(Transaction *tr, Job *j, bool delete_dependencies) { + assert(tr); + assert(j); + + if (j->transaction_prev) + j->transaction_prev->transaction_next = j->transaction_next; + else if (j->transaction_next) + hashmap_replace(tr->jobs, j->unit, j->transaction_next); + else + hashmap_remove_value(tr->jobs, j->unit, j); + + if (j->transaction_next) + j->transaction_next->transaction_prev = j->transaction_prev; + + j->transaction_prev = j->transaction_next = NULL; + + while (j->subject_list) + job_dependency_free(j->subject_list); + + while (j->object_list) { + Job *other = j->object_list->matters ? j->object_list->subject : NULL; + + job_dependency_free(j->object_list); + + if (other && delete_dependencies) { + log_unit_debug(other->unit, + "Deleting job %s/%s as dependency of job %s/%s", + other->unit->id, job_type_to_string(other->type), + j->unit->id, job_type_to_string(j->type)); + transaction_delete_job(tr, other, delete_dependencies); + } + } +} + +int transaction_add_job_and_dependencies( + Transaction *tr, + JobType type, + Unit *unit, + Job *by, + bool matters, + bool conflicts, + bool ignore_requirements, + bool ignore_order, + sd_bus_error *e) { + Job *ret; + Iterator i; + Unit *dep; + int r; + bool is_new; + + assert(tr); + assert(type < _JOB_TYPE_MAX); + assert(type < _JOB_TYPE_MAX_IN_TRANSACTION); + assert(unit); + + /* Before adding jobs for this unit, let's ensure that its state has been loaded + * This matters when jobs are spawned as part of coldplugging itself (see e. g. path_coldplug()). + * This way, we "recursively" coldplug units, ensuring that we do not look at state of + * not-yet-coldplugged units. */ + if (MANAGER_IS_RELOADING(unit->manager)) + unit_coldplug(unit); + + /* log_debug("Pulling in %s/%s from %s/%s", */ + /* unit->id, job_type_to_string(type), */ + /* by ? by->unit->id : "NA", */ + /* by ? job_type_to_string(by->type) : "NA"); */ + + if (!IN_SET(unit->load_state, UNIT_LOADED, UNIT_ERROR, UNIT_NOT_FOUND, UNIT_MASKED)) + return sd_bus_error_setf(e, BUS_ERROR_LOAD_FAILED, "Unit %s is not loaded properly.", unit->id); + + if (type != JOB_STOP) { + r = bus_unit_check_load_state(unit, e); + if (r < 0) + return r; + } + + if (!unit_job_is_applicable(unit, type)) + return sd_bus_error_setf(e, BUS_ERROR_JOB_TYPE_NOT_APPLICABLE, + "Job type %s is not applicable for unit %s.", + job_type_to_string(type), unit->id); + + + /* First add the job. */ + ret = transaction_add_one_job(tr, type, unit, &is_new); + if (!ret) + return -ENOMEM; + + ret->ignore_order = ret->ignore_order || ignore_order; + + /* Then, add a link to the job. */ + if (by) { + if (!job_dependency_new(by, ret, matters, conflicts)) + return -ENOMEM; + } else { + /* If the job has no parent job, it is the anchor job. */ + assert(!tr->anchor_job); + tr->anchor_job = ret; + } + + if (is_new && !ignore_requirements && type != JOB_NOP) { + Set *following; + + /* If we are following some other unit, make sure we + * add all dependencies of everybody following. */ + if (unit_following_set(ret->unit, &following) > 0) { + SET_FOREACH(dep, following, i) { + r = transaction_add_job_and_dependencies(tr, type, dep, ret, false, false, false, ignore_order, e); + if (r < 0) { + log_unit_warning(dep, "Cannot add dependency job for, ignoring: %s", bus_error_message(e, r)); + sd_bus_error_free(e); + } + } + + set_free(following); + } + + /* Finally, recursively add in all dependencies. */ + if (type == JOB_START || type == JOB_RESTART) { + SET_FOREACH(dep, ret->unit->dependencies[UNIT_REQUIRES], i) { + r = transaction_add_job_and_dependencies(tr, JOB_START, dep, ret, true, false, false, ignore_order, e); + if (r < 0) { + if (r != -EBADR) /* job type not applicable */ + goto fail; + + sd_bus_error_free(e); + } + } + + SET_FOREACH(dep, ret->unit->dependencies[UNIT_BINDS_TO], i) { + r = transaction_add_job_and_dependencies(tr, JOB_START, dep, ret, true, false, false, ignore_order, e); + if (r < 0) { + if (r != -EBADR) /* job type not applicable */ + goto fail; + + sd_bus_error_free(e); + } + } + + SET_FOREACH(dep, ret->unit->dependencies[UNIT_WANTS], i) { + r = transaction_add_job_and_dependencies(tr, JOB_START, dep, ret, false, false, false, ignore_order, e); + if (r < 0) { + /* unit masked, job type not applicable and unit not found are not considered as errors. */ + log_unit_full(dep, + IN_SET(r, -ERFKILL, -EBADR, -ENOENT) ? LOG_DEBUG : LOG_WARNING, + r, "Cannot add dependency job, ignoring: %s", + bus_error_message(e, r)); + sd_bus_error_free(e); + } + } + + SET_FOREACH(dep, ret->unit->dependencies[UNIT_REQUISITE], i) { + r = transaction_add_job_and_dependencies(tr, JOB_VERIFY_ACTIVE, dep, ret, true, false, false, ignore_order, e); + if (r < 0) { + if (r != -EBADR) /* job type not applicable */ + goto fail; + + sd_bus_error_free(e); + } + } + + SET_FOREACH(dep, ret->unit->dependencies[UNIT_CONFLICTS], i) { + r = transaction_add_job_and_dependencies(tr, JOB_STOP, dep, ret, true, true, false, ignore_order, e); + if (r < 0) { + if (r != -EBADR) /* job type not applicable */ + goto fail; + + sd_bus_error_free(e); + } + } + + SET_FOREACH(dep, ret->unit->dependencies[UNIT_CONFLICTED_BY], i) { + r = transaction_add_job_and_dependencies(tr, JOB_STOP, dep, ret, false, false, false, ignore_order, e); + if (r < 0) { + log_unit_warning(dep, + "Cannot add dependency job, ignoring: %s", + bus_error_message(e, r)); + sd_bus_error_free(e); + } + } + + } + + if (type == JOB_STOP || type == JOB_RESTART) { + static const UnitDependency propagate_deps[] = { + UNIT_REQUIRED_BY, + UNIT_REQUISITE_OF, + UNIT_BOUND_BY, + UNIT_CONSISTS_OF, + }; + + JobType ptype; + unsigned j; + + /* We propagate STOP as STOP, but RESTART only + * as TRY_RESTART, in order not to start + * dependencies that are not around. */ + ptype = type == JOB_RESTART ? JOB_TRY_RESTART : type; + + for (j = 0; j < ELEMENTSOF(propagate_deps); j++) + SET_FOREACH(dep, ret->unit->dependencies[propagate_deps[j]], i) { + JobType nt; + + nt = job_type_collapse(ptype, dep); + if (nt == JOB_NOP) + continue; + + r = transaction_add_job_and_dependencies(tr, nt, dep, ret, true, false, false, ignore_order, e); + if (r < 0) { + if (r != -EBADR) /* job type not applicable */ + goto fail; + + sd_bus_error_free(e); + } + } + } + + if (type == JOB_RELOAD) { + + SET_FOREACH(dep, ret->unit->dependencies[UNIT_PROPAGATES_RELOAD_TO], i) { + JobType nt; + + nt = job_type_collapse(JOB_TRY_RELOAD, dep); + if (nt == JOB_NOP) + continue; + + r = transaction_add_job_and_dependencies(tr, nt, dep, ret, false, false, false, ignore_order, e); + if (r < 0) { + log_unit_warning(dep, + "Cannot add dependency reload job, ignoring: %s", + bus_error_message(e, r)); + sd_bus_error_free(e); + } + } + } + + /* JOB_VERIFY_STARTED require no dependency handling */ + } + + return 0; + +fail: + return r; +} + +int transaction_add_isolate_jobs(Transaction *tr, Manager *m) { + Iterator i; + Unit *u; + char *k; + int r; + + assert(tr); + assert(m); + + HASHMAP_FOREACH_KEY(u, k, m->units, i) { + + /* ignore aliases */ + if (u->id != k) + continue; + + if (u->ignore_on_isolate) + continue; + + /* No need to stop inactive jobs */ + if (UNIT_IS_INACTIVE_OR_FAILED(unit_active_state(u)) && !u->job) + continue; + + /* Is there already something listed for this? */ + if (hashmap_get(tr->jobs, u)) + continue; + + r = transaction_add_job_and_dependencies(tr, JOB_STOP, u, tr->anchor_job, true, false, false, false, NULL); + if (r < 0) + log_unit_warning_errno(u, r, "Cannot add isolate job, ignoring: %m"); + } + + return 0; +} + +Transaction *transaction_new(bool irreversible) { + Transaction *tr; + + tr = new0(Transaction, 1); + if (!tr) + return NULL; + + tr->jobs = hashmap_new(NULL); + if (!tr->jobs) { + free(tr); + return NULL; + } + + tr->irreversible = irreversible; + + return tr; +} + +void transaction_free(Transaction *tr) { + assert(hashmap_isempty(tr->jobs)); + hashmap_free(tr->jobs); + free(tr); +} diff --git a/src/grp-system/libcore/src/transaction.h b/src/grp-system/libcore/src/transaction.h new file mode 100644 index 0000000000..e4066a0963 --- /dev/null +++ b/src/grp-system/libcore/src/transaction.h @@ -0,0 +1,51 @@ +#pragma once + +/*** + This file is part of systemd. + + Copyright 2010 Lennart Poettering + + systemd is free software; you can redistribute it and/or modify it + under the terms of the GNU Lesser General Public License as published by + the Free Software Foundation; either version 2.1 of the License, or + (at your option) any later version. + + systemd is distributed in the hope that it will be useful, but + WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + Lesser General Public License for more details. + + You should have received a copy of the GNU Lesser General Public License + along with systemd; If not, see . +***/ + +#include "core/job.h" +#include "core/manager.h" +#include "core/unit.h" +#include "systemd-basic/hashmap.h" + +typedef struct Transaction Transaction; + +struct Transaction { + /* Jobs to be added */ + Hashmap *jobs; /* Unit object => Job object list 1:1 */ + Job *anchor_job; /* the job the user asked for */ + bool irreversible; +}; + +Transaction *transaction_new(bool irreversible); +void transaction_free(Transaction *tr); + +int transaction_add_job_and_dependencies( + Transaction *tr, + JobType type, + Unit *unit, + Job *by, + bool matters, + bool conflicts, + bool ignore_requirements, + bool ignore_order, + sd_bus_error *e); +int transaction_activate(Transaction *tr, Manager *m, JobMode mode, sd_bus_error *e); +int transaction_add_isolate_jobs(Transaction *tr, Manager *m); +void transaction_abort(Transaction *tr); diff --git a/src/grp-system/libcore/src/unit-printf.c b/src/grp-system/libcore/src/unit-printf.c new file mode 100644 index 0000000000..2f73214295 --- /dev/null +++ b/src/grp-system/libcore/src/unit-printf.c @@ -0,0 +1,305 @@ +/*** + This file is part of systemd. + + Copyright 2010 Lennart Poettering + + systemd is free software; you can redistribute it and/or modify it + under the terms of the GNU Lesser General Public License as published by + the Free Software Foundation; either version 2.1 of the License, or + (at your option) any later version. + + systemd is distributed in the hope that it will be useful, but + WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + Lesser General Public License for more details. + + You should have received a copy of the GNU Lesser General Public License + along with systemd; If not, see . +***/ + +#include "core/unit.h" +#include "systemd-basic/alloc-util.h" +#include "systemd-basic/cgroup-util.h" +#include "systemd-basic/formats-util.h" +#include "systemd-basic/macro.h" +#include "systemd-basic/string-util.h" +#include "systemd-basic/strv.h" +#include "systemd-basic/unit-name.h" +#include "systemd-basic/user-util.h" +#include "systemd-shared/specifier.h" + +#include "unit-printf.h" + +static int specifier_prefix_and_instance(char specifier, void *data, void *userdata, char **ret) { + Unit *u = userdata; + + assert(u); + + return unit_name_to_prefix_and_instance(u->id, ret); +} + +static int specifier_prefix(char specifier, void *data, void *userdata, char **ret) { + Unit *u = userdata; + + assert(u); + + return unit_name_to_prefix(u->id, ret); +} + +static int specifier_prefix_unescaped(char specifier, void *data, void *userdata, char **ret) { + _cleanup_free_ char *p = NULL; + Unit *u = userdata; + int r; + + assert(u); + + r = unit_name_to_prefix(u->id, &p); + if (r < 0) + return r; + + return unit_name_unescape(p, ret); +} + +static int specifier_instance_unescaped(char specifier, void *data, void *userdata, char **ret) { + Unit *u = userdata; + + assert(u); + + return unit_name_unescape(strempty(u->instance), ret); +} + +static int specifier_filename(char specifier, void *data, void *userdata, char **ret) { + Unit *u = userdata; + + assert(u); + + if (u->instance) + return unit_name_path_unescape(u->instance, ret); + else + return unit_name_to_path(u->id, ret); +} + +static int specifier_cgroup(char specifier, void *data, void *userdata, char **ret) { + Unit *u = userdata; + char *n; + + assert(u); + + if (u->cgroup_path) + n = strdup(u->cgroup_path); + else + n = unit_default_cgroup_path(u); + if (!n) + return -ENOMEM; + + *ret = n; + return 0; +} + +static int specifier_cgroup_root(char specifier, void *data, void *userdata, char **ret) { + Unit *u = userdata; + char *n; + + assert(u); + + n = strdup(u->manager->cgroup_root); + if (!n) + return -ENOMEM; + + *ret = n; + return 0; +} + +static int specifier_cgroup_slice(char specifier, void *data, void *userdata, char **ret) { + Unit *u = userdata; + char *n; + + assert(u); + + if (UNIT_ISSET(u->slice)) { + Unit *slice; + + slice = UNIT_DEREF(u->slice); + + if (slice->cgroup_path) + n = strdup(slice->cgroup_path); + else + n = unit_default_cgroup_path(slice); + } else + n = strdup(u->manager->cgroup_root); + if (!n) + return -ENOMEM; + + *ret = n; + return 0; +} + +static int specifier_runtime(char specifier, void *data, void *userdata, char **ret) { + Unit *u = userdata; + const char *e; + char *n = NULL; + + assert(u); + + e = manager_get_runtime_prefix(u->manager); + if (!e) + return -EOPNOTSUPP; + n = strdup(e); + if (!n) + return -ENOMEM; + + *ret = n; + return 0; +} + +static int specifier_user_name(char specifier, void *data, void *userdata, char **ret) { + char *t; + + /* If we are UID 0 (root), this will not result in NSS, + * otherwise it might. This is good, as we want to be able to + * run this in PID 1, where our user ID is 0, but where NSS + * lookups are not allowed. */ + + t = getusername_malloc(); + if (!t) + return -ENOMEM; + + *ret = t; + return 0; +} + +static int specifier_user_id(char specifier, void *data, void *userdata, char **ret) { + + if (asprintf(ret, UID_FMT, getuid()) < 0) + return -ENOMEM; + + return 0; +} + +static int specifier_user_home(char specifier, void *data, void *userdata, char **ret) { + + /* On PID 1 (which runs as root) this will not result in NSS, + * which is good. See above */ + + return get_home_dir(ret); +} + +static int specifier_user_shell(char specifier, void *data, void *userdata, char **ret) { + + /* On PID 1 (which runs as root) this will not result in NSS, + * which is good. See above */ + + return get_shell(ret); +} + +int unit_name_printf(Unit *u, const char* format, char **ret) { + + /* + * This will use the passed string as format string and + * replace the following specifiers: + * + * %n: the full id of the unit (foo@bar.waldo) + * %N: the id of the unit without the suffix (foo@bar) + * %p: the prefix (foo) + * %i: the instance (bar) + */ + + const Specifier table[] = { + { 'n', specifier_string, u->id }, + { 'N', specifier_prefix_and_instance, NULL }, + { 'p', specifier_prefix, NULL }, + { 'i', specifier_string, u->instance }, + { 0, NULL, NULL } + }; + + assert(u); + assert(format); + assert(ret); + + return specifier_printf(format, table, u, ret); +} + +int unit_full_printf(Unit *u, const char *format, char **ret) { + + /* This is similar to unit_name_printf() but also supports + * unescaping. Also, adds a couple of additional codes: + * + * %f the instance if set, otherwise the id + * %c cgroup path of unit + * %r where units in this slice are placed in the cgroup tree + * %R the root of this systemd's instance tree + * %t the runtime directory to place sockets in (e.g. "/run" or $XDG_RUNTIME_DIR) + * %U the UID of the running user + * %u the username of the running user + * %h the homedir of the running user + * %s the shell of the running user + * %m the machine ID of the running system + * %H the host name of the running system + * %b the boot ID of the running system + * %v `uname -r` of the running system + */ + + const Specifier table[] = { + { 'n', specifier_string, u->id }, + { 'N', specifier_prefix_and_instance, NULL }, + { 'p', specifier_prefix, NULL }, + { 'P', specifier_prefix_unescaped, NULL }, + { 'i', specifier_string, u->instance }, + { 'I', specifier_instance_unescaped, NULL }, + + { 'f', specifier_filename, NULL }, + { 'c', specifier_cgroup, NULL }, + { 'r', specifier_cgroup_slice, NULL }, + { 'R', specifier_cgroup_root, NULL }, + { 't', specifier_runtime, NULL }, + + { 'U', specifier_user_id, NULL }, + { 'u', specifier_user_name, NULL }, + { 'h', specifier_user_home, NULL }, + { 's', specifier_user_shell, NULL }, + + { 'm', specifier_machine_id, NULL }, + { 'H', specifier_host_name, NULL }, + { 'b', specifier_boot_id, NULL }, + { 'v', specifier_kernel_release, NULL }, + {} + }; + + assert(u); + assert(format); + assert(ret); + + return specifier_printf(format, table, u, ret); +} + +int unit_full_printf_strv(Unit *u, char **l, char ***ret) { + size_t n; + char **r, **i, **j; + int q; + + /* Applies unit_full_printf to every entry in l */ + + assert(u); + + n = strv_length(l); + r = new(char*, n+1); + if (!r) + return -ENOMEM; + + for (i = l, j = r; *i; i++, j++) { + q = unit_full_printf(u, *i, j); + if (q < 0) + goto fail; + } + + *j = NULL; + *ret = r; + return 0; + +fail: + for (j--; j >= r; j--) + free(*j); + + free(r); + return q; +} diff --git a/src/grp-system/libcore/src/unit-printf.h b/src/grp-system/libcore/src/unit-printf.h new file mode 100644 index 0000000000..7ef76e5bb9 --- /dev/null +++ b/src/grp-system/libcore/src/unit-printf.h @@ -0,0 +1,26 @@ +#pragma once + +/*** + This file is part of systemd. + + Copyright 2010 Lennart Poettering + + systemd is free software; you can redistribute it and/or modify it + under the terms of the GNU Lesser General Public License as published by + the Free Software Foundation; either version 2.1 of the License, or + (at your option) any later version. + + systemd is distributed in the hope that it will be useful, but + WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + Lesser General Public License for more details. + + You should have received a copy of the GNU Lesser General Public License + along with systemd; If not, see . +***/ + +#include "core/unit.h" + +int unit_name_printf(Unit *u, const char* text, char **ret); +int unit_full_printf(Unit *u, const char *text, char **ret); +int unit_full_printf_strv(Unit *u, char **l, char ***ret); diff --git a/src/grp-system/libcore/src/unit.c b/src/grp-system/libcore/src/unit.c new file mode 100644 index 0000000000..a36e589b73 --- /dev/null +++ b/src/grp-system/libcore/src/unit.c @@ -0,0 +1,3872 @@ +/*** + This file is part of systemd. + + Copyright 2010 Lennart Poettering + + systemd is free software; you can redistribute it and/or modify it + under the terms of the GNU Lesser General Public License as published by + the Free Software Foundation; either version 2.1 of the License, or + (at your option) any later version. + + systemd is distributed in the hope that it will be useful, but + WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + Lesser General Public License for more details. + + You should have received a copy of the GNU Lesser General Public License + along with systemd; If not, see . +***/ + +#include +#include +#include +#include +#include + +#include +#include + +#include "core/execute.h" +#include "core/load-fragment.h" +#include "core/unit.h" +#include "sd-bus/bus-common-errors.h" +#include "sd-bus/bus-util.h" +#include "systemd-basic/alloc-util.h" +#include "systemd-basic/cgroup-util.h" +#include "systemd-basic/escape.h" +#include "systemd-basic/fileio-label.h" +#include "systemd-basic/formats-util.h" +#include "systemd-basic/log.h" +#include "systemd-basic/macro.h" +#include "systemd-basic/missing.h" +#include "systemd-basic/mkdir.h" +#include "systemd-basic/parse-util.h" +#include "systemd-basic/path-util.h" +#include "systemd-basic/process-util.h" +#include "systemd-basic/set.h" +#include "systemd-basic/signal-util.h" +#include "systemd-basic/special.h" +#include "systemd-basic/stat-util.h" +#include "systemd-basic/stdio-util.h" +#include "systemd-basic/string-util.h" +#include "systemd-basic/strv.h" +#include "systemd-basic/umask-util.h" +#include "systemd-basic/unit-name.h" +#include "systemd-basic/user-util.h" +#include "systemd-basic/virt.h" +#include "systemd-shared/dropin.h" + +#include "dbus-unit.h" +#include "dbus.h" +#include "load-dropin.h" + +const UnitVTable * const unit_vtable[_UNIT_TYPE_MAX] = { + [UNIT_SERVICE] = &service_vtable, + [UNIT_SOCKET] = &socket_vtable, + [UNIT_BUSNAME] = &busname_vtable, + [UNIT_TARGET] = &target_vtable, + [UNIT_DEVICE] = &device_vtable, + [UNIT_MOUNT] = &mount_vtable, + [UNIT_AUTOMOUNT] = &automount_vtable, + [UNIT_SWAP] = &swap_vtable, + [UNIT_TIMER] = &timer_vtable, + [UNIT_PATH] = &path_vtable, + [UNIT_SLICE] = &slice_vtable, + [UNIT_SCOPE] = &scope_vtable +}; + +static void maybe_warn_about_dependency(Unit *u, const char *other, UnitDependency dependency); + +Unit *unit_new(Manager *m, size_t size) { + Unit *u; + + assert(m); + assert(size >= sizeof(Unit)); + + u = malloc0(size); + if (!u) + return NULL; + + u->names = set_new(&string_hash_ops); + if (!u->names) { + free(u); + return NULL; + } + + u->manager = m; + u->type = _UNIT_TYPE_INVALID; + u->default_dependencies = true; + u->unit_file_state = _UNIT_FILE_STATE_INVALID; + u->unit_file_preset = -1; + u->on_failure_job_mode = JOB_REPLACE; + u->cgroup_inotify_wd = -1; + u->job_timeout = USEC_INFINITY; + u->sigchldgen = 0; + + RATELIMIT_INIT(u->start_limit, m->default_start_limit_interval, m->default_start_limit_burst); + RATELIMIT_INIT(u->auto_stop_ratelimit, 10 * USEC_PER_SEC, 16); + + return u; +} + +bool unit_has_name(Unit *u, const char *name) { + assert(u); + assert(name); + + return !!set_get(u->names, (char*) name); +} + +static void unit_init(Unit *u) { + CGroupContext *cc; + ExecContext *ec; + KillContext *kc; + + assert(u); + assert(u->manager); + assert(u->type >= 0); + + cc = unit_get_cgroup_context(u); + if (cc) { + cgroup_context_init(cc); + + /* Copy in the manager defaults into the cgroup + * context, _before_ the rest of the settings have + * been initialized */ + + cc->cpu_accounting = u->manager->default_cpu_accounting; + cc->io_accounting = u->manager->default_io_accounting; + cc->blockio_accounting = u->manager->default_blockio_accounting; + cc->memory_accounting = u->manager->default_memory_accounting; + cc->tasks_accounting = u->manager->default_tasks_accounting; + + if (u->type != UNIT_SLICE) + cc->tasks_max = u->manager->default_tasks_max; + } + + ec = unit_get_exec_context(u); + if (ec) + exec_context_init(ec); + + kc = unit_get_kill_context(u); + if (kc) + kill_context_init(kc); + + if (UNIT_VTABLE(u)->init) + UNIT_VTABLE(u)->init(u); +} + +int unit_add_name(Unit *u, const char *text) { + _cleanup_free_ char *s = NULL, *i = NULL; + UnitType t; + int r; + + assert(u); + assert(text); + + if (unit_name_is_valid(text, UNIT_NAME_TEMPLATE)) { + + if (!u->instance) + return -EINVAL; + + r = unit_name_replace_instance(text, u->instance, &s); + if (r < 0) + return r; + } else { + s = strdup(text); + if (!s) + return -ENOMEM; + } + + if (set_contains(u->names, s)) + return 0; + if (hashmap_contains(u->manager->units, s)) + return -EEXIST; + + if (!unit_name_is_valid(s, UNIT_NAME_PLAIN|UNIT_NAME_INSTANCE)) + return -EINVAL; + + t = unit_name_to_type(s); + if (t < 0) + return -EINVAL; + + if (u->type != _UNIT_TYPE_INVALID && t != u->type) + return -EINVAL; + + r = unit_name_to_instance(s, &i); + if (r < 0) + return r; + + if (i && !unit_type_may_template(t)) + return -EINVAL; + + /* Ensure that this unit is either instanced or not instanced, + * but not both. Note that we do allow names with different + * instance names however! */ + if (u->type != _UNIT_TYPE_INVALID && !u->instance != !i) + return -EINVAL; + + if (!unit_type_may_alias(t) && !set_isempty(u->names)) + return -EEXIST; + + if (hashmap_size(u->manager->units) >= MANAGER_MAX_NAMES) + return -E2BIG; + + r = set_put(u->names, s); + if (r < 0) + return r; + assert(r > 0); + + r = hashmap_put(u->manager->units, s, u); + if (r < 0) { + (void) set_remove(u->names, s); + return r; + } + + if (u->type == _UNIT_TYPE_INVALID) { + u->type = t; + u->id = s; + u->instance = i; + + LIST_PREPEND(units_by_type, u->manager->units_by_type[t], u); + + unit_init(u); + + i = NULL; + } + + s = NULL; + + unit_add_to_dbus_queue(u); + return 0; +} + +int unit_choose_id(Unit *u, const char *name) { + _cleanup_free_ char *t = NULL; + char *s, *i; + int r; + + assert(u); + assert(name); + + if (unit_name_is_valid(name, UNIT_NAME_TEMPLATE)) { + + if (!u->instance) + return -EINVAL; + + r = unit_name_replace_instance(name, u->instance, &t); + if (r < 0) + return r; + + name = t; + } + + /* Selects one of the names of this unit as the id */ + s = set_get(u->names, (char*) name); + if (!s) + return -ENOENT; + + /* Determine the new instance from the new id */ + r = unit_name_to_instance(s, &i); + if (r < 0) + return r; + + u->id = s; + + free(u->instance); + u->instance = i; + + unit_add_to_dbus_queue(u); + + return 0; +} + +int unit_set_description(Unit *u, const char *description) { + char *s; + + assert(u); + + if (isempty(description)) + s = NULL; + else { + s = strdup(description); + if (!s) + return -ENOMEM; + } + + free(u->description); + u->description = s; + + unit_add_to_dbus_queue(u); + return 0; +} + +bool unit_check_gc(Unit *u) { + UnitActiveState state; + assert(u); + + if (u->job) + return true; + + if (u->nop_job) + return true; + + state = unit_active_state(u); + + /* If the unit is inactive and failed and no job is queued for + * it, then release its runtime resources */ + if (UNIT_IS_INACTIVE_OR_FAILED(state) && + UNIT_VTABLE(u)->release_resources) + UNIT_VTABLE(u)->release_resources(u); + + /* But we keep the unit object around for longer when it is + * referenced or configured to not be gc'ed */ + if (state != UNIT_INACTIVE) + return true; + + if (u->no_gc) + return true; + + if (u->refs) + return true; + + if (UNIT_VTABLE(u)->check_gc) + if (UNIT_VTABLE(u)->check_gc(u)) + return true; + + return false; +} + +void unit_add_to_load_queue(Unit *u) { + assert(u); + assert(u->type != _UNIT_TYPE_INVALID); + + if (u->load_state != UNIT_STUB || u->in_load_queue) + return; + + LIST_PREPEND(load_queue, u->manager->load_queue, u); + u->in_load_queue = true; +} + +void unit_add_to_cleanup_queue(Unit *u) { + assert(u); + + if (u->in_cleanup_queue) + return; + + LIST_PREPEND(cleanup_queue, u->manager->cleanup_queue, u); + u->in_cleanup_queue = true; +} + +void unit_add_to_gc_queue(Unit *u) { + assert(u); + + if (u->in_gc_queue || u->in_cleanup_queue) + return; + + if (unit_check_gc(u)) + return; + + LIST_PREPEND(gc_queue, u->manager->gc_queue, u); + u->in_gc_queue = true; + + u->manager->n_in_gc_queue++; +} + +void unit_add_to_dbus_queue(Unit *u) { + assert(u); + assert(u->type != _UNIT_TYPE_INVALID); + + if (u->load_state == UNIT_STUB || u->in_dbus_queue) + return; + + /* Shortcut things if nobody cares */ + if (sd_bus_track_count(u->manager->subscribed) <= 0 && + set_isempty(u->manager->private_buses)) { + u->sent_dbus_new_signal = true; + return; + } + + LIST_PREPEND(dbus_queue, u->manager->dbus_unit_queue, u); + u->in_dbus_queue = true; +} + +static void bidi_set_free(Unit *u, Set *s) { + Iterator i; + Unit *other; + + assert(u); + + /* Frees the set and makes sure we are dropped from the + * inverse pointers */ + + SET_FOREACH(other, s, i) { + UnitDependency d; + + for (d = 0; d < _UNIT_DEPENDENCY_MAX; d++) + set_remove(other->dependencies[d], u); + + unit_add_to_gc_queue(other); + } + + set_free(s); +} + +static void unit_remove_transient(Unit *u) { + char **i; + + assert(u); + + if (!u->transient) + return; + + if (u->fragment_path) + (void) unlink(u->fragment_path); + + STRV_FOREACH(i, u->dropin_paths) { + _cleanup_free_ char *p = NULL, *pp = NULL; + + p = dirname_malloc(*i); /* Get the drop-in directory from the drop-in file */ + if (!p) + continue; + + pp = dirname_malloc(p); /* Get the config directory from the drop-in directory */ + if (!pp) + continue; + + /* Only drop transient drop-ins */ + if (!path_equal(u->manager->lookup_paths.transient, pp)) + continue; + + (void) unlink(*i); + (void) rmdir(p); + } +} + +static void unit_free_requires_mounts_for(Unit *u) { + char **j; + + STRV_FOREACH(j, u->requires_mounts_for) { + char s[strlen(*j) + 1]; + + PATH_FOREACH_PREFIX_MORE(s, *j) { + char *y; + Set *x; + + x = hashmap_get2(u->manager->units_requiring_mounts_for, s, (void**) &y); + if (!x) + continue; + + set_remove(x, u); + + if (set_isempty(x)) { + hashmap_remove(u->manager->units_requiring_mounts_for, y); + free(y); + set_free(x); + } + } + } + + u->requires_mounts_for = strv_free(u->requires_mounts_for); +} + +static void unit_done(Unit *u) { + ExecContext *ec; + CGroupContext *cc; + + assert(u); + + if (u->type < 0) + return; + + if (UNIT_VTABLE(u)->done) + UNIT_VTABLE(u)->done(u); + + ec = unit_get_exec_context(u); + if (ec) + exec_context_done(ec); + + cc = unit_get_cgroup_context(u); + if (cc) + cgroup_context_done(cc); +} + +void unit_free(Unit *u) { + UnitDependency d; + Iterator i; + char *t; + + assert(u); + + if (u->transient_file) + fclose(u->transient_file); + + if (!MANAGER_IS_RELOADING(u->manager)) + unit_remove_transient(u); + + bus_unit_send_removed_signal(u); + + unit_done(u); + + sd_bus_slot_unref(u->match_bus_slot); + + unit_free_requires_mounts_for(u); + + SET_FOREACH(t, u->names, i) + hashmap_remove_value(u->manager->units, t, u); + + if (u->job) { + Job *j = u->job; + job_uninstall(j); + job_free(j); + } + + if (u->nop_job) { + Job *j = u->nop_job; + job_uninstall(j); + job_free(j); + } + + for (d = 0; d < _UNIT_DEPENDENCY_MAX; d++) + bidi_set_free(u, u->dependencies[d]); + + if (u->type != _UNIT_TYPE_INVALID) + LIST_REMOVE(units_by_type, u->manager->units_by_type[u->type], u); + + if (u->in_load_queue) + LIST_REMOVE(load_queue, u->manager->load_queue, u); + + if (u->in_dbus_queue) + LIST_REMOVE(dbus_queue, u->manager->dbus_unit_queue, u); + + if (u->in_cleanup_queue) + LIST_REMOVE(cleanup_queue, u->manager->cleanup_queue, u); + + if (u->in_gc_queue) { + LIST_REMOVE(gc_queue, u->manager->gc_queue, u); + u->manager->n_in_gc_queue--; + } + + if (u->in_cgroup_queue) + LIST_REMOVE(cgroup_queue, u->manager->cgroup_queue, u); + + unit_release_cgroup(u); + + (void) manager_update_failed_units(u->manager, u, false); + set_remove(u->manager->startup_units, u); + + free(u->description); + strv_free(u->documentation); + free(u->fragment_path); + free(u->source_path); + strv_free(u->dropin_paths); + free(u->instance); + + free(u->job_timeout_reboot_arg); + + set_free_free(u->names); + + unit_unwatch_all_pids(u); + + condition_free_list(u->conditions); + condition_free_list(u->asserts); + + free(u->reboot_arg); + + unit_ref_unset(&u->slice); + + while (u->refs) + unit_ref_unset(u->refs); + + free(u); +} + +UnitActiveState unit_active_state(Unit *u) { + assert(u); + + if (u->load_state == UNIT_MERGED) + return unit_active_state(unit_follow_merge(u)); + + /* After a reload it might happen that a unit is not correctly + * loaded but still has a process around. That's why we won't + * shortcut failed loading to UNIT_INACTIVE_FAILED. */ + + return UNIT_VTABLE(u)->active_state(u); +} + +const char* unit_sub_state_to_string(Unit *u) { + assert(u); + + return UNIT_VTABLE(u)->sub_state_to_string(u); +} + +static int complete_move(Set **s, Set **other) { + int r; + + assert(s); + assert(other); + + if (!*other) + return 0; + + if (*s) { + r = set_move(*s, *other); + if (r < 0) + return r; + } else { + *s = *other; + *other = NULL; + } + + return 0; +} + +static int merge_names(Unit *u, Unit *other) { + char *t; + Iterator i; + int r; + + assert(u); + assert(other); + + r = complete_move(&u->names, &other->names); + if (r < 0) + return r; + + set_free_free(other->names); + other->names = NULL; + other->id = NULL; + + SET_FOREACH(t, u->names, i) + assert_se(hashmap_replace(u->manager->units, t, u) == 0); + + return 0; +} + +static int reserve_dependencies(Unit *u, Unit *other, UnitDependency d) { + unsigned n_reserve; + + assert(u); + assert(other); + assert(d < _UNIT_DEPENDENCY_MAX); + + /* + * If u does not have this dependency set allocated, there is no need + * to reserve anything. In that case other's set will be transferred + * as a whole to u by complete_move(). + */ + if (!u->dependencies[d]) + return 0; + + /* merge_dependencies() will skip a u-on-u dependency */ + n_reserve = set_size(other->dependencies[d]) - !!set_get(other->dependencies[d], u); + + return set_reserve(u->dependencies[d], n_reserve); +} + +static void merge_dependencies(Unit *u, Unit *other, const char *other_id, UnitDependency d) { + Iterator i; + Unit *back; + int r; + + assert(u); + assert(other); + assert(d < _UNIT_DEPENDENCY_MAX); + + /* Fix backwards pointers */ + SET_FOREACH(back, other->dependencies[d], i) { + UnitDependency k; + + for (k = 0; k < _UNIT_DEPENDENCY_MAX; k++) { + /* Do not add dependencies between u and itself */ + if (back == u) { + if (set_remove(back->dependencies[k], other)) + maybe_warn_about_dependency(u, other_id, k); + } else { + r = set_remove_and_put(back->dependencies[k], other, u); + if (r == -EEXIST) + set_remove(back->dependencies[k], other); + else + assert(r >= 0 || r == -ENOENT); + } + } + } + + /* Also do not move dependencies on u to itself */ + back = set_remove(other->dependencies[d], u); + if (back) + maybe_warn_about_dependency(u, other_id, d); + + /* The move cannot fail. The caller must have performed a reservation. */ + assert_se(complete_move(&u->dependencies[d], &other->dependencies[d]) == 0); + + other->dependencies[d] = set_free(other->dependencies[d]); +} + +int unit_merge(Unit *u, Unit *other) { + UnitDependency d; + const char *other_id = NULL; + int r; + + assert(u); + assert(other); + assert(u->manager == other->manager); + assert(u->type != _UNIT_TYPE_INVALID); + + other = unit_follow_merge(other); + + if (other == u) + return 0; + + if (u->type != other->type) + return -EINVAL; + + if (!u->instance != !other->instance) + return -EINVAL; + + if (!unit_type_may_alias(u->type)) /* Merging only applies to unit names that support aliases */ + return -EEXIST; + + if (other->load_state != UNIT_STUB && + other->load_state != UNIT_NOT_FOUND) + return -EEXIST; + + if (other->job) + return -EEXIST; + + if (other->nop_job) + return -EEXIST; + + if (!UNIT_IS_INACTIVE_OR_FAILED(unit_active_state(other))) + return -EEXIST; + + if (other->id) + other_id = strdupa(other->id); + + /* Make reservations to ensure merge_dependencies() won't fail */ + for (d = 0; d < _UNIT_DEPENDENCY_MAX; d++) { + r = reserve_dependencies(u, other, d); + /* + * We don't rollback reservations if we fail. We don't have + * a way to undo reservations. A reservation is not a leak. + */ + if (r < 0) + return r; + } + + /* Merge names */ + r = merge_names(u, other); + if (r < 0) + return r; + + /* Redirect all references */ + while (other->refs) + unit_ref_set(other->refs, u); + + /* Merge dependencies */ + for (d = 0; d < _UNIT_DEPENDENCY_MAX; d++) + merge_dependencies(u, other, other_id, d); + + other->load_state = UNIT_MERGED; + other->merged_into = u; + + /* If there is still some data attached to the other node, we + * don't need it anymore, and can free it. */ + if (other->load_state != UNIT_STUB) + if (UNIT_VTABLE(other)->done) + UNIT_VTABLE(other)->done(other); + + unit_add_to_dbus_queue(u); + unit_add_to_cleanup_queue(other); + + return 0; +} + +int unit_merge_by_name(Unit *u, const char *name) { + _cleanup_free_ char *s = NULL; + Unit *other; + int r; + + assert(u); + assert(name); + + if (unit_name_is_valid(name, UNIT_NAME_TEMPLATE)) { + if (!u->instance) + return -EINVAL; + + r = unit_name_replace_instance(name, u->instance, &s); + if (r < 0) + return r; + + name = s; + } + + other = manager_get_unit(u->manager, name); + if (other) + return unit_merge(u, other); + + return unit_add_name(u, name); +} + +Unit* unit_follow_merge(Unit *u) { + assert(u); + + while (u->load_state == UNIT_MERGED) + assert_se(u = u->merged_into); + + return u; +} + +int unit_add_exec_dependencies(Unit *u, ExecContext *c) { + int r; + + assert(u); + assert(c); + + if (c->working_directory) { + r = unit_require_mounts_for(u, c->working_directory); + if (r < 0) + return r; + } + + if (c->root_directory) { + r = unit_require_mounts_for(u, c->root_directory); + if (r < 0) + return r; + } + + if (!MANAGER_IS_SYSTEM(u->manager)) + return 0; + + if (c->private_tmp) { + r = unit_require_mounts_for(u, "/tmp"); + if (r < 0) + return r; + + r = unit_require_mounts_for(u, "/var/tmp"); + if (r < 0) + return r; + } + + if (c->std_output != EXEC_OUTPUT_KMSG && + c->std_output != EXEC_OUTPUT_SYSLOG && + c->std_output != EXEC_OUTPUT_JOURNAL && + c->std_output != EXEC_OUTPUT_KMSG_AND_CONSOLE && + c->std_output != EXEC_OUTPUT_SYSLOG_AND_CONSOLE && + c->std_output != EXEC_OUTPUT_JOURNAL_AND_CONSOLE && + c->std_error != EXEC_OUTPUT_KMSG && + c->std_error != EXEC_OUTPUT_SYSLOG && + c->std_error != EXEC_OUTPUT_JOURNAL && + c->std_error != EXEC_OUTPUT_KMSG_AND_CONSOLE && + c->std_error != EXEC_OUTPUT_JOURNAL_AND_CONSOLE && + c->std_error != EXEC_OUTPUT_SYSLOG_AND_CONSOLE) + return 0; + + /* If syslog or kernel logging is requested, make sure our own + * logging daemon is run first. */ + + r = unit_add_dependency_by_name(u, UNIT_AFTER, SPECIAL_JOURNALD_SOCKET, NULL, true); + if (r < 0) + return r; + + return 0; +} + +const char *unit_description(Unit *u) { + assert(u); + + if (u->description) + return u->description; + + return strna(u->id); +} + +void unit_dump(Unit *u, FILE *f, const char *prefix) { + char *t, **j; + UnitDependency d; + Iterator i; + const char *prefix2; + char + timestamp0[FORMAT_TIMESTAMP_MAX], + timestamp1[FORMAT_TIMESTAMP_MAX], + timestamp2[FORMAT_TIMESTAMP_MAX], + timestamp3[FORMAT_TIMESTAMP_MAX], + timestamp4[FORMAT_TIMESTAMP_MAX], + timespan[FORMAT_TIMESPAN_MAX]; + Unit *following; + _cleanup_set_free_ Set *following_set = NULL; + int r; + + assert(u); + assert(u->type >= 0); + + prefix = strempty(prefix); + prefix2 = strjoina(prefix, "\t"); + + fprintf(f, + "%s-> Unit %s:\n" + "%s\tDescription: %s\n" + "%s\tInstance: %s\n" + "%s\tUnit Load State: %s\n" + "%s\tUnit Active State: %s\n" + "%s\tState Change Timestamp: %s\n" + "%s\tInactive Exit Timestamp: %s\n" + "%s\tActive Enter Timestamp: %s\n" + "%s\tActive Exit Timestamp: %s\n" + "%s\tInactive Enter Timestamp: %s\n" + "%s\tGC Check Good: %s\n" + "%s\tNeed Daemon Reload: %s\n" + "%s\tTransient: %s\n" + "%s\tSlice: %s\n" + "%s\tCGroup: %s\n" + "%s\tCGroup realized: %s\n" + "%s\tCGroup mask: 0x%x\n" + "%s\tCGroup members mask: 0x%x\n", + prefix, u->id, + prefix, unit_description(u), + prefix, strna(u->instance), + prefix, unit_load_state_to_string(u->load_state), + prefix, unit_active_state_to_string(unit_active_state(u)), + prefix, strna(format_timestamp(timestamp0, sizeof(timestamp0), u->state_change_timestamp.realtime)), + prefix, strna(format_timestamp(timestamp1, sizeof(timestamp1), u->inactive_exit_timestamp.realtime)), + prefix, strna(format_timestamp(timestamp2, sizeof(timestamp2), u->active_enter_timestamp.realtime)), + prefix, strna(format_timestamp(timestamp3, sizeof(timestamp3), u->active_exit_timestamp.realtime)), + prefix, strna(format_timestamp(timestamp4, sizeof(timestamp4), u->inactive_enter_timestamp.realtime)), + prefix, yes_no(unit_check_gc(u)), + prefix, yes_no(unit_need_daemon_reload(u)), + prefix, yes_no(u->transient), + prefix, strna(unit_slice_name(u)), + prefix, strna(u->cgroup_path), + prefix, yes_no(u->cgroup_realized), + prefix, u->cgroup_realized_mask, + prefix, u->cgroup_members_mask); + + SET_FOREACH(t, u->names, i) + fprintf(f, "%s\tName: %s\n", prefix, t); + + STRV_FOREACH(j, u->documentation) + fprintf(f, "%s\tDocumentation: %s\n", prefix, *j); + + following = unit_following(u); + if (following) + fprintf(f, "%s\tFollowing: %s\n", prefix, following->id); + + r = unit_following_set(u, &following_set); + if (r >= 0) { + Unit *other; + + SET_FOREACH(other, following_set, i) + fprintf(f, "%s\tFollowing Set Member: %s\n", prefix, other->id); + } + + if (u->fragment_path) + fprintf(f, "%s\tFragment Path: %s\n", prefix, u->fragment_path); + + if (u->source_path) + fprintf(f, "%s\tSource Path: %s\n", prefix, u->source_path); + + STRV_FOREACH(j, u->dropin_paths) + fprintf(f, "%s\tDropIn Path: %s\n", prefix, *j); + + if (u->job_timeout != USEC_INFINITY) + fprintf(f, "%s\tJob Timeout: %s\n", prefix, format_timespan(timespan, sizeof(timespan), u->job_timeout, 0)); + + if (u->job_timeout_action != FAILURE_ACTION_NONE) + fprintf(f, "%s\tJob Timeout Action: %s\n", prefix, failure_action_to_string(u->job_timeout_action)); + + if (u->job_timeout_reboot_arg) + fprintf(f, "%s\tJob Timeout Reboot Argument: %s\n", prefix, u->job_timeout_reboot_arg); + + condition_dump_list(u->conditions, f, prefix, condition_type_to_string); + condition_dump_list(u->asserts, f, prefix, assert_type_to_string); + + if (dual_timestamp_is_set(&u->condition_timestamp)) + fprintf(f, + "%s\tCondition Timestamp: %s\n" + "%s\tCondition Result: %s\n", + prefix, strna(format_timestamp(timestamp1, sizeof(timestamp1), u->condition_timestamp.realtime)), + prefix, yes_no(u->condition_result)); + + if (dual_timestamp_is_set(&u->assert_timestamp)) + fprintf(f, + "%s\tAssert Timestamp: %s\n" + "%s\tAssert Result: %s\n", + prefix, strna(format_timestamp(timestamp1, sizeof(timestamp1), u->assert_timestamp.realtime)), + prefix, yes_no(u->assert_result)); + + for (d = 0; d < _UNIT_DEPENDENCY_MAX; d++) { + Unit *other; + + SET_FOREACH(other, u->dependencies[d], i) + fprintf(f, "%s\t%s: %s\n", prefix, unit_dependency_to_string(d), other->id); + } + + if (!strv_isempty(u->requires_mounts_for)) { + fprintf(f, + "%s\tRequiresMountsFor:", prefix); + + STRV_FOREACH(j, u->requires_mounts_for) + fprintf(f, " %s", *j); + + fputs("\n", f); + } + + if (u->load_state == UNIT_LOADED) { + + fprintf(f, + "%s\tStopWhenUnneeded: %s\n" + "%s\tRefuseManualStart: %s\n" + "%s\tRefuseManualStop: %s\n" + "%s\tDefaultDependencies: %s\n" + "%s\tOnFailureJobMode: %s\n" + "%s\tIgnoreOnIsolate: %s\n", + prefix, yes_no(u->stop_when_unneeded), + prefix, yes_no(u->refuse_manual_start), + prefix, yes_no(u->refuse_manual_stop), + prefix, yes_no(u->default_dependencies), + prefix, job_mode_to_string(u->on_failure_job_mode), + prefix, yes_no(u->ignore_on_isolate)); + + if (UNIT_VTABLE(u)->dump) + UNIT_VTABLE(u)->dump(u, f, prefix2); + + } else if (u->load_state == UNIT_MERGED) + fprintf(f, + "%s\tMerged into: %s\n", + prefix, u->merged_into->id); + else if (u->load_state == UNIT_ERROR) + fprintf(f, "%s\tLoad Error Code: %s\n", prefix, strerror(-u->load_error)); + + + if (u->job) + job_dump(u->job, f, prefix2); + + if (u->nop_job) + job_dump(u->nop_job, f, prefix2); + +} + +/* Common implementation for multiple backends */ +int unit_load_fragment_and_dropin(Unit *u) { + int r; + + assert(u); + + /* Load a .{service,socket,...} file */ + r = unit_load_fragment(u); + if (r < 0) + return r; + + if (u->load_state == UNIT_STUB) + return -ENOENT; + + /* Load drop-in directory data */ + r = unit_load_dropin(unit_follow_merge(u)); + if (r < 0) + return r; + + return 0; +} + +/* Common implementation for multiple backends */ +int unit_load_fragment_and_dropin_optional(Unit *u) { + int r; + + assert(u); + + /* Same as unit_load_fragment_and_dropin(), but whether + * something can be loaded or not doesn't matter. */ + + /* Load a .service file */ + r = unit_load_fragment(u); + if (r < 0) + return r; + + if (u->load_state == UNIT_STUB) + u->load_state = UNIT_LOADED; + + /* Load drop-in directory data */ + r = unit_load_dropin(unit_follow_merge(u)); + if (r < 0) + return r; + + return 0; +} + +int unit_add_default_target_dependency(Unit *u, Unit *target) { + assert(u); + assert(target); + + if (target->type != UNIT_TARGET) + return 0; + + /* Only add the dependency if both units are loaded, so that + * that loop check below is reliable */ + if (u->load_state != UNIT_LOADED || + target->load_state != UNIT_LOADED) + return 0; + + /* If either side wants no automatic dependencies, then let's + * skip this */ + if (!u->default_dependencies || + !target->default_dependencies) + return 0; + + /* Don't create loops */ + if (set_get(target->dependencies[UNIT_BEFORE], u)) + return 0; + + return unit_add_dependency(target, UNIT_AFTER, u, true); +} + +static int unit_add_target_dependencies(Unit *u) { + + static const UnitDependency deps[] = { + UNIT_REQUIRED_BY, + UNIT_REQUISITE_OF, + UNIT_WANTED_BY, + UNIT_BOUND_BY + }; + + Unit *target; + Iterator i; + unsigned k; + int r = 0; + + assert(u); + + for (k = 0; k < ELEMENTSOF(deps); k++) + SET_FOREACH(target, u->dependencies[deps[k]], i) { + r = unit_add_default_target_dependency(u, target); + if (r < 0) + return r; + } + + return r; +} + +static int unit_add_slice_dependencies(Unit *u) { + assert(u); + + if (!UNIT_HAS_CGROUP_CONTEXT(u)) + return 0; + + if (UNIT_ISSET(u->slice)) + return unit_add_two_dependencies(u, UNIT_AFTER, UNIT_REQUIRES, UNIT_DEREF(u->slice), true); + + if (unit_has_name(u, SPECIAL_ROOT_SLICE)) + return 0; + + return unit_add_two_dependencies_by_name(u, UNIT_AFTER, UNIT_REQUIRES, SPECIAL_ROOT_SLICE, NULL, true); +} + +static int unit_add_mount_dependencies(Unit *u) { + char **i; + int r; + + assert(u); + + STRV_FOREACH(i, u->requires_mounts_for) { + char prefix[strlen(*i) + 1]; + + PATH_FOREACH_PREFIX_MORE(prefix, *i) { + _cleanup_free_ char *p = NULL; + Unit *m; + + r = unit_name_from_path(prefix, ".mount", &p); + if (r < 0) + return r; + + m = manager_get_unit(u->manager, p); + if (!m) { + /* Make sure to load the mount unit if + * it exists. If so the dependencies + * on this unit will be added later + * during the loading of the mount + * unit. */ + (void) manager_load_unit_prepare(u->manager, p, NULL, NULL, &m); + continue; + } + if (m == u) + continue; + + if (m->load_state != UNIT_LOADED) + continue; + + r = unit_add_dependency(u, UNIT_AFTER, m, true); + if (r < 0) + return r; + + if (m->fragment_path) { + r = unit_add_dependency(u, UNIT_REQUIRES, m, true); + if (r < 0) + return r; + } + } + } + + return 0; +} + +static int unit_add_startup_units(Unit *u) { + CGroupContext *c; + int r; + + c = unit_get_cgroup_context(u); + if (!c) + return 0; + + if (c->startup_cpu_shares == CGROUP_CPU_SHARES_INVALID && + c->startup_io_weight == CGROUP_WEIGHT_INVALID && + c->startup_blockio_weight == CGROUP_BLKIO_WEIGHT_INVALID) + return 0; + + r = set_ensure_allocated(&u->manager->startup_units, NULL); + if (r < 0) + return r; + + return set_put(u->manager->startup_units, u); +} + +int unit_load(Unit *u) { + int r; + + assert(u); + + if (u->in_load_queue) { + LIST_REMOVE(load_queue, u->manager->load_queue, u); + u->in_load_queue = false; + } + + if (u->type == _UNIT_TYPE_INVALID) + return -EINVAL; + + if (u->load_state != UNIT_STUB) + return 0; + + if (u->transient_file) { + r = fflush_and_check(u->transient_file); + if (r < 0) + goto fail; + + fclose(u->transient_file); + u->transient_file = NULL; + + u->fragment_mtime = now(CLOCK_REALTIME); + } + + if (UNIT_VTABLE(u)->load) { + r = UNIT_VTABLE(u)->load(u); + if (r < 0) + goto fail; + } + + if (u->load_state == UNIT_STUB) { + r = -ENOENT; + goto fail; + } + + if (u->load_state == UNIT_LOADED) { + + r = unit_add_target_dependencies(u); + if (r < 0) + goto fail; + + r = unit_add_slice_dependencies(u); + if (r < 0) + goto fail; + + r = unit_add_mount_dependencies(u); + if (r < 0) + goto fail; + + r = unit_add_startup_units(u); + if (r < 0) + goto fail; + + if (u->on_failure_job_mode == JOB_ISOLATE && set_size(u->dependencies[UNIT_ON_FAILURE]) > 1) { + log_unit_error(u, "More than one OnFailure= dependencies specified but OnFailureJobMode=isolate set. Refusing."); + r = -EINVAL; + goto fail; + } + + unit_update_cgroup_members_masks(u); + } + + assert((u->load_state != UNIT_MERGED) == !u->merged_into); + + unit_add_to_dbus_queue(unit_follow_merge(u)); + unit_add_to_gc_queue(u); + + return 0; + +fail: + u->load_state = u->load_state == UNIT_STUB ? UNIT_NOT_FOUND : UNIT_ERROR; + u->load_error = r; + unit_add_to_dbus_queue(u); + unit_add_to_gc_queue(u); + + log_unit_debug_errno(u, r, "Failed to load configuration: %m"); + + return r; +} + +static bool unit_condition_test_list(Unit *u, Condition *first, const char *(*to_string)(ConditionType t)) { + Condition *c; + int triggered = -1; + + assert(u); + assert(to_string); + + /* If the condition list is empty, then it is true */ + if (!first) + return true; + + /* Otherwise, if all of the non-trigger conditions apply and + * if any of the trigger conditions apply (unless there are + * none) we return true */ + LIST_FOREACH(conditions, c, first) { + int r; + + r = condition_test(c); + if (r < 0) + log_unit_warning(u, + "Couldn't determine result for %s=%s%s%s, assuming failed: %m", + to_string(c->type), + c->trigger ? "|" : "", + c->negate ? "!" : "", + c->parameter); + else + log_unit_debug(u, + "%s=%s%s%s %s.", + to_string(c->type), + c->trigger ? "|" : "", + c->negate ? "!" : "", + c->parameter, + condition_result_to_string(c->result)); + + if (!c->trigger && r <= 0) + return false; + + if (c->trigger && triggered <= 0) + triggered = r > 0; + } + + return triggered != 0; +} + +static bool unit_condition_test(Unit *u) { + assert(u); + + dual_timestamp_get(&u->condition_timestamp); + u->condition_result = unit_condition_test_list(u, u->conditions, condition_type_to_string); + + return u->condition_result; +} + +static bool unit_assert_test(Unit *u) { + assert(u); + + dual_timestamp_get(&u->assert_timestamp); + u->assert_result = unit_condition_test_list(u, u->asserts, assert_type_to_string); + + return u->assert_result; +} + +void unit_status_printf(Unit *u, const char *status, const char *unit_status_msg_format) { + DISABLE_WARNING_FORMAT_NONLITERAL; + manager_status_printf(u->manager, STATUS_TYPE_NORMAL, status, unit_status_msg_format, unit_description(u)); + REENABLE_WARNING; +} + +_pure_ static const char* unit_get_status_message_format(Unit *u, JobType t) { + const char *format; + const UnitStatusMessageFormats *format_table; + + assert(u); + assert(IN_SET(t, JOB_START, JOB_STOP, JOB_RELOAD)); + + if (t != JOB_RELOAD) { + format_table = &UNIT_VTABLE(u)->status_message_formats; + if (format_table) { + format = format_table->starting_stopping[t == JOB_STOP]; + if (format) + return format; + } + } + + /* Return generic strings */ + if (t == JOB_START) + return "Starting %s."; + else if (t == JOB_STOP) + return "Stopping %s."; + else + return "Reloading %s."; +} + +static void unit_status_print_starting_stopping(Unit *u, JobType t) { + const char *format; + + assert(u); + + /* Reload status messages have traditionally not been printed to console. */ + if (!IN_SET(t, JOB_START, JOB_STOP)) + return; + + format = unit_get_status_message_format(u, t); + + DISABLE_WARNING_FORMAT_NONLITERAL; + unit_status_printf(u, "", format); + REENABLE_WARNING; +} + +static void unit_status_log_starting_stopping_reloading(Unit *u, JobType t) { + const char *format; + char buf[LINE_MAX]; + sd_id128_t mid; + + assert(u); + + if (!IN_SET(t, JOB_START, JOB_STOP, JOB_RELOAD)) + return; + + if (log_on_console()) + return; + + /* We log status messages for all units and all operations. */ + + format = unit_get_status_message_format(u, t); + + DISABLE_WARNING_FORMAT_NONLITERAL; + xsprintf(buf, format, unit_description(u)); + REENABLE_WARNING; + + mid = t == JOB_START ? SD_MESSAGE_UNIT_STARTING : + t == JOB_STOP ? SD_MESSAGE_UNIT_STOPPING : + SD_MESSAGE_UNIT_RELOADING; + + /* Note that we deliberately use LOG_MESSAGE() instead of + * LOG_UNIT_MESSAGE() here, since this is supposed to mimic + * closely what is written to screen using the status output, + * which is supposed the highest level, friendliest output + * possible, which means we should avoid the low-level unit + * name. */ + log_struct(LOG_INFO, + LOG_MESSAGE_ID(mid), + LOG_UNIT_ID(u), + LOG_MESSAGE("%s", buf), + NULL); +} + +void unit_status_emit_starting_stopping_reloading(Unit *u, JobType t) { + assert(u); + assert(t >= 0); + assert(t < _JOB_TYPE_MAX); + + unit_status_log_starting_stopping_reloading(u, t); + unit_status_print_starting_stopping(u, t); +} + +int unit_start_limit_test(Unit *u) { + assert(u); + + if (ratelimit_test(&u->start_limit)) { + u->start_limit_hit = false; + return 0; + } + + log_unit_warning(u, "Start request repeated too quickly."); + u->start_limit_hit = true; + + return failure_action(u->manager, u->start_limit_action, u->reboot_arg); +} + +/* Errors: + * -EBADR: This unit type does not support starting. + * -EALREADY: Unit is already started. + * -EAGAIN: An operation is already in progress. Retry later. + * -ECANCELED: Too many requests for now. + * -EPROTO: Assert failed + * -EINVAL: Unit not loaded + * -EOPNOTSUPP: Unit type not supported + */ +int unit_start(Unit *u) { + UnitActiveState state; + Unit *following; + + assert(u); + + /* If this is already started, then this will succeed. Note + * that this will even succeed if this unit is not startable + * by the user. This is relied on to detect when we need to + * wait for units and when waiting is finished. */ + state = unit_active_state(u); + if (UNIT_IS_ACTIVE_OR_RELOADING(state)) + return -EALREADY; + + /* Units that aren't loaded cannot be started */ + if (u->load_state != UNIT_LOADED) + return -EINVAL; + + /* If the conditions failed, don't do anything at all. If we + * already are activating this call might still be useful to + * speed up activation in case there is some hold-off time, + * but we don't want to recheck the condition in that case. */ + if (state != UNIT_ACTIVATING && + !unit_condition_test(u)) { + log_unit_debug(u, "Starting requested but condition failed. Not starting unit."); + return -EALREADY; + } + + /* If the asserts failed, fail the entire job */ + if (state != UNIT_ACTIVATING && + !unit_assert_test(u)) { + log_unit_notice(u, "Starting requested but asserts failed."); + return -EPROTO; + } + + /* Units of types that aren't supported cannot be + * started. Note that we do this test only after the condition + * checks, so that we rather return condition check errors + * (which are usually not considered a true failure) than "not + * supported" errors (which are considered a failure). + */ + if (!unit_supported(u)) + return -EOPNOTSUPP; + + /* Forward to the main object, if we aren't it. */ + following = unit_following(u); + if (following) { + log_unit_debug(u, "Redirecting start request from %s to %s.", u->id, following->id); + return unit_start(following); + } + + /* If it is stopped, but we cannot start it, then fail */ + if (!UNIT_VTABLE(u)->start) + return -EBADR; + + /* We don't suppress calls to ->start() here when we are + * already starting, to allow this request to be used as a + * "hurry up" call, for example when the unit is in some "auto + * restart" state where it waits for a holdoff timer to elapse + * before it will start again. */ + + unit_add_to_dbus_queue(u); + + return UNIT_VTABLE(u)->start(u); +} + +bool unit_can_start(Unit *u) { + assert(u); + + if (u->load_state != UNIT_LOADED) + return false; + + if (!unit_supported(u)) + return false; + + return !!UNIT_VTABLE(u)->start; +} + +bool unit_can_isolate(Unit *u) { + assert(u); + + return unit_can_start(u) && + u->allow_isolate; +} + +/* Errors: + * -EBADR: This unit type does not support stopping. + * -EALREADY: Unit is already stopped. + * -EAGAIN: An operation is already in progress. Retry later. + */ +int unit_stop(Unit *u) { + UnitActiveState state; + Unit *following; + + assert(u); + + state = unit_active_state(u); + if (UNIT_IS_INACTIVE_OR_FAILED(state)) + return -EALREADY; + + following = unit_following(u); + if (following) { + log_unit_debug(u, "Redirecting stop request from %s to %s.", u->id, following->id); + return unit_stop(following); + } + + if (!UNIT_VTABLE(u)->stop) + return -EBADR; + + unit_add_to_dbus_queue(u); + + return UNIT_VTABLE(u)->stop(u); +} + +/* Errors: + * -EBADR: This unit type does not support reloading. + * -ENOEXEC: Unit is not started. + * -EAGAIN: An operation is already in progress. Retry later. + */ +int unit_reload(Unit *u) { + UnitActiveState state; + Unit *following; + + assert(u); + + if (u->load_state != UNIT_LOADED) + return -EINVAL; + + if (!unit_can_reload(u)) + return -EBADR; + + state = unit_active_state(u); + if (state == UNIT_RELOADING) + return -EALREADY; + + if (state != UNIT_ACTIVE) { + log_unit_warning(u, "Unit cannot be reloaded because it is inactive."); + return -ENOEXEC; + } + + following = unit_following(u); + if (following) { + log_unit_debug(u, "Redirecting reload request from %s to %s.", u->id, following->id); + return unit_reload(following); + } + + unit_add_to_dbus_queue(u); + + return UNIT_VTABLE(u)->reload(u); +} + +bool unit_can_reload(Unit *u) { + assert(u); + + if (!UNIT_VTABLE(u)->reload) + return false; + + if (!UNIT_VTABLE(u)->can_reload) + return true; + + return UNIT_VTABLE(u)->can_reload(u); +} + +static void unit_check_unneeded(Unit *u) { + + _cleanup_(sd_bus_error_free) sd_bus_error error = SD_BUS_ERROR_NULL; + + static const UnitDependency needed_dependencies[] = { + UNIT_REQUIRED_BY, + UNIT_REQUISITE_OF, + UNIT_WANTED_BY, + UNIT_BOUND_BY, + }; + + Unit *other; + Iterator i; + unsigned j; + int r; + + assert(u); + + /* If this service shall be shut down when unneeded then do + * so. */ + + if (!u->stop_when_unneeded) + return; + + if (!UNIT_IS_ACTIVE_OR_ACTIVATING(unit_active_state(u))) + return; + + for (j = 0; j < ELEMENTSOF(needed_dependencies); j++) + SET_FOREACH(other, u->dependencies[needed_dependencies[j]], i) + if (unit_active_or_pending(other)) + return; + + /* If stopping a unit fails continuously we might enter a stop + * loop here, hence stop acting on the service being + * unnecessary after a while. */ + if (!ratelimit_test(&u->auto_stop_ratelimit)) { + log_unit_warning(u, "Unit not needed anymore, but not stopping since we tried this too often recently."); + return; + } + + log_unit_info(u, "Unit not needed anymore. Stopping."); + + /* Ok, nobody needs us anymore. Sniff. Then let's commit suicide */ + r = manager_add_job(u->manager, JOB_STOP, u, JOB_FAIL, &error, NULL); + if (r < 0) + log_unit_warning_errno(u, r, "Failed to enqueue stop job, ignoring: %s", bus_error_message(&error, r)); +} + +static void unit_check_binds_to(Unit *u) { + _cleanup_(sd_bus_error_free) sd_bus_error error = SD_BUS_ERROR_NULL; + bool stop = false; + Unit *other; + Iterator i; + int r; + + assert(u); + + if (u->job) + return; + + if (unit_active_state(u) != UNIT_ACTIVE) + return; + + SET_FOREACH(other, u->dependencies[UNIT_BINDS_TO], i) { + if (other->job) + continue; + + if (!UNIT_IS_INACTIVE_OR_FAILED(unit_active_state(other))) + continue; + + stop = true; + break; + } + + if (!stop) + return; + + /* If stopping a unit fails continuously we might enter a stop + * loop here, hence stop acting on the service being + * unnecessary after a while. */ + if (!ratelimit_test(&u->auto_stop_ratelimit)) { + log_unit_warning(u, "Unit is bound to inactive unit %s, but not stopping since we tried this too often recently.", other->id); + return; + } + + assert(other); + log_unit_info(u, "Unit is bound to inactive unit %s. Stopping, too.", other->id); + + /* A unit we need to run is gone. Sniff. Let's stop this. */ + r = manager_add_job(u->manager, JOB_STOP, u, JOB_FAIL, &error, NULL); + if (r < 0) + log_unit_warning_errno(u, r, "Failed to enqueue stop job, ignoring: %s", bus_error_message(&error, r)); +} + +static void retroactively_start_dependencies(Unit *u) { + Iterator i; + Unit *other; + + assert(u); + assert(UNIT_IS_ACTIVE_OR_ACTIVATING(unit_active_state(u))); + + SET_FOREACH(other, u->dependencies[UNIT_REQUIRES], i) + if (!set_get(u->dependencies[UNIT_AFTER], other) && + !UNIT_IS_ACTIVE_OR_ACTIVATING(unit_active_state(other))) + manager_add_job(u->manager, JOB_START, other, JOB_REPLACE, NULL, NULL); + + SET_FOREACH(other, u->dependencies[UNIT_BINDS_TO], i) + if (!set_get(u->dependencies[UNIT_AFTER], other) && + !UNIT_IS_ACTIVE_OR_ACTIVATING(unit_active_state(other))) + manager_add_job(u->manager, JOB_START, other, JOB_REPLACE, NULL, NULL); + + SET_FOREACH(other, u->dependencies[UNIT_WANTS], i) + if (!set_get(u->dependencies[UNIT_AFTER], other) && + !UNIT_IS_ACTIVE_OR_ACTIVATING(unit_active_state(other))) + manager_add_job(u->manager, JOB_START, other, JOB_FAIL, NULL, NULL); + + SET_FOREACH(other, u->dependencies[UNIT_CONFLICTS], i) + if (!UNIT_IS_INACTIVE_OR_DEACTIVATING(unit_active_state(other))) + manager_add_job(u->manager, JOB_STOP, other, JOB_REPLACE, NULL, NULL); + + SET_FOREACH(other, u->dependencies[UNIT_CONFLICTED_BY], i) + if (!UNIT_IS_INACTIVE_OR_DEACTIVATING(unit_active_state(other))) + manager_add_job(u->manager, JOB_STOP, other, JOB_REPLACE, NULL, NULL); +} + +static void retroactively_stop_dependencies(Unit *u) { + Iterator i; + Unit *other; + + assert(u); + assert(UNIT_IS_INACTIVE_OR_DEACTIVATING(unit_active_state(u))); + + /* Pull down units which are bound to us recursively if enabled */ + SET_FOREACH(other, u->dependencies[UNIT_BOUND_BY], i) + if (!UNIT_IS_INACTIVE_OR_DEACTIVATING(unit_active_state(other))) + manager_add_job(u->manager, JOB_STOP, other, JOB_REPLACE, NULL, NULL); +} + +static void check_unneeded_dependencies(Unit *u) { + Iterator i; + Unit *other; + + assert(u); + assert(UNIT_IS_INACTIVE_OR_DEACTIVATING(unit_active_state(u))); + + /* Garbage collect services that might not be needed anymore, if enabled */ + SET_FOREACH(other, u->dependencies[UNIT_REQUIRES], i) + if (!UNIT_IS_INACTIVE_OR_DEACTIVATING(unit_active_state(other))) + unit_check_unneeded(other); + SET_FOREACH(other, u->dependencies[UNIT_WANTS], i) + if (!UNIT_IS_INACTIVE_OR_DEACTIVATING(unit_active_state(other))) + unit_check_unneeded(other); + SET_FOREACH(other, u->dependencies[UNIT_REQUISITE], i) + if (!UNIT_IS_INACTIVE_OR_DEACTIVATING(unit_active_state(other))) + unit_check_unneeded(other); + SET_FOREACH(other, u->dependencies[UNIT_BINDS_TO], i) + if (!UNIT_IS_INACTIVE_OR_DEACTIVATING(unit_active_state(other))) + unit_check_unneeded(other); +} + +void unit_start_on_failure(Unit *u) { + Unit *other; + Iterator i; + + assert(u); + + if (set_size(u->dependencies[UNIT_ON_FAILURE]) <= 0) + return; + + log_unit_info(u, "Triggering OnFailure= dependencies."); + + SET_FOREACH(other, u->dependencies[UNIT_ON_FAILURE], i) { + int r; + + r = manager_add_job(u->manager, JOB_START, other, u->on_failure_job_mode, NULL, NULL); + if (r < 0) + log_unit_error_errno(u, r, "Failed to enqueue OnFailure= job: %m"); + } +} + +void unit_trigger_notify(Unit *u) { + Unit *other; + Iterator i; + + assert(u); + + SET_FOREACH(other, u->dependencies[UNIT_TRIGGERED_BY], i) + if (UNIT_VTABLE(other)->trigger_notify) + UNIT_VTABLE(other)->trigger_notify(other, u); +} + +void unit_notify(Unit *u, UnitActiveState os, UnitActiveState ns, bool reload_success) { + Manager *m; + bool unexpected; + + assert(u); + assert(os < _UNIT_ACTIVE_STATE_MAX); + assert(ns < _UNIT_ACTIVE_STATE_MAX); + + /* Note that this is called for all low-level state changes, + * even if they might map to the same high-level + * UnitActiveState! That means that ns == os is an expected + * behavior here. For example: if a mount point is remounted + * this function will be called too! */ + + m = u->manager; + + /* Update timestamps for state changes */ + if (!MANAGER_IS_RELOADING(m)) { + dual_timestamp_get(&u->state_change_timestamp); + + if (UNIT_IS_INACTIVE_OR_FAILED(os) && !UNIT_IS_INACTIVE_OR_FAILED(ns)) + u->inactive_exit_timestamp = u->state_change_timestamp; + else if (!UNIT_IS_INACTIVE_OR_FAILED(os) && UNIT_IS_INACTIVE_OR_FAILED(ns)) + u->inactive_enter_timestamp = u->state_change_timestamp; + + if (!UNIT_IS_ACTIVE_OR_RELOADING(os) && UNIT_IS_ACTIVE_OR_RELOADING(ns)) + u->active_enter_timestamp = u->state_change_timestamp; + else if (UNIT_IS_ACTIVE_OR_RELOADING(os) && !UNIT_IS_ACTIVE_OR_RELOADING(ns)) + u->active_exit_timestamp = u->state_change_timestamp; + } + + /* Keep track of failed units */ + (void) manager_update_failed_units(u->manager, u, ns == UNIT_FAILED); + + /* Make sure the cgroup is always removed when we become inactive */ + if (UNIT_IS_INACTIVE_OR_FAILED(ns)) + unit_prune_cgroup(u); + + /* Note that this doesn't apply to RemainAfterExit services exiting + * successfully, since there's no change of state in that case. Which is + * why it is handled in service_set_state() */ + if (UNIT_IS_INACTIVE_OR_FAILED(os) != UNIT_IS_INACTIVE_OR_FAILED(ns)) { + ExecContext *ec; + + ec = unit_get_exec_context(u); + if (ec && exec_context_may_touch_console(ec)) { + if (UNIT_IS_INACTIVE_OR_FAILED(ns)) { + m->n_on_console--; + + if (m->n_on_console == 0) + /* unset no_console_output flag, since the console is free */ + m->no_console_output = false; + } else + m->n_on_console++; + } + } + + if (u->job) { + unexpected = false; + + if (u->job->state == JOB_WAITING) + + /* So we reached a different state for this + * job. Let's see if we can run it now if it + * failed previously due to EAGAIN. */ + job_add_to_run_queue(u->job); + + /* Let's check whether this state change constitutes a + * finished job, or maybe contradicts a running job and + * hence needs to invalidate jobs. */ + + switch (u->job->type) { + + case JOB_START: + case JOB_VERIFY_ACTIVE: + + if (UNIT_IS_ACTIVE_OR_RELOADING(ns)) + job_finish_and_invalidate(u->job, JOB_DONE, true, false); + else if (u->job->state == JOB_RUNNING && ns != UNIT_ACTIVATING) { + unexpected = true; + + if (UNIT_IS_INACTIVE_OR_FAILED(ns)) + job_finish_and_invalidate(u->job, ns == UNIT_FAILED ? JOB_FAILED : JOB_DONE, true, false); + } + + break; + + case JOB_RELOAD: + case JOB_RELOAD_OR_START: + case JOB_TRY_RELOAD: + + if (u->job->state == JOB_RUNNING) { + if (ns == UNIT_ACTIVE) + job_finish_and_invalidate(u->job, reload_success ? JOB_DONE : JOB_FAILED, true, false); + else if (ns != UNIT_ACTIVATING && ns != UNIT_RELOADING) { + unexpected = true; + + if (UNIT_IS_INACTIVE_OR_FAILED(ns)) + job_finish_and_invalidate(u->job, ns == UNIT_FAILED ? JOB_FAILED : JOB_DONE, true, false); + } + } + + break; + + case JOB_STOP: + case JOB_RESTART: + case JOB_TRY_RESTART: + + if (UNIT_IS_INACTIVE_OR_FAILED(ns)) + job_finish_and_invalidate(u->job, JOB_DONE, true, false); + else if (u->job->state == JOB_RUNNING && ns != UNIT_DEACTIVATING) { + unexpected = true; + job_finish_and_invalidate(u->job, JOB_FAILED, true, false); + } + + break; + + default: + assert_not_reached("Job type unknown"); + } + + } else + unexpected = true; + + if (!MANAGER_IS_RELOADING(m)) { + + /* If this state change happened without being + * requested by a job, then let's retroactively start + * or stop dependencies. We skip that step when + * deserializing, since we don't want to create any + * additional jobs just because something is already + * activated. */ + + if (unexpected) { + if (UNIT_IS_INACTIVE_OR_FAILED(os) && UNIT_IS_ACTIVE_OR_ACTIVATING(ns)) + retroactively_start_dependencies(u); + else if (UNIT_IS_ACTIVE_OR_ACTIVATING(os) && UNIT_IS_INACTIVE_OR_DEACTIVATING(ns)) + retroactively_stop_dependencies(u); + } + + /* stop unneeded units regardless if going down was expected or not */ + if (UNIT_IS_INACTIVE_OR_DEACTIVATING(ns)) + check_unneeded_dependencies(u); + + if (ns != os && ns == UNIT_FAILED) { + log_unit_notice(u, "Unit entered failed state."); + unit_start_on_failure(u); + } + } + + /* Some names are special */ + if (UNIT_IS_ACTIVE_OR_RELOADING(ns)) { + + if (unit_has_name(u, SPECIAL_DBUS_SERVICE)) + /* The bus might have just become available, + * hence try to connect to it, if we aren't + * yet connected. */ + bus_init(m, true); + + if (u->type == UNIT_SERVICE && + !UNIT_IS_ACTIVE_OR_RELOADING(os) && + !MANAGER_IS_RELOADING(m)) { + /* Write audit record if we have just finished starting up */ + manager_send_unit_audit(m, u, AUDIT_SERVICE_START, true); + u->in_audit = true; + } + + if (!UNIT_IS_ACTIVE_OR_RELOADING(os)) + manager_send_unit_plymouth(m, u); + + } else { + + /* We don't care about D-Bus here, since we'll get an + * asynchronous notification for it anyway. */ + + if (u->type == UNIT_SERVICE && + UNIT_IS_INACTIVE_OR_FAILED(ns) && + !UNIT_IS_INACTIVE_OR_FAILED(os) && + !MANAGER_IS_RELOADING(m)) { + + /* Hmm, if there was no start record written + * write it now, so that we always have a nice + * pair */ + if (!u->in_audit) { + manager_send_unit_audit(m, u, AUDIT_SERVICE_START, ns == UNIT_INACTIVE); + + if (ns == UNIT_INACTIVE) + manager_send_unit_audit(m, u, AUDIT_SERVICE_STOP, true); + } else + /* Write audit record if we have just finished shutting down */ + manager_send_unit_audit(m, u, AUDIT_SERVICE_STOP, ns == UNIT_INACTIVE); + + u->in_audit = false; + } + } + + manager_recheck_journal(m); + unit_trigger_notify(u); + + if (!MANAGER_IS_RELOADING(u->manager)) { + /* Maybe we finished startup and are now ready for + * being stopped because unneeded? */ + unit_check_unneeded(u); + + /* Maybe we finished startup, but something we needed + * has vanished? Let's die then. (This happens when + * something BindsTo= to a Type=oneshot unit, as these + * units go directly from starting to inactive, + * without ever entering started.) */ + unit_check_binds_to(u); + } + + unit_add_to_dbus_queue(u); + unit_add_to_gc_queue(u); +} + +int unit_watch_pid(Unit *u, pid_t pid) { + int q, r; + + assert(u); + assert(pid >= 1); + + /* Watch a specific PID. We only support one or two units + * watching each PID for now, not more. */ + + r = set_ensure_allocated(&u->pids, NULL); + if (r < 0) + return r; + + r = hashmap_ensure_allocated(&u->manager->watch_pids1, NULL); + if (r < 0) + return r; + + r = hashmap_put(u->manager->watch_pids1, PID_TO_PTR(pid), u); + if (r == -EEXIST) { + r = hashmap_ensure_allocated(&u->manager->watch_pids2, NULL); + if (r < 0) + return r; + + r = hashmap_put(u->manager->watch_pids2, PID_TO_PTR(pid), u); + } + + q = set_put(u->pids, PID_TO_PTR(pid)); + if (q < 0) + return q; + + return r; +} + +void unit_unwatch_pid(Unit *u, pid_t pid) { + assert(u); + assert(pid >= 1); + + (void) hashmap_remove_value(u->manager->watch_pids1, PID_TO_PTR(pid), u); + (void) hashmap_remove_value(u->manager->watch_pids2, PID_TO_PTR(pid), u); + (void) set_remove(u->pids, PID_TO_PTR(pid)); +} + +void unit_unwatch_all_pids(Unit *u) { + assert(u); + + while (!set_isempty(u->pids)) + unit_unwatch_pid(u, PTR_TO_PID(set_first(u->pids))); + + u->pids = set_free(u->pids); +} + +void unit_tidy_watch_pids(Unit *u, pid_t except1, pid_t except2) { + Iterator i; + void *e; + + assert(u); + + /* Cleans dead PIDs from our list */ + + SET_FOREACH(e, u->pids, i) { + pid_t pid = PTR_TO_PID(e); + + if (pid == except1 || pid == except2) + continue; + + if (!pid_is_unwaited(pid)) + unit_unwatch_pid(u, pid); + } +} + +bool unit_job_is_applicable(Unit *u, JobType j) { + assert(u); + assert(j >= 0 && j < _JOB_TYPE_MAX); + + switch (j) { + + case JOB_VERIFY_ACTIVE: + case JOB_START: + case JOB_STOP: + case JOB_NOP: + return true; + + case JOB_RESTART: + case JOB_TRY_RESTART: + return unit_can_start(u); + + case JOB_RELOAD: + case JOB_TRY_RELOAD: + return unit_can_reload(u); + + case JOB_RELOAD_OR_START: + return unit_can_reload(u) && unit_can_start(u); + + default: + assert_not_reached("Invalid job type"); + } +} + +static void maybe_warn_about_dependency(Unit *u, const char *other, UnitDependency dependency) { + assert(u); + + /* Only warn about some unit types */ + if (!IN_SET(dependency, UNIT_CONFLICTS, UNIT_CONFLICTED_BY, UNIT_BEFORE, UNIT_AFTER, UNIT_ON_FAILURE, UNIT_TRIGGERS, UNIT_TRIGGERED_BY)) + return; + + if (streq_ptr(u->id, other)) + log_unit_warning(u, "Dependency %s=%s dropped", unit_dependency_to_string(dependency), u->id); + else + log_unit_warning(u, "Dependency %s=%s dropped, merged into %s", unit_dependency_to_string(dependency), strna(other), u->id); +} + +int unit_add_dependency(Unit *u, UnitDependency d, Unit *other, bool add_reference) { + + static const UnitDependency inverse_table[_UNIT_DEPENDENCY_MAX] = { + [UNIT_REQUIRES] = UNIT_REQUIRED_BY, + [UNIT_WANTS] = UNIT_WANTED_BY, + [UNIT_REQUISITE] = UNIT_REQUISITE_OF, + [UNIT_BINDS_TO] = UNIT_BOUND_BY, + [UNIT_PART_OF] = UNIT_CONSISTS_OF, + [UNIT_REQUIRED_BY] = UNIT_REQUIRES, + [UNIT_REQUISITE_OF] = UNIT_REQUISITE, + [UNIT_WANTED_BY] = UNIT_WANTS, + [UNIT_BOUND_BY] = UNIT_BINDS_TO, + [UNIT_CONSISTS_OF] = UNIT_PART_OF, + [UNIT_CONFLICTS] = UNIT_CONFLICTED_BY, + [UNIT_CONFLICTED_BY] = UNIT_CONFLICTS, + [UNIT_BEFORE] = UNIT_AFTER, + [UNIT_AFTER] = UNIT_BEFORE, + [UNIT_ON_FAILURE] = _UNIT_DEPENDENCY_INVALID, + [UNIT_REFERENCES] = UNIT_REFERENCED_BY, + [UNIT_REFERENCED_BY] = UNIT_REFERENCES, + [UNIT_TRIGGERS] = UNIT_TRIGGERED_BY, + [UNIT_TRIGGERED_BY] = UNIT_TRIGGERS, + [UNIT_PROPAGATES_RELOAD_TO] = UNIT_RELOAD_PROPAGATED_FROM, + [UNIT_RELOAD_PROPAGATED_FROM] = UNIT_PROPAGATES_RELOAD_TO, + [UNIT_JOINS_NAMESPACE_OF] = UNIT_JOINS_NAMESPACE_OF, + }; + int r, q = 0, v = 0, w = 0; + Unit *orig_u = u, *orig_other = other; + + assert(u); + assert(d >= 0 && d < _UNIT_DEPENDENCY_MAX); + assert(other); + + u = unit_follow_merge(u); + other = unit_follow_merge(other); + + /* We won't allow dependencies on ourselves. We will not + * consider them an error however. */ + if (u == other) { + maybe_warn_about_dependency(orig_u, orig_other->id, d); + return 0; + } + + r = set_ensure_allocated(&u->dependencies[d], NULL); + if (r < 0) + return r; + + if (inverse_table[d] != _UNIT_DEPENDENCY_INVALID) { + r = set_ensure_allocated(&other->dependencies[inverse_table[d]], NULL); + if (r < 0) + return r; + } + + if (add_reference) { + r = set_ensure_allocated(&u->dependencies[UNIT_REFERENCES], NULL); + if (r < 0) + return r; + + r = set_ensure_allocated(&other->dependencies[UNIT_REFERENCED_BY], NULL); + if (r < 0) + return r; + } + + q = set_put(u->dependencies[d], other); + if (q < 0) + return q; + + if (inverse_table[d] != _UNIT_DEPENDENCY_INVALID && inverse_table[d] != d) { + v = set_put(other->dependencies[inverse_table[d]], u); + if (v < 0) { + r = v; + goto fail; + } + } + + if (add_reference) { + w = set_put(u->dependencies[UNIT_REFERENCES], other); + if (w < 0) { + r = w; + goto fail; + } + + r = set_put(other->dependencies[UNIT_REFERENCED_BY], u); + if (r < 0) + goto fail; + } + + unit_add_to_dbus_queue(u); + return 0; + +fail: + if (q > 0) + set_remove(u->dependencies[d], other); + + if (v > 0) + set_remove(other->dependencies[inverse_table[d]], u); + + if (w > 0) + set_remove(u->dependencies[UNIT_REFERENCES], other); + + return r; +} + +int unit_add_two_dependencies(Unit *u, UnitDependency d, UnitDependency e, Unit *other, bool add_reference) { + int r; + + assert(u); + + r = unit_add_dependency(u, d, other, add_reference); + if (r < 0) + return r; + + return unit_add_dependency(u, e, other, add_reference); +} + +static int resolve_template(Unit *u, const char *name, const char*path, char **buf, const char **ret) { + int r; + + assert(u); + assert(name || path); + assert(buf); + assert(ret); + + if (!name) + name = basename(path); + + if (!unit_name_is_valid(name, UNIT_NAME_TEMPLATE)) { + *buf = NULL; + *ret = name; + return 0; + } + + if (u->instance) + r = unit_name_replace_instance(name, u->instance, buf); + else { + _cleanup_free_ char *i = NULL; + + r = unit_name_to_prefix(u->id, &i); + if (r < 0) + return r; + + r = unit_name_replace_instance(name, i, buf); + } + if (r < 0) + return r; + + *ret = *buf; + return 0; +} + +int unit_add_dependency_by_name(Unit *u, UnitDependency d, const char *name, const char *path, bool add_reference) { + _cleanup_free_ char *buf = NULL; + Unit *other; + int r; + + assert(u); + assert(name || path); + + r = resolve_template(u, name, path, &buf, &name); + if (r < 0) + return r; + + r = manager_load_unit(u->manager, name, path, NULL, &other); + if (r < 0) + return r; + + return unit_add_dependency(u, d, other, add_reference); +} + +int unit_add_two_dependencies_by_name(Unit *u, UnitDependency d, UnitDependency e, const char *name, const char *path, bool add_reference) { + _cleanup_free_ char *buf = NULL; + Unit *other; + int r; + + assert(u); + assert(name || path); + + r = resolve_template(u, name, path, &buf, &name); + if (r < 0) + return r; + + r = manager_load_unit(u->manager, name, path, NULL, &other); + if (r < 0) + return r; + + return unit_add_two_dependencies(u, d, e, other, add_reference); +} + +int set_unit_path(const char *p) { + /* This is mostly for debug purposes */ + if (setenv("SYSTEMD_UNIT_PATH", p, 1) < 0) + return -errno; + + return 0; +} + +char *unit_dbus_path(Unit *u) { + assert(u); + + if (!u->id) + return NULL; + + return unit_dbus_path_from_name(u->id); +} + +int unit_set_slice(Unit *u, Unit *slice) { + assert(u); + assert(slice); + + /* Sets the unit slice if it has not been set before. Is extra + * careful, to only allow this for units that actually have a + * cgroup context. Also, we don't allow to set this for slices + * (since the parent slice is derived from the name). Make + * sure the unit we set is actually a slice. */ + + if (!UNIT_HAS_CGROUP_CONTEXT(u)) + return -EOPNOTSUPP; + + if (u->type == UNIT_SLICE) + return -EINVAL; + + if (unit_active_state(u) != UNIT_INACTIVE) + return -EBUSY; + + if (slice->type != UNIT_SLICE) + return -EINVAL; + + if (unit_has_name(u, SPECIAL_INIT_SCOPE) && + !unit_has_name(slice, SPECIAL_ROOT_SLICE)) + return -EPERM; + + if (UNIT_DEREF(u->slice) == slice) + return 0; + + /* Disallow slice changes if @u is already bound to cgroups */ + if (UNIT_ISSET(u->slice) && u->cgroup_realized) + return -EBUSY; + + unit_ref_unset(&u->slice); + unit_ref_set(&u->slice, slice); + return 1; +} + +int unit_set_default_slice(Unit *u) { + _cleanup_free_ char *b = NULL; + const char *slice_name; + Unit *slice; + int r; + + assert(u); + + if (UNIT_ISSET(u->slice)) + return 0; + + if (u->instance) { + _cleanup_free_ char *prefix = NULL, *escaped = NULL; + + /* Implicitly place all instantiated units in their + * own per-template slice */ + + r = unit_name_to_prefix(u->id, &prefix); + if (r < 0) + return r; + + /* The prefix is already escaped, but it might include + * "-" which has a special meaning for slice units, + * hence escape it here extra. */ + escaped = unit_name_escape(prefix); + if (!escaped) + return -ENOMEM; + + if (MANAGER_IS_SYSTEM(u->manager)) + b = strjoin("system-", escaped, ".slice", NULL); + else + b = strappend(escaped, ".slice"); + if (!b) + return -ENOMEM; + + slice_name = b; + } else + slice_name = + MANAGER_IS_SYSTEM(u->manager) && !unit_has_name(u, SPECIAL_INIT_SCOPE) + ? SPECIAL_SYSTEM_SLICE + : SPECIAL_ROOT_SLICE; + + r = manager_load_unit(u->manager, slice_name, NULL, NULL, &slice); + if (r < 0) + return r; + + return unit_set_slice(u, slice); +} + +const char *unit_slice_name(Unit *u) { + assert(u); + + if (!UNIT_ISSET(u->slice)) + return NULL; + + return UNIT_DEREF(u->slice)->id; +} + +int unit_load_related_unit(Unit *u, const char *type, Unit **_found) { + _cleanup_free_ char *t = NULL; + int r; + + assert(u); + assert(type); + assert(_found); + + r = unit_name_change_suffix(u->id, type, &t); + if (r < 0) + return r; + if (unit_has_name(u, t)) + return -EINVAL; + + r = manager_load_unit(u->manager, t, NULL, NULL, _found); + assert(r < 0 || *_found != u); + return r; +} + +static int signal_name_owner_changed(sd_bus_message *message, void *userdata, sd_bus_error *error) { + const char *name, *old_owner, *new_owner; + Unit *u = userdata; + int r; + + assert(message); + assert(u); + + r = sd_bus_message_read(message, "sss", &name, &old_owner, &new_owner); + if (r < 0) { + bus_log_parse_error(r); + return 0; + } + + if (UNIT_VTABLE(u)->bus_name_owner_change) + UNIT_VTABLE(u)->bus_name_owner_change(u, name, old_owner, new_owner); + + return 0; +} + +int unit_install_bus_match(Unit *u, sd_bus *bus, const char *name) { + const char *match; + + assert(u); + assert(bus); + assert(name); + + if (u->match_bus_slot) + return -EBUSY; + + match = strjoina("type='signal'," + "sender='org.freedesktop.DBus'," + "path='/org/freedesktop/DBus'," + "interface='org.freedesktop.DBus'," + "member='NameOwnerChanged'," + "arg0='", name, "'"); + + return sd_bus_add_match(bus, &u->match_bus_slot, match, signal_name_owner_changed, u); +} + +int unit_watch_bus_name(Unit *u, const char *name) { + int r; + + assert(u); + assert(name); + + /* Watch a specific name on the bus. We only support one unit + * watching each name for now. */ + + if (u->manager->api_bus) { + /* If the bus is already available, install the match directly. + * Otherwise, just put the name in the list. bus_setup_api() will take care later. */ + r = unit_install_bus_match(u, u->manager->api_bus, name); + if (r < 0) + return log_warning_errno(r, "Failed to subscribe to NameOwnerChanged signal for '%s': %m", name); + } + + r = hashmap_put(u->manager->watch_bus, name, u); + if (r < 0) { + u->match_bus_slot = sd_bus_slot_unref(u->match_bus_slot); + return log_warning_errno(r, "Failed to put bus name to hashmap: %m"); + } + + return 0; +} + +void unit_unwatch_bus_name(Unit *u, const char *name) { + assert(u); + assert(name); + + hashmap_remove_value(u->manager->watch_bus, name, u); + u->match_bus_slot = sd_bus_slot_unref(u->match_bus_slot); +} + +bool unit_can_serialize(Unit *u) { + assert(u); + + return UNIT_VTABLE(u)->serialize && UNIT_VTABLE(u)->deserialize_item; +} + +int unit_serialize(Unit *u, FILE *f, FDSet *fds, bool serialize_jobs) { + int r; + + assert(u); + assert(f); + assert(fds); + + if (unit_can_serialize(u)) { + ExecRuntime *rt; + + r = UNIT_VTABLE(u)->serialize(u, f, fds); + if (r < 0) + return r; + + rt = unit_get_exec_runtime(u); + if (rt) { + r = exec_runtime_serialize(u, rt, f, fds); + if (r < 0) + return r; + } + } + + dual_timestamp_serialize(f, "state-change-timestamp", &u->state_change_timestamp); + + dual_timestamp_serialize(f, "inactive-exit-timestamp", &u->inactive_exit_timestamp); + dual_timestamp_serialize(f, "active-enter-timestamp", &u->active_enter_timestamp); + dual_timestamp_serialize(f, "active-exit-timestamp", &u->active_exit_timestamp); + dual_timestamp_serialize(f, "inactive-enter-timestamp", &u->inactive_enter_timestamp); + + dual_timestamp_serialize(f, "condition-timestamp", &u->condition_timestamp); + dual_timestamp_serialize(f, "assert-timestamp", &u->assert_timestamp); + + if (dual_timestamp_is_set(&u->condition_timestamp)) + unit_serialize_item(u, f, "condition-result", yes_no(u->condition_result)); + + if (dual_timestamp_is_set(&u->assert_timestamp)) + unit_serialize_item(u, f, "assert-result", yes_no(u->assert_result)); + + unit_serialize_item(u, f, "transient", yes_no(u->transient)); + unit_serialize_item_format(u, f, "cpuacct-usage-base", "%" PRIu64, u->cpuacct_usage_base); + + if (u->cgroup_path) + unit_serialize_item(u, f, "cgroup", u->cgroup_path); + unit_serialize_item(u, f, "cgroup-realized", yes_no(u->cgroup_realized)); + + if (serialize_jobs) { + if (u->job) { + fprintf(f, "job\n"); + job_serialize(u->job, f, fds); + } + + if (u->nop_job) { + fprintf(f, "job\n"); + job_serialize(u->nop_job, f, fds); + } + } + + /* End marker */ + fputc('\n', f); + return 0; +} + +int unit_serialize_item(Unit *u, FILE *f, const char *key, const char *value) { + assert(u); + assert(f); + assert(key); + + if (!value) + return 0; + + fputs(key, f); + fputc('=', f); + fputs(value, f); + fputc('\n', f); + + return 1; +} + +int unit_serialize_item_escaped(Unit *u, FILE *f, const char *key, const char *value) { + _cleanup_free_ char *c = NULL; + + assert(u); + assert(f); + assert(key); + + if (!value) + return 0; + + c = cescape(value); + if (!c) + return -ENOMEM; + + fputs(key, f); + fputc('=', f); + fputs(c, f); + fputc('\n', f); + + return 1; +} + +int unit_serialize_item_fd(Unit *u, FILE *f, FDSet *fds, const char *key, int fd) { + int copy; + + assert(u); + assert(f); + assert(key); + + if (fd < 0) + return 0; + + copy = fdset_put_dup(fds, fd); + if (copy < 0) + return copy; + + fprintf(f, "%s=%i\n", key, copy); + return 1; +} + +void unit_serialize_item_format(Unit *u, FILE *f, const char *key, const char *format, ...) { + va_list ap; + + assert(u); + assert(f); + assert(key); + assert(format); + + fputs(key, f); + fputc('=', f); + + va_start(ap, format); + vfprintf(f, format, ap); + va_end(ap); + + fputc('\n', f); +} + +int unit_deserialize(Unit *u, FILE *f, FDSet *fds) { + ExecRuntime **rt = NULL; + size_t offset; + int r; + + assert(u); + assert(f); + assert(fds); + + offset = UNIT_VTABLE(u)->exec_runtime_offset; + if (offset > 0) + rt = (ExecRuntime**) ((uint8_t*) u + offset); + + for (;;) { + char line[LINE_MAX], *l, *v; + size_t k; + + if (!fgets(line, sizeof(line), f)) { + if (feof(f)) + return 0; + return -errno; + } + + char_array_0(line); + l = strstrip(line); + + /* End marker */ + if (isempty(l)) + break; + + k = strcspn(l, "="); + + if (l[k] == '=') { + l[k] = 0; + v = l+k+1; + } else + v = l+k; + + if (streq(l, "job")) { + if (v[0] == '\0') { + /* new-style serialized job */ + Job *j; + + j = job_new_raw(u); + if (!j) + return log_oom(); + + r = job_deserialize(j, f, fds); + if (r < 0) { + job_free(j); + return r; + } + + r = hashmap_put(u->manager->jobs, UINT32_TO_PTR(j->id), j); + if (r < 0) { + job_free(j); + return r; + } + + r = job_install_deserialized(j); + if (r < 0) { + hashmap_remove(u->manager->jobs, UINT32_TO_PTR(j->id)); + job_free(j); + return r; + } + } else /* legacy for pre-44 */ + log_unit_warning(u, "Update from too old systemd versions are unsupported, cannot deserialize job: %s", v); + continue; + } else if (streq(l, "state-change-timestamp")) { + dual_timestamp_deserialize(v, &u->state_change_timestamp); + continue; + } else if (streq(l, "inactive-exit-timestamp")) { + dual_timestamp_deserialize(v, &u->inactive_exit_timestamp); + continue; + } else if (streq(l, "active-enter-timestamp")) { + dual_timestamp_deserialize(v, &u->active_enter_timestamp); + continue; + } else if (streq(l, "active-exit-timestamp")) { + dual_timestamp_deserialize(v, &u->active_exit_timestamp); + continue; + } else if (streq(l, "inactive-enter-timestamp")) { + dual_timestamp_deserialize(v, &u->inactive_enter_timestamp); + continue; + } else if (streq(l, "condition-timestamp")) { + dual_timestamp_deserialize(v, &u->condition_timestamp); + continue; + } else if (streq(l, "assert-timestamp")) { + dual_timestamp_deserialize(v, &u->assert_timestamp); + continue; + } else if (streq(l, "condition-result")) { + + r = parse_boolean(v); + if (r < 0) + log_unit_debug(u, "Failed to parse condition result value %s, ignoring.", v); + else + u->condition_result = r; + + continue; + + } else if (streq(l, "assert-result")) { + + r = parse_boolean(v); + if (r < 0) + log_unit_debug(u, "Failed to parse assert result value %s, ignoring.", v); + else + u->assert_result = r; + + continue; + + } else if (streq(l, "transient")) { + + r = parse_boolean(v); + if (r < 0) + log_unit_debug(u, "Failed to parse transient bool %s, ignoring.", v); + else + u->transient = r; + + continue; + + } else if (streq(l, "cpuacct-usage-base")) { + + r = safe_atou64(v, &u->cpuacct_usage_base); + if (r < 0) + log_unit_debug(u, "Failed to parse CPU usage %s, ignoring.", v); + + continue; + + } else if (streq(l, "cgroup")) { + + r = unit_set_cgroup_path(u, v); + if (r < 0) + log_unit_debug_errno(u, r, "Failed to set cgroup path %s, ignoring: %m", v); + + (void) unit_watch_cgroup(u); + + continue; + } else if (streq(l, "cgroup-realized")) { + int b; + + b = parse_boolean(v); + if (b < 0) + log_unit_debug(u, "Failed to parse cgroup-realized bool %s, ignoring.", v); + else + u->cgroup_realized = b; + + continue; + } + + if (unit_can_serialize(u)) { + if (rt) { + r = exec_runtime_deserialize_item(u, rt, l, v, fds); + if (r < 0) { + log_unit_warning(u, "Failed to deserialize runtime parameter '%s', ignoring.", l); + continue; + } + + /* Returns positive if key was handled by the call */ + if (r > 0) + continue; + } + + r = UNIT_VTABLE(u)->deserialize_item(u, l, v, fds); + if (r < 0) + log_unit_warning(u, "Failed to deserialize unit parameter '%s', ignoring.", l); + } + } + + /* Versions before 228 did not carry a state change timestamp. In this case, take the current time. This is + * useful, so that timeouts based on this timestamp don't trigger too early, and is in-line with the logic from + * before 228 where the base for timeouts was not persistent across reboots. */ + + if (!dual_timestamp_is_set(&u->state_change_timestamp)) + dual_timestamp_get(&u->state_change_timestamp); + + return 0; +} + +int unit_add_node_link(Unit *u, const char *what, bool wants, UnitDependency dep) { + Unit *device; + _cleanup_free_ char *e = NULL; + int r; + + assert(u); + + /* Adds in links to the device node that this unit is based on */ + if (isempty(what)) + return 0; + + if (!is_device_path(what)) + return 0; + + /* When device units aren't supported (such as in a + * container), don't create dependencies on them. */ + if (!unit_type_supported(UNIT_DEVICE)) + return 0; + + r = unit_name_from_path(what, ".device", &e); + if (r < 0) + return r; + + r = manager_load_unit(u->manager, e, NULL, NULL, &device); + if (r < 0) + return r; + + r = unit_add_two_dependencies(u, UNIT_AFTER, + MANAGER_IS_SYSTEM(u->manager) ? dep : UNIT_WANTS, + device, true); + if (r < 0) + return r; + + if (wants) { + r = unit_add_dependency(device, UNIT_WANTS, u, false); + if (r < 0) + return r; + } + + return 0; +} + +int unit_coldplug(Unit *u) { + int r = 0, q = 0; + + assert(u); + + /* Make sure we don't enter a loop, when coldplugging + * recursively. */ + if (u->coldplugged) + return 0; + + u->coldplugged = true; + + if (UNIT_VTABLE(u)->coldplug) + r = UNIT_VTABLE(u)->coldplug(u); + + if (u->job) + q = job_coldplug(u->job); + + if (r < 0) + return r; + if (q < 0) + return q; + + return 0; +} + +static bool fragment_mtime_newer(const char *path, usec_t mtime) { + struct stat st; + + if (!path) + return false; + + if (stat(path, &st) < 0) + /* What, cannot access this anymore? */ + return true; + + if (mtime > 0) + /* For non-empty files check the mtime */ + return timespec_load(&st.st_mtim) > mtime; + else if (!null_or_empty(&st)) + /* For masked files check if they are still so */ + return true; + + return false; +} + +bool unit_need_daemon_reload(Unit *u) { + _cleanup_strv_free_ char **t = NULL; + char **path; + + assert(u); + + if (fragment_mtime_newer(u->fragment_path, u->fragment_mtime)) + return true; + + if (fragment_mtime_newer(u->source_path, u->source_mtime)) + return true; + + (void) unit_find_dropin_paths(u, &t); + if (!strv_equal(u->dropin_paths, t)) + return true; + + STRV_FOREACH(path, u->dropin_paths) + if (fragment_mtime_newer(*path, u->dropin_mtime)) + return true; + + return false; +} + +void unit_reset_failed(Unit *u) { + assert(u); + + if (UNIT_VTABLE(u)->reset_failed) + UNIT_VTABLE(u)->reset_failed(u); + + RATELIMIT_RESET(u->start_limit); + u->start_limit_hit = false; +} + +Unit *unit_following(Unit *u) { + assert(u); + + if (UNIT_VTABLE(u)->following) + return UNIT_VTABLE(u)->following(u); + + return NULL; +} + +bool unit_stop_pending(Unit *u) { + assert(u); + + /* This call does check the current state of the unit. It's + * hence useful to be called from state change calls of the + * unit itself, where the state isn't updated yet. This is + * different from unit_inactive_or_pending() which checks both + * the current state and for a queued job. */ + + return u->job && u->job->type == JOB_STOP; +} + +bool unit_inactive_or_pending(Unit *u) { + assert(u); + + /* Returns true if the unit is inactive or going down */ + + if (UNIT_IS_INACTIVE_OR_DEACTIVATING(unit_active_state(u))) + return true; + + if (unit_stop_pending(u)) + return true; + + return false; +} + +bool unit_active_or_pending(Unit *u) { + assert(u); + + /* Returns true if the unit is active or going up */ + + if (UNIT_IS_ACTIVE_OR_ACTIVATING(unit_active_state(u))) + return true; + + if (u->job && + (u->job->type == JOB_START || + u->job->type == JOB_RELOAD_OR_START || + u->job->type == JOB_RESTART)) + return true; + + return false; +} + +int unit_kill(Unit *u, KillWho w, int signo, sd_bus_error *error) { + assert(u); + assert(w >= 0 && w < _KILL_WHO_MAX); + assert(SIGNAL_VALID(signo)); + + if (!UNIT_VTABLE(u)->kill) + return -EOPNOTSUPP; + + return UNIT_VTABLE(u)->kill(u, w, signo, error); +} + +static Set *unit_pid_set(pid_t main_pid, pid_t control_pid) { + Set *pid_set; + int r; + + pid_set = set_new(NULL); + if (!pid_set) + return NULL; + + /* Exclude the main/control pids from being killed via the cgroup */ + if (main_pid > 0) { + r = set_put(pid_set, PID_TO_PTR(main_pid)); + if (r < 0) + goto fail; + } + + if (control_pid > 0) { + r = set_put(pid_set, PID_TO_PTR(control_pid)); + if (r < 0) + goto fail; + } + + return pid_set; + +fail: + set_free(pid_set); + return NULL; +} + +int unit_kill_common( + Unit *u, + KillWho who, + int signo, + pid_t main_pid, + pid_t control_pid, + sd_bus_error *error) { + + int r = 0; + bool killed = false; + + if (IN_SET(who, KILL_MAIN, KILL_MAIN_FAIL)) { + if (main_pid < 0) + return sd_bus_error_setf(error, BUS_ERROR_NO_SUCH_PROCESS, "%s units have no main processes", unit_type_to_string(u->type)); + else if (main_pid == 0) + return sd_bus_error_set_const(error, BUS_ERROR_NO_SUCH_PROCESS, "No main process to kill"); + } + + if (IN_SET(who, KILL_CONTROL, KILL_CONTROL_FAIL)) { + if (control_pid < 0) + return sd_bus_error_setf(error, BUS_ERROR_NO_SUCH_PROCESS, "%s units have no control processes", unit_type_to_string(u->type)); + else if (control_pid == 0) + return sd_bus_error_set_const(error, BUS_ERROR_NO_SUCH_PROCESS, "No control process to kill"); + } + + if (IN_SET(who, KILL_CONTROL, KILL_CONTROL_FAIL, KILL_ALL, KILL_ALL_FAIL)) + if (control_pid > 0) { + if (kill(control_pid, signo) < 0) + r = -errno; + else + killed = true; + } + + if (IN_SET(who, KILL_MAIN, KILL_MAIN_FAIL, KILL_ALL, KILL_ALL_FAIL)) + if (main_pid > 0) { + if (kill(main_pid, signo) < 0) + r = -errno; + else + killed = true; + } + + if (IN_SET(who, KILL_ALL, KILL_ALL_FAIL) && u->cgroup_path) { + _cleanup_set_free_ Set *pid_set = NULL; + int q; + + /* Exclude the main/control pids from being killed via the cgroup */ + pid_set = unit_pid_set(main_pid, control_pid); + if (!pid_set) + return -ENOMEM; + + q = cg_kill_recursive(SYSTEMD_CGROUP_CONTROLLER, u->cgroup_path, signo, 0, pid_set, NULL, NULL); + if (q < 0 && q != -EAGAIN && q != -ESRCH && q != -ENOENT) + r = q; + else + killed = true; + } + + if (r == 0 && !killed && IN_SET(who, KILL_ALL_FAIL, KILL_CONTROL_FAIL)) + return -ESRCH; + + return r; +} + +int unit_following_set(Unit *u, Set **s) { + assert(u); + assert(s); + + if (UNIT_VTABLE(u)->following_set) + return UNIT_VTABLE(u)->following_set(u, s); + + *s = NULL; + return 0; +} + +UnitFileState unit_get_unit_file_state(Unit *u) { + int r; + + assert(u); + + if (u->unit_file_state < 0 && u->fragment_path) { + r = unit_file_get_state( + u->manager->unit_file_scope, + NULL, + basename(u->fragment_path), + &u->unit_file_state); + if (r < 0) + u->unit_file_state = UNIT_FILE_BAD; + } + + return u->unit_file_state; +} + +int unit_get_unit_file_preset(Unit *u) { + assert(u); + + if (u->unit_file_preset < 0 && u->fragment_path) + u->unit_file_preset = unit_file_query_preset( + u->manager->unit_file_scope, + NULL, + basename(u->fragment_path)); + + return u->unit_file_preset; +} + +Unit* unit_ref_set(UnitRef *ref, Unit *u) { + assert(ref); + assert(u); + + if (ref->unit) + unit_ref_unset(ref); + + ref->unit = u; + LIST_PREPEND(refs, u->refs, ref); + return u; +} + +void unit_ref_unset(UnitRef *ref) { + assert(ref); + + if (!ref->unit) + return; + + /* We are about to drop a reference to the unit, make sure the garbage collection has a look at it as it might + * be unreferenced now. */ + unit_add_to_gc_queue(ref->unit); + + LIST_REMOVE(refs, ref->unit->refs, ref); + ref->unit = NULL; +} + +int unit_patch_contexts(Unit *u) { + CGroupContext *cc; + ExecContext *ec; + unsigned i; + int r; + + assert(u); + + /* Patch in the manager defaults into the exec and cgroup + * contexts, _after_ the rest of the settings have been + * initialized */ + + ec = unit_get_exec_context(u); + if (ec) { + /* This only copies in the ones that need memory */ + for (i = 0; i < _RLIMIT_MAX; i++) + if (u->manager->rlimit[i] && !ec->rlimit[i]) { + ec->rlimit[i] = newdup(struct rlimit, u->manager->rlimit[i], 1); + if (!ec->rlimit[i]) + return -ENOMEM; + } + + if (MANAGER_IS_USER(u->manager) && + !ec->working_directory) { + + r = get_home_dir(&ec->working_directory); + if (r < 0) + return r; + + /* Allow user services to run, even if the + * home directory is missing */ + ec->working_directory_missing_ok = true; + } + + if (MANAGER_IS_USER(u->manager) && + (ec->syscall_whitelist || + !set_isempty(ec->syscall_filter) || + !set_isempty(ec->syscall_archs) || + ec->address_families_whitelist || + !set_isempty(ec->address_families))) + ec->no_new_privileges = true; + + if (ec->private_devices) + ec->capability_bounding_set &= ~(UINT64_C(1) << CAP_MKNOD); + } + + cc = unit_get_cgroup_context(u); + if (cc) { + + if (ec && + ec->private_devices && + cc->device_policy == CGROUP_AUTO) + cc->device_policy = CGROUP_CLOSED; + } + + return 0; +} + +ExecContext *unit_get_exec_context(Unit *u) { + size_t offset; + assert(u); + + if (u->type < 0) + return NULL; + + offset = UNIT_VTABLE(u)->exec_context_offset; + if (offset <= 0) + return NULL; + + return (ExecContext*) ((uint8_t*) u + offset); +} + +KillContext *unit_get_kill_context(Unit *u) { + size_t offset; + assert(u); + + if (u->type < 0) + return NULL; + + offset = UNIT_VTABLE(u)->kill_context_offset; + if (offset <= 0) + return NULL; + + return (KillContext*) ((uint8_t*) u + offset); +} + +CGroupContext *unit_get_cgroup_context(Unit *u) { + size_t offset; + + if (u->type < 0) + return NULL; + + offset = UNIT_VTABLE(u)->cgroup_context_offset; + if (offset <= 0) + return NULL; + + return (CGroupContext*) ((uint8_t*) u + offset); +} + +ExecRuntime *unit_get_exec_runtime(Unit *u) { + size_t offset; + + if (u->type < 0) + return NULL; + + offset = UNIT_VTABLE(u)->exec_runtime_offset; + if (offset <= 0) + return NULL; + + return *(ExecRuntime**) ((uint8_t*) u + offset); +} + +static const char* unit_drop_in_dir(Unit *u, UnitSetPropertiesMode mode) { + assert(u); + + if (!IN_SET(mode, UNIT_RUNTIME, UNIT_PERSISTENT)) + return NULL; + + if (u->transient) /* Redirect drop-ins for transient units always into the transient directory. */ + return u->manager->lookup_paths.transient; + + if (mode == UNIT_RUNTIME) + return u->manager->lookup_paths.runtime_control; + + if (mode == UNIT_PERSISTENT) + return u->manager->lookup_paths.persistent_control; + + return NULL; +} + +int unit_write_drop_in(Unit *u, UnitSetPropertiesMode mode, const char *name, const char *data) { + _cleanup_free_ char *p = NULL, *q = NULL; + const char *dir, *wrapped; + int r; + + assert(u); + + if (u->transient_file) { + /* When this is a transient unit file in creation, then let's not create a new drop-in but instead + * write to the transient unit file. */ + fputs(data, u->transient_file); + fputc('\n', u->transient_file); + return 0; + } + + if (!IN_SET(mode, UNIT_PERSISTENT, UNIT_RUNTIME)) + return 0; + + dir = unit_drop_in_dir(u, mode); + if (!dir) + return -EINVAL; + + wrapped = strjoina("# This is a drop-in unit file extension, created via \"systemctl set-property\"\n" + "# or an equivalent operation. Do not edit.\n", + data, + "\n"); + + r = drop_in_file(dir, u->id, 50, name, &p, &q); + if (r < 0) + return r; + + (void) mkdir_p(p, 0755); + r = write_string_file_atomic_label(q, wrapped); + if (r < 0) + return r; + + r = strv_push(&u->dropin_paths, q); + if (r < 0) + return r; + q = NULL; + + strv_uniq(u->dropin_paths); + + u->dropin_mtime = now(CLOCK_REALTIME); + + return 0; +} + +int unit_write_drop_in_format(Unit *u, UnitSetPropertiesMode mode, const char *name, const char *format, ...) { + _cleanup_free_ char *p = NULL; + va_list ap; + int r; + + assert(u); + assert(name); + assert(format); + + if (!IN_SET(mode, UNIT_PERSISTENT, UNIT_RUNTIME)) + return 0; + + va_start(ap, format); + r = vasprintf(&p, format, ap); + va_end(ap); + + if (r < 0) + return -ENOMEM; + + return unit_write_drop_in(u, mode, name, p); +} + +int unit_write_drop_in_private(Unit *u, UnitSetPropertiesMode mode, const char *name, const char *data) { + const char *ndata; + + assert(u); + assert(name); + assert(data); + + if (!UNIT_VTABLE(u)->private_section) + return -EINVAL; + + if (!IN_SET(mode, UNIT_PERSISTENT, UNIT_RUNTIME)) + return 0; + + ndata = strjoina("[", UNIT_VTABLE(u)->private_section, "]\n", data); + + return unit_write_drop_in(u, mode, name, ndata); +} + +int unit_write_drop_in_private_format(Unit *u, UnitSetPropertiesMode mode, const char *name, const char *format, ...) { + _cleanup_free_ char *p = NULL; + va_list ap; + int r; + + assert(u); + assert(name); + assert(format); + + if (!IN_SET(mode, UNIT_PERSISTENT, UNIT_RUNTIME)) + return 0; + + va_start(ap, format); + r = vasprintf(&p, format, ap); + va_end(ap); + + if (r < 0) + return -ENOMEM; + + return unit_write_drop_in_private(u, mode, name, p); +} + +int unit_make_transient(Unit *u) { + FILE *f; + char *path; + + assert(u); + + if (!UNIT_VTABLE(u)->can_transient) + return -EOPNOTSUPP; + + path = strjoin(u->manager->lookup_paths.transient, "/", u->id, NULL); + if (!path) + return -ENOMEM; + + /* Let's open the file we'll write the transient settings into. This file is kept open as long as we are + * creating the transient, and is closed in unit_load(), as soon as we start loading the file. */ + + RUN_WITH_UMASK(0022) { + f = fopen(path, "we"); + if (!f) { + free(path); + return -errno; + } + } + + if (u->transient_file) + fclose(u->transient_file); + u->transient_file = f; + + free(u->fragment_path); + u->fragment_path = path; + + u->source_path = mfree(u->source_path); + u->dropin_paths = strv_free(u->dropin_paths); + u->fragment_mtime = u->source_mtime = u->dropin_mtime = 0; + + u->load_state = UNIT_STUB; + u->load_error = 0; + u->transient = true; + + unit_add_to_dbus_queue(u); + unit_add_to_gc_queue(u); + + fputs("# This is a transient unit file, created programmatically via the systemd API. Do not edit.\n", + u->transient_file); + + return 0; +} + +static void log_kill(pid_t pid, int sig, void *userdata) { + _cleanup_free_ char *comm = NULL; + + (void) get_process_comm(pid, &comm); + + /* Don't log about processes marked with brackets, under the assumption that these are temporary processes + only, like for example systemd's own PAM stub process. */ + if (comm && comm[0] == '(') + return; + + log_unit_notice(userdata, + "Killing process " PID_FMT " (%s) with signal SIG%s.", + pid, + strna(comm), + signal_to_string(sig)); +} + +static int operation_to_signal(KillContext *c, KillOperation k) { + assert(c); + + switch (k) { + + case KILL_TERMINATE: + case KILL_TERMINATE_AND_LOG: + return c->kill_signal; + + case KILL_KILL: + return SIGKILL; + + case KILL_ABORT: + return SIGABRT; + + default: + assert_not_reached("KillOperation unknown"); + } +} + +int unit_kill_context( + Unit *u, + KillContext *c, + KillOperation k, + pid_t main_pid, + pid_t control_pid, + bool main_pid_alien) { + + bool wait_for_exit = false, send_sighup; + cg_kill_log_func_t log_func; + int sig, r; + + assert(u); + assert(c); + + /* Kill the processes belonging to this unit, in preparation for shutting the unit down. Returns > 0 if we + * killed something worth waiting for, 0 otherwise. */ + + if (c->kill_mode == KILL_NONE) + return 0; + + sig = operation_to_signal(c, k); + + send_sighup = + c->send_sighup && + IN_SET(k, KILL_TERMINATE, KILL_TERMINATE_AND_LOG) && + sig != SIGHUP; + + log_func = + k != KILL_TERMINATE || + IN_SET(sig, SIGKILL, SIGABRT) ? log_kill : NULL; + + if (main_pid > 0) { + if (log_func) + log_func(main_pid, sig, u); + + r = kill_and_sigcont(main_pid, sig); + if (r < 0 && r != -ESRCH) { + _cleanup_free_ char *comm = NULL; + (void) get_process_comm(main_pid, &comm); + + log_unit_warning_errno(u, r, "Failed to kill main process " PID_FMT " (%s), ignoring: %m", main_pid, strna(comm)); + } else { + if (!main_pid_alien) + wait_for_exit = true; + + if (r != -ESRCH && send_sighup) + (void) kill(main_pid, SIGHUP); + } + } + + if (control_pid > 0) { + if (log_func) + log_func(control_pid, sig, u); + + r = kill_and_sigcont(control_pid, sig); + if (r < 0 && r != -ESRCH) { + _cleanup_free_ char *comm = NULL; + (void) get_process_comm(control_pid, &comm); + + log_unit_warning_errno(u, r, "Failed to kill control process " PID_FMT " (%s), ignoring: %m", control_pid, strna(comm)); + } else { + wait_for_exit = true; + + if (r != -ESRCH && send_sighup) + (void) kill(control_pid, SIGHUP); + } + } + + if (u->cgroup_path && + (c->kill_mode == KILL_CONTROL_GROUP || (c->kill_mode == KILL_MIXED && k == KILL_KILL))) { + _cleanup_set_free_ Set *pid_set = NULL; + + /* Exclude the main/control pids from being killed via the cgroup */ + pid_set = unit_pid_set(main_pid, control_pid); + if (!pid_set) + return -ENOMEM; + + r = cg_kill_recursive(SYSTEMD_CGROUP_CONTROLLER, u->cgroup_path, + sig, + CGROUP_SIGCONT|CGROUP_IGNORE_SELF, + pid_set, + log_func, u); + if (r < 0) { + if (r != -EAGAIN && r != -ESRCH && r != -ENOENT) + log_unit_warning_errno(u, r, "Failed to kill control group %s, ignoring: %m", u->cgroup_path); + + } else if (r > 0) { + + /* FIXME: For now, on the legacy hierarchy, we + * will not wait for the cgroup members to die + * if we are running in a container or if this + * is a delegation unit, simply because cgroup + * notification is unreliable in these + * cases. It doesn't work at all in + * containers, and outside of containers it + * can be confused easily by left-over + * directories in the cgroup — which however + * should not exist in non-delegated units. On + * the unified hierarchy that's different, + * there we get proper events. Hence rely on + * them.*/ + + if (cg_unified() > 0 || + (detect_container() == 0 && !unit_cgroup_delegate(u))) + wait_for_exit = true; + + if (send_sighup) { + set_free(pid_set); + + pid_set = unit_pid_set(main_pid, control_pid); + if (!pid_set) + return -ENOMEM; + + cg_kill_recursive(SYSTEMD_CGROUP_CONTROLLER, u->cgroup_path, + SIGHUP, + CGROUP_IGNORE_SELF, + pid_set, + NULL, NULL); + } + } + } + + return wait_for_exit; +} + +int unit_require_mounts_for(Unit *u, const char *path) { + char prefix[strlen(path) + 1], *p; + int r; + + assert(u); + assert(path); + + /* Registers a unit for requiring a certain path and all its + * prefixes. We keep a simple array of these paths in the + * unit, since its usually short. However, we build a prefix + * table for all possible prefixes so that new appearing mount + * units can easily determine which units to make themselves a + * dependency of. */ + + if (!path_is_absolute(path)) + return -EINVAL; + + p = strdup(path); + if (!p) + return -ENOMEM; + + path_kill_slashes(p); + + if (!path_is_safe(p)) { + free(p); + return -EPERM; + } + + if (strv_contains(u->requires_mounts_for, p)) { + free(p); + return 0; + } + + r = strv_consume(&u->requires_mounts_for, p); + if (r < 0) + return r; + + PATH_FOREACH_PREFIX_MORE(prefix, p) { + Set *x; + + x = hashmap_get(u->manager->units_requiring_mounts_for, prefix); + if (!x) { + char *q; + + r = hashmap_ensure_allocated(&u->manager->units_requiring_mounts_for, &string_hash_ops); + if (r < 0) + return r; + + q = strdup(prefix); + if (!q) + return -ENOMEM; + + x = set_new(NULL); + if (!x) { + free(q); + return -ENOMEM; + } + + r = hashmap_put(u->manager->units_requiring_mounts_for, q, x); + if (r < 0) { + free(q); + set_free(x); + return r; + } + } + + r = set_put(x, u); + if (r < 0) + return r; + } + + return 0; +} + +int unit_setup_exec_runtime(Unit *u) { + ExecRuntime **rt; + size_t offset; + Iterator i; + Unit *other; + + offset = UNIT_VTABLE(u)->exec_runtime_offset; + assert(offset > 0); + + /* Check if there already is an ExecRuntime for this unit? */ + rt = (ExecRuntime**) ((uint8_t*) u + offset); + if (*rt) + return 0; + + /* Try to get it from somebody else */ + SET_FOREACH(other, u->dependencies[UNIT_JOINS_NAMESPACE_OF], i) { + + *rt = unit_get_exec_runtime(other); + if (*rt) { + exec_runtime_ref(*rt); + return 0; + } + } + + return exec_runtime_make(rt, unit_get_exec_context(u), u->id); +} + +bool unit_type_supported(UnitType t) { + if (_unlikely_(t < 0)) + return false; + if (_unlikely_(t >= _UNIT_TYPE_MAX)) + return false; + + if (!unit_vtable[t]->supported) + return true; + + return unit_vtable[t]->supported(); +} + +void unit_warn_if_dir_nonempty(Unit *u, const char* where) { + int r; + + assert(u); + assert(where); + + r = dir_is_empty(where); + if (r > 0) + return; + if (r < 0) { + log_unit_warning_errno(u, r, "Failed to check directory %s: %m", where); + return; + } + + log_struct(LOG_NOTICE, + LOG_MESSAGE_ID(SD_MESSAGE_OVERMOUNTING), + LOG_UNIT_ID(u), + LOG_UNIT_MESSAGE(u, "Directory %s to mount over is not empty, mounting anyway.", where), + "WHERE=%s", where, + NULL); +} + +int unit_fail_if_symlink(Unit *u, const char* where) { + int r; + + assert(u); + assert(where); + + r = is_symlink(where); + if (r < 0) { + log_unit_debug_errno(u, r, "Failed to check symlink %s, ignoring: %m", where); + return 0; + } + if (r == 0) + return 0; + + log_struct(LOG_ERR, + LOG_MESSAGE_ID(SD_MESSAGE_OVERMOUNTING), + LOG_UNIT_ID(u), + LOG_UNIT_MESSAGE(u, "Mount on symlink %s not allowed.", where), + "WHERE=%s", where, + NULL); + + return -ELOOP; +} + +bool unit_is_pristine(Unit *u) { + assert(u); + + /* Check if the unit already exists or is already around, + * in a number of different ways. Note that to cater for unit + * types such as slice, we are generally fine with units that + * are marked UNIT_LOADED even though nothing was + * actually loaded, as those unit types don't require a file + * on disk to validly load. */ + + return !(!IN_SET(u->load_state, UNIT_NOT_FOUND, UNIT_LOADED) || + u->fragment_path || + u->source_path || + !strv_isempty(u->dropin_paths) || + u->job || + u->merged_into); +} + +pid_t unit_control_pid(Unit *u) { + assert(u); + + if (UNIT_VTABLE(u)->control_pid) + return UNIT_VTABLE(u)->control_pid(u); + + return 0; +} + +pid_t unit_main_pid(Unit *u) { + assert(u); + + if (UNIT_VTABLE(u)->main_pid) + return UNIT_VTABLE(u)->main_pid(u); + + return 0; +} diff --git a/src/grp-system/systemctl/.gitignore b/src/grp-system/systemctl/.gitignore new file mode 100644 index 0000000000..ebd59d3c9e --- /dev/null +++ b/src/grp-system/systemctl/.gitignore @@ -0,0 +1,2 @@ +/systemctl +/_systemctl diff --git a/src/grp-system/systemctl/Makefile b/src/grp-system/systemctl/Makefile new file mode 100644 index 0000000000..3551118514 --- /dev/null +++ b/src/grp-system/systemctl/Makefile @@ -0,0 +1,33 @@ +# -*- Mode: makefile; indent-tabs-mode: t -*- +# +# This file is part of systemd. +# +# Copyright 2010-2012 Lennart Poettering +# Copyright 2010-2012 Kay Sievers +# Copyright 2013 Zbigniew Jędrzejewski-Szmek +# Copyright 2013 David Strauss +# Copyright 2016 Luke Shumaker +# +# systemd is free software; you can redistribute it and/or modify it +# under the terms of the GNU Lesser General Public License as published by +# the Free Software Foundation; either version 2.1 of the License, or +# (at your option) any later version. +# +# systemd is distributed in the hope that it will be useful, but +# WITHOUT ANY WARRANTY; without even the implied warranty of +# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU +# Lesser General Public License for more details. +# +# You should have received a copy of the GNU Lesser General Public License +# along with systemd; If not, see . +include $(dir $(lastword $(MAKEFILE_LIST)))/../../../config.mk +include $(topsrcdir)/build-aux/Makefile.head.mk + +rootbin_PROGRAMS += systemctl +systemctl_SOURCES = \ + src/systemctl/systemctl.c + +systemctl_LDADD = \ + libsystemd-shared.la + +include $(topsrcdir)/build-aux/Makefile.tail.mk diff --git a/src/grp-system/systemctl/halt.xml b/src/grp-system/systemctl/halt.xml new file mode 100644 index 0000000000..e3fa60a915 --- /dev/null +++ b/src/grp-system/systemctl/halt.xml @@ -0,0 +1,176 @@ + + + + + + + + + halt + systemd + + + + Developer + Lennart + Poettering + lennart@poettering.net + + + + + + halt + 8 + + + + halt + poweroff + reboot + Halt, power-off or reboot the machine + + + + + halt + OPTIONS + + + poweroff + OPTIONS + + + reboot + OPTIONS + + + + + Description + + halt, poweroff, + reboot may be used to halt, power-off or reboot + the machine. + + + + + Options + + The following options are understood: + + + + + + + + + + + + Halt the machine, regardless of which one of + the three commands is invoked. + + + + + + + Power-off the machine, regardless of which one + of the three commands is invoked. + + + + + + Reboot the machine, regardless of which one of + the three commands is invoked. + + + + + + + Force immediate halt, power-off, reboot. Do + not contact the init system. + + + + + + + Only write wtmp shutdown entry, do not + actually halt, power-off, reboot. + + + + + + + Do not write wtmp shutdown + entry. + + + + + + + Don't sync hard disks/storage media before + halt, power-off, reboot. + + + + + + Do not send wall message before halt, + power-off, reboot. + + + + + + Exit status + + On success, 0 is returned, a non-zero failure code + otherwise. + + + + Notes + + These are legacy commands available for compatibility + only. + + + + See Also + + systemd1, + systemctl1, + shutdown8, + wall1 + + + + diff --git a/src/grp-system/systemctl/runlevel.xml b/src/grp-system/systemctl/runlevel.xml new file mode 100644 index 0000000000..ca29c7c22c --- /dev/null +++ b/src/grp-system/systemctl/runlevel.xml @@ -0,0 +1,192 @@ + + + + + + + + + runlevel + systemd + + + + Developer + Lennart + Poettering + lennart@poettering.net + + + + + + runlevel + 8 + + + + runlevel + Print previous and current SysV runlevel + + + + + runlevel + options + + + + + Overview + + "Runlevels" are an obsolete way to start and stop groups of + services used in SysV init. systemd provides a compatibility layer + that maps runlevels to targets, and associated binaries like + runlevel. Nevertheless, only one runlevel can + be "active" at a given time, while systemd can activate multiple + targets concurrently, so the mapping to runlevels is confusing + and only approximate. Runlevels should not be used in new code, + and are mostly useful as a shorthand way to refer the matching + systemd targets in kernel boot parameters. + + + Mapping between runlevels and systemd targets + + + + + + Runlevel + Target + + + + + 0 + poweroff.target + + + 1 + rescue.target + + + 2, 3, 4 + multi-user.target + + + 5 + graphical.target + + + 6 + reboot.target + + + +
+ + + + Description + + runlevel prints the previous and current + SysV runlevel if they are known. + + The two runlevel characters are separated by a single space + character. If a runlevel cannot be determined, N is printed + instead. If neither can be determined, the word "unknown" is + printed. + + Unless overridden in the environment, this will check the + utmp database for recent runlevel changes. + + + + Options + + The following option is understood: + + + + + + + + + + + + + Exit status + + If one or both runlevels could be determined, 0 is returned, + a non-zero failure code otherwise. + + + + + Environment + + + + $RUNLEVEL + + If $RUNLEVEL is set, + runlevel will print this value as current + runlevel and ignore utmp. + + + + $PREVLEVEL + + If $PREVLEVEL is set, + runlevel will print this value as previous + runlevel and ignore utmp. + + + + + + Files + + + + /var/run/utmp + + The utmp database runlevel + reads the previous and current runlevel + from. + + + + + + See Also + + systemd1, + systemd.target5, + systemctl1 + + + + diff --git a/src/grp-system/systemctl/shutdown.xml b/src/grp-system/systemctl/shutdown.xml new file mode 100644 index 0000000000..a8af387c67 --- /dev/null +++ b/src/grp-system/systemctl/shutdown.xml @@ -0,0 +1,175 @@ + + + + + + + + + shutdown + systemd + + + + Developer + Lennart + Poettering + lennart@poettering.net + + + + + + shutdown + 8 + + + + shutdown + Halt, power-off or reboot the machine + + + + + shutdown + OPTIONS + TIME + WALL + + + + + Description + + shutdown may be used to halt, power-off + or reboot the machine. + + The first argument may be a time string (which is usually + now). Optionally, this may be followed by a + wall message to be sent to all logged-in users before going + down. + + The time string may either be in the format + hh:mm for hour/minutes specifying the time to + execute the shutdown at, specified in 24h clock format. + Alternatively it may be in the syntax +m + referring to the specified number of minutes m from now. + now is an alias for +0, i.e. + for triggering an immediate shutdown. If no time argument is + specified, +1 is implied. + + Note that to specify a wall message you must specify a time + argument, too. + + If the time argument is used, 5 minutes before the system + goes down the /run/nologin file is created to + ensure that further logins shall not be allowed. + + + + Options + + The following options are understood: + + + + + + + + + + + + + Halt the machine. + + + + + + + Power-off the machine (the + default). + + + + + + + Reboot the + machine. + + + + + + Equivalent to , + unless is specified. + + + + + + Do not halt, power-off, reboot, just write + wall message. + + + + + + Do not send wall + message before + halt, power-off, reboot. + + + + + + Cancel a pending shutdown. This may be used + cancel the effect of an invocation of + shutdown with a time argument that is not + +0 or + now. + + + + + + + Exit status + + On success, 0 is returned, a non-zero failure code + otherwise. + + + + See Also + + systemd1, + systemctl1, + halt8, + wall1 + + + + diff --git a/src/grp-system/systemctl/systemctl.c b/src/grp-system/systemctl/systemctl.c new file mode 100644 index 0000000000..ab9fc4a427 --- /dev/null +++ b/src/grp-system/systemctl/systemctl.c @@ -0,0 +1,8079 @@ +/*** + This file is part of systemd. + + Copyright 2010 Lennart Poettering + Copyright 2013 Marc-Antoine Perennou + + systemd is free software; you can redistribute it and/or modify it + under the terms of the GNU Lesser General Public License as published by + the Free Software Foundation; either version 2.1 of the License, or + (at your option) any later version. + + systemd is distributed in the hope that it will be useful, but + WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + Lesser General Public License for more details. + + You should have received a copy of the GNU Lesser General Public License + along with systemd; If not, see . +***/ + +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include + +#include + +#include +#include +#include + +#include "sd-bus/bus-common-errors.h" +#include "sd-bus/bus-error.h" +#include "sd-bus/bus-message.h" +#include "sd-bus/bus-util.h" +#include "systemd-basic/alloc-util.h" +#include "systemd-basic/cgroup-util.h" +#include "systemd-basic/copy.h" +#include "systemd-basic/env-util.h" +#include "systemd-basic/exit-status.h" +#include "systemd-basic/fd-util.h" +#include "systemd-basic/fileio.h" +#include "systemd-basic/formats-util.h" +#include "systemd-basic/fs-util.h" +#include "systemd-basic/glob-util.h" +#include "systemd-basic/hostname-util.h" +#include "systemd-basic/io-util.h" +#include "systemd-basic/list.h" +#include "systemd-basic/locale-util.h" +#include "systemd-basic/log.h" +#include "systemd-basic/macro.h" +#include "systemd-basic/mkdir.h" +#include "systemd-basic/parse-util.h" +#include "systemd-basic/path-util.h" +#include "systemd-basic/process-util.h" +#include "systemd-basic/rlimit-util.h" +#include "systemd-basic/set.h" +#include "systemd-basic/sigbus.h" +#include "systemd-basic/signal-util.h" +#include "systemd-basic/socket-util.h" +#include "systemd-basic/special.h" +#include "systemd-basic/stat-util.h" +#include "systemd-basic/strv.h" +#include "systemd-basic/terminal-util.h" +#include "systemd-basic/unit-name.h" +#include "systemd-basic/user-util.h" +#include "systemd-basic/util.h" +#include "systemd-basic/verbs.h" +#include "systemd-basic/virt.h" +#include "systemd-shared/bus-unit-util.h" +#include "systemd-shared/cgroup-show.h" +#include "systemd-shared/dropin.h" +#include "systemd-shared/efivars.h" +#include "systemd-shared/initreq.h" +#include "systemd-shared/install.h" +#include "systemd-shared/logs-show.h" +#include "systemd-shared/pager.h" +#include "systemd-shared/path-lookup.h" +#include "systemd-shared/spawn-ask-password-agent.h" +#include "systemd-shared/spawn-polkit-agent.h" +#include "systemd-shared/utmp-wtmp.h" + +/* The init script exit status codes + 0 program is running or service is OK + 1 program is dead and /var/run pid file exists + 2 program is dead and /var/lock lock file exists + 3 program is not running + 4 program or service status is unknown + 5-99 reserved for future LSB use + 100-149 reserved for distribution use + 150-199 reserved for application use + 200-254 reserved +*/ +enum { + EXIT_PROGRAM_RUNNING_OR_SERVICE_OK = 0, + EXIT_PROGRAM_DEAD_AND_PID_EXISTS = 1, + EXIT_PROGRAM_DEAD_AND_LOCK_FILE_EXISTS = 2, + EXIT_PROGRAM_NOT_RUNNING = 3, + EXIT_PROGRAM_OR_SERVICES_STATUS_UNKNOWN = 4, +}; + +static char **arg_types = NULL; +static char **arg_states = NULL; +static char **arg_properties = NULL; +static bool arg_all = false; +static enum dependency { + DEPENDENCY_FORWARD, + DEPENDENCY_REVERSE, + DEPENDENCY_AFTER, + DEPENDENCY_BEFORE, + _DEPENDENCY_MAX +} arg_dependency = DEPENDENCY_FORWARD; +static const char *arg_job_mode = "replace"; +static UnitFileScope arg_scope = UNIT_FILE_SYSTEM; +static bool arg_no_block = false; +static bool arg_no_legend = false; +static bool arg_no_pager = false; +static bool arg_no_wtmp = false; +static bool arg_no_sync = false; +static bool arg_no_wall = false; +static bool arg_no_reload = false; +static bool arg_value = false; +static bool arg_show_types = false; +static bool arg_ignore_inhibitors = false; +static bool arg_dry = false; +static bool arg_quiet = false; +static bool arg_full = false; +static bool arg_recursive = false; +static int arg_force = 0; +static bool arg_ask_password = false; +static bool arg_runtime = false; +static UnitFilePresetMode arg_preset_mode = UNIT_FILE_PRESET_FULL; +static char **arg_wall = NULL; +static const char *arg_kill_who = NULL; +static int arg_signal = SIGTERM; +static char *arg_root = NULL; +static usec_t arg_when = 0; +static enum action { + _ACTION_INVALID, + ACTION_SYSTEMCTL, + ACTION_HALT, + ACTION_POWEROFF, + ACTION_REBOOT, + ACTION_KEXEC, + ACTION_EXIT, + ACTION_SUSPEND, + ACTION_HIBERNATE, + ACTION_HYBRID_SLEEP, + ACTION_RUNLEVEL2, + ACTION_RUNLEVEL3, + ACTION_RUNLEVEL4, + ACTION_RUNLEVEL5, + ACTION_RESCUE, + ACTION_EMERGENCY, + ACTION_DEFAULT, + ACTION_RELOAD, + ACTION_REEXEC, + ACTION_RUNLEVEL, + ACTION_CANCEL_SHUTDOWN, + _ACTION_MAX +} arg_action = ACTION_SYSTEMCTL; +static BusTransport arg_transport = BUS_TRANSPORT_LOCAL; +static const char *arg_host = NULL; +static unsigned arg_lines = 10; +static OutputMode arg_output = OUTPUT_SHORT; +static bool arg_plain = false; +static bool arg_firmware_setup = false; +static bool arg_now = false; + +static int daemon_reload(int argc, char *argv[], void* userdata); +static int trivial_method(int argc, char *argv[], void *userdata); +static int halt_now(enum action a); +static int get_state_one_unit(sd_bus *bus, const char *name, UnitActiveState *active_state); + +static bool original_stdout_is_tty; + +typedef enum BusFocus { + BUS_FULL, /* The full bus indicated via --system or --user */ + BUS_MANAGER, /* The manager itself, possibly directly, possibly via the bus */ + _BUS_FOCUS_MAX +} BusFocus; + +static sd_bus *busses[_BUS_FOCUS_MAX] = {}; + +static int acquire_bus(BusFocus focus, sd_bus **ret) { + int r; + + assert(focus < _BUS_FOCUS_MAX); + assert(ret); + + /* We only go directly to the manager, if we are using a local transport */ + if (arg_transport != BUS_TRANSPORT_LOCAL) + focus = BUS_FULL; + + if (!busses[focus]) { + bool user; + + user = arg_scope != UNIT_FILE_SYSTEM; + + if (focus == BUS_MANAGER) + r = bus_connect_transport_systemd(arg_transport, arg_host, user, &busses[focus]); + else + r = bus_connect_transport(arg_transport, arg_host, user, &busses[focus]); + if (r < 0) + return log_error_errno(r, "Failed to connect to bus: %m"); + + (void) sd_bus_set_allow_interactive_authorization(busses[focus], arg_ask_password); + } + + *ret = busses[focus]; + return 0; +} + +static void release_busses(void) { + BusFocus w; + + for (w = 0; w < _BUS_FOCUS_MAX; w++) + busses[w] = sd_bus_flush_close_unref(busses[w]); +} + +static int map_string_no_copy(sd_bus *bus, const char *member, sd_bus_message *m, sd_bus_error *error, void *userdata) { + char *s; + const char **p = userdata; + int r; + + r = sd_bus_message_read_basic(m, SD_BUS_TYPE_STRING, &s); + if (r < 0) + return r; + + if (!isempty(s)) + *p = s; + + return 0; +} + +static void ask_password_agent_open_if_enabled(void) { + + /* Open the password agent as a child process if necessary */ + + if (!arg_ask_password) + return; + + if (arg_scope != UNIT_FILE_SYSTEM) + return; + + if (arg_transport != BUS_TRANSPORT_LOCAL) + return; + + ask_password_agent_open(); +} + +static void polkit_agent_open_if_enabled(void) { + + /* Open the polkit agent as a child process if necessary */ + + if (!arg_ask_password) + return; + + if (arg_scope != UNIT_FILE_SYSTEM) + return; + + if (arg_transport != BUS_TRANSPORT_LOCAL) + return; + + polkit_agent_open(); +} + +static OutputFlags get_output_flags(void) { + return + arg_all * OUTPUT_SHOW_ALL | + arg_full * OUTPUT_FULL_WIDTH | + (!on_tty() || pager_have()) * OUTPUT_FULL_WIDTH | + colors_enabled() * OUTPUT_COLOR | + !arg_quiet * OUTPUT_WARN_CUTOFF; +} + +static int translate_bus_error_to_exit_status(int r, const sd_bus_error *error) { + assert(error); + + if (!sd_bus_error_is_set(error)) + return r; + + if (sd_bus_error_has_name(error, SD_BUS_ERROR_ACCESS_DENIED) || + sd_bus_error_has_name(error, BUS_ERROR_ONLY_BY_DEPENDENCY) || + sd_bus_error_has_name(error, BUS_ERROR_NO_ISOLATION) || + sd_bus_error_has_name(error, BUS_ERROR_TRANSACTION_IS_DESTRUCTIVE)) + return EXIT_NOPERMISSION; + + if (sd_bus_error_has_name(error, BUS_ERROR_NO_SUCH_UNIT)) + return EXIT_NOTINSTALLED; + + if (sd_bus_error_has_name(error, BUS_ERROR_JOB_TYPE_NOT_APPLICABLE) || + sd_bus_error_has_name(error, SD_BUS_ERROR_NOT_SUPPORTED)) + return EXIT_NOTIMPLEMENTED; + + if (sd_bus_error_has_name(error, BUS_ERROR_LOAD_FAILED)) + return EXIT_NOTCONFIGURED; + + if (r != 0) + return r; + + return EXIT_FAILURE; +} + +static bool install_client_side(void) { + + /* Decides when to execute enable/disable/... operations + * client-side rather than server-side. */ + + if (running_in_chroot() > 0) + return true; + + if (sd_booted() <= 0) + return true; + + if (!isempty(arg_root)) + return true; + + if (arg_scope == UNIT_FILE_GLOBAL) + return true; + + /* Unsupported environment variable, mostly for debugging purposes */ + if (getenv_bool("SYSTEMCTL_INSTALL_CLIENT_SIDE") > 0) + return true; + + return false; +} + +static int compare_unit_info(const void *a, const void *b) { + const UnitInfo *u = a, *v = b; + const char *d1, *d2; + int r; + + /* First, order by machine */ + if (!u->machine && v->machine) + return -1; + if (u->machine && !v->machine) + return 1; + if (u->machine && v->machine) { + r = strcasecmp(u->machine, v->machine); + if (r != 0) + return r; + } + + /* Second, order by unit type */ + d1 = strrchr(u->id, '.'); + d2 = strrchr(v->id, '.'); + if (d1 && d2) { + r = strcasecmp(d1, d2); + if (r != 0) + return r; + } + + /* Third, order by name */ + return strcasecmp(u->id, v->id); +} + +static bool output_show_unit(const UnitInfo *u, char **patterns) { + assert(u); + + if (!strv_fnmatch_or_empty(patterns, u->id, FNM_NOESCAPE)) + return false; + + if (arg_types) { + const char *dot; + + dot = strrchr(u->id, '.'); + if (!dot) + return false; + + if (!strv_find(arg_types, dot+1)) + return false; + } + + if (arg_all) + return true; + + /* Note that '--all' is not purely a state filter, but also a + * filter that hides units that "follow" other units (which is + * used for device units that appear under different names). */ + if (!isempty(u->following)) + return false; + + if (!strv_isempty(arg_states)) + return true; + + /* By default show all units except the ones in inactive + * state and with no pending job */ + if (u->job_id > 0) + return true; + + if (streq(u->active_state, "inactive")) + return false; + + return true; +} + +static int output_units_list(const UnitInfo *unit_infos, unsigned c) { + unsigned circle_len = 0, id_len, max_id_len, load_len, active_len, sub_len, job_len, desc_len; + const UnitInfo *u; + unsigned n_shown = 0; + int job_count = 0; + + max_id_len = strlen("UNIT"); + load_len = strlen("LOAD"); + active_len = strlen("ACTIVE"); + sub_len = strlen("SUB"); + job_len = strlen("JOB"); + desc_len = 0; + + for (u = unit_infos; u < unit_infos + c; u++) { + max_id_len = MAX(max_id_len, strlen(u->id) + (u->machine ? strlen(u->machine)+1 : 0)); + load_len = MAX(load_len, strlen(u->load_state)); + active_len = MAX(active_len, strlen(u->active_state)); + sub_len = MAX(sub_len, strlen(u->sub_state)); + + if (u->job_id != 0) { + job_len = MAX(job_len, strlen(u->job_type)); + job_count++; + } + + if (!arg_no_legend && + (streq(u->active_state, "failed") || + STR_IN_SET(u->load_state, "error", "not-found", "masked"))) + circle_len = 2; + } + + if (!arg_full && original_stdout_is_tty) { + unsigned basic_len; + + id_len = MIN(max_id_len, 25u); + basic_len = circle_len + 5 + id_len + 5 + active_len + sub_len; + + if (job_count) + basic_len += job_len + 1; + + if (basic_len < (unsigned) columns()) { + unsigned extra_len, incr; + extra_len = columns() - basic_len; + + /* Either UNIT already got 25, or is fully satisfied. + * Grant up to 25 to DESC now. */ + incr = MIN(extra_len, 25u); + desc_len += incr; + extra_len -= incr; + + /* split the remaining space between UNIT and DESC, + * but do not give UNIT more than it needs. */ + if (extra_len > 0) { + incr = MIN(extra_len / 2, max_id_len - id_len); + id_len += incr; + desc_len += extra_len - incr; + } + } + } else + id_len = max_id_len; + + for (u = unit_infos; u < unit_infos + c; u++) { + _cleanup_free_ char *e = NULL, *j = NULL; + const char *on_loaded = "", *off_loaded = ""; + const char *on_active = "", *off_active = ""; + const char *on_circle = "", *off_circle = ""; + const char *id; + bool circle = false; + + if (!n_shown && !arg_no_legend) { + + if (circle_len > 0) + fputs(" ", stdout); + + printf("%-*s %-*s %-*s %-*s ", + id_len, "UNIT", + load_len, "LOAD", + active_len, "ACTIVE", + sub_len, "SUB"); + + if (job_count) + printf("%-*s ", job_len, "JOB"); + + if (!arg_full && arg_no_pager) + printf("%.*s\n", desc_len, "DESCRIPTION"); + else + printf("%s\n", "DESCRIPTION"); + } + + n_shown++; + + if (STR_IN_SET(u->load_state, "error", "not-found", "masked") && !arg_plain) { + on_loaded = ansi_highlight_red(); + on_circle = ansi_highlight_yellow(); + off_loaded = off_circle = ansi_normal(); + circle = true; + } else if (streq(u->active_state, "failed") && !arg_plain) { + on_circle = on_active = ansi_highlight_red(); + off_circle = off_active = ansi_normal(); + circle = true; + } + + if (u->machine) { + j = strjoin(u->machine, ":", u->id, NULL); + if (!j) + return log_oom(); + + id = j; + } else + id = u->id; + + if (arg_full) { + e = ellipsize(id, id_len, 33); + if (!e) + return log_oom(); + + id = e; + } + + if (circle_len > 0) + printf("%s%s%s ", on_circle, circle ? special_glyph(BLACK_CIRCLE) : " ", off_circle); + + printf("%s%-*s%s %s%-*s%s %s%-*s %-*s%s %-*s", + on_active, id_len, id, off_active, + on_loaded, load_len, u->load_state, off_loaded, + on_active, active_len, u->active_state, + sub_len, u->sub_state, off_active, + job_count ? job_len + 1 : 0, u->job_id ? u->job_type : ""); + + if (desc_len > 0) + printf("%.*s\n", desc_len, u->description); + else + printf("%s\n", u->description); + } + + if (!arg_no_legend) { + const char *on, *off; + + if (n_shown) { + puts("\n" + "LOAD = Reflects whether the unit definition was properly loaded.\n" + "ACTIVE = The high-level unit activation state, i.e. generalization of SUB.\n" + "SUB = The low-level unit activation state, values depend on unit type."); + puts(job_count ? "JOB = Pending job for the unit.\n" : ""); + on = ansi_highlight(); + off = ansi_normal(); + } else { + on = ansi_highlight_red(); + off = ansi_normal(); + } + + if (arg_all) + printf("%s%u loaded units listed.%s\n" + "To show all installed unit files use 'systemctl list-unit-files'.\n", + on, n_shown, off); + else + printf("%s%u loaded units listed.%s Pass --all to see loaded but inactive units, too.\n" + "To show all installed unit files use 'systemctl list-unit-files'.\n", + on, n_shown, off); + } + + return 0; +} + +static int get_unit_list( + sd_bus *bus, + const char *machine, + char **patterns, + UnitInfo **unit_infos, + int c, + sd_bus_message **_reply) { + + _cleanup_(sd_bus_message_unrefp) sd_bus_message *m = NULL; + _cleanup_(sd_bus_error_free) sd_bus_error error = SD_BUS_ERROR_NULL; + _cleanup_(sd_bus_message_unrefp) sd_bus_message *reply = NULL; + size_t size = c; + int r; + UnitInfo u; + bool fallback = false; + + assert(bus); + assert(unit_infos); + assert(_reply); + + r = sd_bus_message_new_method_call( + bus, + &m, + "org.freedesktop.systemd1", + "/org/freedesktop/systemd1", + "org.freedesktop.systemd1.Manager", + "ListUnitsByPatterns"); + if (r < 0) + return bus_log_create_error(r); + + r = sd_bus_message_append_strv(m, arg_states); + if (r < 0) + return bus_log_create_error(r); + + r = sd_bus_message_append_strv(m, patterns); + if (r < 0) + return bus_log_create_error(r); + + r = sd_bus_call(bus, m, 0, &error, &reply); + if (r < 0 && (sd_bus_error_has_name(&error, SD_BUS_ERROR_UNKNOWN_METHOD) || + sd_bus_error_has_name(&error, SD_BUS_ERROR_ACCESS_DENIED))) { + /* Fallback to legacy ListUnitsFiltered method */ + fallback = true; + log_debug_errno(r, "Failed to list units: %s Falling back to ListUnitsFiltered method.", bus_error_message(&error, r)); + m = sd_bus_message_unref(m); + sd_bus_error_free(&error); + + r = sd_bus_message_new_method_call( + bus, + &m, + "org.freedesktop.systemd1", + "/org/freedesktop/systemd1", + "org.freedesktop.systemd1.Manager", + "ListUnitsFiltered"); + if (r < 0) + return bus_log_create_error(r); + + r = sd_bus_message_append_strv(m, arg_states); + if (r < 0) + return bus_log_create_error(r); + + r = sd_bus_call(bus, m, 0, &error, &reply); + } + if (r < 0) + return log_error_errno(r, "Failed to list units: %s", bus_error_message(&error, r)); + + r = sd_bus_message_enter_container(reply, SD_BUS_TYPE_ARRAY, "(ssssssouso)"); + if (r < 0) + return bus_log_parse_error(r); + + while ((r = bus_parse_unit_info(reply, &u)) > 0) { + u.machine = machine; + + if (!output_show_unit(&u, fallback ? patterns : NULL)) + continue; + + if (!GREEDY_REALLOC(*unit_infos, size, c+1)) + return log_oom(); + + (*unit_infos)[c++] = u; + } + if (r < 0) + return bus_log_parse_error(r); + + r = sd_bus_message_exit_container(reply); + if (r < 0) + return bus_log_parse_error(r); + + *_reply = reply; + reply = NULL; + + return c; +} + +static void message_set_freep(Set **set) { + sd_bus_message *m; + + while ((m = set_steal_first(*set))) + sd_bus_message_unref(m); + + set_free(*set); +} + +static int get_unit_list_recursive( + sd_bus *bus, + char **patterns, + UnitInfo **_unit_infos, + Set **_replies, + char ***_machines) { + + _cleanup_free_ UnitInfo *unit_infos = NULL; + _cleanup_(message_set_freep) Set *replies; + sd_bus_message *reply; + int c, r; + + assert(bus); + assert(_replies); + assert(_unit_infos); + assert(_machines); + + replies = set_new(NULL); + if (!replies) + return log_oom(); + + c = get_unit_list(bus, NULL, patterns, &unit_infos, 0, &reply); + if (c < 0) + return c; + + r = set_put(replies, reply); + if (r < 0) { + sd_bus_message_unref(reply); + return log_oom(); + } + + if (arg_recursive) { + _cleanup_strv_free_ char **machines = NULL; + char **i; + + r = sd_get_machine_names(&machines); + if (r < 0) + return log_error_errno(r, "Failed to get machine names: %m"); + + STRV_FOREACH(i, machines) { + _cleanup_(sd_bus_flush_close_unrefp) sd_bus *container = NULL; + int k; + + r = sd_bus_open_system_machine(&container, *i); + if (r < 0) { + log_warning_errno(r, "Failed to connect to container %s, ignoring: %m", *i); + continue; + } + + k = get_unit_list(container, *i, patterns, &unit_infos, c, &reply); + if (k < 0) + return k; + + c = k; + + r = set_put(replies, reply); + if (r < 0) { + sd_bus_message_unref(reply); + return log_oom(); + } + } + + *_machines = machines; + machines = NULL; + } else + *_machines = NULL; + + *_unit_infos = unit_infos; + unit_infos = NULL; + + *_replies = replies; + replies = NULL; + + return c; +} + +static int list_units(int argc, char *argv[], void *userdata) { + _cleanup_free_ UnitInfo *unit_infos = NULL; + _cleanup_(message_set_freep) Set *replies = NULL; + _cleanup_strv_free_ char **machines = NULL; + sd_bus *bus; + int r; + + r = acquire_bus(BUS_MANAGER, &bus); + if (r < 0) + return r; + + pager_open(arg_no_pager, false); + + r = get_unit_list_recursive(bus, strv_skip(argv, 1), &unit_infos, &replies, &machines); + if (r < 0) + return r; + + qsort_safe(unit_infos, r, sizeof(UnitInfo), compare_unit_info); + return output_units_list(unit_infos, r); +} + +static int get_triggered_units( + sd_bus *bus, + const char* path, + char*** ret) { + + _cleanup_(sd_bus_error_free) sd_bus_error error = SD_BUS_ERROR_NULL; + int r; + + assert(bus); + assert(path); + assert(ret); + + r = sd_bus_get_property_strv( + bus, + "org.freedesktop.systemd1", + path, + "org.freedesktop.systemd1.Unit", + "Triggers", + &error, + ret); + if (r < 0) + return log_error_errno(r, "Failed to determine triggers: %s", bus_error_message(&error, r)); + + return 0; +} + +static int get_listening( + sd_bus *bus, + const char* unit_path, + char*** listening) { + + _cleanup_(sd_bus_error_free) sd_bus_error error = SD_BUS_ERROR_NULL; + _cleanup_(sd_bus_message_unrefp) sd_bus_message *reply = NULL; + const char *type, *path; + int r, n = 0; + + r = sd_bus_get_property( + bus, + "org.freedesktop.systemd1", + unit_path, + "org.freedesktop.systemd1.Socket", + "Listen", + &error, + &reply, + "a(ss)"); + if (r < 0) + return log_error_errno(r, "Failed to get list of listening sockets: %s", bus_error_message(&error, r)); + + r = sd_bus_message_enter_container(reply, SD_BUS_TYPE_ARRAY, "(ss)"); + if (r < 0) + return bus_log_parse_error(r); + + while ((r = sd_bus_message_read(reply, "(ss)", &type, &path)) > 0) { + + r = strv_extend(listening, type); + if (r < 0) + return log_oom(); + + r = strv_extend(listening, path); + if (r < 0) + return log_oom(); + + n++; + } + if (r < 0) + return bus_log_parse_error(r); + + r = sd_bus_message_exit_container(reply); + if (r < 0) + return bus_log_parse_error(r); + + return n; +} + +struct socket_info { + const char *machine; + const char* id; + + char* type; + char* path; + + /* Note: triggered is a list here, although it almost certainly + * will always be one unit. Nevertheless, dbus API allows for multiple + * values, so let's follow that. */ + char** triggered; + + /* The strv above is shared. free is set only in the first one. */ + bool own_triggered; +}; + +static int socket_info_compare(const struct socket_info *a, const struct socket_info *b) { + int o; + + assert(a); + assert(b); + + if (!a->machine && b->machine) + return -1; + if (a->machine && !b->machine) + return 1; + if (a->machine && b->machine) { + o = strcasecmp(a->machine, b->machine); + if (o != 0) + return o; + } + + o = strcmp(a->path, b->path); + if (o == 0) + o = strcmp(a->type, b->type); + + return o; +} + +static int output_sockets_list(struct socket_info *socket_infos, unsigned cs) { + struct socket_info *s; + unsigned pathlen = strlen("LISTEN"), + typelen = strlen("TYPE") * arg_show_types, + socklen = strlen("UNIT"), + servlen = strlen("ACTIVATES"); + const char *on, *off; + + for (s = socket_infos; s < socket_infos + cs; s++) { + unsigned tmp = 0; + char **a; + + socklen = MAX(socklen, strlen(s->id)); + if (arg_show_types) + typelen = MAX(typelen, strlen(s->type)); + pathlen = MAX(pathlen, strlen(s->path) + (s->machine ? strlen(s->machine)+1 : 0)); + + STRV_FOREACH(a, s->triggered) + tmp += strlen(*a) + 2*(a != s->triggered); + servlen = MAX(servlen, tmp); + } + + if (cs) { + if (!arg_no_legend) + printf("%-*s %-*.*s%-*s %s\n", + pathlen, "LISTEN", + typelen + arg_show_types, typelen + arg_show_types, "TYPE ", + socklen, "UNIT", + "ACTIVATES"); + + for (s = socket_infos; s < socket_infos + cs; s++) { + _cleanup_free_ char *j = NULL; + const char *path; + char **a; + + if (s->machine) { + j = strjoin(s->machine, ":", s->path, NULL); + if (!j) + return log_oom(); + path = j; + } else + path = s->path; + + if (arg_show_types) + printf("%-*s %-*s %-*s", + pathlen, path, typelen, s->type, socklen, s->id); + else + printf("%-*s %-*s", + pathlen, path, socklen, s->id); + STRV_FOREACH(a, s->triggered) + printf("%s %s", + a == s->triggered ? "" : ",", *a); + printf("\n"); + } + + on = ansi_highlight(); + off = ansi_normal(); + if (!arg_no_legend) + printf("\n"); + } else { + on = ansi_highlight_red(); + off = ansi_normal(); + } + + if (!arg_no_legend) { + printf("%s%u sockets listed.%s\n", on, cs, off); + if (!arg_all) + printf("Pass --all to see loaded but inactive sockets, too.\n"); + } + + return 0; +} + +static int list_sockets(int argc, char *argv[], void *userdata) { + _cleanup_(message_set_freep) Set *replies = NULL; + _cleanup_strv_free_ char **machines = NULL; + _cleanup_free_ UnitInfo *unit_infos = NULL; + _cleanup_free_ struct socket_info *socket_infos = NULL; + const UnitInfo *u; + struct socket_info *s; + unsigned cs = 0; + size_t size = 0; + int r = 0, n; + sd_bus *bus; + + r = acquire_bus(BUS_MANAGER, &bus); + if (r < 0) + return r; + + pager_open(arg_no_pager, false); + + n = get_unit_list_recursive(bus, strv_skip(argv, 1), &unit_infos, &replies, &machines); + if (n < 0) + return n; + + for (u = unit_infos; u < unit_infos + n; u++) { + _cleanup_strv_free_ char **listening = NULL, **triggered = NULL; + int i, c; + + if (!endswith(u->id, ".socket")) + continue; + + r = get_triggered_units(bus, u->unit_path, &triggered); + if (r < 0) + goto cleanup; + + c = get_listening(bus, u->unit_path, &listening); + if (c < 0) { + r = c; + goto cleanup; + } + + if (!GREEDY_REALLOC(socket_infos, size, cs + c)) { + r = log_oom(); + goto cleanup; + } + + for (i = 0; i < c; i++) + socket_infos[cs + i] = (struct socket_info) { + .machine = u->machine, + .id = u->id, + .type = listening[i*2], + .path = listening[i*2 + 1], + .triggered = triggered, + .own_triggered = i==0, + }; + + /* from this point on we will cleanup those socket_infos */ + cs += c; + free(listening); + listening = triggered = NULL; /* avoid cleanup */ + } + + qsort_safe(socket_infos, cs, sizeof(struct socket_info), + (__compar_fn_t) socket_info_compare); + + output_sockets_list(socket_infos, cs); + + cleanup: + assert(cs == 0 || socket_infos); + for (s = socket_infos; s < socket_infos + cs; s++) { + free(s->type); + free(s->path); + if (s->own_triggered) + strv_free(s->triggered); + } + + return r; +} + +static int get_next_elapse( + sd_bus *bus, + const char *path, + dual_timestamp *next) { + + _cleanup_(sd_bus_error_free) sd_bus_error error = SD_BUS_ERROR_NULL; + dual_timestamp t; + int r; + + assert(bus); + assert(path); + assert(next); + + r = sd_bus_get_property_trivial( + bus, + "org.freedesktop.systemd1", + path, + "org.freedesktop.systemd1.Timer", + "NextElapseUSecMonotonic", + &error, + 't', + &t.monotonic); + if (r < 0) + return log_error_errno(r, "Failed to get next elapsation time: %s", bus_error_message(&error, r)); + + r = sd_bus_get_property_trivial( + bus, + "org.freedesktop.systemd1", + path, + "org.freedesktop.systemd1.Timer", + "NextElapseUSecRealtime", + &error, + 't', + &t.realtime); + if (r < 0) + return log_error_errno(r, "Failed to get next elapsation time: %s", bus_error_message(&error, r)); + + *next = t; + return 0; +} + +static int get_last_trigger( + sd_bus *bus, + const char *path, + usec_t *last) { + + _cleanup_(sd_bus_error_free) sd_bus_error error = SD_BUS_ERROR_NULL; + int r; + + assert(bus); + assert(path); + assert(last); + + r = sd_bus_get_property_trivial( + bus, + "org.freedesktop.systemd1", + path, + "org.freedesktop.systemd1.Timer", + "LastTriggerUSec", + &error, + 't', + last); + if (r < 0) + return log_error_errno(r, "Failed to get last trigger time: %s", bus_error_message(&error, r)); + + return 0; +} + +struct timer_info { + const char* machine; + const char* id; + usec_t next_elapse; + usec_t last_trigger; + char** triggered; +}; + +static int timer_info_compare(const struct timer_info *a, const struct timer_info *b) { + int o; + + assert(a); + assert(b); + + if (!a->machine && b->machine) + return -1; + if (a->machine && !b->machine) + return 1; + if (a->machine && b->machine) { + o = strcasecmp(a->machine, b->machine); + if (o != 0) + return o; + } + + if (a->next_elapse < b->next_elapse) + return -1; + if (a->next_elapse > b->next_elapse) + return 1; + + return strcmp(a->id, b->id); +} + +static int output_timers_list(struct timer_info *timer_infos, unsigned n) { + struct timer_info *t; + unsigned + nextlen = strlen("NEXT"), + leftlen = strlen("LEFT"), + lastlen = strlen("LAST"), + passedlen = strlen("PASSED"), + unitlen = strlen("UNIT"), + activatelen = strlen("ACTIVATES"); + + const char *on, *off; + + assert(timer_infos || n == 0); + + for (t = timer_infos; t < timer_infos + n; t++) { + unsigned ul = 0; + char **a; + + if (t->next_elapse > 0) { + char tstamp[FORMAT_TIMESTAMP_MAX] = "", trel[FORMAT_TIMESTAMP_RELATIVE_MAX] = ""; + + format_timestamp(tstamp, sizeof(tstamp), t->next_elapse); + nextlen = MAX(nextlen, strlen(tstamp) + 1); + + format_timestamp_relative(trel, sizeof(trel), t->next_elapse); + leftlen = MAX(leftlen, strlen(trel)); + } + + if (t->last_trigger > 0) { + char tstamp[FORMAT_TIMESTAMP_MAX] = "", trel[FORMAT_TIMESTAMP_RELATIVE_MAX] = ""; + + format_timestamp(tstamp, sizeof(tstamp), t->last_trigger); + lastlen = MAX(lastlen, strlen(tstamp) + 1); + + format_timestamp_relative(trel, sizeof(trel), t->last_trigger); + passedlen = MAX(passedlen, strlen(trel)); + } + + unitlen = MAX(unitlen, strlen(t->id) + (t->machine ? strlen(t->machine)+1 : 0)); + + STRV_FOREACH(a, t->triggered) + ul += strlen(*a) + 2*(a != t->triggered); + + activatelen = MAX(activatelen, ul); + } + + if (n > 0) { + if (!arg_no_legend) + printf("%-*s %-*s %-*s %-*s %-*s %s\n", + nextlen, "NEXT", + leftlen, "LEFT", + lastlen, "LAST", + passedlen, "PASSED", + unitlen, "UNIT", + "ACTIVATES"); + + for (t = timer_infos; t < timer_infos + n; t++) { + _cleanup_free_ char *j = NULL; + const char *unit; + char tstamp1[FORMAT_TIMESTAMP_MAX] = "n/a", trel1[FORMAT_TIMESTAMP_RELATIVE_MAX] = "n/a"; + char tstamp2[FORMAT_TIMESTAMP_MAX] = "n/a", trel2[FORMAT_TIMESTAMP_RELATIVE_MAX] = "n/a"; + char **a; + + format_timestamp(tstamp1, sizeof(tstamp1), t->next_elapse); + format_timestamp_relative(trel1, sizeof(trel1), t->next_elapse); + + format_timestamp(tstamp2, sizeof(tstamp2), t->last_trigger); + format_timestamp_relative(trel2, sizeof(trel2), t->last_trigger); + + if (t->machine) { + j = strjoin(t->machine, ":", t->id, NULL); + if (!j) + return log_oom(); + unit = j; + } else + unit = t->id; + + printf("%-*s %-*s %-*s %-*s %-*s", + nextlen, tstamp1, leftlen, trel1, lastlen, tstamp2, passedlen, trel2, unitlen, unit); + + STRV_FOREACH(a, t->triggered) + printf("%s %s", + a == t->triggered ? "" : ",", *a); + printf("\n"); + } + + on = ansi_highlight(); + off = ansi_normal(); + if (!arg_no_legend) + printf("\n"); + } else { + on = ansi_highlight_red(); + off = ansi_normal(); + } + + if (!arg_no_legend) { + printf("%s%u timers listed.%s\n", on, n, off); + if (!arg_all) + printf("Pass --all to see loaded but inactive timers, too.\n"); + } + + return 0; +} + +static usec_t calc_next_elapse(dual_timestamp *nw, dual_timestamp *next) { + usec_t next_elapse; + + assert(nw); + assert(next); + + if (next->monotonic != USEC_INFINITY && next->monotonic > 0) { + usec_t converted; + + if (next->monotonic > nw->monotonic) + converted = nw->realtime + (next->monotonic - nw->monotonic); + else + converted = nw->realtime - (nw->monotonic - next->monotonic); + + if (next->realtime != USEC_INFINITY && next->realtime > 0) + next_elapse = MIN(converted, next->realtime); + else + next_elapse = converted; + + } else + next_elapse = next->realtime; + + return next_elapse; +} + +static int list_timers(int argc, char *argv[], void *userdata) { + _cleanup_(message_set_freep) Set *replies = NULL; + _cleanup_strv_free_ char **machines = NULL; + _cleanup_free_ struct timer_info *timer_infos = NULL; + _cleanup_free_ UnitInfo *unit_infos = NULL; + struct timer_info *t; + const UnitInfo *u; + size_t size = 0; + int n, c = 0; + dual_timestamp nw; + sd_bus *bus; + int r = 0; + + r = acquire_bus(BUS_MANAGER, &bus); + if (r < 0) + return r; + + pager_open(arg_no_pager, false); + + n = get_unit_list_recursive(bus, strv_skip(argv, 1), &unit_infos, &replies, &machines); + if (n < 0) + return n; + + dual_timestamp_get(&nw); + + for (u = unit_infos; u < unit_infos + n; u++) { + _cleanup_strv_free_ char **triggered = NULL; + dual_timestamp next = DUAL_TIMESTAMP_NULL; + usec_t m, last = 0; + + if (!endswith(u->id, ".timer")) + continue; + + r = get_triggered_units(bus, u->unit_path, &triggered); + if (r < 0) + goto cleanup; + + r = get_next_elapse(bus, u->unit_path, &next); + if (r < 0) + goto cleanup; + + get_last_trigger(bus, u->unit_path, &last); + + if (!GREEDY_REALLOC(timer_infos, size, c+1)) { + r = log_oom(); + goto cleanup; + } + + m = calc_next_elapse(&nw, &next); + + timer_infos[c++] = (struct timer_info) { + .machine = u->machine, + .id = u->id, + .next_elapse = m, + .last_trigger = last, + .triggered = triggered, + }; + + triggered = NULL; /* avoid cleanup */ + } + + qsort_safe(timer_infos, c, sizeof(struct timer_info), + (__compar_fn_t) timer_info_compare); + + output_timers_list(timer_infos, c); + + cleanup: + for (t = timer_infos; t < timer_infos + c; t++) + strv_free(t->triggered); + + return r; +} + +static int compare_unit_file_list(const void *a, const void *b) { + const char *d1, *d2; + const UnitFileList *u = a, *v = b; + + d1 = strrchr(u->path, '.'); + d2 = strrchr(v->path, '.'); + + if (d1 && d2) { + int r; + + r = strcasecmp(d1, d2); + if (r != 0) + return r; + } + + return strcasecmp(basename(u->path), basename(v->path)); +} + +static bool output_show_unit_file(const UnitFileList *u, char **states, char **patterns) { + assert(u); + + if (!strv_fnmatch_or_empty(patterns, basename(u->path), FNM_NOESCAPE)) + return false; + + if (!strv_isempty(arg_types)) { + const char *dot; + + dot = strrchr(u->path, '.'); + if (!dot) + return false; + + if (!strv_find(arg_types, dot+1)) + return false; + } + + if (!strv_isempty(states) && + !strv_find(states, unit_file_state_to_string(u->state))) + return false; + + return true; +} + +static void output_unit_file_list(const UnitFileList *units, unsigned c) { + unsigned max_id_len, id_cols, state_cols; + const UnitFileList *u; + + max_id_len = strlen("UNIT FILE"); + state_cols = strlen("STATE"); + + for (u = units; u < units + c; u++) { + max_id_len = MAX(max_id_len, strlen(basename(u->path))); + state_cols = MAX(state_cols, strlen(unit_file_state_to_string(u->state))); + } + + if (!arg_full) { + unsigned basic_cols; + + id_cols = MIN(max_id_len, 25u); + basic_cols = 1 + id_cols + state_cols; + if (basic_cols < (unsigned) columns()) + id_cols += MIN(columns() - basic_cols, max_id_len - id_cols); + } else + id_cols = max_id_len; + + if (!arg_no_legend && c > 0) + printf("%-*s %-*s\n", + id_cols, "UNIT FILE", + state_cols, "STATE"); + + for (u = units; u < units + c; u++) { + _cleanup_free_ char *e = NULL; + const char *on, *off; + const char *id; + + if (IN_SET(u->state, + UNIT_FILE_MASKED, + UNIT_FILE_MASKED_RUNTIME, + UNIT_FILE_DISABLED, + UNIT_FILE_BAD)) { + on = ansi_highlight_red(); + off = ansi_normal(); + } else if (u->state == UNIT_FILE_ENABLED) { + on = ansi_highlight_green(); + off = ansi_normal(); + } else + on = off = ""; + + id = basename(u->path); + + e = arg_full ? NULL : ellipsize(id, id_cols, 33); + + printf("%-*s %s%-*s%s\n", + id_cols, e ? e : id, + on, state_cols, unit_file_state_to_string(u->state), off); + } + + if (!arg_no_legend) + printf("\n%u unit files listed.\n", c); +} + +static int list_unit_files(int argc, char *argv[], void *userdata) { + _cleanup_(sd_bus_message_unrefp) sd_bus_message *reply = NULL; + _cleanup_free_ UnitFileList *units = NULL; + UnitFileList *unit; + size_t size = 0; + unsigned c = 0; + const char *state; + char *path; + int r; + bool fallback = false; + + if (install_client_side()) { + Hashmap *h; + UnitFileList *u; + Iterator i; + unsigned n_units; + + h = hashmap_new(&string_hash_ops); + if (!h) + return log_oom(); + + r = unit_file_get_list(arg_scope, arg_root, h, arg_states, strv_skip(argv, 1)); + if (r < 0) { + unit_file_list_free(h); + return log_error_errno(r, "Failed to get unit file list: %m"); + } + + n_units = hashmap_size(h); + + units = new(UnitFileList, n_units ?: 1); /* avoid malloc(0) */ + if (!units) { + unit_file_list_free(h); + return log_oom(); + } + + HASHMAP_FOREACH(u, h, i) { + if (!output_show_unit_file(u, NULL, NULL)) + continue; + + units[c++] = *u; + free(u); + } + + assert(c <= n_units); + hashmap_free(h); + + r = 0; + } else { + _cleanup_(sd_bus_message_unrefp) sd_bus_message *m = NULL; + _cleanup_(sd_bus_error_free) sd_bus_error error = SD_BUS_ERROR_NULL; + sd_bus *bus; + + r = acquire_bus(BUS_MANAGER, &bus); + if (r < 0) + return r; + + r = sd_bus_message_new_method_call( + bus, + &m, + "org.freedesktop.systemd1", + "/org/freedesktop/systemd1", + "org.freedesktop.systemd1.Manager", + "ListUnitFilesByPatterns"); + if (r < 0) + return bus_log_create_error(r); + + r = sd_bus_message_append_strv(m, arg_states); + if (r < 0) + return bus_log_create_error(r); + + r = sd_bus_message_append_strv(m, strv_skip(argv, 1)); + if (r < 0) + return bus_log_create_error(r); + + r = sd_bus_call(bus, m, 0, &error, &reply); + if (r < 0 && sd_bus_error_has_name(&error, SD_BUS_ERROR_UNKNOWN_METHOD)) { + /* Fallback to legacy ListUnitFiles method */ + fallback = true; + log_debug_errno(r, "Failed to list unit files: %s Falling back to ListUnitsFiles method.", bus_error_message(&error, r)); + m = sd_bus_message_unref(m); + sd_bus_error_free(&error); + + r = sd_bus_message_new_method_call( + bus, + &m, + "org.freedesktop.systemd1", + "/org/freedesktop/systemd1", + "org.freedesktop.systemd1.Manager", + "ListUnitFiles"); + if (r < 0) + return bus_log_create_error(r); + + r = sd_bus_call(bus, m, 0, &error, &reply); + } + if (r < 0) + return log_error_errno(r, "Failed to list unit files: %s", bus_error_message(&error, r)); + + r = sd_bus_message_enter_container(reply, SD_BUS_TYPE_ARRAY, "(ss)"); + if (r < 0) + return bus_log_parse_error(r); + + while ((r = sd_bus_message_read(reply, "(ss)", &path, &state)) > 0) { + + if (!GREEDY_REALLOC(units, size, c + 1)) + return log_oom(); + + units[c] = (struct UnitFileList) { + path, + unit_file_state_from_string(state) + }; + + if (output_show_unit_file(&units[c], + fallback ? arg_states : NULL, + fallback ? strv_skip(argv, 1) : NULL)) + c++; + + } + if (r < 0) + return bus_log_parse_error(r); + + r = sd_bus_message_exit_container(reply); + if (r < 0) + return bus_log_parse_error(r); + } + + pager_open(arg_no_pager, false); + + qsort_safe(units, c, sizeof(UnitFileList), compare_unit_file_list); + output_unit_file_list(units, c); + + if (install_client_side()) + for (unit = units; unit < units + c; unit++) + free(unit->path); + + return 0; +} + +static int list_dependencies_print(const char *name, int level, unsigned int branches, bool last) { + _cleanup_free_ char *n = NULL; + size_t max_len = MAX(columns(),20u); + size_t len = 0; + int i; + + if (!arg_plain) { + + for (i = level - 1; i >= 0; i--) { + len += 2; + if (len > max_len - 3 && !arg_full) { + printf("%s...\n",max_len % 2 ? "" : " "); + return 0; + } + printf("%s", special_glyph(branches & (1 << i) ? TREE_VERTICAL : TREE_SPACE)); + } + len += 2; + + if (len > max_len - 3 && !arg_full) { + printf("%s...\n",max_len % 2 ? "" : " "); + return 0; + } + + printf("%s", special_glyph(last ? TREE_RIGHT : TREE_BRANCH)); + } + + if (arg_full) { + printf("%s\n", name); + return 0; + } + + n = ellipsize(name, max_len-len, 100); + if (!n) + return log_oom(); + + printf("%s\n", n); + return 0; +} + +static int list_dependencies_get_dependencies(sd_bus *bus, const char *name, char ***deps) { + + static const char *dependencies[_DEPENDENCY_MAX] = { + [DEPENDENCY_FORWARD] = "Requires\0" + "Requisite\0" + "Wants\0" + "ConsistsOf\0" + "BindsTo\0", + [DEPENDENCY_REVERSE] = "RequiredBy\0" + "RequisiteOf\0" + "WantedBy\0" + "PartOf\0" + "BoundBy\0", + [DEPENDENCY_AFTER] = "After\0", + [DEPENDENCY_BEFORE] = "Before\0", + }; + + _cleanup_(sd_bus_error_free) sd_bus_error error = SD_BUS_ERROR_NULL; + _cleanup_(sd_bus_message_unrefp) sd_bus_message *reply = NULL; + _cleanup_strv_free_ char **ret = NULL; + _cleanup_free_ char *path = NULL; + int r; + + assert(bus); + assert(name); + assert(deps); + assert_cc(ELEMENTSOF(dependencies) == _DEPENDENCY_MAX); + + path = unit_dbus_path_from_name(name); + if (!path) + return log_oom(); + + r = sd_bus_call_method( + bus, + "org.freedesktop.systemd1", + path, + "org.freedesktop.DBus.Properties", + "GetAll", + &error, + &reply, + "s", "org.freedesktop.systemd1.Unit"); + if (r < 0) + return log_error_errno(r, "Failed to get properties of %s: %s", name, bus_error_message(&error, r)); + + r = sd_bus_message_enter_container(reply, SD_BUS_TYPE_ARRAY, "{sv}"); + if (r < 0) + return bus_log_parse_error(r); + + while ((r = sd_bus_message_enter_container(reply, SD_BUS_TYPE_DICT_ENTRY, "sv")) > 0) { + const char *prop; + + r = sd_bus_message_read(reply, "s", &prop); + if (r < 0) + return bus_log_parse_error(r); + + if (!nulstr_contains(dependencies[arg_dependency], prop)) { + r = sd_bus_message_skip(reply, "v"); + if (r < 0) + return bus_log_parse_error(r); + } else { + + r = sd_bus_message_enter_container(reply, SD_BUS_TYPE_VARIANT, "as"); + if (r < 0) + return bus_log_parse_error(r); + + r = bus_message_read_strv_extend(reply, &ret); + if (r < 0) + return bus_log_parse_error(r); + + r = sd_bus_message_exit_container(reply); + if (r < 0) + return bus_log_parse_error(r); + } + + r = sd_bus_message_exit_container(reply); + if (r < 0) + return bus_log_parse_error(r); + + } + if (r < 0) + return bus_log_parse_error(r); + + r = sd_bus_message_exit_container(reply); + if (r < 0) + return bus_log_parse_error(r); + + *deps = ret; + ret = NULL; + + return 0; +} + +static int list_dependencies_compare(const void *_a, const void *_b) { + const char **a = (const char**) _a, **b = (const char**) _b; + + if (unit_name_to_type(*a) == UNIT_TARGET && unit_name_to_type(*b) != UNIT_TARGET) + return 1; + if (unit_name_to_type(*a) != UNIT_TARGET && unit_name_to_type(*b) == UNIT_TARGET) + return -1; + + return strcasecmp(*a, *b); +} + +static int list_dependencies_one( + sd_bus *bus, + const char *name, + int level, + char ***units, + unsigned int branches) { + + _cleanup_strv_free_ char **deps = NULL; + char **c; + int r = 0; + + assert(bus); + assert(name); + assert(units); + + r = strv_extend(units, name); + if (r < 0) + return log_oom(); + + r = list_dependencies_get_dependencies(bus, name, &deps); + if (r < 0) + return r; + + qsort_safe(deps, strv_length(deps), sizeof (char*), list_dependencies_compare); + + STRV_FOREACH(c, deps) { + if (strv_contains(*units, *c)) { + if (!arg_plain) { + r = list_dependencies_print("...", level + 1, (branches << 1) | (c[1] == NULL ? 0 : 1), 1); + if (r < 0) + return r; + } + continue; + } + + if (arg_plain) + printf(" "); + else { + UnitActiveState active_state = _UNIT_ACTIVE_STATE_INVALID; + const char *on; + + (void) get_state_one_unit(bus, *c, &active_state); + + switch (active_state) { + case UNIT_ACTIVE: + case UNIT_RELOADING: + case UNIT_ACTIVATING: + on = ansi_highlight_green(); + break; + + case UNIT_INACTIVE: + case UNIT_DEACTIVATING: + on = ansi_normal(); + break; + + default: + on = ansi_highlight_red(); + break; + } + + printf("%s%s%s ", on, special_glyph(BLACK_CIRCLE), ansi_normal()); + } + + r = list_dependencies_print(*c, level, branches, c[1] == NULL); + if (r < 0) + return r; + + if (arg_all || unit_name_to_type(*c) == UNIT_TARGET) { + r = list_dependencies_one(bus, *c, level + 1, units, (branches << 1) | (c[1] == NULL ? 0 : 1)); + if (r < 0) + return r; + } + } + + if (!arg_plain) + strv_remove(*units, name); + + return 0; +} + +static int list_dependencies(int argc, char *argv[], void *userdata) { + _cleanup_strv_free_ char **units = NULL; + _cleanup_free_ char *unit = NULL; + const char *u; + sd_bus *bus; + int r; + + if (argv[1]) { + r = unit_name_mangle(argv[1], UNIT_NAME_NOGLOB, &unit); + if (r < 0) + return log_error_errno(r, "Failed to mangle unit name: %m"); + + u = unit; + } else + u = SPECIAL_DEFAULT_TARGET; + + r = acquire_bus(BUS_MANAGER, &bus); + if (r < 0) + return r; + + pager_open(arg_no_pager, false); + + puts(u); + + return list_dependencies_one(bus, u, 0, &units, 0); +} + +struct machine_info { + bool is_host; + char *name; + char *state; + char *control_group; + uint32_t n_failed_units; + uint32_t n_jobs; + usec_t timestamp; +}; + +static const struct bus_properties_map machine_info_property_map[] = { + { "SystemState", "s", NULL, offsetof(struct machine_info, state) }, + { "NJobs", "u", NULL, offsetof(struct machine_info, n_jobs) }, + { "NFailedUnits", "u", NULL, offsetof(struct machine_info, n_failed_units) }, + { "ControlGroup", "s", NULL, offsetof(struct machine_info, control_group) }, + { "UserspaceTimestamp", "t", NULL, offsetof(struct machine_info, timestamp) }, + {} +}; + +static void machine_info_clear(struct machine_info *info) { + assert(info); + + free(info->name); + free(info->state); + free(info->control_group); + zero(*info); +} + +static void free_machines_list(struct machine_info *machine_infos, int n) { + int i; + + if (!machine_infos) + return; + + for (i = 0; i < n; i++) + machine_info_clear(&machine_infos[i]); + + free(machine_infos); +} + +static int compare_machine_info(const void *a, const void *b) { + const struct machine_info *u = a, *v = b; + + if (u->is_host != v->is_host) + return u->is_host > v->is_host ? -1 : 1; + + return strcasecmp(u->name, v->name); +} + +static int get_machine_properties(sd_bus *bus, struct machine_info *mi) { + _cleanup_(sd_bus_flush_close_unrefp) sd_bus *container = NULL; + int r; + + assert(mi); + + if (!bus) { + r = sd_bus_open_system_machine(&container, mi->name); + if (r < 0) + return r; + + bus = container; + } + + r = bus_map_all_properties(bus, "org.freedesktop.systemd1", "/org/freedesktop/systemd1", machine_info_property_map, mi); + if (r < 0) + return r; + + return 0; +} + +static bool output_show_machine(const char *name, char **patterns) { + return strv_fnmatch_or_empty(patterns, name, FNM_NOESCAPE); +} + +static int get_machine_list( + sd_bus *bus, + struct machine_info **_machine_infos, + char **patterns) { + + struct machine_info *machine_infos = NULL; + _cleanup_strv_free_ char **m = NULL; + _cleanup_free_ char *hn = NULL; + size_t sz = 0; + char **i; + int c = 0, r; + + hn = gethostname_malloc(); + if (!hn) + return log_oom(); + + if (output_show_machine(hn, patterns)) { + if (!GREEDY_REALLOC0(machine_infos, sz, c+1)) + return log_oom(); + + machine_infos[c].is_host = true; + machine_infos[c].name = hn; + hn = NULL; + + get_machine_properties(bus, &machine_infos[c]); + c++; + } + + r = sd_get_machine_names(&m); + if (r < 0) + return log_error_errno(r, "Failed to get machine list: %m"); + + STRV_FOREACH(i, m) { + _cleanup_free_ char *class = NULL; + + if (!output_show_machine(*i, patterns)) + continue; + + sd_machine_get_class(*i, &class); + if (!streq_ptr(class, "container")) + continue; + + if (!GREEDY_REALLOC0(machine_infos, sz, c+1)) { + free_machines_list(machine_infos, c); + return log_oom(); + } + + machine_infos[c].is_host = false; + machine_infos[c].name = strdup(*i); + if (!machine_infos[c].name) { + free_machines_list(machine_infos, c); + return log_oom(); + } + + get_machine_properties(NULL, &machine_infos[c]); + c++; + } + + *_machine_infos = machine_infos; + return c; +} + +static void output_machines_list(struct machine_info *machine_infos, unsigned n) { + struct machine_info *m; + unsigned + circle_len = 0, + namelen = sizeof("NAME") - 1, + statelen = sizeof("STATE") - 1, + failedlen = sizeof("FAILED") - 1, + jobslen = sizeof("JOBS") - 1; + + assert(machine_infos || n == 0); + + for (m = machine_infos; m < machine_infos + n; m++) { + namelen = MAX(namelen, strlen(m->name) + (m->is_host ? sizeof(" (host)") - 1 : 0)); + statelen = MAX(statelen, m->state ? strlen(m->state) : 0); + failedlen = MAX(failedlen, DECIMAL_STR_WIDTH(m->n_failed_units)); + jobslen = MAX(jobslen, DECIMAL_STR_WIDTH(m->n_jobs)); + + if (!arg_plain && !streq_ptr(m->state, "running")) + circle_len = 2; + } + + if (!arg_no_legend) { + if (circle_len > 0) + fputs(" ", stdout); + + printf("%-*s %-*s %-*s %-*s\n", + namelen, "NAME", + statelen, "STATE", + failedlen, "FAILED", + jobslen, "JOBS"); + } + + for (m = machine_infos; m < machine_infos + n; m++) { + const char *on_state = "", *off_state = ""; + const char *on_failed = "", *off_failed = ""; + bool circle = false; + + if (streq_ptr(m->state, "degraded")) { + on_state = ansi_highlight_red(); + off_state = ansi_normal(); + circle = true; + } else if (!streq_ptr(m->state, "running")) { + on_state = ansi_highlight_yellow(); + off_state = ansi_normal(); + circle = true; + } + + if (m->n_failed_units > 0) { + on_failed = ansi_highlight_red(); + off_failed = ansi_normal(); + } else + on_failed = off_failed = ""; + + if (circle_len > 0) + printf("%s%s%s ", on_state, circle ? special_glyph(BLACK_CIRCLE) : " ", off_state); + + if (m->is_host) + printf("%-*s (host) %s%-*s%s %s%*" PRIu32 "%s %*" PRIu32 "\n", + (int) (namelen - (sizeof(" (host)")-1)), strna(m->name), + on_state, statelen, strna(m->state), off_state, + on_failed, failedlen, m->n_failed_units, off_failed, + jobslen, m->n_jobs); + else + printf("%-*s %s%-*s%s %s%*" PRIu32 "%s %*" PRIu32 "\n", + namelen, strna(m->name), + on_state, statelen, strna(m->state), off_state, + on_failed, failedlen, m->n_failed_units, off_failed, + jobslen, m->n_jobs); + } + + if (!arg_no_legend) + printf("\n%u machines listed.\n", n); +} + +static int list_machines(int argc, char *argv[], void *userdata) { + struct machine_info *machine_infos = NULL; + sd_bus *bus; + int r; + + if (geteuid() != 0) { + log_error("Must be root."); + return -EPERM; + } + + r = acquire_bus(BUS_MANAGER, &bus); + if (r < 0) + return r; + + r = get_machine_list(bus, &machine_infos, strv_skip(argv, 1)); + if (r < 0) + return r; + + pager_open(arg_no_pager, false); + + qsort_safe(machine_infos, r, sizeof(struct machine_info), compare_machine_info); + output_machines_list(machine_infos, r); + free_machines_list(machine_infos, r); + + return 0; +} + +static int get_default(int argc, char *argv[], void *userdata) { + _cleanup_(sd_bus_message_unrefp) sd_bus_message *reply = NULL; + _cleanup_free_ char *_path = NULL; + const char *path; + int r; + + if (install_client_side()) { + r = unit_file_get_default(arg_scope, arg_root, &_path); + if (r < 0) + return log_error_errno(r, "Failed to get default target: %m"); + path = _path; + + r = 0; + } else { + _cleanup_(sd_bus_error_free) sd_bus_error error = SD_BUS_ERROR_NULL; + sd_bus *bus; + + r = acquire_bus(BUS_MANAGER, &bus); + if (r < 0) + return r; + + r = sd_bus_call_method( + bus, + "org.freedesktop.systemd1", + "/org/freedesktop/systemd1", + "org.freedesktop.systemd1.Manager", + "GetDefaultTarget", + &error, + &reply, + NULL); + if (r < 0) + return log_error_errno(r, "Failed to get default target: %s", bus_error_message(&error, r)); + + r = sd_bus_message_read(reply, "s", &path); + if (r < 0) + return bus_log_parse_error(r); + } + + if (path) + printf("%s\n", path); + + return 0; +} + +static int set_default(int argc, char *argv[], void *userdata) { + _cleanup_free_ char *unit = NULL; + UnitFileChange *changes = NULL; + unsigned n_changes = 0; + int r; + + assert(argc >= 2); + assert(argv); + + r = unit_name_mangle_with_suffix(argv[1], UNIT_NAME_NOGLOB, ".target", &unit); + if (r < 0) + return log_error_errno(r, "Failed to mangle unit name: %m"); + + if (install_client_side()) { + r = unit_file_set_default(arg_scope, arg_root, unit, true, &changes, &n_changes); + unit_file_dump_changes(r, "set default", changes, n_changes, arg_quiet); + + if (r > 0) + r = 0; + } else { + _cleanup_(sd_bus_error_free) sd_bus_error error = SD_BUS_ERROR_NULL; + _cleanup_(sd_bus_message_unrefp) sd_bus_message *reply = NULL; + sd_bus *bus; + + polkit_agent_open_if_enabled(); + + r = acquire_bus(BUS_MANAGER, &bus); + if (r < 0) + return r; + + r = sd_bus_call_method( + bus, + "org.freedesktop.systemd1", + "/org/freedesktop/systemd1", + "org.freedesktop.systemd1.Manager", + "SetDefaultTarget", + &error, + &reply, + "sb", unit, 1); + if (r < 0) + return log_error_errno(r, "Failed to set default target: %s", bus_error_message(&error, r)); + + r = bus_deserialize_and_dump_unit_file_changes(reply, arg_quiet, &changes, &n_changes); + if (r < 0) + goto finish; + + /* Try to reload if enabled */ + if (!arg_no_reload) + r = daemon_reload(argc, argv, userdata); + else + r = 0; + } + +finish: + unit_file_changes_free(changes, n_changes); + + return r; +} + +struct job_info { + uint32_t id; + const char *name, *type, *state; +}; + +static void output_jobs_list(const struct job_info* jobs, unsigned n, bool skipped) { + unsigned id_len, unit_len, type_len, state_len; + const struct job_info *j; + const char *on, *off; + bool shorten = false; + + assert(n == 0 || jobs); + + if (n == 0) { + if (!arg_no_legend) { + on = ansi_highlight_green(); + off = ansi_normal(); + + printf("%sNo jobs %s.%s\n", on, skipped ? "listed" : "running", off); + } + return; + } + + pager_open(arg_no_pager, false); + + id_len = strlen("JOB"); + unit_len = strlen("UNIT"); + type_len = strlen("TYPE"); + state_len = strlen("STATE"); + + for (j = jobs; j < jobs + n; j++) { + uint32_t id = j->id; + assert(j->name && j->type && j->state); + + id_len = MAX(id_len, DECIMAL_STR_WIDTH(id)); + unit_len = MAX(unit_len, strlen(j->name)); + type_len = MAX(type_len, strlen(j->type)); + state_len = MAX(state_len, strlen(j->state)); + } + + if (!arg_full && id_len + 1 + unit_len + type_len + 1 + state_len > columns()) { + unit_len = MAX(33u, columns() - id_len - type_len - state_len - 3); + shorten = true; + } + + if (!arg_no_legend) + printf("%*s %-*s %-*s %-*s\n", + id_len, "JOB", + unit_len, "UNIT", + type_len, "TYPE", + state_len, "STATE"); + + for (j = jobs; j < jobs + n; j++) { + _cleanup_free_ char *e = NULL; + + if (streq(j->state, "running")) { + on = ansi_highlight(); + off = ansi_normal(); + } else + on = off = ""; + + e = shorten ? ellipsize(j->name, unit_len, 33) : NULL; + printf("%*u %s%-*s%s %-*s %s%-*s%s\n", + id_len, j->id, + on, unit_len, e ? e : j->name, off, + type_len, j->type, + on, state_len, j->state, off); + } + + if (!arg_no_legend) { + on = ansi_highlight(); + off = ansi_normal(); + + printf("\n%s%u jobs listed%s.\n", on, n, off); + } +} + +static bool output_show_job(struct job_info *job, char **patterns) { + return strv_fnmatch_or_empty(patterns, job->name, FNM_NOESCAPE); +} + +static int list_jobs(int argc, char *argv[], void *userdata) { + _cleanup_(sd_bus_error_free) sd_bus_error error = SD_BUS_ERROR_NULL; + _cleanup_(sd_bus_message_unrefp) sd_bus_message *reply = NULL; + const char *name, *type, *state, *job_path, *unit_path; + _cleanup_free_ struct job_info *jobs = NULL; + size_t size = 0; + unsigned c = 0; + sd_bus *bus; + uint32_t id; + int r; + bool skipped = false; + + r = acquire_bus(BUS_MANAGER, &bus); + if (r < 0) + return r; + + r = sd_bus_call_method( + bus, + "org.freedesktop.systemd1", + "/org/freedesktop/systemd1", + "org.freedesktop.systemd1.Manager", + "ListJobs", + &error, + &reply, + NULL); + if (r < 0) + return log_error_errno(r, "Failed to list jobs: %s", bus_error_message(&error, r)); + + r = sd_bus_message_enter_container(reply, 'a', "(usssoo)"); + if (r < 0) + return bus_log_parse_error(r); + + while ((r = sd_bus_message_read(reply, "(usssoo)", &id, &name, &type, &state, &job_path, &unit_path)) > 0) { + struct job_info job = { id, name, type, state }; + + if (!output_show_job(&job, strv_skip(argv, 1))) { + skipped = true; + continue; + } + + if (!GREEDY_REALLOC(jobs, size, c + 1)) + return log_oom(); + + jobs[c++] = job; + } + if (r < 0) + return bus_log_parse_error(r); + + r = sd_bus_message_exit_container(reply); + if (r < 0) + return bus_log_parse_error(r); + + pager_open(arg_no_pager, false); + + output_jobs_list(jobs, c, skipped); + return 0; +} + +static int cancel_job(int argc, char *argv[], void *userdata) { + sd_bus *bus; + char **name; + int r = 0; + + if (argc <= 1) + return trivial_method(argc, argv, userdata); + + r = acquire_bus(BUS_MANAGER, &bus); + if (r < 0) + return r; + + polkit_agent_open_if_enabled(); + + STRV_FOREACH(name, strv_skip(argv, 1)) { + _cleanup_(sd_bus_error_free) sd_bus_error error = SD_BUS_ERROR_NULL; + uint32_t id; + int q; + + q = safe_atou32(*name, &id); + if (q < 0) + return log_error_errno(q, "Failed to parse job id \"%s\": %m", *name); + + q = sd_bus_call_method( + bus, + "org.freedesktop.systemd1", + "/org/freedesktop/systemd1", + "org.freedesktop.systemd1.Manager", + "CancelJob", + &error, + NULL, + "u", id); + if (q < 0) { + log_error_errno(q, "Failed to cancel job %"PRIu32": %s", id, bus_error_message(&error, q)); + if (r == 0) + r = q; + } + } + + return r; +} + +static int need_daemon_reload(sd_bus *bus, const char *unit) { + _cleanup_(sd_bus_message_unrefp) sd_bus_message *reply = NULL; + const char *path; + int b, r; + + /* We ignore all errors here, since this is used to show a + * warning only */ + + /* We don't use unit_dbus_path_from_name() directly since we + * don't want to load the unit if it isn't loaded. */ + + r = sd_bus_call_method( + bus, + "org.freedesktop.systemd1", + "/org/freedesktop/systemd1", + "org.freedesktop.systemd1.Manager", + "GetUnit", + NULL, + &reply, + "s", unit); + if (r < 0) + return r; + + r = sd_bus_message_read(reply, "o", &path); + if (r < 0) + return r; + + r = sd_bus_get_property_trivial( + bus, + "org.freedesktop.systemd1", + path, + "org.freedesktop.systemd1.Unit", + "NeedDaemonReload", + NULL, + 'b', &b); + if (r < 0) + return r; + + return b; +} + +static void warn_unit_file_changed(const char *name) { + assert(name); + + log_warning("%sWarning:%s %s changed on disk. Run 'systemctl%s daemon-reload' to reload units.", + ansi_highlight_red(), + ansi_normal(), + name, + arg_scope == UNIT_FILE_SYSTEM ? "" : " --user"); +} + +static int unit_file_find_path(LookupPaths *lp, const char *unit_name, char **unit_path) { + char **p; + + assert(lp); + assert(unit_name); + assert(unit_path); + + STRV_FOREACH(p, lp->search_path) { + _cleanup_free_ char *path; + + path = path_join(arg_root, *p, unit_name); + if (!path) + return log_oom(); + + if (access(path, F_OK) == 0) { + *unit_path = path; + path = NULL; + return 1; + } + } + + return 0; +} + +static int unit_find_paths( + sd_bus *bus, + const char *unit_name, + LookupPaths *lp, + char **fragment_path, + char ***dropin_paths) { + + _cleanup_free_ char *path = NULL; + _cleanup_strv_free_ char **dropins = NULL; + int r; + + /** + * Finds where the unit is defined on disk. Returns 0 if the unit + * is not found. Returns 1 if it is found, and sets + * - the path to the unit in *path, if it exists on disk, + * - and a strv of existing drop-ins in *dropins, + * if the arg is not NULL and any dropins were found. + */ + + assert(unit_name); + assert(fragment_path); + assert(lp); + + if (!install_client_side() && !unit_name_is_valid(unit_name, UNIT_NAME_TEMPLATE)) { + _cleanup_(sd_bus_error_free) sd_bus_error error = SD_BUS_ERROR_NULL; + _cleanup_free_ char *unit = NULL; + + unit = unit_dbus_path_from_name(unit_name); + if (!unit) + return log_oom(); + + r = sd_bus_get_property_string( + bus, + "org.freedesktop.systemd1", + unit, + "org.freedesktop.systemd1.Unit", + "FragmentPath", + &error, + &path); + if (r < 0) + return log_error_errno(r, "Failed to get FragmentPath: %s", bus_error_message(&error, r)); + + if (dropin_paths) { + r = sd_bus_get_property_strv( + bus, + "org.freedesktop.systemd1", + unit, + "org.freedesktop.systemd1.Unit", + "DropInPaths", + &error, + &dropins); + if (r < 0) + return log_error_errno(r, "Failed to get DropInPaths: %s", bus_error_message(&error, r)); + } + } else { + _cleanup_set_free_ Set *names; + + names = set_new(NULL); + if (!names) + return log_oom(); + + r = set_put(names, unit_name); + if (r < 0) + return log_error_errno(r, "Failed to add unit name: %m"); + + r = unit_file_find_path(lp, unit_name, &path); + if (r < 0) + return r; + + if (r == 0) { + _cleanup_free_ char *template = NULL; + + r = unit_name_template(unit_name, &template); + if (r < 0 && r != -EINVAL) + return log_error_errno(r, "Failed to determine template name: %m"); + if (r >= 0) { + r = unit_file_find_path(lp, template, &path); + if (r < 0) + return r; + } + } + + if (dropin_paths) { + r = unit_file_find_dropin_paths(lp->search_path, NULL, names, &dropins); + if (r < 0) + return r; + } + } + + r = 0; + + if (!isempty(path)) { + *fragment_path = path; + path = NULL; + r = 1; + } + + if (dropin_paths && !strv_isempty(dropins)) { + *dropin_paths = dropins; + dropins = NULL; + r = 1; + } + + if (r == 0 && !arg_force) + log_error("No files found for %s.", unit_name); + + return r; +} + +static int get_state_one_unit(sd_bus *bus, const char *name, UnitActiveState *active_state) { + _cleanup_(sd_bus_error_free) sd_bus_error error = SD_BUS_ERROR_NULL; + _cleanup_(sd_bus_message_unrefp) sd_bus_message *reply = NULL; + _cleanup_free_ char *buf = NULL; + UnitActiveState state; + const char *path; + int r; + + assert(name); + assert(active_state); + + /* We don't use unit_dbus_path_from_name() directly since we don't want to load the unit unnecessarily, if it + * isn't loaded. */ + r = sd_bus_call_method( + bus, + "org.freedesktop.systemd1", + "/org/freedesktop/systemd1", + "org.freedesktop.systemd1.Manager", + "GetUnit", + &error, + &reply, + "s", name); + if (r < 0) { + if (!sd_bus_error_has_name(&error, BUS_ERROR_NO_SUCH_UNIT)) + return log_error_errno(r, "Failed to retrieve unit: %s", bus_error_message(&error, r)); + + /* The unit is currently not loaded, hence say it's "inactive", since all units that aren't loaded are + * considered inactive. */ + state = UNIT_INACTIVE; + + } else { + r = sd_bus_message_read(reply, "o", &path); + if (r < 0) + return bus_log_parse_error(r); + + r = sd_bus_get_property_string( + bus, + "org.freedesktop.systemd1", + path, + "org.freedesktop.systemd1.Unit", + "ActiveState", + &error, + &buf); + if (r < 0) + return log_error_errno(r, "Failed to retrieve unit state: %s", bus_error_message(&error, r)); + + state = unit_active_state_from_string(buf); + if (state == _UNIT_ACTIVE_STATE_INVALID) { + log_error("Invalid unit state '%s' for: %s", buf, name); + return -EINVAL; + } + } + + *active_state = state; + return 0; +} + +static int check_triggering_units( + sd_bus *bus, + const char *name) { + + _cleanup_(sd_bus_error_free) sd_bus_error error = SD_BUS_ERROR_NULL; + _cleanup_free_ char *path = NULL, *n = NULL, *load_state = NULL; + _cleanup_strv_free_ char **triggered_by = NULL; + bool print_warning_label = true; + UnitActiveState active_state; + char **i; + int r; + + r = unit_name_mangle(name, UNIT_NAME_NOGLOB, &n); + if (r < 0) + return log_error_errno(r, "Failed to mangle unit name: %m"); + + path = unit_dbus_path_from_name(n); + if (!path) + return log_oom(); + + r = sd_bus_get_property_string( + bus, + "org.freedesktop.systemd1", + path, + "org.freedesktop.systemd1.Unit", + "LoadState", + &error, + &load_state); + if (r < 0) + return log_error_errno(r, "Failed to get load state of %s: %s", n, bus_error_message(&error, r)); + + if (streq(load_state, "masked")) + return 0; + + r = sd_bus_get_property_strv( + bus, + "org.freedesktop.systemd1", + path, + "org.freedesktop.systemd1.Unit", + "TriggeredBy", + &error, + &triggered_by); + if (r < 0) + return log_error_errno(r, "Failed to get triggered by array of %s: %s", n, bus_error_message(&error, r)); + + STRV_FOREACH(i, triggered_by) { + r = get_state_one_unit(bus, *i, &active_state); + if (r < 0) + return r; + + if (!IN_SET(active_state, UNIT_ACTIVE, UNIT_RELOADING)) + continue; + + if (print_warning_label) { + log_warning("Warning: Stopping %s, but it can still be activated by:", n); + print_warning_label = false; + } + + log_warning(" %s", *i); + } + + return 0; +} + +static const struct { + const char *verb; + const char *method; +} unit_actions[] = { + { "start", "StartUnit" }, + { "stop", "StopUnit" }, + { "condstop", "StopUnit" }, + { "reload", "ReloadUnit" }, + { "restart", "RestartUnit" }, + { "try-restart", "TryRestartUnit" }, + { "condrestart", "TryRestartUnit" }, + { "reload-or-restart", "ReloadOrRestartUnit" }, + { "try-reload-or-restart", "ReloadOrTryRestartUnit" }, + { "reload-or-try-restart", "ReloadOrTryRestartUnit" }, + { "condreload", "ReloadOrTryRestartUnit" }, + { "force-reload", "ReloadOrTryRestartUnit" } +}; + +static const char *verb_to_method(const char *verb) { + uint i; + + for (i = 0; i < ELEMENTSOF(unit_actions); i++) + if (streq_ptr(unit_actions[i].verb, verb)) + return unit_actions[i].method; + + return "StartUnit"; +} + +static const char *method_to_verb(const char *method) { + uint i; + + for (i = 0; i < ELEMENTSOF(unit_actions); i++) + if (streq_ptr(unit_actions[i].method, method)) + return unit_actions[i].verb; + + return "n/a"; +} + +static int start_unit_one( + sd_bus *bus, + const char *method, + const char *name, + const char *mode, + sd_bus_error *error, + BusWaitForJobs *w) { + + _cleanup_(sd_bus_message_unrefp) sd_bus_message *reply = NULL; + const char *path; + int r; + + assert(method); + assert(name); + assert(mode); + assert(error); + + log_debug("Calling manager for %s on %s, %s", method, name, mode); + + r = sd_bus_call_method( + bus, + "org.freedesktop.systemd1", + "/org/freedesktop/systemd1", + "org.freedesktop.systemd1.Manager", + method, + error, + &reply, + "ss", name, mode); + if (r < 0) { + const char *verb; + + /* There's always a fallback possible for legacy actions. */ + if (arg_action != ACTION_SYSTEMCTL) + return r; + + verb = method_to_verb(method); + + log_error("Failed to %s %s: %s", verb, name, bus_error_message(error, r)); + + if (!sd_bus_error_has_name(error, BUS_ERROR_NO_SUCH_UNIT) && + !sd_bus_error_has_name(error, BUS_ERROR_UNIT_MASKED)) + log_error("See %s logs and 'systemctl%s status %s' for details.", + arg_scope == UNIT_FILE_SYSTEM ? "system" : "user", + arg_scope == UNIT_FILE_SYSTEM ? "" : " --user", + name); + + return r; + } + + r = sd_bus_message_read(reply, "o", &path); + if (r < 0) + return bus_log_parse_error(r); + + if (need_daemon_reload(bus, name) > 0) + warn_unit_file_changed(name); + + if (w) { + log_debug("Adding %s to the set", path); + r = bus_wait_for_jobs_add(w, path); + if (r < 0) + return log_oom(); + } + + return 0; +} + +static int expand_names(sd_bus *bus, char **names, const char* suffix, char ***ret) { + _cleanup_strv_free_ char **mangled = NULL, **globs = NULL; + char **name; + int r, i; + + assert(bus); + assert(ret); + + STRV_FOREACH(name, names) { + char *t; + + if (suffix) + r = unit_name_mangle_with_suffix(*name, UNIT_NAME_GLOB, suffix, &t); + else + r = unit_name_mangle(*name, UNIT_NAME_GLOB, &t); + if (r < 0) + return log_error_errno(r, "Failed to mangle name: %m"); + + if (string_is_glob(t)) + r = strv_consume(&globs, t); + else + r = strv_consume(&mangled, t); + if (r < 0) + return log_oom(); + } + + /* Query the manager only if any of the names are a glob, since + * this is fairly expensive */ + if (!strv_isempty(globs)) { + _cleanup_(sd_bus_message_unrefp) sd_bus_message *reply = NULL; + _cleanup_free_ UnitInfo *unit_infos = NULL; + size_t allocated, n; + + r = get_unit_list(bus, NULL, globs, &unit_infos, 0, &reply); + if (r < 0) + return r; + + n = strv_length(mangled); + allocated = n + 1; + + for (i = 0; i < r; i++) { + if (!GREEDY_REALLOC(mangled, allocated, n+2)) + return log_oom(); + + mangled[n] = strdup(unit_infos[i].id); + if (!mangled[n]) + return log_oom(); + + mangled[++n] = NULL; + } + } + + *ret = mangled; + mangled = NULL; /* do not free */ + + return 0; +} + +static const struct { + const char *target; + const char *verb; + const char *mode; +} action_table[_ACTION_MAX] = { + [ACTION_HALT] = { SPECIAL_HALT_TARGET, "halt", "replace-irreversibly" }, + [ACTION_POWEROFF] = { SPECIAL_POWEROFF_TARGET, "poweroff", "replace-irreversibly" }, + [ACTION_REBOOT] = { SPECIAL_REBOOT_TARGET, "reboot", "replace-irreversibly" }, + [ACTION_KEXEC] = { SPECIAL_KEXEC_TARGET, "kexec", "replace-irreversibly" }, + [ACTION_RUNLEVEL2] = { SPECIAL_MULTI_USER_TARGET, NULL, "isolate" }, + [ACTION_RUNLEVEL3] = { SPECIAL_MULTI_USER_TARGET, NULL, "isolate" }, + [ACTION_RUNLEVEL4] = { SPECIAL_MULTI_USER_TARGET, NULL, "isolate" }, + [ACTION_RUNLEVEL5] = { SPECIAL_GRAPHICAL_TARGET, NULL, "isolate" }, + [ACTION_RESCUE] = { SPECIAL_RESCUE_TARGET, "rescue", "isolate" }, + [ACTION_EMERGENCY] = { SPECIAL_EMERGENCY_TARGET, "emergency", "isolate" }, + [ACTION_DEFAULT] = { SPECIAL_DEFAULT_TARGET, "default", "isolate" }, + [ACTION_EXIT] = { SPECIAL_EXIT_TARGET, "exit", "replace-irreversibly" }, + [ACTION_SUSPEND] = { SPECIAL_SUSPEND_TARGET, "suspend", "replace-irreversibly" }, + [ACTION_HIBERNATE] = { SPECIAL_HIBERNATE_TARGET, "hibernate", "replace-irreversibly" }, + [ACTION_HYBRID_SLEEP] = { SPECIAL_HYBRID_SLEEP_TARGET, "hybrid-sleep", "replace-irreversibly" }, +}; + +static enum action verb_to_action(const char *verb) { + enum action i; + + for (i = _ACTION_INVALID; i < _ACTION_MAX; i++) + if (streq_ptr(action_table[i].verb, verb)) + return i; + + return _ACTION_INVALID; +} + +static int start_unit(int argc, char *argv[], void *userdata) { + _cleanup_(bus_wait_for_jobs_freep) BusWaitForJobs *w = NULL; + const char *method, *mode, *one_name, *suffix = NULL; + _cleanup_strv_free_ char **names = NULL; + sd_bus *bus; + char **name; + int r = 0; + + r = acquire_bus(BUS_MANAGER, &bus); + if (r < 0) + return r; + + ask_password_agent_open_if_enabled(); + polkit_agent_open_if_enabled(); + + if (arg_action == ACTION_SYSTEMCTL) { + enum action action; + + method = verb_to_method(argv[0]); + action = verb_to_action(argv[0]); + + if (streq(argv[0], "isolate")) { + mode = "isolate"; + suffix = ".target"; + } else + mode = action_table[action].mode ?: arg_job_mode; + + one_name = action_table[action].target; + } else { + assert(arg_action < ELEMENTSOF(action_table)); + assert(action_table[arg_action].target); + + method = "StartUnit"; + + mode = action_table[arg_action].mode; + one_name = action_table[arg_action].target; + } + + if (one_name) + names = strv_new(one_name, NULL); + else { + r = expand_names(bus, strv_skip(argv, 1), suffix, &names); + if (r < 0) + return log_error_errno(r, "Failed to expand names: %m"); + } + + if (!arg_no_block) { + r = bus_wait_for_jobs_new(bus, &w); + if (r < 0) + return log_error_errno(r, "Could not watch jobs: %m"); + } + + STRV_FOREACH(name, names) { + _cleanup_(sd_bus_error_free) sd_bus_error error = SD_BUS_ERROR_NULL; + int q; + + q = start_unit_one(bus, method, *name, mode, &error, w); + if (r >= 0 && q < 0) + r = translate_bus_error_to_exit_status(q, &error); + } + + if (!arg_no_block) { + int q, arg_count = 0; + const char* extra_args[4] = {}; + + if (arg_scope != UNIT_FILE_SYSTEM) + extra_args[arg_count++] = "--user"; + + assert(IN_SET(arg_transport, BUS_TRANSPORT_LOCAL, BUS_TRANSPORT_REMOTE, BUS_TRANSPORT_MACHINE)); + if (arg_transport == BUS_TRANSPORT_REMOTE) { + extra_args[arg_count++] = "-H"; + extra_args[arg_count++] = arg_host; + } else if (arg_transport == BUS_TRANSPORT_MACHINE) { + extra_args[arg_count++] = "-M"; + extra_args[arg_count++] = arg_host; + } + + q = bus_wait_for_jobs(w, arg_quiet, extra_args); + if (q < 0) + return q; + + /* When stopping units, warn if they can still be triggered by + * another active unit (socket, path, timer) */ + if (!arg_quiet && streq(method, "StopUnit")) + STRV_FOREACH(name, names) + check_triggering_units(bus, *name); + } + + return r; +} + +static int logind_set_wall_message(void) { +#ifdef HAVE_LOGIND + _cleanup_(sd_bus_error_free) sd_bus_error error = SD_BUS_ERROR_NULL; + sd_bus *bus; + _cleanup_free_ char *m = NULL; + int r; + + r = acquire_bus(BUS_FULL, &bus); + if (r < 0) + return r; + + m = strv_join(arg_wall, " "); + if (!m) + return log_oom(); + + r = sd_bus_call_method( + bus, + "org.freedesktop.login1", + "/org/freedesktop/login1", + "org.freedesktop.login1.Manager", + "SetWallMessage", + &error, + NULL, + "sb", + m, + !arg_no_wall); + + if (r < 0) + return log_warning_errno(r, "Failed to set wall message, ignoring: %s", bus_error_message(&error, r)); + +#endif + return 0; +} + +/* Ask systemd-logind, which might grant access to unprivileged users + * through PolicyKit */ +static int logind_reboot(enum action a) { +#ifdef HAVE_LOGIND + _cleanup_(sd_bus_error_free) sd_bus_error error = SD_BUS_ERROR_NULL; + const char *method, *description; + sd_bus *bus; + int r; + + r = acquire_bus(BUS_FULL, &bus); + if (r < 0) + return r; + + switch (a) { + + case ACTION_REBOOT: + method = "Reboot"; + description = "reboot system"; + break; + + case ACTION_POWEROFF: + method = "PowerOff"; + description = "power off system"; + break; + + case ACTION_SUSPEND: + method = "Suspend"; + description = "suspend system"; + break; + + case ACTION_HIBERNATE: + method = "Hibernate"; + description = "hibernate system"; + break; + + case ACTION_HYBRID_SLEEP: + method = "HybridSleep"; + description = "put system into hybrid sleep"; + break; + + default: + return -EINVAL; + } + + polkit_agent_open_if_enabled(); + (void) logind_set_wall_message(); + + r = sd_bus_call_method( + bus, + "org.freedesktop.login1", + "/org/freedesktop/login1", + "org.freedesktop.login1.Manager", + method, + &error, + NULL, + "b", arg_ask_password); + if (r < 0) + return log_error_errno(r, "Failed to %s via logind: %s", description, bus_error_message(&error, r)); + + return 0; +#else + return -ENOSYS; +#endif +} + +static int logind_check_inhibitors(enum action a) { +#ifdef HAVE_LOGIND + _cleanup_(sd_bus_message_unrefp) sd_bus_message *reply = NULL; + _cleanup_strv_free_ char **sessions = NULL; + const char *what, *who, *why, *mode; + uint32_t uid, pid; + sd_bus *bus; + unsigned c = 0; + char **s; + int r; + + if (arg_ignore_inhibitors || arg_force > 0) + return 0; + + if (arg_when > 0) + return 0; + + if (geteuid() == 0) + return 0; + + if (!on_tty()) + return 0; + + if (arg_transport != BUS_TRANSPORT_LOCAL) + return 0; + + r = acquire_bus(BUS_FULL, &bus); + if (r < 0) + return r; + + r = sd_bus_call_method( + bus, + "org.freedesktop.login1", + "/org/freedesktop/login1", + "org.freedesktop.login1.Manager", + "ListInhibitors", + NULL, + &reply, + NULL); + if (r < 0) + /* If logind is not around, then there are no inhibitors... */ + return 0; + + r = sd_bus_message_enter_container(reply, SD_BUS_TYPE_ARRAY, "(ssssuu)"); + if (r < 0) + return bus_log_parse_error(r); + + while ((r = sd_bus_message_read(reply, "(ssssuu)", &what, &who, &why, &mode, &uid, &pid)) > 0) { + _cleanup_free_ char *comm = NULL, *user = NULL; + _cleanup_strv_free_ char **sv = NULL; + + if (!streq(mode, "block")) + continue; + + sv = strv_split(what, ":"); + if (!sv) + return log_oom(); + + if ((pid_t) pid < 0) + return log_error_errno(ERANGE, "Bad PID %"PRIu32": %m", pid); + + if (!strv_contains(sv, + IN_SET(a, + ACTION_HALT, + ACTION_POWEROFF, + ACTION_REBOOT, + ACTION_KEXEC) ? "shutdown" : "sleep")) + continue; + + get_process_comm(pid, &comm); + user = uid_to_name(uid); + + log_warning("Operation inhibited by \"%s\" (PID "PID_FMT" \"%s\", user %s), reason is \"%s\".", + who, (pid_t) pid, strna(comm), strna(user), why); + + c++; + } + if (r < 0) + return bus_log_parse_error(r); + + r = sd_bus_message_exit_container(reply); + if (r < 0) + return bus_log_parse_error(r); + + /* Check for current sessions */ + sd_get_sessions(&sessions); + STRV_FOREACH(s, sessions) { + _cleanup_free_ char *type = NULL, *tty = NULL, *seat = NULL, *user = NULL, *service = NULL, *class = NULL; + + if (sd_session_get_uid(*s, &uid) < 0 || uid == getuid()) + continue; + + if (sd_session_get_class(*s, &class) < 0 || !streq(class, "user")) + continue; + + if (sd_session_get_type(*s, &type) < 0 || (!streq(type, "x11") && !streq(type, "tty"))) + continue; + + sd_session_get_tty(*s, &tty); + sd_session_get_seat(*s, &seat); + sd_session_get_service(*s, &service); + user = uid_to_name(uid); + + log_warning("User %s is logged in on %s.", strna(user), isempty(tty) ? (isempty(seat) ? strna(service) : seat) : tty); + c++; + } + + if (c <= 0) + return 0; + + log_error("Please retry operation after closing inhibitors and logging out other users.\nAlternatively, ignore inhibitors and users with 'systemctl %s -i'.", + action_table[a].verb); + + return -EPERM; +#else + return 0; +#endif +} + +static int logind_prepare_firmware_setup(void) { +#ifdef HAVE_LOGIND + _cleanup_(sd_bus_error_free) sd_bus_error error = SD_BUS_ERROR_NULL; + sd_bus *bus; + int r; + + r = acquire_bus(BUS_FULL, &bus); + if (r < 0) + return r; + + r = sd_bus_call_method( + bus, + "org.freedesktop.login1", + "/org/freedesktop/login1", + "org.freedesktop.login1.Manager", + "SetRebootToFirmwareSetup", + &error, + NULL, + "b", true); + if (r < 0) + return log_error_errno(r, "Cannot indicate to EFI to boot into setup mode: %s", bus_error_message(&error, r)); + + return 0; +#else + log_error("Cannot remotely indicate to EFI to boot into setup mode."); + return -ENOSYS; +#endif +} + +static int prepare_firmware_setup(void) { + int r; + + if (!arg_firmware_setup) + return 0; + + if (arg_transport == BUS_TRANSPORT_LOCAL) { + + r = efi_set_reboot_to_firmware(true); + if (r < 0) + log_debug_errno(r, "Cannot indicate to EFI to boot into setup mode, will retry via logind: %m"); + else + return r; + } + + return logind_prepare_firmware_setup(); +} + +static int set_exit_code(uint8_t code) { + _cleanup_(sd_bus_error_free) sd_bus_error error = SD_BUS_ERROR_NULL; + sd_bus *bus; + int r; + + r = acquire_bus(BUS_MANAGER, &bus); + if (r < 0) + return r; + + r = sd_bus_call_method( + bus, + "org.freedesktop.systemd1", + "/org/freedesktop/systemd1", + "org.freedesktop.systemd1.Manager", + "SetExitCode", + &error, + NULL, + "y", code); + if (r < 0) + return log_error_errno(r, "Failed to set exit code: %s", bus_error_message(&error, r)); + + return 0; +} + +static int start_special(int argc, char *argv[], void *userdata) { + enum action a; + int r; + + assert(argv); + + a = verb_to_action(argv[0]); + + r = logind_check_inhibitors(a); + if (r < 0) + return r; + + if (arg_force >= 2 && geteuid() != 0) { + log_error("Must be root."); + return -EPERM; + } + + r = prepare_firmware_setup(); + if (r < 0) + return r; + + if (a == ACTION_REBOOT && argc > 1) { + r = update_reboot_parameter_and_warn(argv[1]); + if (r < 0) + return r; + + } else if (a == ACTION_EXIT && argc > 1) { + uint8_t code; + + /* If the exit code is not given on the command line, + * don't reset it to zero: just keep it as it might + * have been set previously. */ + + r = safe_atou8(argv[1], &code); + if (r < 0) + return log_error_errno(r, "Invalid exit code."); + + r = set_exit_code(code); + if (r < 0) + return r; + } + + if (arg_force >= 2 && + IN_SET(a, + ACTION_HALT, + ACTION_POWEROFF, + ACTION_REBOOT)) + return halt_now(a); + + if (arg_force >= 1 && + IN_SET(a, + ACTION_HALT, + ACTION_POWEROFF, + ACTION_REBOOT, + ACTION_KEXEC, + ACTION_EXIT)) + return trivial_method(argc, argv, userdata); + + /* First try logind, to allow authentication with polkit */ + if (IN_SET(a, + ACTION_POWEROFF, + ACTION_REBOOT, + ACTION_SUSPEND, + ACTION_HIBERNATE, + ACTION_HYBRID_SLEEP)) { + r = logind_reboot(a); + if (r >= 0) + return r; + if (IN_SET(r, -EOPNOTSUPP, -EINPROGRESS)) + /* requested operation is not supported or already in progress */ + return r; + + /* On all other errors, try low-level operation */ + } + + return start_unit(argc, argv, userdata); +} + +static int start_system_special(int argc, char *argv[], void *userdata) { + /* Like start_special above, but raises an error when running in user mode */ + + if (arg_scope != UNIT_FILE_SYSTEM) { + log_error("Bad action for %s mode.", + arg_scope == UNIT_FILE_GLOBAL ? "--global" : "--user"); + return -EINVAL; + } + + return start_special(argc, argv, userdata); +} + +static int check_unit_generic(int code, const UnitActiveState good_states[], int nb_states, char **args) { + _cleanup_strv_free_ char **names = NULL; + UnitActiveState active_state; + sd_bus *bus; + char **name; + int r, i; + bool found = false; + + r = acquire_bus(BUS_MANAGER, &bus); + if (r < 0) + return r; + + r = expand_names(bus, args, NULL, &names); + if (r < 0) + return log_error_errno(r, "Failed to expand names: %m"); + + STRV_FOREACH(name, names) { + r = get_state_one_unit(bus, *name, &active_state); + if (r < 0) + return r; + + if (!arg_quiet) + puts(unit_active_state_to_string(active_state)); + + for (i = 0; i < nb_states; ++i) + if (good_states[i] == active_state) + found = true; + } + + /* use the given return code for the case that we won't find + * any unit which matches the list */ + return found ? 0 : code; +} + +static int check_unit_active(int argc, char *argv[], void *userdata) { + const UnitActiveState states[] = { UNIT_ACTIVE, UNIT_RELOADING }; + /* According to LSB: 3, "program is not running" */ + return check_unit_generic(EXIT_PROGRAM_NOT_RUNNING, states, ELEMENTSOF(states), strv_skip(argv, 1)); +} + +static int check_unit_failed(int argc, char *argv[], void *userdata) { + const UnitActiveState states[] = { UNIT_FAILED }; + return check_unit_generic(EXIT_PROGRAM_DEAD_AND_PID_EXISTS, states, ELEMENTSOF(states), strv_skip(argv, 1)); +} + +static int kill_unit(int argc, char *argv[], void *userdata) { + _cleanup_strv_free_ char **names = NULL; + char *kill_who = NULL, **name; + sd_bus *bus; + int r, q; + + r = acquire_bus(BUS_MANAGER, &bus); + if (r < 0) + return r; + + polkit_agent_open_if_enabled(); + + if (!arg_kill_who) + arg_kill_who = "all"; + + /* --fail was specified */ + if (streq(arg_job_mode, "fail")) + kill_who = strjoina(arg_kill_who, "-fail"); + + r = expand_names(bus, strv_skip(argv, 1), NULL, &names); + if (r < 0) + return log_error_errno(r, "Failed to expand names: %m"); + + STRV_FOREACH(name, names) { + _cleanup_(sd_bus_error_free) sd_bus_error error = SD_BUS_ERROR_NULL; + + q = sd_bus_call_method( + bus, + "org.freedesktop.systemd1", + "/org/freedesktop/systemd1", + "org.freedesktop.systemd1.Manager", + "KillUnit", + &error, + NULL, + "ssi", *names, kill_who ? kill_who : arg_kill_who, arg_signal); + if (q < 0) { + log_error_errno(q, "Failed to kill unit %s: %s", *names, bus_error_message(&error, q)); + if (r == 0) + r = q; + } + } + + return r; +} + +typedef struct ExecStatusInfo { + char *name; + + char *path; + char **argv; + + bool ignore; + + usec_t start_timestamp; + usec_t exit_timestamp; + pid_t pid; + int code; + int status; + + LIST_FIELDS(struct ExecStatusInfo, exec); +} ExecStatusInfo; + +static void exec_status_info_free(ExecStatusInfo *i) { + assert(i); + + free(i->name); + free(i->path); + strv_free(i->argv); + free(i); +} + +static int exec_status_info_deserialize(sd_bus_message *m, ExecStatusInfo *i) { + uint64_t start_timestamp, exit_timestamp, start_timestamp_monotonic, exit_timestamp_monotonic; + const char *path; + uint32_t pid; + int32_t code, status; + int ignore, r; + + assert(m); + assert(i); + + r = sd_bus_message_enter_container(m, SD_BUS_TYPE_STRUCT, "sasbttttuii"); + if (r < 0) + return bus_log_parse_error(r); + else if (r == 0) + return 0; + + r = sd_bus_message_read(m, "s", &path); + if (r < 0) + return bus_log_parse_error(r); + + i->path = strdup(path); + if (!i->path) + return log_oom(); + + r = sd_bus_message_read_strv(m, &i->argv); + if (r < 0) + return bus_log_parse_error(r); + + r = sd_bus_message_read(m, + "bttttuii", + &ignore, + &start_timestamp, &start_timestamp_monotonic, + &exit_timestamp, &exit_timestamp_monotonic, + &pid, + &code, &status); + if (r < 0) + return bus_log_parse_error(r); + + i->ignore = ignore; + i->start_timestamp = (usec_t) start_timestamp; + i->exit_timestamp = (usec_t) exit_timestamp; + i->pid = (pid_t) pid; + i->code = code; + i->status = status; + + r = sd_bus_message_exit_container(m); + if (r < 0) + return bus_log_parse_error(r); + + return 1; +} + +typedef struct UnitCondition { + char *name; + char *param; + bool trigger; + bool negate; + int tristate; + + LIST_FIELDS(struct UnitCondition, conditions); +} UnitCondition; + +static void unit_condition_free(UnitCondition *c) { + if (!c) + return; + + free(c->name); + free(c->param); + free(c); +} + +DEFINE_TRIVIAL_CLEANUP_FUNC(UnitCondition*, unit_condition_free); + +typedef struct UnitStatusInfo { + const char *id; + const char *load_state; + const char *active_state; + const char *sub_state; + const char *unit_file_state; + const char *unit_file_preset; + + const char *description; + const char *following; + + char **documentation; + + const char *fragment_path; + const char *source_path; + const char *control_group; + + char **dropin_paths; + + const char *load_error; + const char *result; + + usec_t inactive_exit_timestamp; + usec_t inactive_exit_timestamp_monotonic; + usec_t active_enter_timestamp; + usec_t active_exit_timestamp; + usec_t inactive_enter_timestamp; + + bool need_daemon_reload; + bool transient; + + /* Service */ + pid_t main_pid; + pid_t control_pid; + const char *status_text; + const char *pid_file; + bool running:1; + int status_errno; + + usec_t start_timestamp; + usec_t exit_timestamp; + + int exit_code, exit_status; + + usec_t condition_timestamp; + bool condition_result; + LIST_HEAD(UnitCondition, conditions); + + usec_t assert_timestamp; + bool assert_result; + bool failed_assert_trigger; + bool failed_assert_negate; + const char *failed_assert; + const char *failed_assert_parameter; + + /* Socket */ + unsigned n_accepted; + unsigned n_connections; + bool accept; + + /* Pairs of type, path */ + char **listen; + + /* Device */ + const char *sysfs_path; + + /* Mount, Automount */ + const char *where; + + /* Swap */ + const char *what; + + /* CGroup */ + uint64_t memory_current; + uint64_t memory_low; + uint64_t memory_high; + uint64_t memory_max; + uint64_t memory_limit; + uint64_t cpu_usage_nsec; + uint64_t tasks_current; + uint64_t tasks_max; + + LIST_HEAD(ExecStatusInfo, exec); +} UnitStatusInfo; + +static void unit_status_info_free(UnitStatusInfo *info) { + ExecStatusInfo *p; + UnitCondition *c; + + strv_free(info->documentation); + strv_free(info->dropin_paths); + strv_free(info->listen); + + while ((c = info->conditions)) { + LIST_REMOVE(conditions, info->conditions, c); + unit_condition_free(c); + } + + while ((p = info->exec)) { + LIST_REMOVE(exec, info->exec, p); + exec_status_info_free(p); + } +} + +static void print_status_info( + sd_bus *bus, + UnitStatusInfo *i, + bool *ellipsized) { + + ExecStatusInfo *p; + const char *active_on, *active_off, *on, *off, *ss; + usec_t timestamp; + char since1[FORMAT_TIMESTAMP_RELATIVE_MAX], *s1; + char since2[FORMAT_TIMESTAMP_MAX], *s2; + const char *path; + char **t, **t2; + int r; + + assert(i); + + /* This shows pretty information about a unit. See + * print_property() for a low-level property printer */ + + if (streq_ptr(i->active_state, "failed")) { + active_on = ansi_highlight_red(); + active_off = ansi_normal(); + } else if (streq_ptr(i->active_state, "active") || streq_ptr(i->active_state, "reloading")) { + active_on = ansi_highlight_green(); + active_off = ansi_normal(); + } else + active_on = active_off = ""; + + printf("%s%s%s %s", active_on, special_glyph(BLACK_CIRCLE), active_off, strna(i->id)); + + if (i->description && !streq_ptr(i->id, i->description)) + printf(" - %s", i->description); + + printf("\n"); + + if (i->following) + printf(" Follow: unit currently follows state of %s\n", i->following); + + if (streq_ptr(i->load_state, "error")) { + on = ansi_highlight_red(); + off = ansi_normal(); + } else + on = off = ""; + + path = i->source_path ? i->source_path : i->fragment_path; + + if (i->load_error != 0) + printf(" Loaded: %s%s%s (Reason: %s)\n", + on, strna(i->load_state), off, i->load_error); + else if (path && !isempty(i->unit_file_state) && !isempty(i->unit_file_preset)) + printf(" Loaded: %s%s%s (%s; %s; vendor preset: %s)\n", + on, strna(i->load_state), off, path, i->unit_file_state, i->unit_file_preset); + else if (path && !isempty(i->unit_file_state)) + printf(" Loaded: %s%s%s (%s; %s)\n", + on, strna(i->load_state), off, path, i->unit_file_state); + else if (path) + printf(" Loaded: %s%s%s (%s)\n", + on, strna(i->load_state), off, path); + else + printf(" Loaded: %s%s%s\n", + on, strna(i->load_state), off); + + if (i->transient) + printf("Transient: yes\n"); + + if (!strv_isempty(i->dropin_paths)) { + _cleanup_free_ char *dir = NULL; + bool last = false; + char ** dropin; + + STRV_FOREACH(dropin, i->dropin_paths) { + if (! dir || last) { + printf(dir ? " " : " Drop-In: "); + + dir = mfree(dir); + + dir = dirname_malloc(*dropin); + if (!dir) { + log_oom(); + return; + } + + printf("%s\n %s", dir, + special_glyph(TREE_RIGHT)); + } + + last = ! (*(dropin + 1) && startswith(*(dropin + 1), dir)); + + printf("%s%s", basename(*dropin), last ? "\n" : ", "); + } + } + + ss = streq_ptr(i->active_state, i->sub_state) ? NULL : i->sub_state; + if (ss) + printf(" Active: %s%s (%s)%s", + active_on, strna(i->active_state), ss, active_off); + else + printf(" Active: %s%s%s", + active_on, strna(i->active_state), active_off); + + if (!isempty(i->result) && !streq(i->result, "success")) + printf(" (Result: %s)", i->result); + + timestamp = (streq_ptr(i->active_state, "active") || + streq_ptr(i->active_state, "reloading")) ? i->active_enter_timestamp : + (streq_ptr(i->active_state, "inactive") || + streq_ptr(i->active_state, "failed")) ? i->inactive_enter_timestamp : + streq_ptr(i->active_state, "activating") ? i->inactive_exit_timestamp : + i->active_exit_timestamp; + + s1 = format_timestamp_relative(since1, sizeof(since1), timestamp); + s2 = format_timestamp(since2, sizeof(since2), timestamp); + + if (s1) + printf(" since %s; %s\n", s2, s1); + else if (s2) + printf(" since %s\n", s2); + else + printf("\n"); + + if (!i->condition_result && i->condition_timestamp > 0) { + UnitCondition *c; + int n = 0; + + s1 = format_timestamp_relative(since1, sizeof(since1), i->condition_timestamp); + s2 = format_timestamp(since2, sizeof(since2), i->condition_timestamp); + + printf("Condition: start %scondition failed%s at %s%s%s\n", + ansi_highlight_yellow(), ansi_normal(), + s2, s1 ? "; " : "", strempty(s1)); + + LIST_FOREACH(conditions, c, i->conditions) + if (c->tristate < 0) + n++; + + LIST_FOREACH(conditions, c, i->conditions) + if (c->tristate < 0) + printf(" %s %s=%s%s%s was not met\n", + --n ? special_glyph(TREE_BRANCH) : special_glyph(TREE_RIGHT), + c->name, + c->trigger ? "|" : "", + c->negate ? "!" : "", + c->param); + } + + if (!i->assert_result && i->assert_timestamp > 0) { + s1 = format_timestamp_relative(since1, sizeof(since1), i->assert_timestamp); + s2 = format_timestamp(since2, sizeof(since2), i->assert_timestamp); + + printf(" Assert: start %sassertion failed%s at %s%s%s\n", + ansi_highlight_red(), ansi_normal(), + s2, s1 ? "; " : "", strempty(s1)); + if (i->failed_assert_trigger) + printf(" none of the trigger assertions were met\n"); + else if (i->failed_assert) + printf(" %s=%s%s was not met\n", + i->failed_assert, + i->failed_assert_negate ? "!" : "", + i->failed_assert_parameter); + } + + if (i->sysfs_path) + printf(" Device: %s\n", i->sysfs_path); + if (i->where) + printf(" Where: %s\n", i->where); + if (i->what) + printf(" What: %s\n", i->what); + + STRV_FOREACH(t, i->documentation) + printf(" %*s %s\n", 9, t == i->documentation ? "Docs:" : "", *t); + + STRV_FOREACH_PAIR(t, t2, i->listen) + printf(" %*s %s (%s)\n", 9, t == i->listen ? "Listen:" : "", *t2, *t); + + if (i->accept) + printf(" Accepted: %u; Connected: %u\n", i->n_accepted, i->n_connections); + + LIST_FOREACH(exec, p, i->exec) { + _cleanup_free_ char *argv = NULL; + bool good; + + /* Only show exited processes here */ + if (p->code == 0) + continue; + + argv = strv_join(p->argv, " "); + printf(" Process: "PID_FMT" %s=%s ", p->pid, p->name, strna(argv)); + + good = is_clean_exit_lsb(p->code, p->status, NULL); + if (!good) { + on = ansi_highlight_red(); + off = ansi_normal(); + } else + on = off = ""; + + printf("%s(code=%s, ", on, sigchld_code_to_string(p->code)); + + if (p->code == CLD_EXITED) { + const char *c; + + printf("status=%i", p->status); + + c = exit_status_to_string(p->status, EXIT_STATUS_SYSTEMD); + if (c) + printf("/%s", c); + + } else + printf("signal=%s", signal_to_string(p->status)); + + printf(")%s\n", off); + + if (i->main_pid == p->pid && + i->start_timestamp == p->start_timestamp && + i->exit_timestamp == p->start_timestamp) + /* Let's not show this twice */ + i->main_pid = 0; + + if (p->pid == i->control_pid) + i->control_pid = 0; + } + + if (i->main_pid > 0 || i->control_pid > 0) { + if (i->main_pid > 0) { + printf(" Main PID: "PID_FMT, i->main_pid); + + if (i->running) { + _cleanup_free_ char *comm = NULL; + (void) get_process_comm(i->main_pid, &comm); + if (comm) + printf(" (%s)", comm); + } else if (i->exit_code > 0) { + printf(" (code=%s, ", sigchld_code_to_string(i->exit_code)); + + if (i->exit_code == CLD_EXITED) { + const char *c; + + printf("status=%i", i->exit_status); + + c = exit_status_to_string(i->exit_status, EXIT_STATUS_SYSTEMD); + if (c) + printf("/%s", c); + + } else + printf("signal=%s", signal_to_string(i->exit_status)); + printf(")"); + } + } + + if (i->control_pid > 0) { + _cleanup_free_ char *c = NULL; + + if (i->main_pid > 0) + fputs("; Control PID: ", stdout); + else + fputs("Cntrl PID: ", stdout); /* if first in column, abbreviated so it fits alignment */ + + printf(PID_FMT, i->control_pid); + + (void) get_process_comm(i->control_pid, &c); + if (c) + printf(" (%s)", c); + } + + printf("\n"); + } + + if (i->status_text) + printf(" Status: \"%s\"\n", i->status_text); + if (i->status_errno > 0) + printf(" Error: %i (%s)\n", i->status_errno, strerror(i->status_errno)); + + if (i->tasks_current != (uint64_t) -1) { + printf(" Tasks: %" PRIu64, i->tasks_current); + + if (i->tasks_max != (uint64_t) -1) + printf(" (limit: %" PRIu64 ")\n", i->tasks_max); + else + printf("\n"); + } + + if (i->memory_current != (uint64_t) -1) { + char buf[FORMAT_BYTES_MAX]; + + printf(" Memory: %s", format_bytes(buf, sizeof(buf), i->memory_current)); + + if (i->memory_low > 0 || i->memory_high != CGROUP_LIMIT_MAX || i->memory_max != CGROUP_LIMIT_MAX || + i->memory_limit != CGROUP_LIMIT_MAX) { + const char *prefix = ""; + + printf(" ("); + if (i->memory_low > 0) { + printf("%slow: %s", prefix, format_bytes(buf, sizeof(buf), i->memory_low)); + prefix = " "; + } + if (i->memory_high != CGROUP_LIMIT_MAX) { + printf("%shigh: %s", prefix, format_bytes(buf, sizeof(buf), i->memory_high)); + prefix = " "; + } + if (i->memory_max != CGROUP_LIMIT_MAX) { + printf("%smax: %s", prefix, format_bytes(buf, sizeof(buf), i->memory_max)); + prefix = " "; + } + if (i->memory_limit != CGROUP_LIMIT_MAX) { + printf("%slimit: %s", prefix, format_bytes(buf, sizeof(buf), i->memory_limit)); + prefix = " "; + } + printf(")"); + } + printf("\n"); + } + + if (i->cpu_usage_nsec != (uint64_t) -1) { + char buf[FORMAT_TIMESPAN_MAX]; + printf(" CPU: %s\n", format_timespan(buf, sizeof(buf), i->cpu_usage_nsec / NSEC_PER_USEC, USEC_PER_MSEC)); + } + + if (i->control_group) { + _cleanup_(sd_bus_error_free) sd_bus_error error = SD_BUS_ERROR_NULL; + static const char prefix[] = " "; + unsigned c; + + printf(" CGroup: %s\n", i->control_group); + + c = columns(); + if (c > sizeof(prefix) - 1) + c -= sizeof(prefix) - 1; + else + c = 0; + + r = unit_show_processes(bus, i->id, i->control_group, prefix, c, get_output_flags(), &error); + if (r == -EBADR) { + unsigned k = 0; + pid_t extra[2]; + + /* Fallback for older systemd versions where the GetUnitProcesses() call is not yet available */ + + if (i->main_pid > 0) + extra[k++] = i->main_pid; + + if (i->control_pid > 0) + extra[k++] = i->control_pid; + + show_cgroup_and_extra(SYSTEMD_CGROUP_CONTROLLER, i->control_group, prefix, c, extra, k, get_output_flags()); + } else if (r < 0) + log_warning_errno(r, "Failed to dump process list, ignoring: %s", bus_error_message(&error, r)); + } + + if (i->id && arg_transport == BUS_TRANSPORT_LOCAL) + show_journal_by_unit( + stdout, + i->id, + arg_output, + 0, + i->inactive_exit_timestamp_monotonic, + arg_lines, + getuid(), + get_output_flags() | OUTPUT_BEGIN_NEWLINE, + SD_JOURNAL_LOCAL_ONLY, + arg_scope == UNIT_FILE_SYSTEM, + ellipsized); + + if (i->need_daemon_reload) + warn_unit_file_changed(i->id); +} + +static void show_unit_help(UnitStatusInfo *i) { + char **p; + + assert(i); + + if (!i->documentation) { + log_info("Documentation for %s not known.", i->id); + return; + } + + STRV_FOREACH(p, i->documentation) + if (startswith(*p, "man:")) + show_man_page(*p + 4, false); + else + log_info("Can't show: %s", *p); +} + +static int status_property(const char *name, sd_bus_message *m, UnitStatusInfo *i, const char *contents) { + int r; + + assert(name); + assert(m); + assert(i); + + switch (contents[0]) { + + case SD_BUS_TYPE_STRING: { + const char *s; + + r = sd_bus_message_read(m, "s", &s); + if (r < 0) + return bus_log_parse_error(r); + + if (!isempty(s)) { + if (streq(name, "Id")) + i->id = s; + else if (streq(name, "LoadState")) + i->load_state = s; + else if (streq(name, "ActiveState")) + i->active_state = s; + else if (streq(name, "SubState")) + i->sub_state = s; + else if (streq(name, "Description")) + i->description = s; + else if (streq(name, "FragmentPath")) + i->fragment_path = s; + else if (streq(name, "SourcePath")) + i->source_path = s; +#ifndef NOLEGACY + else if (streq(name, "DefaultControlGroup")) { + const char *e; + e = startswith(s, SYSTEMD_CGROUP_CONTROLLER ":"); + if (e) + i->control_group = e; + } +#endif + else if (streq(name, "ControlGroup")) + i->control_group = s; + else if (streq(name, "StatusText")) + i->status_text = s; + else if (streq(name, "PIDFile")) + i->pid_file = s; + else if (streq(name, "SysFSPath")) + i->sysfs_path = s; + else if (streq(name, "Where")) + i->where = s; + else if (streq(name, "What")) + i->what = s; + else if (streq(name, "Following")) + i->following = s; + else if (streq(name, "UnitFileState")) + i->unit_file_state = s; + else if (streq(name, "UnitFilePreset")) + i->unit_file_preset = s; + else if (streq(name, "Result")) + i->result = s; + } + + break; + } + + case SD_BUS_TYPE_BOOLEAN: { + int b; + + r = sd_bus_message_read(m, "b", &b); + if (r < 0) + return bus_log_parse_error(r); + + if (streq(name, "Accept")) + i->accept = b; + else if (streq(name, "NeedDaemonReload")) + i->need_daemon_reload = b; + else if (streq(name, "ConditionResult")) + i->condition_result = b; + else if (streq(name, "AssertResult")) + i->assert_result = b; + else if (streq(name, "Transient")) + i->transient = b; + + break; + } + + case SD_BUS_TYPE_UINT32: { + uint32_t u; + + r = sd_bus_message_read(m, "u", &u); + if (r < 0) + return bus_log_parse_error(r); + + if (streq(name, "MainPID")) { + if (u > 0) { + i->main_pid = (pid_t) u; + i->running = true; + } + } else if (streq(name, "ControlPID")) + i->control_pid = (pid_t) u; + else if (streq(name, "ExecMainPID")) { + if (u > 0) + i->main_pid = (pid_t) u; + } else if (streq(name, "NAccepted")) + i->n_accepted = u; + else if (streq(name, "NConnections")) + i->n_connections = u; + + break; + } + + case SD_BUS_TYPE_INT32: { + int32_t j; + + r = sd_bus_message_read(m, "i", &j); + if (r < 0) + return bus_log_parse_error(r); + + if (streq(name, "ExecMainCode")) + i->exit_code = (int) j; + else if (streq(name, "ExecMainStatus")) + i->exit_status = (int) j; + else if (streq(name, "StatusErrno")) + i->status_errno = (int) j; + + break; + } + + case SD_BUS_TYPE_UINT64: { + uint64_t u; + + r = sd_bus_message_read(m, "t", &u); + if (r < 0) + return bus_log_parse_error(r); + + if (streq(name, "ExecMainStartTimestamp")) + i->start_timestamp = (usec_t) u; + else if (streq(name, "ExecMainExitTimestamp")) + i->exit_timestamp = (usec_t) u; + else if (streq(name, "ActiveEnterTimestamp")) + i->active_enter_timestamp = (usec_t) u; + else if (streq(name, "InactiveEnterTimestamp")) + i->inactive_enter_timestamp = (usec_t) u; + else if (streq(name, "InactiveExitTimestamp")) + i->inactive_exit_timestamp = (usec_t) u; + else if (streq(name, "InactiveExitTimestampMonotonic")) + i->inactive_exit_timestamp_monotonic = (usec_t) u; + else if (streq(name, "ActiveExitTimestamp")) + i->active_exit_timestamp = (usec_t) u; + else if (streq(name, "ConditionTimestamp")) + i->condition_timestamp = (usec_t) u; + else if (streq(name, "AssertTimestamp")) + i->assert_timestamp = (usec_t) u; + else if (streq(name, "MemoryCurrent")) + i->memory_current = u; + else if (streq(name, "MemoryLow")) + i->memory_low = u; + else if (streq(name, "MemoryHigh")) + i->memory_high = u; + else if (streq(name, "MemoryMax")) + i->memory_max = u; + else if (streq(name, "MemoryLimit")) + i->memory_limit = u; + else if (streq(name, "TasksCurrent")) + i->tasks_current = u; + else if (streq(name, "TasksMax")) + i->tasks_max = u; + else if (streq(name, "CPUUsageNSec")) + i->cpu_usage_nsec = u; + + break; + } + + case SD_BUS_TYPE_ARRAY: + + if (contents[1] == SD_BUS_TYPE_STRUCT_BEGIN && startswith(name, "Exec")) { + _cleanup_free_ ExecStatusInfo *info = NULL; + + r = sd_bus_message_enter_container(m, SD_BUS_TYPE_ARRAY, "(sasbttttuii)"); + if (r < 0) + return bus_log_parse_error(r); + + info = new0(ExecStatusInfo, 1); + if (!info) + return log_oom(); + + while ((r = exec_status_info_deserialize(m, info)) > 0) { + + info->name = strdup(name); + if (!info->name) + return log_oom(); + + LIST_PREPEND(exec, i->exec, info); + + info = new0(ExecStatusInfo, 1); + if (!info) + return log_oom(); + } + + if (r < 0) + return bus_log_parse_error(r); + + r = sd_bus_message_exit_container(m); + if (r < 0) + return bus_log_parse_error(r); + + return 0; + + } else if (contents[1] == SD_BUS_TYPE_STRUCT_BEGIN && streq(name, "Listen")) { + const char *type, *path; + + r = sd_bus_message_enter_container(m, SD_BUS_TYPE_ARRAY, "(ss)"); + if (r < 0) + return bus_log_parse_error(r); + + while ((r = sd_bus_message_read(m, "(ss)", &type, &path)) > 0) { + + r = strv_extend(&i->listen, type); + if (r < 0) + return r; + + r = strv_extend(&i->listen, path); + if (r < 0) + return r; + } + if (r < 0) + return bus_log_parse_error(r); + + r = sd_bus_message_exit_container(m); + if (r < 0) + return bus_log_parse_error(r); + + return 0; + + } else if (contents[1] == SD_BUS_TYPE_STRING && streq(name, "DropInPaths")) { + + r = sd_bus_message_read_strv(m, &i->dropin_paths); + if (r < 0) + return bus_log_parse_error(r); + + } else if (contents[1] == SD_BUS_TYPE_STRING && streq(name, "Documentation")) { + + r = sd_bus_message_read_strv(m, &i->documentation); + if (r < 0) + return bus_log_parse_error(r); + + } else if (contents[1] == SD_BUS_TYPE_STRUCT_BEGIN && streq(name, "Conditions")) { + const char *cond, *param; + int trigger, negate; + int32_t state; + + r = sd_bus_message_enter_container(m, SD_BUS_TYPE_ARRAY, "(sbbsi)"); + if (r < 0) + return bus_log_parse_error(r); + + while ((r = sd_bus_message_read(m, "(sbbsi)", &cond, &trigger, &negate, ¶m, &state)) > 0) { + _cleanup_(unit_condition_freep) UnitCondition *c = NULL; + + log_debug("%s trigger=%d negate=%d %s →%d", cond, trigger, negate, param, state); + + c = new0(UnitCondition, 1); + if (!c) + return log_oom(); + + c->name = strdup(cond); + c->param = strdup(param); + if (!c->name || !c->param) + return log_oom(); + + c->trigger = trigger; + c->negate = negate; + c->tristate = state; + + LIST_PREPEND(conditions, i->conditions, c); + c = NULL; + } + if (r < 0) + return bus_log_parse_error(r); + + r = sd_bus_message_exit_container(m); + if (r < 0) + return bus_log_parse_error(r); + + } else if (contents[1] == SD_BUS_TYPE_STRUCT_BEGIN && streq(name, "Asserts")) { + const char *cond, *param; + int trigger, negate; + int32_t state; + + r = sd_bus_message_enter_container(m, SD_BUS_TYPE_ARRAY, "(sbbsi)"); + if (r < 0) + return bus_log_parse_error(r); + + while ((r = sd_bus_message_read(m, "(sbbsi)", &cond, &trigger, &negate, ¶m, &state)) > 0) { + log_debug("%s %d %d %s %d", cond, trigger, negate, param, state); + if (state < 0 && (!trigger || !i->failed_assert)) { + i->failed_assert = cond; + i->failed_assert_trigger = trigger; + i->failed_assert_negate = negate; + i->failed_assert_parameter = param; + } + } + if (r < 0) + return bus_log_parse_error(r); + + r = sd_bus_message_exit_container(m); + if (r < 0) + return bus_log_parse_error(r); + + } else + goto skip; + + break; + + case SD_BUS_TYPE_STRUCT_BEGIN: + + if (streq(name, "LoadError")) { + const char *n, *message; + + r = sd_bus_message_read(m, "(ss)", &n, &message); + if (r < 0) + return bus_log_parse_error(r); + + if (!isempty(message)) + i->load_error = message; + } else + goto skip; + + break; + + default: + goto skip; + } + + return 0; + +skip: + r = sd_bus_message_skip(m, contents); + if (r < 0) + return bus_log_parse_error(r); + + return 0; +} + +#define print_prop(name, fmt, ...) \ + do { \ + if (arg_value) \ + printf(fmt "\n", __VA_ARGS__); \ + else \ + printf("%s=" fmt "\n", name, __VA_ARGS__); \ + } while(0) + +static int print_property(const char *name, sd_bus_message *m, const char *contents) { + int r; + + assert(name); + assert(m); + + /* This is a low-level property printer, see + * print_status_info() for the nicer output */ + + if (arg_properties && !strv_find(arg_properties, name)) { + /* skip what we didn't read */ + r = sd_bus_message_skip(m, contents); + return r; + } + + switch (contents[0]) { + + case SD_BUS_TYPE_STRUCT_BEGIN: + + if (contents[1] == SD_BUS_TYPE_UINT32 && streq(name, "Job")) { + uint32_t u; + + r = sd_bus_message_read(m, "(uo)", &u, NULL); + if (r < 0) + return bus_log_parse_error(r); + + if (u > 0) + print_prop(name, "%"PRIu32, u); + else if (arg_all) + print_prop(name, "%s", ""); + + return 0; + + } else if (contents[1] == SD_BUS_TYPE_STRING && streq(name, "Unit")) { + const char *s; + + r = sd_bus_message_read(m, "(so)", &s, NULL); + if (r < 0) + return bus_log_parse_error(r); + + if (arg_all || !isempty(s)) + print_prop(name, "%s", s); + + return 0; + + } else if (contents[1] == SD_BUS_TYPE_STRING && streq(name, "LoadError")) { + const char *a = NULL, *b = NULL; + + r = sd_bus_message_read(m, "(ss)", &a, &b); + if (r < 0) + return bus_log_parse_error(r); + + if (arg_all || !isempty(a) || !isempty(b)) + print_prop(name, "%s \"%s\"", strempty(a), strempty(b)); + + return 0; + } else if (streq_ptr(name, "SystemCallFilter")) { + _cleanup_strv_free_ char **l = NULL; + int whitelist; + + r = sd_bus_message_enter_container(m, 'r', "bas"); + if (r < 0) + return bus_log_parse_error(r); + + r = sd_bus_message_read(m, "b", &whitelist); + if (r < 0) + return bus_log_parse_error(r); + + r = sd_bus_message_read_strv(m, &l); + if (r < 0) + return bus_log_parse_error(r); + + r = sd_bus_message_exit_container(m); + if (r < 0) + return bus_log_parse_error(r); + + if (arg_all || whitelist || !strv_isempty(l)) { + bool first = true; + char **i; + + if (!arg_value) { + fputs(name, stdout); + fputc('=', stdout); + } + + if (!whitelist) + fputc('~', stdout); + + STRV_FOREACH(i, l) { + if (first) + first = false; + else + fputc(' ', stdout); + + fputs(*i, stdout); + } + fputc('\n', stdout); + } + + return 0; + } + + break; + + case SD_BUS_TYPE_ARRAY: + + if (contents[1] == SD_BUS_TYPE_STRUCT_BEGIN && streq(name, "EnvironmentFiles")) { + const char *path; + int ignore; + + r = sd_bus_message_enter_container(m, SD_BUS_TYPE_ARRAY, "(sb)"); + if (r < 0) + return bus_log_parse_error(r); + + while ((r = sd_bus_message_read(m, "(sb)", &path, &ignore)) > 0) + print_prop("EnvironmentFile", "%s (ignore_errors=%s)", path, yes_no(ignore)); + + if (r < 0) + return bus_log_parse_error(r); + + r = sd_bus_message_exit_container(m); + if (r < 0) + return bus_log_parse_error(r); + + return 0; + + } else if (contents[1] == SD_BUS_TYPE_STRUCT_BEGIN && streq(name, "Paths")) { + const char *type, *path; + + r = sd_bus_message_enter_container(m, SD_BUS_TYPE_ARRAY, "(ss)"); + if (r < 0) + return bus_log_parse_error(r); + + while ((r = sd_bus_message_read(m, "(ss)", &type, &path)) > 0) + print_prop(type, "%s", path); + if (r < 0) + return bus_log_parse_error(r); + + r = sd_bus_message_exit_container(m); + if (r < 0) + return bus_log_parse_error(r); + + return 0; + + } else if (contents[1] == SD_BUS_TYPE_STRUCT_BEGIN && streq(name, "Listen")) { + const char *type, *path; + + r = sd_bus_message_enter_container(m, SD_BUS_TYPE_ARRAY, "(ss)"); + if (r < 0) + return bus_log_parse_error(r); + + while ((r = sd_bus_message_read(m, "(ss)", &type, &path)) > 0) + if (arg_value) + puts(path); + else + printf("Listen%s=%s\n", type, path); + if (r < 0) + return bus_log_parse_error(r); + + r = sd_bus_message_exit_container(m); + if (r < 0) + return bus_log_parse_error(r); + + return 0; + + } else if (contents[1] == SD_BUS_TYPE_STRUCT_BEGIN && streq(name, "Timers")) { + const char *base; + uint64_t value, next_elapse; + + r = sd_bus_message_enter_container(m, SD_BUS_TYPE_ARRAY, "(stt)"); + if (r < 0) + return bus_log_parse_error(r); + + while ((r = sd_bus_message_read(m, "(stt)", &base, &value, &next_elapse)) > 0) { + char timespan1[FORMAT_TIMESPAN_MAX], timespan2[FORMAT_TIMESPAN_MAX]; + + print_prop(base, "{ value=%s ; next_elapse=%s }", + format_timespan(timespan1, sizeof(timespan1), value, 0), + format_timespan(timespan2, sizeof(timespan2), next_elapse, 0)); + } + if (r < 0) + return bus_log_parse_error(r); + + r = sd_bus_message_exit_container(m); + if (r < 0) + return bus_log_parse_error(r); + + return 0; + + } else if (contents[1] == SD_BUS_TYPE_STRUCT_BEGIN && startswith(name, "Exec")) { + ExecStatusInfo info = {}; + + r = sd_bus_message_enter_container(m, SD_BUS_TYPE_ARRAY, "(sasbttttuii)"); + if (r < 0) + return bus_log_parse_error(r); + + while ((r = exec_status_info_deserialize(m, &info)) > 0) { + char timestamp1[FORMAT_TIMESTAMP_MAX], timestamp2[FORMAT_TIMESTAMP_MAX]; + _cleanup_free_ char *tt; + + tt = strv_join(info.argv, " "); + + print_prop(name, + "{ path=%s ; argv[]=%s ; ignore_errors=%s ; start_time=[%s] ; stop_time=[%s] ; pid="PID_FMT" ; code=%s ; status=%i%s%s }", + strna(info.path), + strna(tt), + yes_no(info.ignore), + strna(format_timestamp(timestamp1, sizeof(timestamp1), info.start_timestamp)), + strna(format_timestamp(timestamp2, sizeof(timestamp2), info.exit_timestamp)), + info.pid, + sigchld_code_to_string(info.code), + info.status, + info.code == CLD_EXITED ? "" : "/", + strempty(info.code == CLD_EXITED ? NULL : signal_to_string(info.status))); + + free(info.path); + strv_free(info.argv); + zero(info); + } + + r = sd_bus_message_exit_container(m); + if (r < 0) + return bus_log_parse_error(r); + + return 0; + + } else if (contents[1] == SD_BUS_TYPE_STRUCT_BEGIN && streq(name, "DeviceAllow")) { + const char *path, *rwm; + + r = sd_bus_message_enter_container(m, SD_BUS_TYPE_ARRAY, "(ss)"); + if (r < 0) + return bus_log_parse_error(r); + + while ((r = sd_bus_message_read(m, "(ss)", &path, &rwm)) > 0) + print_prop(name, "%s %s", strna(path), strna(rwm)); + if (r < 0) + return bus_log_parse_error(r); + + r = sd_bus_message_exit_container(m); + if (r < 0) + return bus_log_parse_error(r); + + return 0; + + } else if (contents[1] == SD_BUS_TYPE_STRUCT_BEGIN && (streq(name, "IODeviceWeight") || streq(name, "BlockIODeviceWeight"))) { + const char *path; + uint64_t weight; + + r = sd_bus_message_enter_container(m, SD_BUS_TYPE_ARRAY, "(st)"); + if (r < 0) + return bus_log_parse_error(r); + + while ((r = sd_bus_message_read(m, "(st)", &path, &weight)) > 0) + print_prop(name, "%s %"PRIu64, strna(path), weight); + if (r < 0) + return bus_log_parse_error(r); + + r = sd_bus_message_exit_container(m); + if (r < 0) + return bus_log_parse_error(r); + + return 0; + + } else if (contents[1] == SD_BUS_TYPE_STRUCT_BEGIN && (cgroup_io_limit_type_from_string(name) >= 0 || + streq(name, "BlockIOReadBandwidth") || streq(name, "BlockIOWriteBandwidth"))) { + const char *path; + uint64_t bandwidth; + + r = sd_bus_message_enter_container(m, SD_BUS_TYPE_ARRAY, "(st)"); + if (r < 0) + return bus_log_parse_error(r); + + while ((r = sd_bus_message_read(m, "(st)", &path, &bandwidth)) > 0) + print_prop(name, "%s %"PRIu64, strna(path), bandwidth); + if (r < 0) + return bus_log_parse_error(r); + + r = sd_bus_message_exit_container(m); + if (r < 0) + return bus_log_parse_error(r); + + return 0; + } + + break; + } + + r = bus_print_property(name, m, arg_value, arg_all); + if (r < 0) + return bus_log_parse_error(r); + + if (r == 0) { + r = sd_bus_message_skip(m, contents); + if (r < 0) + return bus_log_parse_error(r); + + if (arg_all) + printf("%s=[unprintable]\n", name); + } + + return 0; +} + +static int show_one( + const char *verb, + sd_bus *bus, + const char *path, + const char *unit, + bool show_properties, + bool *new_line, + bool *ellipsized) { + + static const struct bus_properties_map property_map[] = { + { "LoadState", "s", map_string_no_copy, offsetof(UnitStatusInfo, load_state) }, + { "ActiveState", "s", map_string_no_copy, offsetof(UnitStatusInfo, active_state) }, + {} + }; + + _cleanup_(sd_bus_message_unrefp) sd_bus_message *reply = NULL; + _cleanup_(sd_bus_error_free) sd_bus_error error = SD_BUS_ERROR_NULL; + _cleanup_set_free_ Set *found_properties = NULL; + _cleanup_(unit_status_info_free) UnitStatusInfo info = { + .memory_current = (uint64_t) -1, + .memory_high = CGROUP_LIMIT_MAX, + .memory_max = CGROUP_LIMIT_MAX, + .memory_limit = (uint64_t) -1, + .cpu_usage_nsec = (uint64_t) -1, + .tasks_current = (uint64_t) -1, + .tasks_max = (uint64_t) -1, + }; + int r; + + assert(path); + assert(new_line); + + log_debug("Showing one %s", path); + + r = sd_bus_call_method( + bus, + "org.freedesktop.systemd1", + path, + "org.freedesktop.DBus.Properties", + "GetAll", + &error, + &reply, + "s", ""); + if (r < 0) + return log_error_errno(r, "Failed to get properties: %s", bus_error_message(&error, r)); + + if (unit) { + r = bus_message_map_all_properties(reply, property_map, &info); + if (r < 0) + return log_error_errno(r, "Failed to map properties: %s", bus_error_message(&error, r)); + + if (streq_ptr(info.load_state, "not-found") && streq_ptr(info.active_state, "inactive")) { + log_full(streq(verb, "status") ? LOG_ERR : LOG_DEBUG, + "Unit %s could not be found.", unit); + + if (streq(verb, "status")) + return EXIT_PROGRAM_OR_SERVICES_STATUS_UNKNOWN; + + if (!streq(verb, "show")) + return -ENOENT; + } + + r = sd_bus_message_rewind(reply, true); + if (r < 0) + return log_error_errno(r, "Failed to rewind: %s", bus_error_message(&error, r)); + } + + r = sd_bus_message_enter_container(reply, SD_BUS_TYPE_ARRAY, "{sv}"); + if (r < 0) + return bus_log_parse_error(r); + + if (*new_line) + printf("\n"); + + *new_line = true; + + while ((r = sd_bus_message_enter_container(reply, SD_BUS_TYPE_DICT_ENTRY, "sv")) > 0) { + const char *name, *contents; + + r = sd_bus_message_read(reply, "s", &name); + if (r < 0) + return bus_log_parse_error(r); + + r = sd_bus_message_peek_type(reply, NULL, &contents); + if (r < 0) + return bus_log_parse_error(r); + + r = sd_bus_message_enter_container(reply, SD_BUS_TYPE_VARIANT, contents); + if (r < 0) + return bus_log_parse_error(r); + + if (show_properties) { + r = set_ensure_allocated(&found_properties, &string_hash_ops); + if (r < 0) + return log_oom(); + + r = set_put(found_properties, name); + if (r < 0 && r != EEXIST) + return log_oom(); + + r = print_property(name, reply, contents); + } else + r = status_property(name, reply, &info, contents); + if (r < 0) + return r; + + r = sd_bus_message_exit_container(reply); + if (r < 0) + return bus_log_parse_error(r); + + r = sd_bus_message_exit_container(reply); + if (r < 0) + return bus_log_parse_error(r); + } + if (r < 0) + return bus_log_parse_error(r); + + r = sd_bus_message_exit_container(reply); + if (r < 0) + return bus_log_parse_error(r); + + r = 0; + if (show_properties) { + char **pp; + int not_found_level = streq(verb, "show") ? LOG_DEBUG : LOG_WARNING; + + STRV_FOREACH(pp, arg_properties) + if (!set_contains(found_properties, *pp)) { + log_full(not_found_level, "Property %s does not exist.", *pp); + r = -ENXIO; + } + + } else if (streq(verb, "help")) + show_unit_help(&info); + else if (streq(verb, "status")) { + print_status_info(bus, &info, ellipsized); + + if (info.active_state && STR_IN_SET(info.active_state, "inactive", "failed")) + r = EXIT_PROGRAM_NOT_RUNNING; + else + r = EXIT_PROGRAM_RUNNING_OR_SERVICE_OK; + } + + return r; +} + +static int get_unit_dbus_path_by_pid( + sd_bus *bus, + uint32_t pid, + char **unit) { + + _cleanup_(sd_bus_error_free) sd_bus_error error = SD_BUS_ERROR_NULL; + _cleanup_(sd_bus_message_unrefp) sd_bus_message *reply = NULL; + char *u; + int r; + + r = sd_bus_call_method( + bus, + "org.freedesktop.systemd1", + "/org/freedesktop/systemd1", + "org.freedesktop.systemd1.Manager", + "GetUnitByPID", + &error, + &reply, + "u", pid); + if (r < 0) + return log_error_errno(r, "Failed to get unit for PID %"PRIu32": %s", pid, bus_error_message(&error, r)); + + r = sd_bus_message_read(reply, "o", &u); + if (r < 0) + return bus_log_parse_error(r); + + u = strdup(u); + if (!u) + return log_oom(); + + *unit = u; + return 0; +} + +static int show_all( + const char* verb, + sd_bus *bus, + bool show_properties, + bool *new_line, + bool *ellipsized) { + + _cleanup_(sd_bus_message_unrefp) sd_bus_message *reply = NULL; + _cleanup_free_ UnitInfo *unit_infos = NULL; + const UnitInfo *u; + unsigned c; + int r, ret = 0; + + r = get_unit_list(bus, NULL, NULL, &unit_infos, 0, &reply); + if (r < 0) + return r; + + pager_open(arg_no_pager, false); + + c = (unsigned) r; + + qsort_safe(unit_infos, c, sizeof(UnitInfo), compare_unit_info); + + for (u = unit_infos; u < unit_infos + c; u++) { + _cleanup_free_ char *p = NULL; + + p = unit_dbus_path_from_name(u->id); + if (!p) + return log_oom(); + + r = show_one(verb, bus, p, u->id, show_properties, new_line, ellipsized); + if (r < 0) + return r; + else if (r > 0 && ret == 0) + ret = r; + } + + return ret; +} + +static int show_system_status(sd_bus *bus) { + char since1[FORMAT_TIMESTAMP_RELATIVE_MAX], since2[FORMAT_TIMESTAMP_MAX]; + _cleanup_free_ char *hn = NULL; + _cleanup_(machine_info_clear) struct machine_info mi = {}; + const char *on, *off; + int r; + + hn = gethostname_malloc(); + if (!hn) + return log_oom(); + + r = bus_map_all_properties(bus, "org.freedesktop.systemd1", "/org/freedesktop/systemd1", machine_info_property_map, &mi); + if (r < 0) + return log_error_errno(r, "Failed to read server status: %m"); + + if (streq_ptr(mi.state, "degraded")) { + on = ansi_highlight_red(); + off = ansi_normal(); + } else if (!streq_ptr(mi.state, "running")) { + on = ansi_highlight_yellow(); + off = ansi_normal(); + } else + on = off = ""; + + printf("%s%s%s %s\n", on, special_glyph(BLACK_CIRCLE), off, arg_host ? arg_host : hn); + + printf(" State: %s%s%s\n", + on, strna(mi.state), off); + + printf(" Jobs: %" PRIu32 " queued\n", mi.n_jobs); + printf(" Failed: %" PRIu32 " units\n", mi.n_failed_units); + + printf(" Since: %s; %s\n", + format_timestamp(since2, sizeof(since2), mi.timestamp), + format_timestamp_relative(since1, sizeof(since1), mi.timestamp)); + + printf(" CGroup: %s\n", mi.control_group ?: "/"); + if (IN_SET(arg_transport, + BUS_TRANSPORT_LOCAL, + BUS_TRANSPORT_MACHINE)) { + static const char prefix[] = " "; + unsigned c; + + c = columns(); + if (c > sizeof(prefix) - 1) + c -= sizeof(prefix) - 1; + else + c = 0; + + show_cgroup(SYSTEMD_CGROUP_CONTROLLER, strempty(mi.control_group), prefix, c, get_output_flags()); + } + + return 0; +} + +static int show(int argc, char *argv[], void *userdata) { + bool show_properties, show_status, show_help, new_line = false; + bool ellipsized = false; + int r, ret = 0; + sd_bus *bus; + + assert(argv); + + show_properties = streq(argv[0], "show"); + show_status = streq(argv[0], "status"); + show_help = streq(argv[0], "help"); + + if (show_help && argc <= 1) { + log_error("This command expects one or more unit names. Did you mean --help?"); + return -EINVAL; + } + + r = acquire_bus(BUS_MANAGER, &bus); + if (r < 0) + return r; + + pager_open(arg_no_pager, false); + + if (show_status) + /* Increase max number of open files to 16K if we can, we + * might needs this when browsing journal files, which might + * be split up into many files. */ + setrlimit_closest(RLIMIT_NOFILE, &RLIMIT_MAKE_CONST(16384)); + + /* If no argument is specified inspect the manager itself */ + if (show_properties && argc <= 1) + return show_one(argv[0], bus, "/org/freedesktop/systemd1", NULL, show_properties, &new_line, &ellipsized); + + if (show_status && argc <= 1) { + + show_system_status(bus); + new_line = true; + + if (arg_all) + ret = show_all(argv[0], bus, false, &new_line, &ellipsized); + } else { + _cleanup_free_ char **patterns = NULL; + char **name; + + STRV_FOREACH(name, strv_skip(argv, 1)) { + _cleanup_free_ char *path = NULL, *unit = NULL; + uint32_t id; + + if (safe_atou32(*name, &id) < 0) { + if (strv_push(&patterns, *name) < 0) + return log_oom(); + + continue; + } else if (show_properties) { + /* Interpret as job id */ + if (asprintf(&path, "/org/freedesktop/systemd1/job/%u", id) < 0) + return log_oom(); + + } else { + /* Interpret as PID */ + r = get_unit_dbus_path_by_pid(bus, id, &path); + if (r < 0) { + ret = r; + continue; + } + + r = unit_name_from_dbus_path(path, &unit); + if (r < 0) + return log_oom(); + } + + r = show_one(argv[0], bus, path, unit, show_properties, &new_line, &ellipsized); + if (r < 0) + return r; + else if (r > 0 && ret == 0) + ret = r; + } + + if (!strv_isempty(patterns)) { + _cleanup_strv_free_ char **names = NULL; + + r = expand_names(bus, patterns, NULL, &names); + if (r < 0) + return log_error_errno(r, "Failed to expand names: %m"); + + STRV_FOREACH(name, names) { + _cleanup_free_ char *path; + + path = unit_dbus_path_from_name(*name); + if (!path) + return log_oom(); + + r = show_one(argv[0], bus, path, *name, show_properties, &new_line, &ellipsized); + if (r < 0) + return r; + if (r > 0 && ret == 0) + ret = r; + } + } + } + + if (ellipsized && !arg_quiet) + printf("Hint: Some lines were ellipsized, use -l to show in full.\n"); + + return ret; +} + +static int cat_file(const char *filename, bool newline) { + _cleanup_close_ int fd; + + fd = open(filename, O_RDONLY|O_CLOEXEC|O_NOCTTY); + if (fd < 0) + return -errno; + + printf("%s%s# %s%s\n", + newline ? "\n" : "", + ansi_highlight_blue(), + filename, + ansi_normal()); + fflush(stdout); + + return copy_bytes(fd, STDOUT_FILENO, (uint64_t) -1, false); +} + +static int cat(int argc, char *argv[], void *userdata) { + _cleanup_lookup_paths_free_ LookupPaths lp = {}; + _cleanup_strv_free_ char **names = NULL; + char **name; + sd_bus *bus; + bool first = true; + int r; + + if (arg_transport != BUS_TRANSPORT_LOCAL) { + log_error("Cannot remotely cat units."); + return -EINVAL; + } + + r = lookup_paths_init(&lp, arg_scope, 0, arg_root); + if (r < 0) + return log_error_errno(r, "Failed to determine unit paths: %m"); + + r = acquire_bus(BUS_MANAGER, &bus); + if (r < 0) + return r; + + r = expand_names(bus, strv_skip(argv, 1), NULL, &names); + if (r < 0) + return log_error_errno(r, "Failed to expand names: %m"); + + pager_open(arg_no_pager, false); + + STRV_FOREACH(name, names) { + _cleanup_free_ char *fragment_path = NULL; + _cleanup_strv_free_ char **dropin_paths = NULL; + char **path; + + r = unit_find_paths(bus, *name, &lp, &fragment_path, &dropin_paths); + if (r < 0) + return r; + else if (r == 0) + return -ENOENT; + + if (first) + first = false; + else + puts(""); + + if (fragment_path) { + r = cat_file(fragment_path, false); + if (r < 0) + return log_warning_errno(r, "Failed to cat %s: %m", fragment_path); + } + + STRV_FOREACH(path, dropin_paths) { + r = cat_file(*path, path == dropin_paths); + if (r < 0) + return log_warning_errno(r, "Failed to cat %s: %m", *path); + } + } + + return 0; +} + +static int set_property(int argc, char *argv[], void *userdata) { + _cleanup_(sd_bus_message_unrefp) sd_bus_message *m = NULL; + _cleanup_(sd_bus_error_free) sd_bus_error error = SD_BUS_ERROR_NULL; + _cleanup_free_ char *n = NULL; + sd_bus *bus; + char **i; + int r; + + r = acquire_bus(BUS_MANAGER, &bus); + if (r < 0) + return r; + + polkit_agent_open_if_enabled(); + + r = sd_bus_message_new_method_call( + bus, + &m, + "org.freedesktop.systemd1", + "/org/freedesktop/systemd1", + "org.freedesktop.systemd1.Manager", + "SetUnitProperties"); + if (r < 0) + return bus_log_create_error(r); + + r = unit_name_mangle(argv[1], UNIT_NAME_NOGLOB, &n); + if (r < 0) + return log_error_errno(r, "Failed to mangle unit name: %m"); + + r = sd_bus_message_append(m, "sb", n, arg_runtime); + if (r < 0) + return bus_log_create_error(r); + + r = sd_bus_message_open_container(m, SD_BUS_TYPE_ARRAY, "(sv)"); + if (r < 0) + return bus_log_create_error(r); + + STRV_FOREACH(i, strv_skip(argv, 2)) { + r = bus_append_unit_property_assignment(m, *i); + if (r < 0) + return r; + } + + r = sd_bus_message_close_container(m); + if (r < 0) + return bus_log_create_error(r); + + r = sd_bus_call(bus, m, 0, &error, NULL); + if (r < 0) + return log_error_errno(r, "Failed to set unit properties on %s: %s", n, bus_error_message(&error, r)); + + return 0; +} + +static int daemon_reload(int argc, char *argv[], void *userdata) { + _cleanup_(sd_bus_error_free) sd_bus_error error = SD_BUS_ERROR_NULL; + _cleanup_(sd_bus_message_unrefp) sd_bus_message *m = NULL; + const char *method; + sd_bus *bus; + int r; + + r = acquire_bus(BUS_MANAGER, &bus); + if (r < 0) + return r; + + polkit_agent_open_if_enabled(); + + switch (arg_action) { + + case ACTION_RELOAD: + method = "Reload"; + break; + + case ACTION_REEXEC: + method = "Reexecute"; + break; + + case ACTION_SYSTEMCTL: + method = streq(argv[0], "daemon-reexec") ? "Reexecute" : + /* "daemon-reload" */ "Reload"; + break; + + default: + assert_not_reached("Unexpected action"); + } + + r = sd_bus_message_new_method_call( + bus, + &m, + "org.freedesktop.systemd1", + "/org/freedesktop/systemd1", + "org.freedesktop.systemd1.Manager", + method); + if (r < 0) + return bus_log_create_error(r); + + /* Note we use an extra-long timeout here. This is because a reload or reexec means generators are rerun which + * are timed out after DEFAULT_TIMEOUT_USEC. Let's use twice that time here, so that the generators can have + * their timeout, and for everything else there's the same time budget in place. */ + + r = sd_bus_call(bus, m, DEFAULT_TIMEOUT_USEC * 2, &error, NULL); + + /* On reexecution, we expect a disconnect, not a reply */ + if (IN_SET(r, -ETIMEDOUT, -ECONNRESET) && streq(method, "Reexecute")) + r = 0; + + if (r < 0 && arg_action == ACTION_SYSTEMCTL) + return log_error_errno(r, "Failed to reload daemon: %s", bus_error_message(&error, r)); + + /* Note that for the legacy commands (i.e. those with action != ACTION_SYSTEMCTL) we support fallbacks to the + * old ways of doing things, hence don't log any error in that case here. */ + + return r < 0 ? r : 0; +} + +static int trivial_method(int argc, char *argv[], void *userdata) { + _cleanup_(sd_bus_error_free) sd_bus_error error = SD_BUS_ERROR_NULL; + const char *method; + sd_bus *bus; + int r; + + r = acquire_bus(BUS_MANAGER, &bus); + if (r < 0) + return r; + + polkit_agent_open_if_enabled(); + + method = + streq(argv[0], "clear-jobs") || + streq(argv[0], "cancel") ? "ClearJobs" : + streq(argv[0], "reset-failed") ? "ResetFailed" : + streq(argv[0], "halt") ? "Halt" : + streq(argv[0], "reboot") ? "Reboot" : + streq(argv[0], "kexec") ? "KExec" : + streq(argv[0], "exit") ? "Exit" : + /* poweroff */ "PowerOff"; + + r = sd_bus_call_method( + bus, + "org.freedesktop.systemd1", + "/org/freedesktop/systemd1", + "org.freedesktop.systemd1.Manager", + method, + &error, + NULL, + NULL); + if (r < 0 && arg_action == ACTION_SYSTEMCTL) + return log_error_errno(r, "Failed to execute operation: %s", bus_error_message(&error, r)); + + /* Note that for the legacy commands (i.e. those with action != ACTION_SYSTEMCTL) we support fallbacks to the + * old ways of doing things, hence don't log any error in that case here. */ + + return r < 0 ? r : 0; +} + +static int reset_failed(int argc, char *argv[], void *userdata) { + _cleanup_strv_free_ char **names = NULL; + sd_bus *bus; + char **name; + int r, q; + + if (argc <= 1) + return trivial_method(argc, argv, userdata); + + r = acquire_bus(BUS_MANAGER, &bus); + if (r < 0) + return r; + + polkit_agent_open_if_enabled(); + + r = expand_names(bus, strv_skip(argv, 1), NULL, &names); + if (r < 0) + return log_error_errno(r, "Failed to expand names: %m"); + + STRV_FOREACH(name, names) { + _cleanup_(sd_bus_error_free) sd_bus_error error = SD_BUS_ERROR_NULL; + + q = sd_bus_call_method( + bus, + "org.freedesktop.systemd1", + "/org/freedesktop/systemd1", + "org.freedesktop.systemd1.Manager", + "ResetFailedUnit", + &error, + NULL, + "s", *name); + if (q < 0) { + log_error_errno(q, "Failed to reset failed state of unit %s: %s", *name, bus_error_message(&error, q)); + if (r == 0) + r = q; + } + } + + return r; +} + +static int show_environment(int argc, char *argv[], void *userdata) { + _cleanup_(sd_bus_error_free) sd_bus_error error = SD_BUS_ERROR_NULL; + _cleanup_(sd_bus_message_unrefp) sd_bus_message *reply = NULL; + const char *text; + sd_bus *bus; + int r; + + r = acquire_bus(BUS_MANAGER, &bus); + if (r < 0) + return r; + + pager_open(arg_no_pager, false); + + r = sd_bus_get_property( + bus, + "org.freedesktop.systemd1", + "/org/freedesktop/systemd1", + "org.freedesktop.systemd1.Manager", + "Environment", + &error, + &reply, + "as"); + if (r < 0) + return log_error_errno(r, "Failed to get environment: %s", bus_error_message(&error, r)); + + r = sd_bus_message_enter_container(reply, SD_BUS_TYPE_ARRAY, "s"); + if (r < 0) + return bus_log_parse_error(r); + + while ((r = sd_bus_message_read_basic(reply, SD_BUS_TYPE_STRING, &text)) > 0) + puts(text); + if (r < 0) + return bus_log_parse_error(r); + + r = sd_bus_message_exit_container(reply); + if (r < 0) + return bus_log_parse_error(r); + + return 0; +} + +static int switch_root(int argc, char *argv[], void *userdata) { + _cleanup_(sd_bus_error_free) sd_bus_error error = SD_BUS_ERROR_NULL; + _cleanup_free_ char *cmdline_init = NULL; + const char *root, *init; + sd_bus *bus; + int r; + + if (arg_transport != BUS_TRANSPORT_LOCAL) { + log_error("Cannot switch root remotely."); + return -EINVAL; + } + + if (argc < 2 || argc > 3) { + log_error("Wrong number of arguments."); + return -EINVAL; + } + + root = argv[1]; + + if (argc >= 3) + init = argv[2]; + else { + r = parse_env_file("/proc/cmdline", WHITESPACE, + "init", &cmdline_init, + NULL); + if (r < 0) + log_debug_errno(r, "Failed to parse /proc/cmdline: %m"); + + init = cmdline_init; + } + + init = empty_to_null(init); + if (init) { + const char *root_systemd_path = NULL, *root_init_path = NULL; + + root_systemd_path = strjoina(root, "/" SYSTEMD_BINARY_PATH); + root_init_path = strjoina(root, "/", init); + + /* If the passed init is actually the same as the + * systemd binary, then let's suppress it. */ + if (files_same(root_init_path, root_systemd_path) > 0) + init = NULL; + } + + r = acquire_bus(BUS_MANAGER, &bus); + if (r < 0) + return r; + + log_debug("Switching root - root: %s; init: %s", root, strna(init)); + + r = sd_bus_call_method( + bus, + "org.freedesktop.systemd1", + "/org/freedesktop/systemd1", + "org.freedesktop.systemd1.Manager", + "SwitchRoot", + &error, + NULL, + "ss", root, init); + if (r < 0) + return log_error_errno(r, "Failed to switch root: %s", bus_error_message(&error, r)); + + return 0; +} + +static int set_environment(int argc, char *argv[], void *userdata) { + _cleanup_(sd_bus_error_free) sd_bus_error error = SD_BUS_ERROR_NULL; + _cleanup_(sd_bus_message_unrefp) sd_bus_message *m = NULL; + const char *method; + sd_bus *bus; + int r; + + assert(argc > 1); + assert(argv); + + r = acquire_bus(BUS_MANAGER, &bus); + if (r < 0) + return r; + + polkit_agent_open_if_enabled(); + + method = streq(argv[0], "set-environment") + ? "SetEnvironment" + : "UnsetEnvironment"; + + r = sd_bus_message_new_method_call( + bus, + &m, + "org.freedesktop.systemd1", + "/org/freedesktop/systemd1", + "org.freedesktop.systemd1.Manager", + method); + if (r < 0) + return bus_log_create_error(r); + + r = sd_bus_message_append_strv(m, strv_skip(argv, 1)); + if (r < 0) + return bus_log_create_error(r); + + r = sd_bus_call(bus, m, 0, &error, NULL); + if (r < 0) + return log_error_errno(r, "Failed to set environment: %s", bus_error_message(&error, r)); + + return 0; +} + +static int import_environment(int argc, char *argv[], void *userdata) { + _cleanup_(sd_bus_error_free) sd_bus_error error = SD_BUS_ERROR_NULL; + _cleanup_(sd_bus_message_unrefp) sd_bus_message *m = NULL; + sd_bus *bus; + int r; + + r = acquire_bus(BUS_MANAGER, &bus); + if (r < 0) + return r; + + polkit_agent_open_if_enabled(); + + r = sd_bus_message_new_method_call( + bus, + &m, + "org.freedesktop.systemd1", + "/org/freedesktop/systemd1", + "org.freedesktop.systemd1.Manager", + "SetEnvironment"); + if (r < 0) + return bus_log_create_error(r); + + if (argc < 2) + r = sd_bus_message_append_strv(m, environ); + else { + char **a, **b; + + r = sd_bus_message_open_container(m, 'a', "s"); + if (r < 0) + return bus_log_create_error(r); + + STRV_FOREACH(a, strv_skip(argv, 1)) { + + if (!env_name_is_valid(*a)) { + log_error("Not a valid environment variable name: %s", *a); + return -EINVAL; + } + + STRV_FOREACH(b, environ) { + const char *eq; + + eq = startswith(*b, *a); + if (eq && *eq == '=') { + + r = sd_bus_message_append(m, "s", *b); + if (r < 0) + return bus_log_create_error(r); + + break; + } + } + } + + r = sd_bus_message_close_container(m); + } + if (r < 0) + return bus_log_create_error(r); + + r = sd_bus_call(bus, m, 0, &error, NULL); + if (r < 0) + return log_error_errno(r, "Failed to import environment: %s", bus_error_message(&error, r)); + + return 0; +} + +static int enable_sysv_units(const char *verb, char **args) { + int r = 0; + +#if defined(HAVE_SYSV_COMPAT) + _cleanup_lookup_paths_free_ LookupPaths paths = {}; + unsigned f = 0; + + /* Processes all SysV units, and reshuffles the array so that afterwards only the native units remain */ + + if (arg_scope != UNIT_FILE_SYSTEM) + return 0; + + if (getenv_bool("SYSTEMCTL_SKIP_SYSV") > 0) + return 0; + + if (!STR_IN_SET(verb, + "enable", + "disable", + "is-enabled")) + return 0; + + r = lookup_paths_init(&paths, arg_scope, LOOKUP_PATHS_EXCLUDE_GENERATED, arg_root); + if (r < 0) + return r; + + r = 0; + while (args[f]) { + + const char *argv[] = { + ROOTLIBEXECDIR "/systemd-sysv-install", + NULL, + NULL, + NULL, + NULL, + }; + + _cleanup_free_ char *p = NULL, *q = NULL, *l = NULL; + bool found_native = false, found_sysv; + siginfo_t status; + const char *name; + unsigned c = 1; + pid_t pid; + int j; + + name = args[f++]; + + if (!endswith(name, ".service")) + continue; + + if (path_is_absolute(name)) + continue; + + j = unit_file_exists(arg_scope, &paths, name); + if (j < 0 && !IN_SET(j, -ELOOP, -ERFKILL, -EADDRNOTAVAIL)) + return log_error_errno(j, "Failed to lookup unit file state: %m"); + found_native = j != 0; + + /* If we have both a native unit and a SysV script, enable/disable them both (below); for is-enabled, + * prefer the native unit */ + if (found_native && streq(verb, "is-enabled")) + continue; + + p = path_join(arg_root, SYSTEM_SYSVINIT_PATH, name); + if (!p) + return log_oom(); + + p[strlen(p) - strlen(".service")] = 0; + found_sysv = access(p, F_OK) >= 0; + if (!found_sysv) + continue; + + if (found_native) + log_info("Synchronizing state of %s with SysV service script with %s.", name, argv[0]); + else + log_info("%s is not a native service, redirecting to systemd-sysv-install.", name); + + if (!isempty(arg_root)) + argv[c++] = q = strappend("--root=", arg_root); + + argv[c++] = verb; + argv[c++] = basename(p); + argv[c] = NULL; + + l = strv_join((char**)argv, " "); + if (!l) + return log_oom(); + + log_info("Executing: %s", l); + + pid = fork(); + if (pid < 0) + return log_error_errno(errno, "Failed to fork: %m"); + else if (pid == 0) { + /* Child */ + + (void) reset_all_signal_handlers(); + (void) reset_signal_mask(); + + execv(argv[0], (char**) argv); + log_error_errno(errno, "Failed to execute %s: %m", argv[0]); + _exit(EXIT_FAILURE); + } + + j = wait_for_terminate(pid, &status); + if (j < 0) + return log_error_errno(j, "Failed to wait for child: %m"); + + if (status.si_code == CLD_EXITED) { + if (streq(verb, "is-enabled")) { + if (status.si_status == 0) { + if (!arg_quiet) + puts("enabled"); + r = 1; + } else { + if (!arg_quiet) + puts("disabled"); + } + + } else if (status.si_status != 0) + return -EBADE; /* We don't warn here, under the assumption the script already showed an explanation */ + } else { + log_error("Unexpected waitid() result."); + return -EPROTO; + } + + if (found_native) + continue; + + /* Remove this entry, so that we don't try enabling it as native unit */ + assert(f > 0); + f--; + assert(args[f] == name); + strv_remove(args, name); + } + +#endif + return r; +} + +static int mangle_names(char **original_names, char ***mangled_names) { + char **i, **l, **name; + int r; + + l = i = new(char*, strv_length(original_names) + 1); + if (!l) + return log_oom(); + + STRV_FOREACH(name, original_names) { + + /* When enabling units qualified path names are OK, + * too, hence allow them explicitly. */ + + if (is_path(*name)) { + *i = strdup(*name); + if (!*i) { + strv_free(l); + return log_oom(); + } + } else { + r = unit_name_mangle(*name, UNIT_NAME_NOGLOB, i); + if (r < 0) { + strv_free(l); + return log_error_errno(r, "Failed to mangle unit name: %m"); + } + } + + i++; + } + + *i = NULL; + *mangled_names = l; + + return 0; +} + +static int unit_exists(const char *unit) { + _cleanup_(sd_bus_message_unrefp) sd_bus_message *reply = NULL; + _cleanup_(sd_bus_error_free) sd_bus_error error = SD_BUS_ERROR_NULL; + _cleanup_free_ char *path = NULL; + static const struct bus_properties_map property_map[] = { + { "LoadState", "s", map_string_no_copy, offsetof(UnitStatusInfo, load_state) }, + { "ActiveState", "s", map_string_no_copy, offsetof(UnitStatusInfo, active_state)}, + {}, + }; + UnitStatusInfo info = {}; + sd_bus *bus; + int r; + + path = unit_dbus_path_from_name(unit); + if (!path) + return log_oom(); + + r = acquire_bus(BUS_MANAGER, &bus); + if (r < 0) + return r; + + r = sd_bus_call_method( + bus, + "org.freedesktop.systemd1", + path, + "org.freedesktop.DBus.Properties", + "GetAll", + &error, + &reply, + "s", ""); + if (r < 0) + return log_error_errno(r, "Failed to get properties: %s", bus_error_message(&error, r)); + + r = bus_message_map_all_properties(reply, property_map, &info); + if (r < 0) + return log_error_errno(r, "Failed to map properties: %s", bus_error_message(&error, r)); + + return !streq_ptr(info.load_state, "not-found") || !streq_ptr(info.active_state, "inactive"); +} + +static int enable_unit(int argc, char *argv[], void *userdata) { + _cleanup_strv_free_ char **names = NULL; + const char *verb = argv[0]; + UnitFileChange *changes = NULL; + unsigned n_changes = 0; + int carries_install_info = -1; + bool ignore_carries_install_info = arg_quiet; + int r; + + if (!argv[1]) + return 0; + + r = mangle_names(strv_skip(argv, 1), &names); + if (r < 0) + return r; + + r = enable_sysv_units(verb, names); + if (r < 0) + return r; + + /* If the operation was fully executed by the SysV compat, let's finish early */ + if (strv_isempty(names)) { + if (arg_no_reload || install_client_side()) + return 0; + return daemon_reload(argc, argv, userdata); + } + + if (install_client_side()) { + if (streq(verb, "enable")) { + r = unit_file_enable(arg_scope, arg_runtime, arg_root, names, arg_force, &changes, &n_changes); + carries_install_info = r; + } else if (streq(verb, "disable")) + r = unit_file_disable(arg_scope, arg_runtime, arg_root, names, &changes, &n_changes); + else if (streq(verb, "reenable")) { + r = unit_file_reenable(arg_scope, arg_runtime, arg_root, names, arg_force, &changes, &n_changes); + carries_install_info = r; + } else if (streq(verb, "link")) + r = unit_file_link(arg_scope, arg_runtime, arg_root, names, arg_force, &changes, &n_changes); + else if (streq(verb, "preset")) { + r = unit_file_preset(arg_scope, arg_runtime, arg_root, names, arg_preset_mode, arg_force, &changes, &n_changes); + } else if (streq(verb, "mask")) + r = unit_file_mask(arg_scope, arg_runtime, arg_root, names, arg_force, &changes, &n_changes); + else if (streq(verb, "unmask")) + r = unit_file_unmask(arg_scope, arg_runtime, arg_root, names, &changes, &n_changes); + else if (streq(verb, "revert")) + r = unit_file_revert(arg_scope, arg_root, names, &changes, &n_changes); + else + assert_not_reached("Unknown verb"); + + unit_file_dump_changes(r, verb, changes, n_changes, arg_quiet); + if (r < 0) + goto finish; + r = 0; + } else { + _cleanup_(sd_bus_message_unrefp) sd_bus_message *reply = NULL, *m = NULL; + _cleanup_(sd_bus_error_free) sd_bus_error error = SD_BUS_ERROR_NULL; + bool expect_carries_install_info = false; + bool send_runtime = true, send_force = true, send_preset_mode = false; + const char *method; + sd_bus *bus; + + if (STR_IN_SET(verb, "mask", "unmask")) { + r = unit_exists(*names); + if (r < 0) + return r; + if (r == 0) + log_notice("Unit %s does not exist, proceeding anyway.", *names); + } + + r = acquire_bus(BUS_MANAGER, &bus); + if (r < 0) + return r; + + polkit_agent_open_if_enabled(); + + if (streq(verb, "enable")) { + method = "EnableUnitFiles"; + expect_carries_install_info = true; + } else if (streq(verb, "disable")) { + method = "DisableUnitFiles"; + send_force = false; + } else if (streq(verb, "reenable")) { + method = "ReenableUnitFiles"; + expect_carries_install_info = true; + } else if (streq(verb, "link")) + method = "LinkUnitFiles"; + else if (streq(verb, "preset")) { + + if (arg_preset_mode != UNIT_FILE_PRESET_FULL) { + method = "PresetUnitFilesWithMode"; + send_preset_mode = true; + } else + method = "PresetUnitFiles"; + + expect_carries_install_info = true; + ignore_carries_install_info = true; + } else if (streq(verb, "mask")) + method = "MaskUnitFiles"; + else if (streq(verb, "unmask")) { + method = "UnmaskUnitFiles"; + send_force = false; + } else if (streq(verb, "revert")) { + method = "RevertUnitFiles"; + send_runtime = send_force = false; + } else + assert_not_reached("Unknown verb"); + + r = sd_bus_message_new_method_call( + bus, + &m, + "org.freedesktop.systemd1", + "/org/freedesktop/systemd1", + "org.freedesktop.systemd1.Manager", + method); + if (r < 0) + return bus_log_create_error(r); + + r = sd_bus_message_append_strv(m, names); + if (r < 0) + return bus_log_create_error(r); + + if (send_preset_mode) { + r = sd_bus_message_append(m, "s", unit_file_preset_mode_to_string(arg_preset_mode)); + if (r < 0) + return bus_log_create_error(r); + } + + if (send_runtime) { + r = sd_bus_message_append(m, "b", arg_runtime); + if (r < 0) + return bus_log_create_error(r); + } + + if (send_force) { + r = sd_bus_message_append(m, "b", arg_force); + if (r < 0) + return bus_log_create_error(r); + } + + r = sd_bus_call(bus, m, 0, &error, &reply); + if (r < 0) + return log_error_errno(r, "Failed to %s unit: %s", verb, bus_error_message(&error, r)); + + if (expect_carries_install_info) { + r = sd_bus_message_read(reply, "b", &carries_install_info); + if (r < 0) + return bus_log_parse_error(r); + } + + r = bus_deserialize_and_dump_unit_file_changes(reply, arg_quiet, &changes, &n_changes); + if (r < 0) + goto finish; + + /* Try to reload if enabled */ + if (!arg_no_reload) + r = daemon_reload(argc, argv, userdata); + else + r = 0; + } + + if (carries_install_info == 0 && !ignore_carries_install_info) + log_warning("The unit files have no installation config (WantedBy, RequiredBy, Also, Alias\n" + "settings in the [Install] section, and DefaultInstance for template units).\n" + "This means they are not meant to be enabled using systemctl.\n" + "Possible reasons for having this kind of units are:\n" + "1) A unit may be statically enabled by being symlinked from another unit's\n" + " .wants/ or .requires/ directory.\n" + "2) A unit's purpose may be to act as a helper for some other unit which has\n" + " a requirement dependency on it.\n" + "3) A unit may be started when needed via activation (socket, path, timer,\n" + " D-Bus, udev, scripted systemctl call, ...).\n" + "4) In case of template units, the unit is meant to be enabled with some\n" + " instance name specified."); + + if (arg_now && n_changes > 0 && STR_IN_SET(argv[0], "enable", "disable", "mask")) { + char *new_args[n_changes + 2]; + sd_bus *bus; + unsigned i; + + r = acquire_bus(BUS_MANAGER, &bus); + if (r < 0) + goto finish; + + new_args[0] = (char*) (streq(argv[0], "enable") ? "start" : "stop"); + for (i = 0; i < n_changes; i++) + new_args[i + 1] = basename(changes[i].path); + new_args[i + 1] = NULL; + + r = start_unit(strv_length(new_args), new_args, userdata); + } + +finish: + unit_file_changes_free(changes, n_changes); + + return r; +} + +static int add_dependency(int argc, char *argv[], void *userdata) { + _cleanup_strv_free_ char **names = NULL; + _cleanup_free_ char *target = NULL; + const char *verb = argv[0]; + UnitFileChange *changes = NULL; + unsigned n_changes = 0; + UnitDependency dep; + int r = 0; + + if (!argv[1]) + return 0; + + r = unit_name_mangle_with_suffix(argv[1], UNIT_NAME_NOGLOB, ".target", &target); + if (r < 0) + return log_error_errno(r, "Failed to mangle unit name: %m"); + + r = mangle_names(strv_skip(argv, 2), &names); + if (r < 0) + return r; + + if (streq(verb, "add-wants")) + dep = UNIT_WANTS; + else if (streq(verb, "add-requires")) + dep = UNIT_REQUIRES; + else + assert_not_reached("Unknown verb"); + + if (install_client_side()) { + r = unit_file_add_dependency(arg_scope, arg_runtime, arg_root, names, target, dep, arg_force, &changes, &n_changes); + unit_file_dump_changes(r, "add dependency on", changes, n_changes, arg_quiet); + + if (r > 0) + r = 0; + } else { + _cleanup_(sd_bus_message_unrefp) sd_bus_message *reply = NULL, *m = NULL; + _cleanup_(sd_bus_error_free) sd_bus_error error = SD_BUS_ERROR_NULL; + sd_bus *bus; + + r = acquire_bus(BUS_MANAGER, &bus); + if (r < 0) + return r; + + polkit_agent_open_if_enabled(); + + r = sd_bus_message_new_method_call( + bus, + &m, + "org.freedesktop.systemd1", + "/org/freedesktop/systemd1", + "org.freedesktop.systemd1.Manager", + "AddDependencyUnitFiles"); + if (r < 0) + return bus_log_create_error(r); + + r = sd_bus_message_append_strv(m, names); + if (r < 0) + return bus_log_create_error(r); + + r = sd_bus_message_append(m, "ssbb", target, unit_dependency_to_string(dep), arg_runtime, arg_force); + if (r < 0) + return bus_log_create_error(r); + + r = sd_bus_call(bus, m, 0, &error, &reply); + if (r < 0) + return log_error_errno(r, "Failed to add dependency: %s", bus_error_message(&error, r)); + + r = bus_deserialize_and_dump_unit_file_changes(reply, arg_quiet, &changes, &n_changes); + if (r < 0) + goto finish; + + if (arg_no_reload) { + r = 0; + goto finish; + } + + r = daemon_reload(argc, argv, userdata); + } + +finish: + unit_file_changes_free(changes, n_changes); + + return r; +} + +static int preset_all(int argc, char *argv[], void *userdata) { + UnitFileChange *changes = NULL; + unsigned n_changes = 0; + int r; + + if (install_client_side()) { + r = unit_file_preset_all(arg_scope, arg_runtime, arg_root, arg_preset_mode, arg_force, &changes, &n_changes); + unit_file_dump_changes(r, "preset", changes, n_changes, arg_quiet); + + if (r > 0) + r = 0; + } else { + _cleanup_(sd_bus_error_free) sd_bus_error error = SD_BUS_ERROR_NULL; + _cleanup_(sd_bus_message_unrefp) sd_bus_message *reply = NULL; + sd_bus *bus; + + r = acquire_bus(BUS_MANAGER, &bus); + if (r < 0) + return r; + + polkit_agent_open_if_enabled(); + + r = sd_bus_call_method( + bus, + "org.freedesktop.systemd1", + "/org/freedesktop/systemd1", + "org.freedesktop.systemd1.Manager", + "PresetAllUnitFiles", + &error, + &reply, + "sbb", + unit_file_preset_mode_to_string(arg_preset_mode), + arg_runtime, + arg_force); + if (r < 0) + return log_error_errno(r, "Failed to preset all units: %s", bus_error_message(&error, r)); + + r = bus_deserialize_and_dump_unit_file_changes(reply, arg_quiet, &changes, &n_changes); + if (r < 0) + goto finish; + + if (arg_no_reload) { + r = 0; + goto finish; + } + + r = daemon_reload(argc, argv, userdata); + } + +finish: + unit_file_changes_free(changes, n_changes); + + return r; +} + +static int unit_is_enabled(int argc, char *argv[], void *userdata) { + + _cleanup_strv_free_ char **names = NULL; + bool enabled; + char **name; + int r; + + r = mangle_names(strv_skip(argv, 1), &names); + if (r < 0) + return r; + + r = enable_sysv_units(argv[0], names); + if (r < 0) + return r; + + enabled = r > 0; + + if (install_client_side()) { + + STRV_FOREACH(name, names) { + UnitFileState state; + + r = unit_file_get_state(arg_scope, arg_root, *name, &state); + if (r < 0) + return log_error_errno(state, "Failed to get unit file state for %s: %m", *name); + + if (IN_SET(state, + UNIT_FILE_ENABLED, + UNIT_FILE_ENABLED_RUNTIME, + UNIT_FILE_STATIC, + UNIT_FILE_INDIRECT, + UNIT_FILE_GENERATED)) + enabled = true; + + if (!arg_quiet) + puts(unit_file_state_to_string(state)); + } + + r = 0; + } else { + _cleanup_(sd_bus_error_free) sd_bus_error error = SD_BUS_ERROR_NULL; + sd_bus *bus; + + r = acquire_bus(BUS_MANAGER, &bus); + if (r < 0) + return r; + + STRV_FOREACH(name, names) { + _cleanup_(sd_bus_message_unrefp) sd_bus_message *reply = NULL; + const char *s; + + r = sd_bus_call_method( + bus, + "org.freedesktop.systemd1", + "/org/freedesktop/systemd1", + "org.freedesktop.systemd1.Manager", + "GetUnitFileState", + &error, + &reply, + "s", *name); + if (r < 0) + return log_error_errno(r, "Failed to get unit file state for %s: %s", *name, bus_error_message(&error, r)); + + r = sd_bus_message_read(reply, "s", &s); + if (r < 0) + return bus_log_parse_error(r); + + if (STR_IN_SET(s, "enabled", "enabled-runtime", "static", "indirect", "generated")) + enabled = true; + + if (!arg_quiet) + puts(s); + } + } + + return enabled ? EXIT_SUCCESS : EXIT_FAILURE; +} + +static int is_system_running(int argc, char *argv[], void *userdata) { + _cleanup_free_ char *state = NULL; + sd_bus *bus; + int r; + + if (running_in_chroot() > 0 || (arg_transport == BUS_TRANSPORT_LOCAL && !sd_booted())) { + if (!arg_quiet) + puts("offline"); + return EXIT_FAILURE; + } + + r = acquire_bus(BUS_MANAGER, &bus); + if (r < 0) + return r; + + r = sd_bus_get_property_string( + bus, + "org.freedesktop.systemd1", + "/org/freedesktop/systemd1", + "org.freedesktop.systemd1.Manager", + "SystemState", + NULL, + &state); + if (r < 0) { + if (!arg_quiet) + puts("unknown"); + return 0; + } + + if (!arg_quiet) + puts(state); + + return streq(state, "running") ? EXIT_SUCCESS : EXIT_FAILURE; +} + +static int create_edit_temp_file(const char *new_path, const char *original_path, char **ret_tmp_fn) { + _cleanup_free_ char *t = NULL; + int r; + + assert(new_path); + assert(original_path); + assert(ret_tmp_fn); + + r = tempfn_random(new_path, NULL, &t); + if (r < 0) + return log_error_errno(r, "Failed to determine temporary filename for \"%s\": %m", new_path); + + r = mkdir_parents(new_path, 0755); + if (r < 0) + return log_error_errno(r, "Failed to create directories for \"%s\": %m", new_path); + + r = copy_file(original_path, t, 0, 0644, 0); + if (r == -ENOENT) { + + r = touch(t); + if (r < 0) + return log_error_errno(r, "Failed to create temporary file \"%s\": %m", t); + + } else if (r < 0) + return log_error_errno(r, "Failed to create temporary file for \"%s\": %m", new_path); + + *ret_tmp_fn = t; + t = NULL; + + return 0; +} + +static int get_file_to_edit( + const LookupPaths *paths, + const char *name, + char **ret_path) { + + _cleanup_free_ char *path = NULL, *run = NULL; + + assert(name); + assert(ret_path); + + path = strjoin(paths->persistent_config, "/", name, NULL); + if (!path) + return log_oom(); + + if (arg_runtime) { + run = strjoin(paths->runtime_config, "/", name, NULL); + if (!run) + return log_oom(); + } + + if (arg_runtime) { + if (access(path, F_OK) >= 0) { + log_error("Refusing to create \"%s\" because it would be overridden by \"%s\" anyway.", run, path); + return -EEXIST; + } + + *ret_path = run; + run = NULL; + } else { + *ret_path = path; + path = NULL; + } + + return 0; +} + +static int unit_file_create_new( + const LookupPaths *paths, + const char *unit_name, + const char *suffix, + char **ret_new_path, + char **ret_tmp_path) { + + char *tmp_new_path, *tmp_tmp_path, *ending; + int r; + + assert(unit_name); + assert(ret_new_path); + assert(ret_tmp_path); + + ending = strjoina(unit_name, suffix); + r = get_file_to_edit(paths, ending, &tmp_new_path); + if (r < 0) + return r; + + r = create_edit_temp_file(tmp_new_path, tmp_new_path, &tmp_tmp_path); + if (r < 0) { + free(tmp_new_path); + return r; + } + + *ret_new_path = tmp_new_path; + *ret_tmp_path = tmp_tmp_path; + + return 0; +} + +static int unit_file_create_copy( + const LookupPaths *paths, + const char *unit_name, + const char *fragment_path, + char **ret_new_path, + char **ret_tmp_path) { + + char *tmp_new_path, *tmp_tmp_path; + int r; + + assert(fragment_path); + assert(unit_name); + assert(ret_new_path); + assert(ret_tmp_path); + + r = get_file_to_edit(paths, unit_name, &tmp_new_path); + if (r < 0) + return r; + + if (!path_equal(fragment_path, tmp_new_path) && access(tmp_new_path, F_OK) == 0) { + char response; + + r = ask_char(&response, "yn", "\"%s\" already exists. Overwrite with \"%s\"? [(y)es, (n)o] ", tmp_new_path, fragment_path); + if (r < 0) { + free(tmp_new_path); + return r; + } + if (response != 'y') { + log_warning("%s ignored", unit_name); + free(tmp_new_path); + return -EKEYREJECTED; + } + } + + r = create_edit_temp_file(tmp_new_path, fragment_path, &tmp_tmp_path); + if (r < 0) { + free(tmp_new_path); + return r; + } + + *ret_new_path = tmp_new_path; + *ret_tmp_path = tmp_tmp_path; + + return 0; +} + +static int run_editor(char **paths) { + pid_t pid; + int r; + + assert(paths); + + pid = fork(); + if (pid < 0) + return log_error_errno(errno, "Failed to fork: %m"); + + if (pid == 0) { + const char **args; + char *editor, **editor_args = NULL; + char **tmp_path, **original_path, *p; + unsigned n_editor_args = 0, i = 1; + size_t argc; + + (void) reset_all_signal_handlers(); + (void) reset_signal_mask(); + + argc = strv_length(paths)/2 + 1; + + /* SYSTEMD_EDITOR takes precedence over EDITOR which takes precedence over VISUAL + * If neither SYSTEMD_EDITOR nor EDITOR nor VISUAL are present, + * we try to execute well known editors + */ + editor = getenv("SYSTEMD_EDITOR"); + if (!editor) + editor = getenv("EDITOR"); + if (!editor) + editor = getenv("VISUAL"); + + if (!isempty(editor)) { + editor_args = strv_split(editor, WHITESPACE); + if (!editor_args) { + (void) log_oom(); + _exit(EXIT_FAILURE); + } + n_editor_args = strv_length(editor_args); + argc += n_editor_args - 1; + } + args = newa(const char*, argc + 1); + + if (n_editor_args > 0) { + args[0] = editor_args[0]; + for (; i < n_editor_args; i++) + args[i] = editor_args[i]; + } + + STRV_FOREACH_PAIR(original_path, tmp_path, paths) { + args[i] = *tmp_path; + i++; + } + args[i] = NULL; + + if (n_editor_args > 0) + execvp(args[0], (char* const*) args); + + FOREACH_STRING(p, "editor", "nano", "vim", "vi") { + args[0] = p; + execvp(p, (char* const*) args); + /* We do not fail if the editor doesn't exist + * because we want to try each one of them before + * failing. + */ + if (errno != ENOENT) { + log_error_errno(errno, "Failed to execute %s: %m", editor); + _exit(EXIT_FAILURE); + } + } + + log_error("Cannot edit unit(s), no editor available. Please set either $SYSTEMD_EDITOR, $EDITOR or $VISUAL."); + _exit(EXIT_FAILURE); + } + + r = wait_for_terminate_and_warn("editor", pid, true); + if (r < 0) + return log_error_errno(r, "Failed to wait for child: %m"); + + return 0; +} + +static int find_paths_to_edit(sd_bus *bus, char **names, char ***paths) { + _cleanup_lookup_paths_free_ LookupPaths lp = {}; + char **name; + int r; + + assert(names); + assert(paths); + + r = lookup_paths_init(&lp, arg_scope, 0, arg_root); + if (r < 0) + return r; + + STRV_FOREACH(name, names) { + _cleanup_free_ char *path = NULL, *new_path = NULL, *tmp_path = NULL; + + r = unit_find_paths(bus, *name, &lp, &path, NULL); + if (r < 0) + return r; + else if (!arg_force) { + if (r == 0) { + log_error("Run 'systemctl edit --force %s' to create a new unit.", *name); + return -ENOENT; + } else if (!path) { + // FIXME: support units with path==NULL (no FragmentPath) + log_error("No fragment exists for %s.", *name); + return -ENOENT; + } + } + + if (path) { + if (arg_full) + r = unit_file_create_copy(&lp, *name, path, &new_path, &tmp_path); + else + r = unit_file_create_new(&lp, *name, ".d/override.conf", &new_path, &tmp_path); + } else + r = unit_file_create_new(&lp, *name, NULL, &new_path, &tmp_path); + if (r < 0) + return r; + + r = strv_push_pair(paths, new_path, tmp_path); + if (r < 0) + return log_oom(); + new_path = tmp_path = NULL; + } + + return 0; +} + +static int edit(int argc, char *argv[], void *userdata) { + _cleanup_strv_free_ char **names = NULL; + _cleanup_strv_free_ char **paths = NULL; + char **original, **tmp; + sd_bus *bus; + int r; + + if (!on_tty()) { + log_error("Cannot edit units if not on a tty."); + return -EINVAL; + } + + if (arg_transport != BUS_TRANSPORT_LOCAL) { + log_error("Cannot edit units remotely."); + return -EINVAL; + } + + r = acquire_bus(BUS_MANAGER, &bus); + if (r < 0) + return r; + + r = expand_names(bus, strv_skip(argv, 1), NULL, &names); + if (r < 0) + return log_error_errno(r, "Failed to expand names: %m"); + + r = find_paths_to_edit(bus, names, &paths); + if (r < 0) + return r; + + if (strv_isempty(paths)) + return -ENOENT; + + r = run_editor(paths); + if (r < 0) + goto end; + + STRV_FOREACH_PAIR(original, tmp, paths) { + /* If the temporary file is empty we ignore it. It's + * useful if the user wants to cancel its modification + */ + if (null_or_empty_path(*tmp)) { + log_warning("Editing \"%s\" canceled: temporary file is empty.", *original); + continue; + } + + r = rename(*tmp, *original); + if (r < 0) { + r = log_error_errno(errno, "Failed to rename \"%s\" to \"%s\": %m", *tmp, *original); + goto end; + } + } + + r = 0; + + if (!arg_no_reload && !install_client_side()) + r = daemon_reload(argc, argv, userdata); + +end: + STRV_FOREACH_PAIR(original, tmp, paths) { + (void) unlink(*tmp); + + /* Removing empty dropin dirs */ + if (!arg_full) { + _cleanup_free_ char *dir; + + dir = dirname_malloc(*original); + if (!dir) + return log_oom(); + + /* no need to check if the dir is empty, rmdir + * does nothing if it is not the case. + */ + (void) rmdir(dir); + } + } + + return r; +} + +static void systemctl_help(void) { + + pager_open(arg_no_pager, false); + + printf("%s [OPTIONS...] {COMMAND} ...\n\n" + "Query or send control commands to the systemd manager.\n\n" + " -h --help Show this help\n" + " --version Show package version\n" + " --system Connect to system manager\n" + " --user Connect to user service manager\n" + " -H --host=[USER@]HOST\n" + " Operate on remote host\n" + " -M --machine=CONTAINER\n" + " Operate on local container\n" + " -t --type=TYPE List units of a particular type\n" + " --state=STATE List units with particular LOAD or SUB or ACTIVE state\n" + " -p --property=NAME Show only properties by this name\n" + " -a --all Show all loaded units/properties, including dead/empty\n" + " ones. To list all units installed on the system, use\n" + " the 'list-unit-files' command instead.\n" + " -l --full Don't ellipsize unit names on output\n" + " -r --recursive Show unit list of host and local containers\n" + " --reverse Show reverse dependencies with 'list-dependencies'\n" + " --job-mode=MODE Specify how to deal with already queued jobs, when\n" + " queueing a new job\n" + " --show-types When showing sockets, explicitly show their type\n" + " --value When showing properties, only print the value\n" + " -i --ignore-inhibitors\n" + " When shutting down or sleeping, ignore inhibitors\n" + " --kill-who=WHO Who to send signal to\n" + " -s --signal=SIGNAL Which signal to send\n" + " --now Start or stop unit in addition to enabling or disabling it\n" + " -q --quiet Suppress output\n" + " --no-block Do not wait until operation finished\n" + " --no-wall Don't send wall message before halt/power-off/reboot\n" + " --no-reload Don't reload daemon after en-/dis-abling unit files\n" + " --no-legend Do not print a legend (column headers and hints)\n" + " --no-pager Do not pipe output into a pager\n" + " --no-ask-password\n" + " Do not ask for system passwords\n" + " --global Enable/disable unit files globally\n" + " --runtime Enable unit files only temporarily until next reboot\n" + " -f --force When enabling unit files, override existing symlinks\n" + " When shutting down, execute action immediately\n" + " --preset-mode= Apply only enable, only disable, or all presets\n" + " --root=PATH Enable unit files in the specified root directory\n" + " -n --lines=INTEGER Number of journal entries to show\n" + " -o --output=STRING Change journal output mode (short, short-iso,\n" + " short-precise, short-monotonic, verbose,\n" + " export, json, json-pretty, json-sse, cat)\n" + " --firmware-setup Tell the firmware to show the setup menu on next boot\n" + " --plain Print unit dependencies as a list instead of a tree\n\n" + "Unit Commands:\n" + " list-units [PATTERN...] List loaded units\n" + " list-sockets [PATTERN...] List loaded sockets ordered by address\n" + " list-timers [PATTERN...] List loaded timers ordered by next elapse\n" + " start NAME... Start (activate) one or more units\n" + " stop NAME... Stop (deactivate) one or more units\n" + " reload NAME... Reload one or more units\n" + " restart NAME... Start or restart one or more units\n" + " try-restart NAME... Restart one or more units if active\n" + " reload-or-restart NAME... Reload one or more units if possible,\n" + " otherwise start or restart\n" + " try-reload-or-restart NAME... If active, reload one or more units,\n" + " if supported, otherwise restart\n" + " isolate NAME Start one unit and stop all others\n" + " kill NAME... Send signal to processes of a unit\n" + " is-active PATTERN... Check whether units are active\n" + " is-failed PATTERN... Check whether units are failed\n" + " status [PATTERN...|PID...] Show runtime status of one or more units\n" + " show [PATTERN...|JOB...] Show properties of one or more\n" + " units/jobs or the manager\n" + " cat PATTERN... Show files and drop-ins of one or more units\n" + " set-property NAME ASSIGNMENT... Sets one or more properties of a unit\n" + " help PATTERN...|PID... Show manual for one or more units\n" + " reset-failed [PATTERN...] Reset failed state for all, one, or more\n" + " units\n" + " list-dependencies [NAME] Recursively show units which are required\n" + " or wanted by this unit or by which this\n" + " unit is required or wanted\n\n" + "Unit File Commands:\n" + " list-unit-files [PATTERN...] List installed unit files\n" + " enable [NAME...|PATH...] Enable one or more unit files\n" + " disable NAME... Disable one or more unit files\n" + " reenable NAME... Reenable one or more unit files\n" + " preset NAME... Enable/disable one or more unit files\n" + " based on preset configuration\n" + " preset-all Enable/disable all unit files based on\n" + " preset configuration\n" + " is-enabled NAME... Check whether unit files are enabled\n" + " mask NAME... Mask one or more units\n" + " unmask NAME... Unmask one or more units\n" + " link PATH... Link one or more units files into\n" + " the search path\n" + " revert NAME... Revert one or more unit files to vendor\n" + " version\n" + " add-wants TARGET NAME... Add 'Wants' dependency for the target\n" + " on specified one or more units\n" + " add-requires TARGET NAME... Add 'Requires' dependency for the target\n" + " on specified one or more units\n" + " edit NAME... Edit one or more unit files\n" + " get-default Get the name of the default target\n" + " set-default NAME Set the default target\n\n" + "Machine Commands:\n" + " list-machines [PATTERN...] List local containers and host\n\n" + "Job Commands:\n" + " list-jobs [PATTERN...] List jobs\n" + " cancel [JOB...] Cancel all, one, or more jobs\n\n" + "Environment Commands:\n" + " show-environment Dump environment\n" + " set-environment NAME=VALUE... Set one or more environment variables\n" + " unset-environment NAME... Unset one or more environment variables\n" + " import-environment [NAME...] Import all or some environment variables\n\n" + "Manager Lifecycle Commands:\n" + " daemon-reload Reload systemd manager configuration\n" + " daemon-reexec Reexecute systemd manager\n\n" + "System Commands:\n" + " is-system-running Check whether system is fully running\n" + " default Enter system default mode\n" + " rescue Enter system rescue mode\n" + " emergency Enter system emergency mode\n" + " halt Shut down and halt the system\n" + " poweroff Shut down and power-off the system\n" + " reboot [ARG] Shut down and reboot the system\n" + " kexec Shut down and reboot the system with kexec\n" + " exit [EXIT_CODE] Request user instance or container exit\n" + " switch-root ROOT [INIT] Change to a different root file system\n" + " suspend Suspend the system\n" + " hibernate Hibernate the system\n" + " hybrid-sleep Hibernate and suspend the system\n", + program_invocation_short_name); +} + +static void halt_help(void) { + printf("%s [OPTIONS...]%s\n\n" + "%s the system.\n\n" + " --help Show this help\n" + " --halt Halt the machine\n" + " -p --poweroff Switch off the machine\n" + " --reboot Reboot the machine\n" + " -f --force Force immediate halt/power-off/reboot\n" + " -w --wtmp-only Don't halt/power-off/reboot, just write wtmp record\n" + " -d --no-wtmp Don't write wtmp record\n" + " --no-wall Don't send wall message before halt/power-off/reboot\n", + program_invocation_short_name, + arg_action == ACTION_REBOOT ? " [ARG]" : "", + arg_action == ACTION_REBOOT ? "Reboot" : + arg_action == ACTION_POWEROFF ? "Power off" : + "Halt"); +} + +static void shutdown_help(void) { + printf("%s [OPTIONS...] [TIME] [WALL...]\n\n" + "Shut down the system.\n\n" + " --help Show this help\n" + " -H --halt Halt the machine\n" + " -P --poweroff Power-off the machine\n" + " -r --reboot Reboot the machine\n" + " -h Equivalent to --poweroff, overridden by --halt\n" + " -k Don't halt/power-off/reboot, just send warnings\n" + " --no-wall Don't send wall message before halt/power-off/reboot\n" + " -c Cancel a pending shutdown\n", + program_invocation_short_name); +} + +static void telinit_help(void) { + printf("%s [OPTIONS...] {COMMAND}\n\n" + "Send control commands to the init daemon.\n\n" + " --help Show this help\n" + " --no-wall Don't send wall message before halt/power-off/reboot\n\n" + "Commands:\n" + " 0 Power-off the machine\n" + " 6 Reboot the machine\n" + " 2, 3, 4, 5 Start runlevelX.target unit\n" + " 1, s, S Enter rescue mode\n" + " q, Q Reload init daemon configuration\n" + " u, U Reexecute init daemon\n", + program_invocation_short_name); +} + +static void runlevel_help(void) { + printf("%s [OPTIONS...]\n\n" + "Prints the previous and current runlevel of the init system.\n\n" + " --help Show this help\n", + program_invocation_short_name); +} + +static void help_types(void) { + int i; + + if (!arg_no_legend) + puts("Available unit types:"); + for (i = 0; i < _UNIT_TYPE_MAX; i++) + puts(unit_type_to_string(i)); +} + +static void help_states(void) { + int i; + + if (!arg_no_legend) + puts("Available unit load states:"); + for (i = 0; i < _UNIT_LOAD_STATE_MAX; i++) + puts(unit_load_state_to_string(i)); + + if (!arg_no_legend) + puts("\nAvailable unit active states:"); + for (i = 0; i < _UNIT_ACTIVE_STATE_MAX; i++) + puts(unit_active_state_to_string(i)); + + if (!arg_no_legend) + puts("\nAvailable automount unit substates:"); + for (i = 0; i < _AUTOMOUNT_STATE_MAX; i++) + puts(automount_state_to_string(i)); + + if (!arg_no_legend) + puts("\nAvailable busname unit substates:"); + for (i = 0; i < _BUSNAME_STATE_MAX; i++) + puts(busname_state_to_string(i)); + + if (!arg_no_legend) + puts("\nAvailable device unit substates:"); + for (i = 0; i < _DEVICE_STATE_MAX; i++) + puts(device_state_to_string(i)); + + if (!arg_no_legend) + puts("\nAvailable mount unit substates:"); + for (i = 0; i < _MOUNT_STATE_MAX; i++) + puts(mount_state_to_string(i)); + + if (!arg_no_legend) + puts("\nAvailable path unit substates:"); + for (i = 0; i < _PATH_STATE_MAX; i++) + puts(path_state_to_string(i)); + + if (!arg_no_legend) + puts("\nAvailable scope unit substates:"); + for (i = 0; i < _SCOPE_STATE_MAX; i++) + puts(scope_state_to_string(i)); + + if (!arg_no_legend) + puts("\nAvailable service unit substates:"); + for (i = 0; i < _SERVICE_STATE_MAX; i++) + puts(service_state_to_string(i)); + + if (!arg_no_legend) + puts("\nAvailable slice unit substates:"); + for (i = 0; i < _SLICE_STATE_MAX; i++) + puts(slice_state_to_string(i)); + + if (!arg_no_legend) + puts("\nAvailable socket unit substates:"); + for (i = 0; i < _SOCKET_STATE_MAX; i++) + puts(socket_state_to_string(i)); + + if (!arg_no_legend) + puts("\nAvailable swap unit substates:"); + for (i = 0; i < _SWAP_STATE_MAX; i++) + puts(swap_state_to_string(i)); + + if (!arg_no_legend) + puts("\nAvailable target unit substates:"); + for (i = 0; i < _TARGET_STATE_MAX; i++) + puts(target_state_to_string(i)); + + if (!arg_no_legend) + puts("\nAvailable timer unit substates:"); + for (i = 0; i < _TIMER_STATE_MAX; i++) + puts(timer_state_to_string(i)); +} + +static int systemctl_parse_argv(int argc, char *argv[]) { + + enum { + ARG_FAIL = 0x100, + ARG_REVERSE, + ARG_AFTER, + ARG_BEFORE, + ARG_SHOW_TYPES, + ARG_IRREVERSIBLE, + ARG_IGNORE_DEPENDENCIES, + ARG_VALUE, + ARG_VERSION, + ARG_USER, + ARG_SYSTEM, + ARG_GLOBAL, + ARG_NO_BLOCK, + ARG_NO_LEGEND, + ARG_NO_PAGER, + ARG_NO_WALL, + ARG_ROOT, + ARG_NO_RELOAD, + ARG_KILL_WHO, + ARG_NO_ASK_PASSWORD, + ARG_FAILED, + ARG_RUNTIME, + ARG_FORCE, + ARG_PLAIN, + ARG_STATE, + ARG_JOB_MODE, + ARG_PRESET_MODE, + ARG_FIRMWARE_SETUP, + ARG_NOW, + ARG_MESSAGE, + }; + + static const struct option options[] = { + { "help", no_argument, NULL, 'h' }, + { "version", no_argument, NULL, ARG_VERSION }, + { "type", required_argument, NULL, 't' }, + { "property", required_argument, NULL, 'p' }, + { "all", no_argument, NULL, 'a' }, + { "reverse", no_argument, NULL, ARG_REVERSE }, + { "after", no_argument, NULL, ARG_AFTER }, + { "before", no_argument, NULL, ARG_BEFORE }, + { "show-types", no_argument, NULL, ARG_SHOW_TYPES }, + { "failed", no_argument, NULL, ARG_FAILED }, /* compatibility only */ + { "full", no_argument, NULL, 'l' }, + { "job-mode", required_argument, NULL, ARG_JOB_MODE }, + { "fail", no_argument, NULL, ARG_FAIL }, /* compatibility only */ + { "irreversible", no_argument, NULL, ARG_IRREVERSIBLE }, /* compatibility only */ + { "ignore-dependencies", no_argument, NULL, ARG_IGNORE_DEPENDENCIES }, /* compatibility only */ + { "ignore-inhibitors", no_argument, NULL, 'i' }, + { "value", no_argument, NULL, ARG_VALUE }, + { "user", no_argument, NULL, ARG_USER }, + { "system", no_argument, NULL, ARG_SYSTEM }, + { "global", no_argument, NULL, ARG_GLOBAL }, + { "no-block", no_argument, NULL, ARG_NO_BLOCK }, + { "no-legend", no_argument, NULL, ARG_NO_LEGEND }, + { "no-pager", no_argument, NULL, ARG_NO_PAGER }, + { "no-wall", no_argument, NULL, ARG_NO_WALL }, + { "quiet", no_argument, NULL, 'q' }, + { "root", required_argument, NULL, ARG_ROOT }, + { "force", no_argument, NULL, ARG_FORCE }, + { "no-reload", no_argument, NULL, ARG_NO_RELOAD }, + { "kill-who", required_argument, NULL, ARG_KILL_WHO }, + { "signal", required_argument, NULL, 's' }, + { "no-ask-password", no_argument, NULL, ARG_NO_ASK_PASSWORD }, + { "host", required_argument, NULL, 'H' }, + { "machine", required_argument, NULL, 'M' }, + { "runtime", no_argument, NULL, ARG_RUNTIME }, + { "lines", required_argument, NULL, 'n' }, + { "output", required_argument, NULL, 'o' }, + { "plain", no_argument, NULL, ARG_PLAIN }, + { "state", required_argument, NULL, ARG_STATE }, + { "recursive", no_argument, NULL, 'r' }, + { "preset-mode", required_argument, NULL, ARG_PRESET_MODE }, + { "firmware-setup", no_argument, NULL, ARG_FIRMWARE_SETUP }, + { "now", no_argument, NULL, ARG_NOW }, + { "message", required_argument, NULL, ARG_MESSAGE }, + {} + }; + + const char *p; + int c, r; + + assert(argc >= 0); + assert(argv); + + /* we default to allowing interactive authorization only in systemctl (not in the legacy commands) */ + arg_ask_password = true; + + while ((c = getopt_long(argc, argv, "ht:p:alqfs:H:M:n:o:ir", options, NULL)) >= 0) + + switch (c) { + + case 'h': + systemctl_help(); + return 0; + + case ARG_VERSION: + return version(); + + case 't': { + if (isempty(optarg)) { + log_error("--type requires arguments."); + return -EINVAL; + } + + p = optarg; + for (;;) { + _cleanup_free_ char *type = NULL; + + r = extract_first_word(&p, &type, ",", 0); + if (r < 0) + return log_error_errno(r, "Failed to parse type: %s", optarg); + + if (r == 0) + break; + + if (streq(type, "help")) { + help_types(); + return 0; + } + + if (unit_type_from_string(type) >= 0) { + if (strv_push(&arg_types, type) < 0) + return log_oom(); + type = NULL; + continue; + } + + /* It's much nicer to use --state= for + * load states, but let's support this + * in --types= too for compatibility + * with old versions */ + if (unit_load_state_from_string(type) >= 0) { + if (strv_push(&arg_states, type) < 0) + return log_oom(); + type = NULL; + continue; + } + + log_error("Unknown unit type or load state '%s'.", type); + log_info("Use -t help to see a list of allowed values."); + return -EINVAL; + } + + break; + } + + case 'p': { + /* Make sure that if the empty property list + was specified, we won't show any properties. */ + if (isempty(optarg) && !arg_properties) { + arg_properties = new0(char*, 1); + if (!arg_properties) + return log_oom(); + } else { + p = optarg; + for (;;) { + _cleanup_free_ char *prop = NULL; + + r = extract_first_word(&p, &prop, ",", 0); + if (r < 0) + return log_error_errno(r, "Failed to parse property: %s", optarg); + + if (r == 0) + break; + + if (strv_push(&arg_properties, prop) < 0) + return log_oom(); + + prop = NULL; + } + } + + /* If the user asked for a particular + * property, show it to him, even if it is + * empty. */ + arg_all = true; + + break; + } + + case 'a': + arg_all = true; + break; + + case ARG_REVERSE: + arg_dependency = DEPENDENCY_REVERSE; + break; + + case ARG_AFTER: + arg_dependency = DEPENDENCY_AFTER; + break; + + case ARG_BEFORE: + arg_dependency = DEPENDENCY_BEFORE; + break; + + case ARG_SHOW_TYPES: + arg_show_types = true; + break; + + case ARG_VALUE: + arg_value = true; + break; + + case ARG_JOB_MODE: + arg_job_mode = optarg; + break; + + case ARG_FAIL: + arg_job_mode = "fail"; + break; + + case ARG_IRREVERSIBLE: + arg_job_mode = "replace-irreversibly"; + break; + + case ARG_IGNORE_DEPENDENCIES: + arg_job_mode = "ignore-dependencies"; + break; + + case ARG_USER: + arg_scope = UNIT_FILE_USER; + break; + + case ARG_SYSTEM: + arg_scope = UNIT_FILE_SYSTEM; + break; + + case ARG_GLOBAL: + arg_scope = UNIT_FILE_GLOBAL; + break; + + case ARG_NO_BLOCK: + arg_no_block = true; + break; + + case ARG_NO_LEGEND: + arg_no_legend = true; + break; + + case ARG_NO_PAGER: + arg_no_pager = true; + break; + + case ARG_NO_WALL: + arg_no_wall = true; + break; + + case ARG_ROOT: + r = parse_path_argument_and_warn(optarg, false, &arg_root); + if (r < 0) + return r; + break; + + case 'l': + arg_full = true; + break; + + case ARG_FAILED: + if (strv_extend(&arg_states, "failed") < 0) + return log_oom(); + + break; + + case 'q': + arg_quiet = true; + break; + + case ARG_FORCE: + arg_force++; + break; + + case 'f': + arg_force++; + break; + + case ARG_NO_RELOAD: + arg_no_reload = true; + break; + + case ARG_KILL_WHO: + arg_kill_who = optarg; + break; + + case 's': + arg_signal = signal_from_string_try_harder(optarg); + if (arg_signal < 0) { + log_error("Failed to parse signal string %s.", optarg); + return -EINVAL; + } + break; + + case ARG_NO_ASK_PASSWORD: + arg_ask_password = false; + break; + + case 'H': + arg_transport = BUS_TRANSPORT_REMOTE; + arg_host = optarg; + break; + + case 'M': + arg_transport = BUS_TRANSPORT_MACHINE; + arg_host = optarg; + break; + + case ARG_RUNTIME: + arg_runtime = true; + break; + + case 'n': + if (safe_atou(optarg, &arg_lines) < 0) { + log_error("Failed to parse lines '%s'", optarg); + return -EINVAL; + } + break; + + case 'o': + arg_output = output_mode_from_string(optarg); + if (arg_output < 0) { + log_error("Unknown output '%s'.", optarg); + return -EINVAL; + } + break; + + case 'i': + arg_ignore_inhibitors = true; + break; + + case ARG_PLAIN: + arg_plain = true; + break; + + case ARG_FIRMWARE_SETUP: + arg_firmware_setup = true; + break; + + case ARG_STATE: { + if (isempty(optarg)) { + log_error("--signal requires arguments."); + return -EINVAL; + } + + p = optarg; + for (;;) { + _cleanup_free_ char *s = NULL; + + r = extract_first_word(&p, &s, ",", 0); + if (r < 0) + return log_error_errno(r, "Failed to parse signal: %s", optarg); + + if (r == 0) + break; + + if (streq(s, "help")) { + help_states(); + return 0; + } + + if (strv_push(&arg_states, s) < 0) + return log_oom(); + + s = NULL; + } + break; + } + + case 'r': + if (geteuid() != 0) { + log_error("--recursive requires root privileges."); + return -EPERM; + } + + arg_recursive = true; + break; + + case ARG_PRESET_MODE: + + arg_preset_mode = unit_file_preset_mode_from_string(optarg); + if (arg_preset_mode < 0) { + log_error("Failed to parse preset mode: %s.", optarg); + return -EINVAL; + } + + break; + + case ARG_NOW: + arg_now = true; + break; + + case ARG_MESSAGE: + if (strv_extend(&arg_wall, optarg) < 0) + return log_oom(); + break; + + case '?': + return -EINVAL; + + default: + assert_not_reached("Unhandled option"); + } + + if (arg_transport != BUS_TRANSPORT_LOCAL && arg_scope != UNIT_FILE_SYSTEM) { + log_error("Cannot access user instance remotely."); + return -EINVAL; + } + + return 1; +} + +static int halt_parse_argv(int argc, char *argv[]) { + + enum { + ARG_HELP = 0x100, + ARG_HALT, + ARG_REBOOT, + ARG_NO_WALL + }; + + static const struct option options[] = { + { "help", no_argument, NULL, ARG_HELP }, + { "halt", no_argument, NULL, ARG_HALT }, + { "poweroff", no_argument, NULL, 'p' }, + { "reboot", no_argument, NULL, ARG_REBOOT }, + { "force", no_argument, NULL, 'f' }, + { "wtmp-only", no_argument, NULL, 'w' }, + { "no-wtmp", no_argument, NULL, 'd' }, + { "no-sync", no_argument, NULL, 'n' }, + { "no-wall", no_argument, NULL, ARG_NO_WALL }, + {} + }; + + int c, r, runlevel; + + assert(argc >= 0); + assert(argv); + + if (utmp_get_runlevel(&runlevel, NULL) >= 0) + if (runlevel == '0' || runlevel == '6') + arg_force = 2; + + while ((c = getopt_long(argc, argv, "pfwdnih", options, NULL)) >= 0) + switch (c) { + + case ARG_HELP: + halt_help(); + return 0; + + case ARG_HALT: + arg_action = ACTION_HALT; + break; + + case 'p': + if (arg_action != ACTION_REBOOT) + arg_action = ACTION_POWEROFF; + break; + + case ARG_REBOOT: + arg_action = ACTION_REBOOT; + break; + + case 'f': + arg_force = 2; + break; + + case 'w': + arg_dry = true; + break; + + case 'd': + arg_no_wtmp = true; + break; + + case 'n': + arg_no_sync = true; + break; + + case ARG_NO_WALL: + arg_no_wall = true; + break; + + case 'i': + case 'h': + /* Compatibility nops */ + break; + + case '?': + return -EINVAL; + + default: + assert_not_reached("Unhandled option"); + } + + if (arg_action == ACTION_REBOOT && (argc == optind || argc == optind + 1)) { + r = update_reboot_parameter_and_warn(argc == optind + 1 ? argv[optind] : NULL); + if (r < 0) + return r; + } else if (optind < argc) { + log_error("Too many arguments."); + return -EINVAL; + } + + return 1; +} + +static int parse_shutdown_time_spec(const char *t, usec_t *_u) { + assert(t); + assert(_u); + + if (streq(t, "now")) + *_u = 0; + else if (!strchr(t, ':')) { + uint64_t u; + + if (safe_atou64(t, &u) < 0) + return -EINVAL; + + *_u = now(CLOCK_REALTIME) + USEC_PER_MINUTE * u; + } else { + char *e = NULL; + long hour, minute; + struct tm tm = {}; + time_t s; + usec_t n; + + errno = 0; + hour = strtol(t, &e, 10); + if (errno > 0 || *e != ':' || hour < 0 || hour > 23) + return -EINVAL; + + minute = strtol(e+1, &e, 10); + if (errno > 0 || *e != 0 || minute < 0 || minute > 59) + return -EINVAL; + + n = now(CLOCK_REALTIME); + s = (time_t) (n / USEC_PER_SEC); + + assert_se(localtime_r(&s, &tm)); + + tm.tm_hour = (int) hour; + tm.tm_min = (int) minute; + tm.tm_sec = 0; + + assert_se(s = mktime(&tm)); + + *_u = (usec_t) s * USEC_PER_SEC; + + while (*_u <= n) + *_u += USEC_PER_DAY; + } + + return 0; +} + +static int shutdown_parse_argv(int argc, char *argv[]) { + + enum { + ARG_HELP = 0x100, + ARG_NO_WALL + }; + + static const struct option options[] = { + { "help", no_argument, NULL, ARG_HELP }, + { "halt", no_argument, NULL, 'H' }, + { "poweroff", no_argument, NULL, 'P' }, + { "reboot", no_argument, NULL, 'r' }, + { "kexec", no_argument, NULL, 'K' }, /* not documented extension */ + { "no-wall", no_argument, NULL, ARG_NO_WALL }, + {} + }; + + char **wall = NULL; + int c, r; + + assert(argc >= 0); + assert(argv); + + while ((c = getopt_long(argc, argv, "HPrhkKtafFc", options, NULL)) >= 0) + switch (c) { + + case ARG_HELP: + shutdown_help(); + return 0; + + case 'H': + arg_action = ACTION_HALT; + break; + + case 'P': + arg_action = ACTION_POWEROFF; + break; + + case 'r': + if (kexec_loaded()) + arg_action = ACTION_KEXEC; + else + arg_action = ACTION_REBOOT; + break; + + case 'K': + arg_action = ACTION_KEXEC; + break; + + case 'h': + if (arg_action != ACTION_HALT) + arg_action = ACTION_POWEROFF; + break; + + case 'k': + arg_dry = true; + break; + + case ARG_NO_WALL: + arg_no_wall = true; + break; + + case 't': + case 'a': + case 'f': + case 'F': + /* Compatibility nops */ + break; + + case 'c': + arg_action = ACTION_CANCEL_SHUTDOWN; + break; + + case '?': + return -EINVAL; + + default: + assert_not_reached("Unhandled option"); + } + + if (argc > optind && arg_action != ACTION_CANCEL_SHUTDOWN) { + r = parse_shutdown_time_spec(argv[optind], &arg_when); + if (r < 0) { + log_error("Failed to parse time specification: %s", argv[optind]); + return r; + } + } else + arg_when = now(CLOCK_REALTIME) + USEC_PER_MINUTE; + + if (argc > optind && arg_action == ACTION_CANCEL_SHUTDOWN) + /* No time argument for shutdown cancel */ + wall = argv + optind; + else if (argc > optind + 1) + /* We skip the time argument */ + wall = argv + optind + 1; + + if (wall) { + arg_wall = strv_copy(wall); + if (!arg_wall) + return log_oom(); + } + + optind = argc; + + return 1; +} + +static int telinit_parse_argv(int argc, char *argv[]) { + + enum { + ARG_HELP = 0x100, + ARG_NO_WALL + }; + + static const struct option options[] = { + { "help", no_argument, NULL, ARG_HELP }, + { "no-wall", no_argument, NULL, ARG_NO_WALL }, + {} + }; + + static const struct { + char from; + enum action to; + } table[] = { + { '0', ACTION_POWEROFF }, + { '6', ACTION_REBOOT }, + { '1', ACTION_RESCUE }, + { '2', ACTION_RUNLEVEL2 }, + { '3', ACTION_RUNLEVEL3 }, + { '4', ACTION_RUNLEVEL4 }, + { '5', ACTION_RUNLEVEL5 }, + { 's', ACTION_RESCUE }, + { 'S', ACTION_RESCUE }, + { 'q', ACTION_RELOAD }, + { 'Q', ACTION_RELOAD }, + { 'u', ACTION_REEXEC }, + { 'U', ACTION_REEXEC } + }; + + unsigned i; + int c; + + assert(argc >= 0); + assert(argv); + + while ((c = getopt_long(argc, argv, "", options, NULL)) >= 0) + switch (c) { + + case ARG_HELP: + telinit_help(); + return 0; + + case ARG_NO_WALL: + arg_no_wall = true; + break; + + case '?': + return -EINVAL; + + default: + assert_not_reached("Unhandled option"); + } + + if (optind >= argc) { + log_error("%s: required argument missing.", program_invocation_short_name); + return -EINVAL; + } + + if (optind + 1 < argc) { + log_error("Too many arguments."); + return -EINVAL; + } + + if (strlen(argv[optind]) != 1) { + log_error("Expected single character argument."); + return -EINVAL; + } + + for (i = 0; i < ELEMENTSOF(table); i++) + if (table[i].from == argv[optind][0]) + break; + + if (i >= ELEMENTSOF(table)) { + log_error("Unknown command '%s'.", argv[optind]); + return -EINVAL; + } + + arg_action = table[i].to; + + optind++; + + return 1; +} + +static int runlevel_parse_argv(int argc, char *argv[]) { + + enum { + ARG_HELP = 0x100, + }; + + static const struct option options[] = { + { "help", no_argument, NULL, ARG_HELP }, + {} + }; + + int c; + + assert(argc >= 0); + assert(argv); + + while ((c = getopt_long(argc, argv, "", options, NULL)) >= 0) + switch (c) { + + case ARG_HELP: + runlevel_help(); + return 0; + + case '?': + return -EINVAL; + + default: + assert_not_reached("Unhandled option"); + } + + if (optind < argc) { + log_error("Too many arguments."); + return -EINVAL; + } + + return 1; +} + +static int parse_argv(int argc, char *argv[]) { + assert(argc >= 0); + assert(argv); + + if (program_invocation_short_name) { + + if (strstr(program_invocation_short_name, "halt")) { + arg_action = ACTION_HALT; + return halt_parse_argv(argc, argv); + } else if (strstr(program_invocation_short_name, "poweroff")) { + arg_action = ACTION_POWEROFF; + return halt_parse_argv(argc, argv); + } else if (strstr(program_invocation_short_name, "reboot")) { + if (kexec_loaded()) + arg_action = ACTION_KEXEC; + else + arg_action = ACTION_REBOOT; + return halt_parse_argv(argc, argv); + } else if (strstr(program_invocation_short_name, "shutdown")) { + arg_action = ACTION_POWEROFF; + return shutdown_parse_argv(argc, argv); + } else if (strstr(program_invocation_short_name, "init")) { + + if (sd_booted() > 0) { + arg_action = _ACTION_INVALID; + return telinit_parse_argv(argc, argv); + } else { + /* Hmm, so some other init system is + * running, we need to forward this + * request to it. For now we simply + * guess that it is Upstart. */ + + execv(TELINIT, argv); + + log_error("Couldn't find an alternative telinit implementation to spawn."); + return -EIO; + } + + } else if (strstr(program_invocation_short_name, "runlevel")) { + arg_action = ACTION_RUNLEVEL; + return runlevel_parse_argv(argc, argv); + } + } + + arg_action = ACTION_SYSTEMCTL; + return systemctl_parse_argv(argc, argv); +} + +#ifdef HAVE_SYSV_COMPAT +_pure_ static int action_to_runlevel(void) { + + static const char table[_ACTION_MAX] = { + [ACTION_HALT] = '0', + [ACTION_POWEROFF] = '0', + [ACTION_REBOOT] = '6', + [ACTION_RUNLEVEL2] = '2', + [ACTION_RUNLEVEL3] = '3', + [ACTION_RUNLEVEL4] = '4', + [ACTION_RUNLEVEL5] = '5', + [ACTION_RESCUE] = '1' + }; + + assert(arg_action < _ACTION_MAX); + + return table[arg_action]; +} +#endif + +static int talk_initctl(void) { +#ifdef HAVE_SYSV_COMPAT + struct init_request request = { + .magic = INIT_MAGIC, + .sleeptime = 0, + .cmd = INIT_CMD_RUNLVL + }; + + _cleanup_close_ int fd = -1; + char rl; + int r; + + rl = action_to_runlevel(); + if (!rl) + return 0; + + request.runlevel = rl; + + fd = open(INIT_FIFO, O_WRONLY|O_NDELAY|O_CLOEXEC|O_NOCTTY); + if (fd < 0) { + if (errno == ENOENT) + return 0; + + return log_error_errno(errno, "Failed to open "INIT_FIFO": %m"); + } + + r = loop_write(fd, &request, sizeof(request), false); + if (r < 0) + return log_error_errno(r, "Failed to write to "INIT_FIFO": %m"); + + return 1; +#else + return 0; +#endif +} + +static int systemctl_main(int argc, char *argv[]) { + + static const Verb verbs[] = { + { "list-units", VERB_ANY, VERB_ANY, VERB_DEFAULT|VERB_NOCHROOT, list_units }, + { "list-unit-files", VERB_ANY, VERB_ANY, 0, list_unit_files }, + { "list-sockets", VERB_ANY, VERB_ANY, VERB_NOCHROOT, list_sockets }, + { "list-timers", VERB_ANY, VERB_ANY, VERB_NOCHROOT, list_timers }, + { "list-jobs", VERB_ANY, VERB_ANY, VERB_NOCHROOT, list_jobs }, + { "list-machines", VERB_ANY, VERB_ANY, VERB_NOCHROOT, list_machines }, + { "clear-jobs", VERB_ANY, 1, VERB_NOCHROOT, trivial_method }, + { "cancel", VERB_ANY, VERB_ANY, VERB_NOCHROOT, cancel_job }, + { "start", 2, VERB_ANY, VERB_NOCHROOT, start_unit }, + { "stop", 2, VERB_ANY, VERB_NOCHROOT, start_unit }, + { "condstop", 2, VERB_ANY, VERB_NOCHROOT, start_unit }, /* For compatibility with ALTLinux */ + { "reload", 2, VERB_ANY, VERB_NOCHROOT, start_unit }, + { "restart", 2, VERB_ANY, VERB_NOCHROOT, start_unit }, + { "try-restart", 2, VERB_ANY, VERB_NOCHROOT, start_unit }, + { "reload-or-restart", 2, VERB_ANY, VERB_NOCHROOT, start_unit }, + { "reload-or-try-restart", 2, VERB_ANY, VERB_NOCHROOT, start_unit }, /* For compatbility with old systemctl <= 228 */ + { "try-reload-or-restart", 2, VERB_ANY, VERB_NOCHROOT, start_unit }, + { "force-reload", 2, VERB_ANY, VERB_NOCHROOT, start_unit }, /* For compatibility with SysV */ + { "condreload", 2, VERB_ANY, VERB_NOCHROOT, start_unit }, /* For compatibility with ALTLinux */ + { "condrestart", 2, VERB_ANY, VERB_NOCHROOT, start_unit }, /* For compatibility with RH */ + { "isolate", 2, 2, VERB_NOCHROOT, start_unit }, + { "kill", 2, VERB_ANY, VERB_NOCHROOT, kill_unit }, + { "is-active", 2, VERB_ANY, VERB_NOCHROOT, check_unit_active }, + { "check", 2, VERB_ANY, VERB_NOCHROOT, check_unit_active }, + { "is-failed", 2, VERB_ANY, VERB_NOCHROOT, check_unit_failed }, + { "show", VERB_ANY, VERB_ANY, VERB_NOCHROOT, show }, + { "cat", 2, VERB_ANY, VERB_NOCHROOT, cat }, + { "status", VERB_ANY, VERB_ANY, VERB_NOCHROOT, show }, + { "help", VERB_ANY, VERB_ANY, VERB_NOCHROOT, show }, + { "daemon-reload", VERB_ANY, 1, VERB_NOCHROOT, daemon_reload }, + { "daemon-reexec", VERB_ANY, 1, VERB_NOCHROOT, daemon_reload }, + { "show-environment", VERB_ANY, 1, VERB_NOCHROOT, show_environment }, + { "set-environment", 2, VERB_ANY, VERB_NOCHROOT, set_environment }, + { "unset-environment", 2, VERB_ANY, VERB_NOCHROOT, set_environment }, + { "import-environment", VERB_ANY, VERB_ANY, VERB_NOCHROOT, import_environment }, + { "halt", VERB_ANY, 1, VERB_NOCHROOT, start_system_special }, + { "poweroff", VERB_ANY, 1, VERB_NOCHROOT, start_system_special }, + { "reboot", VERB_ANY, 2, VERB_NOCHROOT, start_system_special }, + { "kexec", VERB_ANY, 1, VERB_NOCHROOT, start_system_special }, + { "suspend", VERB_ANY, 1, VERB_NOCHROOT, start_system_special }, + { "hibernate", VERB_ANY, 1, VERB_NOCHROOT, start_system_special }, + { "hybrid-sleep", VERB_ANY, 1, VERB_NOCHROOT, start_system_special }, + { "default", VERB_ANY, 1, VERB_NOCHROOT, start_special }, + { "rescue", VERB_ANY, 1, VERB_NOCHROOT, start_system_special }, + { "emergency", VERB_ANY, 1, VERB_NOCHROOT, start_system_special }, + { "exit", VERB_ANY, 2, VERB_NOCHROOT, start_special }, + { "reset-failed", VERB_ANY, VERB_ANY, VERB_NOCHROOT, reset_failed }, + { "enable", 2, VERB_ANY, 0, enable_unit }, + { "disable", 2, VERB_ANY, 0, enable_unit }, + { "is-enabled", 2, VERB_ANY, 0, unit_is_enabled }, + { "reenable", 2, VERB_ANY, 0, enable_unit }, + { "preset", 2, VERB_ANY, 0, enable_unit }, + { "preset-all", VERB_ANY, 1, 0, preset_all }, + { "mask", 2, VERB_ANY, 0, enable_unit }, + { "unmask", 2, VERB_ANY, 0, enable_unit }, + { "link", 2, VERB_ANY, 0, enable_unit }, + { "revert", 2, VERB_ANY, 0, enable_unit }, + { "switch-root", 2, VERB_ANY, VERB_NOCHROOT, switch_root }, + { "list-dependencies", VERB_ANY, 2, VERB_NOCHROOT, list_dependencies }, + { "set-default", 2, 2, 0, set_default }, + { "get-default", VERB_ANY, 1, 0, get_default }, + { "set-property", 3, VERB_ANY, VERB_NOCHROOT, set_property }, + { "is-system-running", VERB_ANY, 1, 0, is_system_running }, + { "add-wants", 3, VERB_ANY, 0, add_dependency }, + { "add-requires", 3, VERB_ANY, 0, add_dependency }, + { "edit", 2, VERB_ANY, VERB_NOCHROOT, edit }, + {} + }; + + return dispatch_verb(argc, argv, verbs, NULL); +} + +static int reload_with_fallback(void) { + + /* First, try systemd via D-Bus. */ + if (daemon_reload(0, NULL, NULL) >= 0) + return 0; + + /* Nothing else worked, so let's try signals */ + assert(IN_SET(arg_action, ACTION_RELOAD, ACTION_REEXEC)); + + if (kill(1, arg_action == ACTION_RELOAD ? SIGHUP : SIGTERM) < 0) + return log_error_errno(errno, "kill() failed: %m"); + + return 0; +} + +static int start_with_fallback(void) { + + /* First, try systemd via D-Bus. */ + if (start_unit(0, NULL, NULL) >= 0) + return 0; + + /* Nothing else worked, so let's try /dev/initctl */ + if (talk_initctl() > 0) + return 0; + + log_error("Failed to talk to init daemon."); + return -EIO; +} + +static int halt_now(enum action a) { + int r; + + /* The kernel will automaticall flush ATA disks and suchlike + * on reboot(), but the file systems need to be synce'd + * explicitly in advance. */ + if (!arg_no_sync) + (void) sync(); + + /* Make sure C-A-D is handled by the kernel from this point + * on... */ + (void) reboot(RB_ENABLE_CAD); + + switch (a) { + + case ACTION_HALT: + log_info("Halting."); + (void) reboot(RB_HALT_SYSTEM); + return -errno; + + case ACTION_POWEROFF: + log_info("Powering off."); + (void) reboot(RB_POWER_OFF); + return -errno; + + case ACTION_KEXEC: + case ACTION_REBOOT: { + _cleanup_free_ char *param = NULL; + + r = read_one_line_file("/run/systemd/reboot-param", ¶m); + if (r < 0) + log_warning_errno(r, "Failed to read reboot parameter file: %m"); + + if (!isempty(param)) { + log_info("Rebooting with argument '%s'.", param); + (void) syscall(SYS_reboot, LINUX_REBOOT_MAGIC1, LINUX_REBOOT_MAGIC2, LINUX_REBOOT_CMD_RESTART2, param); + log_warning_errno(errno, "Failed to reboot with parameter, retrying without: %m"); + } + + log_info("Rebooting."); + (void) reboot(RB_AUTOBOOT); + return -errno; + } + + default: + assert_not_reached("Unknown action."); + } +} + +static int logind_schedule_shutdown(void) { + +#ifdef HAVE_LOGIND + _cleanup_(sd_bus_error_free) sd_bus_error error = SD_BUS_ERROR_NULL; + char date[FORMAT_TIMESTAMP_MAX]; + const char *action; + sd_bus *bus; + int r; + + r = acquire_bus(BUS_FULL, &bus); + if (r < 0) + return r; + + switch (arg_action) { + case ACTION_HALT: + action = "halt"; + break; + case ACTION_POWEROFF: + action = "poweroff"; + break; + case ACTION_KEXEC: + action = "kexec"; + break; + case ACTION_EXIT: + action = "exit"; + break; + case ACTION_REBOOT: + default: + action = "reboot"; + break; + } + + if (arg_dry) + action = strjoina("dry-", action); + + (void) logind_set_wall_message(); + + r = sd_bus_call_method( + bus, + "org.freedesktop.login1", + "/org/freedesktop/login1", + "org.freedesktop.login1.Manager", + "ScheduleShutdown", + &error, + NULL, + "st", + action, + arg_when); + if (r < 0) + return log_warning_errno(r, "Failed to call ScheduleShutdown in logind, proceeding with immediate shutdown: %s", bus_error_message(&error, r)); + + log_info("Shutdown scheduled for %s, use 'shutdown -c' to cancel.", format_timestamp(date, sizeof(date), arg_when)); + return 0; +#else + log_error("Cannot schedule shutdown without logind support, proceeding with immediate shutdown."); + return -ENOSYS; +#endif +} + +static int halt_main(void) { + int r; + + r = logind_check_inhibitors(arg_action); + if (r < 0) + return r; + + if (arg_when > 0) + return logind_schedule_shutdown(); + + if (geteuid() != 0) { + if (arg_dry || arg_force > 0) { + log_error("Must be root."); + return -EPERM; + } + + /* Try logind if we are a normal user and no special + * mode applies. Maybe PolicyKit allows us to shutdown + * the machine. */ + if (IN_SET(arg_action, ACTION_POWEROFF, ACTION_REBOOT)) { + r = logind_reboot(arg_action); + if (r >= 0) + return r; + if (IN_SET(r, -EOPNOTSUPP, -EINPROGRESS)) + /* requested operation is not + * supported on the local system or + * already in progress */ + return r; + /* on all other errors, try low-level operation */ + } + } + + if (!arg_dry && !arg_force) + return start_with_fallback(); + + assert(geteuid() == 0); + + if (!arg_no_wtmp) { + if (sd_booted() > 0) + log_debug("Not writing utmp record, assuming that systemd-update-utmp is used."); + else { + r = utmp_put_shutdown(); + if (r < 0) + log_warning_errno(r, "Failed to write utmp record: %m"); + } + } + + if (arg_dry) + return 0; + + r = halt_now(arg_action); + return log_error_errno(r, "Failed to reboot: %m"); +} + +static int runlevel_main(void) { + int r, runlevel, previous; + + r = utmp_get_runlevel(&runlevel, &previous); + if (r < 0) { + puts("unknown"); + return r; + } + + printf("%c %c\n", + previous <= 0 ? 'N' : previous, + runlevel <= 0 ? 'N' : runlevel); + + return 0; +} + +static int logind_cancel_shutdown(void) { +#ifdef HAVE_LOGIND + _cleanup_(sd_bus_error_free) sd_bus_error error = SD_BUS_ERROR_NULL; + sd_bus *bus; + int r; + + r = acquire_bus(BUS_FULL, &bus); + if (r < 0) + return r; + + (void) logind_set_wall_message(); + + r = sd_bus_call_method( + bus, + "org.freedesktop.login1", + "/org/freedesktop/login1", + "org.freedesktop.login1.Manager", + "CancelScheduledShutdown", + &error, + NULL, NULL); + if (r < 0) + return log_warning_errno(r, "Failed to talk to logind, shutdown hasn't been cancelled: %s", bus_error_message(&error, r)); + + return 0; +#else + log_error("Not compiled with logind support, cannot cancel scheduled shutdowns."); + return -ENOSYS; +#endif +} + +int main(int argc, char*argv[]) { + int r; + + setlocale(LC_ALL, ""); + log_parse_environment(); + log_open(); + sigbus_install(); + + /* Explicitly not on_tty() to avoid setting cached value. + * This becomes relevant for piping output which might be + * ellipsized. */ + original_stdout_is_tty = isatty(STDOUT_FILENO); + + r = parse_argv(argc, argv); + if (r <= 0) + goto finish; + + if (arg_action != ACTION_SYSTEMCTL && running_in_chroot() > 0) { + log_info("Running in chroot, ignoring request."); + r = 0; + goto finish; + } + + /* systemctl_main() will print an error message for the bus + * connection, but only if it needs to */ + + switch (arg_action) { + + case ACTION_SYSTEMCTL: + r = systemctl_main(argc, argv); + break; + + case ACTION_HALT: + case ACTION_POWEROFF: + case ACTION_REBOOT: + case ACTION_KEXEC: + r = halt_main(); + break; + + case ACTION_RUNLEVEL2: + case ACTION_RUNLEVEL3: + case ACTION_RUNLEVEL4: + case ACTION_RUNLEVEL5: + case ACTION_RESCUE: + case ACTION_EMERGENCY: + case ACTION_DEFAULT: + r = start_with_fallback(); + break; + + case ACTION_RELOAD: + case ACTION_REEXEC: + r = reload_with_fallback(); + break; + + case ACTION_CANCEL_SHUTDOWN: + r = logind_cancel_shutdown(); + break; + + case ACTION_RUNLEVEL: + r = runlevel_main(); + break; + + case _ACTION_INVALID: + default: + assert_not_reached("Unknown action"); + } + +finish: + release_busses(); + + pager_close(); + ask_password_agent_close(); + polkit_agent_close(); + + strv_free(arg_types); + strv_free(arg_states); + strv_free(arg_properties); + + strv_free(arg_wall); + free(arg_root); + + /* Note that we return r here, not EXIT_SUCCESS, so that we can implement the LSB-like return codes */ + return r < 0 ? EXIT_FAILURE : r; +} diff --git a/src/grp-system/systemctl/systemctl.completion.bash.in b/src/grp-system/systemctl/systemctl.completion.bash.in new file mode 100644 index 0000000000..6f2b3f122c --- /dev/null +++ b/src/grp-system/systemctl/systemctl.completion.bash.in @@ -0,0 +1,284 @@ +# systemctl(1) completion -*- shell-script -*- +# +# This file is part of systemd. +# +# Copyright 2010 Ran Benita +# +# systemd is free software; you can redistribute it and/or modify it +# under the terms of the GNU Lesser General Public License as published by +# the Free Software Foundation; either version 2.1 of the License, or +# (at your option) any later version. +# +# systemd is distributed in the hope that it will be useful, but +# WITHOUT ANY WARRANTY; without even the implied warranty of +# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU +# General Public License for more details. +# +# You should have received a copy of the GNU Lesser General Public License +# along with systemd; If not, see . + +__systemctl() { + local mode=$1; shift 1 + systemctl $mode --full --no-legend "$@" +} + +__systemd_properties() { + local mode=$1 + { __systemctl $mode show --all; + @rootlibexecdir@/systemd --dump-configuration-items; } | + while IFS='=' read -r key value; do + [[ $value ]] && echo "$key" + done +} + +__contains_word () { + local w word=$1; shift + for w in "$@"; do + [[ $w = "$word" ]] && return + done +} + +__filter_units_by_property () { + local mode=$1 property=$2 value=$3 ; shift 3 + local units=("$@") + local props + IFS=$'\n' read -rd '' -a props < \ + <(__systemctl $mode show --property "$property" -- "${units[@]}") + for ((i=0; $i < ${#units[*]}; i++)); do + if [[ "${props[i]}" = "$property=$value" ]]; then + echo " ${units[i]}" + fi + done +} + +__get_all_units () { { __systemctl $1 list-unit-files; __systemctl $1 list-units --all; } \ + | { while read -r a b; do [[ $a =~ @\. ]] || echo " $a"; done; }; } +__get_template_names () { __systemctl $1 list-unit-files \ + | { while read -r a b; do [[ $a =~ @\. ]] && echo " ${a%%@.*}@"; done; }; } + +__get_active_units () { __systemctl $1 list-units \ + | { while read -r a b; do echo " $a"; done; }; } +__get_startable_units () { + # find startable inactive units + __filter_units_by_property $mode ActiveState inactive $( + __filter_units_by_property $mode CanStart yes $( + __systemctl $mode list-unit-files --state enabled,disabled,static | \ + { while read -r a b; do [[ $a =~ @\. ]] || echo " $a"; done; } + __systemctl $mode list-units --state inactive,failed | \ + { while read -r a b; do echo " $a"; done; } )) +} +__get_restartable_units () { + # filter out masked and not-found + __filter_units_by_property $mode CanStart yes $( + __systemctl $mode list-unit-files --state enabled,disabled,static | \ + { while read -r a b; do [[ $a =~ @\. ]] || echo " $a"; done; } + __systemctl $mode list-units | \ + { while read -r a b; do echo " $a"; done; } ) +} +__get_failed_units () { __systemctl $1 list-units \ + | { while read -r a b c d; do [[ $c == "failed" ]] && echo " $a"; done; }; } +__get_enabled_units () { __systemctl $1 list-unit-files \ + | { while read -r a b c ; do [[ $b == "enabled" ]] && echo " $a"; done; }; } +__get_disabled_units () { __systemctl $1 list-unit-files \ + | { while read -r a b c ; do [[ $b == "disabled" ]] && echo " $a"; done; }; } +__get_masked_units () { __systemctl $1 list-unit-files \ + | { while read -r a b c ; do [[ $b == "masked" ]] && echo " $a"; done; }; } +__get_all_unit_files () { { __systemctl $1 list-unit-files; } | { while read -r a b; do echo " $a"; done; }; } + +__get_machines() { + local a b + { machinectl list-images --no-legend --no-pager; machinectl list --no-legend --no-pager; } | \ + { while read a b; do echo " $a"; done; } +} + +_systemctl () { + local cur=${COMP_WORDS[COMP_CWORD]} prev=${COMP_WORDS[COMP_CWORD-1]} + local i verb comps mode + + local -A OPTS=( + [STANDALONE]='--all -a --reverse --after --before --defaults --force -f --full -l --global + --help -h --no-ask-password --no-block --no-legend --no-pager --no-reload --no-wall + --quiet -q --privileged -P --system --user --version --runtime --recursive -r --firmware-setup + --show-types -i --ignore-inhibitors --plain' + [ARG]='--host -H --kill-who --property -p --signal -s --type -t --state --job-mode --root + --preset-mode -n --lines -o --output -M --machine' + ) + + if __contains_word "--user" ${COMP_WORDS[*]}; then + mode=--user + elif __contains_word "--global" ${COMP_WORDS[*]}; then + mode=--user + else + mode=--system + fi + + if __contains_word "$prev" ${OPTS[ARG]}; then + case $prev in + --signal|-s) + _signals + return + ;; + --type|-t) + comps=$(__systemctl $mode -t help) + ;; + --state) + comps=$(__systemctl $mode --state=help) + ;; + --job-mode) + comps='fail replace replace-irreversibly isolate + ignore-dependencies ignore-requirements flush' + ;; + --kill-who) + comps='all control main' + ;; + --root) + comps=$(compgen -A directory -- "$cur" ) + compopt -o filenames + ;; + --host|-H) + comps=$(compgen -A hostname) + ;; + --property|-p) + comps=$(__systemd_properties $mode) + ;; + --preset-mode) + comps='full enable-only disable-only' + ;; + --output|-o) + comps='short short-iso short-precise short-monotonic verbose export json + json-pretty json-sse cat' + ;; + --machine|-M) + comps=$( __get_machines ) + ;; + esac + COMPREPLY=( $(compgen -W '$comps' -- "$cur") ) + return 0 + fi + + if [[ "$cur" = -* ]]; then + COMPREPLY=( $(compgen -W '${OPTS[*]}' -- "$cur") ) + return 0 + fi + + local -A VERBS=( + [ALL_UNITS]='is-active is-failed is-enabled status show cat mask preset help list-dependencies edit set-property' + [ENABLED_UNITS]='disable' + [DISABLED_UNITS]='enable' + [REENABLABLE_UNITS]='reenable' + [FAILED_UNITS]='reset-failed' + [STARTABLE_UNITS]='start' + [STOPPABLE_UNITS]='stop condstop kill try-restart condrestart' + [ISOLATABLE_UNITS]='isolate' + [RELOADABLE_UNITS]='reload condreload try-reload-or-restart force-reload' + [RESTARTABLE_UNITS]='restart reload-or-restart' + [TARGET_AND_UNITS]='add-wants add-requires' + [MASKED_UNITS]='unmask' + [JOBS]='cancel' + [ENVS]='set-environment unset-environment' + [STANDALONE]='daemon-reexec daemon-reload default + emergency exit halt hibernate hybrid-sleep kexec list-jobs + list-sockets list-timers list-units list-unit-files poweroff + reboot rescue show-environment suspend get-default + is-system-running' + [FILE]='link switch-root' + [TARGETS]='set-default' + ) + + for ((i=0; i < COMP_CWORD; i++)); do + if __contains_word "${COMP_WORDS[i]}" ${VERBS[*]} && + ! __contains_word "${COMP_WORDS[i-1]}" ${OPTS[ARG]}; then + verb=${COMP_WORDS[i]} + break + fi + done + + if [[ -z $verb ]]; then + comps="${VERBS[*]}" + + elif __contains_word "$verb" ${VERBS[ALL_UNITS]}; then + comps=$( __get_all_units $mode ) + compopt -o filenames + + elif __contains_word "$verb" ${VERBS[ENABLED_UNITS]}; then + comps=$( __get_enabled_units $mode ) + compopt -o filenames + + elif __contains_word "$verb" ${VERBS[DISABLED_UNITS]}; then + comps=$( __get_disabled_units $mode; + __get_template_names $mode) + compopt -o filenames + + elif __contains_word "$verb" ${VERBS[REENABLABLE_UNITS]}; then + comps=$( __get_disabled_units $mode; + __get_enabled_units $mode; + __get_template_names $mode) + compopt -o filenames + + elif __contains_word "$verb" ${VERBS[STARTABLE_UNITS]}; then + comps=$( __get_startable_units $mode; + __get_template_names $mode) + compopt -o filenames + + elif __contains_word "$verb" ${VERBS[RESTARTABLE_UNITS]}; then + comps=$( __get_restartable_units $mode; + __get_template_names $mode) + compopt -o filenames + + elif __contains_word "$verb" ${VERBS[STOPPABLE_UNITS]}; then + comps=$( __filter_units_by_property $mode CanStop yes \ + $( __get_active_units $mode ) ) + compopt -o filenames + + elif __contains_word "$verb" ${VERBS[RELOADABLE_UNITS]}; then + comps=$( __filter_units_by_property $mode CanReload yes \ + $( __get_active_units $mode ) ) + compopt -o filenames + + elif __contains_word "$verb" ${VERBS[ISOLATABLE_UNITS]}; then + comps=$( __filter_units_by_property $mode AllowIsolate yes \ + $( __get_all_units $mode ) ) + compopt -o filenames + + elif __contains_word "$verb" ${VERBS[FAILED_UNITS]}; then + comps=$( __get_failed_units $mode ) + compopt -o filenames + + elif __contains_word "$verb" ${VERBS[MASKED_UNITS]}; then + comps=$( __get_masked_units $mode ) + compopt -o filenames + + elif __contains_word "$verb" ${VERBS[TARGET_AND_UNITS]}; then + if __contains_word "$prev" ${VERBS[TARGET_AND_UNITS]} \ + || __contains_word "$prev" ${OPTS[STANDALONE]}; then + comps=$( __systemctl $mode list-unit-files --type target --all \ + | { while read -r a b; do echo " $a"; done; } ) + else + comps=$( __get_all_unit_files $mode ) + fi + compopt -o filenames + + elif __contains_word "$verb" ${VERBS[STANDALONE]}; then + comps='' + + elif __contains_word "$verb" ${VERBS[JOBS]}; then + comps=$( __systemctl $mode list-jobs | { while read -r a b; do echo " $a"; done; } ) + + elif __contains_word "$verb" ${VERBS[ENVS]}; then + comps=$( __systemctl $mode show-environment \ + | while read -r line; do echo " ${line%%=*}=";done ) + compopt -o nospace + + elif __contains_word "$verb" ${VERBS[FILE]}; then + comps=$( compgen -A file -- "$cur" ) + compopt -o filenames + elif __contains_word "$verb" ${VERBS[TARGETS]}; then + comps=$( __systemctl $mode list-unit-files --type target --full --all \ + | { while read -r a b; do echo " $a"; done; } ) + fi + + COMPREPLY=( $(compgen -o filenames -W '$comps' -- "$cur") ) + return 0 +} + +complete -F _systemctl systemctl diff --git a/src/grp-system/systemctl/systemctl.completion.zsh.in b/src/grp-system/systemctl/systemctl.completion.zsh.in new file mode 100644 index 0000000000..44c31b7833 --- /dev/null +++ b/src/grp-system/systemctl/systemctl.completion.zsh.in @@ -0,0 +1,391 @@ +#compdef systemctl + +(( $+functions[_systemctl_command] )) || _systemctl_command() +{ + local -a _systemctl_cmds + _systemctl_cmds=( + "list-sockets:List sockets" + "list-timers:List timers" + "list-units:List units" + "start:Start (activate) one or more units" + "stop:Stop (deactivate) one or more units" + "reload:Reload one or more units" + "restart:Start or restart one or more units" + "condrestart:Restart one or more units if active" + "try-restart:Restart one or more units if active" + "reload-or-restart:Reload one or more units if possible, otherwise start or restart" + "force-reload:Reload one or more units if possible, otherwise restart if active" + "hibernate:Hibernate the system" + "hybrid-sleep:Hibernate and suspend the system" + "try-reload-or-restart:Reload one or more units if possible, otherwise restart if active" + "isolate:Start one unit and stop all others" + "kill:Send signal to processes of a unit" + "is-active:Check whether units are active" + "is-failed:Check whether units are failed" + "status:Show runtime status of one or more units" + "show:Show properties of one or more units/jobs or the manager" + "cat:Show the source unit files and drop-ins" + "reset-failed:Reset failed state for all, one, or more units" + "list-unit-files:List installed unit files" + "enable:Enable one or more unit files" + "disable:Disable one or more unit files" + "reenable:Reenable one or more unit files" + "preset:Enable/disable one or more unit files based on preset configuration" + "set-default:Set the default target" + "get-default:Query the default target" + "edit:Edit one or more unit files" + "is-system-running:Query overall status of the system" + "help:Show documentation for specified units" + "list-dependencies:Show unit dependency tree" + "mask:Mask one or more units" + "unmask:Unmask one or more units" + "link:Link one or more units files into the search path" + "is-enabled:Check whether unit files are enabled" + "list-jobs:List jobs" + "cancel:Cancel all, one, or more jobs" + "show-environment:Dump environment" + "set-environment:Set one or more environment variables" + "unset-environment:Unset one or more environment variables" + "daemon-reload:Reload systemd manager configuration" + "daemon-reexec:Reexecute systemd manager" + "default:Enter system default mode" + "rescue:Enter system rescue mode" + "emergency:Enter system emergency mode" + "halt:Shut down and halt the system" + "suspend:Suspend the system" + "poweroff:Shut down and power-off the system" + "reboot:Shut down and reboot the system" + "kexec:Shut down and reboot the system with kexec" + "exit:Ask for user instance termination" + "switch-root:Change root directory" + ) + + if (( CURRENT == 1 )); then + _describe -t commands 'systemctl command' _systemctl_cmds || compadd "$@" + else + local curcontext="$curcontext" expl + + cmd="${${_systemctl_cmds[(r)$words[1]:*]%%:*}}" + # Deal with any aliases + case $cmd in + condrestart) cmd="try-restart";; + force-reload) cmd="try-reload-or-restart";; + esac + + if (( $#cmd )); then + curcontext="${curcontext%:*:*}:systemctl-${cmd}:" + + local update_policy + zstyle -s ":completion:${curcontext}:" cache-policy update_policy + if [[ -z "$update_policy" ]]; then + zstyle ":completion:${curcontext}:" cache-policy _systemctl_caching_policy + fi + + _call_function ret _systemctl_$cmd || _message 'no more arguments' + else + _message "unknown systemctl command: $words[1]" + fi + return ret + fi +} + +__systemctl() +{ + systemctl $_sys_service_mgr --full --no-legend --no-pager "$@" +} + + +# Fills the unit list +_systemctl_all_units() +{ + if ( [[ ${+_sys_all_units} -eq 0 ]] || _cache_invalid SYS_ALL_UNITS ) && + ! _retrieve_cache SYS_ALL_UNITS; + then + _sys_all_units=( ${${(f)"$(__systemctl list-units --all)"}%% *} ) + _store_cache SYS_ALL_UNITS _sys_all_units + fi +} + +# Fills the unit list including all file units +_systemctl_really_all_units() +{ + local -a all_unit_files; + local -a really_all_units; + if ( [[ ${+_sys_really_all_units} -eq 0 ]] || _cache_invalid SYS_REALLY_ALL_UNITS ) && + ! _retrieve_cache SYS_REALLY_ALL_UNITS; + then + all_unit_files=( ${${(f)"$(__systemctl list-unit-files)"}%% *} ) + _systemctl_all_units + really_all_units=($_sys_all_units $all_unit_files) + _sys_really_all_units=(${(u)really_all_units}) + _store_cache SYS_REALLY_ALL_UNITS _sys_really_all_units + fi +} + +_filter_units_by_property() { + local property=$1 value=$2 ; shift ; shift + local -a units ; units=($*) + local props + for props in ${(ps:\n\n:)"$(_call_program units "$service show --no-pager --property="Id,$property" -- ${units} 2>/dev/null")"}; do + props=(${(f)props}) + if [[ "${props[2]}" = "$property=$value" ]]; then + echo -E - " ${props[1]#Id=}" + fi + done +} + +_systemctl_get_template_names() { echo -E - ${^${(M)${(f)"$(__systemctl list-unit-files)"}##*@.[^[:space:]]##}%%@.*}\@ } + + +_systemctl_active_units() {_sys_active_units=( ${${(f)"$(__systemctl list-units)"}%% *} )} + +_systemctl_startable_units(){ + _sys_startable_units=( $( _filter_units_by_property ActiveState inactive $( + _filter_units_by_property CanStart yes $( + __systemctl $mode list-unit-files --state enabled,disabled,static | \ + { while read -r a b; do [[ $a =~ @\. ]] || echo -E - " $a"; done; } + __systemctl $mode list-units --state inactive,failed | \ + { while read -r a b; do echo -E - " $a"; done; } )) ) ) +} + +_systemctl_restartable_units(){ + _sys_restartable_units=( $(_filter_units_by_property CanStart yes $( + __systemctl $mode list-unit-files --state enabled,disabled,static | \ + { while read -r a b; do [[ $a =~ @\. ]] || echo -E - " $a"; done; } + __systemctl $mode list-units | \ + { while read -r a b; do echo -E - " $a"; done; } )) ) +} + +_systemctl_failed_units() {_sys_failed_units=( ${${(f)"$(__systemctl list-units --state=failed)"}%% *} ) } +_systemctl_unit_state() { typeset -gA _sys_unit_state; _sys_unit_state=( $(__systemctl list-unit-files) ) } + +local fun +# Completion functions for ALL_UNITS +for fun in is-active is-failed is-enabled status show cat mask preset help list-dependencies edit ; do + (( $+functions[_systemctl_$fun] )) || _systemctl_$fun() + { + _systemctl_really_all_units + _wanted systemd-units expl unit \ + compadd "$@" -a - _sys_really_all_units + } +done + +# Completion functions for ENABLED_UNITS +(( $+functions[_systemctl_disable] )) || _systemctl_disable() +{ + local _sys_unit_state; _systemctl_unit_state + _wanted systemd-units expl 'enabled unit' \ + compadd "$@" - ${(k)_sys_unit_state[(R)enabled]} +} + +(( $+functions[_systemctl_reenable] )) || _systemctl_reenable() +{ + local _sys_unit_state; _systemctl_unit_state + _wanted systemd-units expl 'enabled/disabled unit' \ + compadd "$@" - ${(k)_sys_unit_state[(R)(enabled|disabled)]} $(_systemctl_get_template_names) +} + +# Completion functions for DISABLED_UNITS +(( $+functions[_systemctl_enable] )) || _systemctl_enable() +{ + local _sys_unit_state; _systemctl_unit_state + _wanted systemd-units expl 'disabled unit' \ + compadd "$@" - ${(k)_sys_unit_state[(R)disabled]} $(_systemctl_get_template_names) +} + +# Completion functions for FAILED_UNITS +(( $+functions[_systemctl_reset-failed] )) || _systemctl_reset-failed() +{ + local _sys_failed_units; _systemctl_failed_units + _wanted systemd-units expl 'failed unit' \ + compadd "$@" -a - _sys_failed_units || _message "no failed unit found" +} + +# Completion functions for STARTABLE_UNITS +(( $+functions[_systemctl_start] )) || _systemctl_start() +{ + local _sys_startable_units; _systemctl_startable_units + _wanted systemd-units expl 'startable unit' \ + compadd "$@" - ${_sys_startable_units[*]} $(_systemctl_get_template_names) +} + +# Completion functions for STOPPABLE_UNITS +for fun in stop kill try-restart condrestart ; do + (( $+functions[_systemctl_$fun] )) || _systemctl_$fun() + { + local _sys_active_units; _systemctl_active_units + _wanted systemd-units expl 'stoppable unit' \ + compadd "$@" - $( _filter_units_by_property CanStop yes \ + ${_sys_active_units[*]} ) + } +done + +# Completion functions for ISOLATABLE_UNITS +(( $+functions[_systemctl_isolate] )) || _systemctl_isolate() +{ + _systemctl_all_units + _wanted systemd-units expl 'isolatable unit' \ + compadd "$@" - $( _filter_units_by_property AllowIsolate yes \ + ${_sys_all_units[*]} ) +} + +# Completion functions for RELOADABLE_UNITS +for fun in reload try-reload-or-restart force-reload ; do + (( $+functions[_systemctl_$fun] )) || _systemctl_$fun() + { + local _sys_active_units; _systemctl_active_units + _wanted systemd-units expl 'reloadable unit' \ + compadd "$@" - $( _filter_units_by_property CanReload yes \ + ${_sys_active_units[*]} ) + } +done + +# Completion functions for RESTARTABLE_UNITS +for fun in restart reload-or-restart ; do + (( $+functions[_systemctl_$fun] )) || _systemctl_$fun() + { + local _sys_restartable_units; _systemctl_restartable_units + _wanted systemd-units expl 'restartable unit' \ + compadd "$@" - ${_sys_restartable_units[*]} $(_systemctl_get_template_names) + } +done + +# Completion functions for MASKED_UNITS +(( $+functions[_systemctl_unmask] )) || _systemctl_unmask() +{ + local _sys_unit_state; _systemctl_unit_state + _wanted systemd-units expl 'masked unit' \ + compadd "$@" - ${(k)_sys_unit_state[(R)masked]} || _message "no masked units found" +} + +# Completion functions for JOBS +(( $+functions[_systemctl_cancel] )) || _systemctl_cancel() +{ + _wanted systemd-jobs expl job \ + compadd "$@" - ${${(f)"$(__systemctl list-jobs)"}%% *} || + _message "no jobs found" +} + +# Completion functions for TARGETS +(( $+functions[_systemctl_set-default] )) || _systemctl_set-default() +{ + _wanted systemd-targets expl target \ + compadd "$@" - ${${(f)"$(__systemctl list-unit-files --type target --all)"}%% *} || + _message "no targets found" +} + +# Completion functions for ENVS +for fun in set-environment unset-environment ; do + (( $+functions[_systemctl_$fun] )) || _systemctl_$fun() + { + local fun=$0 ; fun=${fun##_systemctl_} + local suf + if [[ "${fun}" = "set-environment" ]]; then + suf='-S=' + fi + _wanted systemd-environment expl 'environment variable' \ + compadd "$@" ${suf} - ${${(f)"$(systemctl show-environment)"}%%=*} + } +done + +(( $+functions[_systemctl_link] )) || _systemctl_link() { + _sd_unit_files +} + +(( $+functions[_systemctl_switch-root] )) || _systemctl_switch-root() { + _files +} + +# no systemctl completion for: +# [STANDALONE]='daemon-reexec daemon-reload default +# emergency exit halt kexec list-jobs list-units +# list-unit-files poweroff reboot rescue show-environment' + +_systemctl_caching_policy() +{ + local _sysunits + local -a oldcache + + # rebuild if cache is more than a day old + oldcache=( "$1"(mh+1) ) + (( $#oldcache )) && return 0 + + _sysunits=(${${(f)"$(__systemctl --all)"}%% *}) + + if (( $#_sysunits )); then + for unit in $_sysunits; do + [[ "$unit" -nt "$1" ]] && return 0 + done + fi + + return 1 +} + +_unit_states() { + local -a _states + _states=("${(fo)$(__systemctl --state=help)}") + _values -s , "${_states[@]}" +} + +_unit_types() { + local -a _types + _types=("${(fo)$(__systemctl -t help)}") + _values -s , "${_types[@]}" +} + +_unit_properties() { + if ( [[ ${+_sys_all_properties} -eq 0 ]] || _cache_invalid SYS_ALL_PROPERTIES ) && + ! _retrieve_cache SYS_ALL_PROPERTIES; + then + _sys_all_properties=( ${${(M)${(f)"$(__systemctl show --all; + @rootlibexecdir@/systemd --dump-configuration-items)"}##[[:alnum:]]##=*}%%=*} + ) + _store_cache SYS_ALL_PROPRTIES _sys_all_properties + fi + _values -s , "${_sys_all_properties[@]}" +} + +_job_modes() { + local -a _modes + _modes=(fail replace replace-irreversibly isolate ignore-dependencies ignore-requirements flush) + _values -s , "${_modes[@]}" +} + +local -a _modes; _modes=("--user" "--system") +local _sys_service_mgr=${${words:*_modes}[(R)(${(j.|.)_modes})]:---system} +_arguments -s \ + {-h,--help}'[Show help]' \ + '--version[Show package version]' \ + {-t+,--type=}'[List only units of a particular type]:unit type:_unit_types' \ + '--state=[Display units in the specified state]:unit state:_unit_states' \ + '--job-mode=[Specify how to deal with other jobs]:mode:_job_modes' \ + {-p+,--property=}'[Show only properties by specific name]:unit property:_unit_properties' \ + {-a,--all}'[Show all units/properties, including dead/empty ones]' \ + '--reverse[Show reverse dependencies]' \ + '--after[Show units ordered after]' \ + '--before[Show units ordered before]' \ + {-l,--full}"[Don't ellipsize unit names on output]" \ + '--show-types[When showing sockets, show socket type]' \ + {-i,--ignore-inhibitors}'[When executing a job, ignore jobs dependencies]' \ + {-q,--quiet}'[Suppress output]' \ + '--no-block[Do not wait until operation finished]' \ + '--no-legend[Do not print a legend, i.e. the column headers and the footer with hints]' \ + '--no-pager[Do not pipe output into a pager]' \ + '--system[Connect to system manager]' \ + '--user[Connect to user service manager]' \ + "--no-wall[Don't send wall message before halt/power-off/reboot]" \ + '--global[Enable/disable unit files globally]' \ + "--no-reload[When enabling/disabling unit files, don't reload daemon configuration]" \ + '--no-ask-password[Do not ask for system passwords]' \ + '--kill-who=[Who to send signal to]:killwho:(main control all)' \ + {-s+,--signal=}'[Which signal to send]:signal:_signals' \ + {-f,--force}'[When enabling unit files, override existing symlinks. When shutting down, execute action immediately]' \ + '--root=[Enable unit files in the specified root directory]:directory:_directories' \ + '--runtime[Enable unit files only temporarily until next reboot]' \ + {-H+,--host=}'[Operate on remote host]:userathost:_sd_hosts_or_user_at_host' \ + {-P,--privileged}'[Acquire privileges before execution]' \ + {-n+,--lines=}'[Journal entries to show]:number of entries' \ + {-o+,--output=}'[Change journal output mode]:modes:_sd_outputmodes' \ + '--firmware-setup[Tell the firmware to show the setup menu on next boot]' \ + '--plain[When used with list-dependencies, print output as a list]' \ + '*::systemctl command:_systemctl_command' diff --git a/src/grp-system/systemctl/systemctl.xml b/src/grp-system/systemctl/systemctl.xml new file mode 100644 index 0000000000..e7880d24f7 --- /dev/null +++ b/src/grp-system/systemctl/systemctl.xml @@ -0,0 +1,1826 @@ + + +%entities; +]> + + + + + + + systemctl + systemd + + + + Developer + Lennart + Poettering + lennart@poettering.net + + + + + + systemctl + 1 + + + + systemctl + Control the systemd system and service manager + + + + + systemctl + OPTIONS + COMMAND + NAME + + + + + Description + + systemctl may be used to introspect and + control the state of the systemd system and + service manager. Please refer to + systemd1 + for an introduction into the basic concepts and functionality this + tool manages. + + + + Options + + The following options are understood: + + + + + + + + The argument should be a comma-separated list of unit + types such as and + . + + + If one of the arguments is a unit type, when listing + units, limit display to certain unit types. Otherwise, units + of all types will be shown. + + As a special case, if one of the arguments is + , a list of allowed values will be + printed and the program will exit. + + + + + + + + The argument should be a comma-separated list of unit + LOAD, SUB, or ACTIVE states. When listing units, show only + those in the specified states. Use + to show only failed units. + + As a special case, if one of the arguments is + , a list of allowed values will be + printed and the program will exit. + + + + + + + + + When showing unit/job/manager properties with the + show command, limit display to properties + specified in the argument. The argument should be a + comma-separated list of property names, such as + MainPID. Unless specified, all known + properties are shown. If specified more than once, all + properties with the specified names are shown. Shell + completion is implemented for property names. + + For the manager itself, + systemctl show will show all available + properties. Those properties are documented in + systemd-system.conf5. + + + Properties for units vary by unit type, so showing any + unit (even a non-existent one) is a way to list properties + pertaining to this type. Similarly, showing any job will list + properties pertaining to all jobs. Properties for units are + documented in + systemd.unit5, + and the pages for individual unit types + systemd.service5, + systemd.socket5, + etc. + + + + + + + + + When listing units with list-units, also show inactive units and + units which are following other units. When showing unit/job/manager properties, show all + properties regardless whether they are set or not. + + To list all units installed in the file system, use the + list-unit-files command instead. + + + + + + + + + When listing units, also show units of local + containers. Units of local containers will be prefixed with + the container name, separated by a single colon character + (:). + + + + + + + + Show reverse dependencies between units with + list-dependencies, i.e. follow + dependencies of type WantedBy=, + RequiredBy=, + PartOf=, BoundBy=, + instead of Wants= and similar. + + + + + + + + + With list-dependencies, show the + units that are ordered before the specified unit. In other + words, recursively list units following the + After= dependency. + + Note that any After= dependency is + automatically mirrored to create a + Before= dependency. Temporal dependencies + may be specified explicitly, but are also created implicitly + for units which are WantedBy= targets + (see + systemd.target5), + and as a result of other directives (for example + RequiresMountsFor=). Both explicitly + and implicitly introduced dependencies are shown with + list-dependencies. + + + + + + + + With list-dependencies, show the + units that are ordered after the specified unit. In other + words, recursively list units following the + Before= dependency. + + + + + + + + + Do not ellipsize unit names, process tree entries, + journal output, or truncate unit descriptions in the output + of status, list-units, + list-jobs, and + list-timers. + + + + + + + + When printing properties with show, + only print the value, and skip the property name and + =. + + + + + + + + When showing sockets, show the type of the socket. + + + + + + + + When queuing a new job, this option controls how to deal with + already queued jobs. It takes one of fail, + replace, + replace-irreversibly, + isolate, + ignore-dependencies, + ignore-requirements or + flush. Defaults to + replace, except when the + isolate command is used which implies the + isolate job mode. + + If fail is specified and a requested + operation conflicts with a pending job (more specifically: + causes an already pending start job to be reversed into a stop + job or vice versa), cause the operation to fail. + + If replace (the default) is + specified, any conflicting pending job will be replaced, as + necessary. + + If replace-irreversibly is specified, + operate like replace, but also mark the new + jobs as irreversible. This prevents future conflicting + transactions from replacing these jobs (or even being enqueued + while the irreversible jobs are still pending). Irreversible + jobs can still be cancelled using the cancel + command. + + isolate is only valid for start + operations and causes all other units to be stopped when the + specified unit is started. This mode is always used when the + isolate command is used. + + flush will cause all queued jobs to + be canceled when the new job is enqueued. + + If ignore-dependencies is specified, + then all unit dependencies are ignored for this new job and + the operation is executed immediately. If passed, no required + units of the unit passed will be pulled in, and no ordering + dependencies will be honored. This is mostly a debugging and + rescue tool for the administrator and should not be used by + applications. + + ignore-requirements is similar to + ignore-dependencies, but only causes the + requirement dependencies to be ignored, the ordering + dependencies will still be honoured. + + + + + + + + + Shorthand for fail. + When used with the kill command, + if no units were killed, the operation results in an error. + + + + + + + + + + When system shutdown or a sleep state is requested, + ignore inhibitor locks. Applications can establish inhibitor + locks to avoid that certain important operations (such as CD + burning or suchlike) are interrupted by system shutdown or a + sleep state. Any user may take these locks and privileged + users may override these locks. If any locks are taken, + shutdown and sleep state requests will normally fail + (regardless of whether privileged or not) and a list of active locks + is printed. However, if + is specified, the locks are ignored and not printed, and the + operation attempted anyway, possibly requiring additional + privileges. + + + + + + + + + Suppress printing of the results of various commands + and also the hints about truncated log lines. This does not + suppress output of commands for which the printed output is + the only result (like show). Errors are + always printed. + + + + + + + + Do not synchronously wait for the requested operation + to finish. If this is not specified, the job will be + verified, enqueued and systemctl will + wait until the unit's start-up is completed. By passing this + argument, it is only verified and enqueued. + + + + + + + + + + + + + Do not send wall message before halt, power-off, + reboot. + + + + + + + + When used with enable and + disable, operate on the global user + configuration directory, thus enabling or disabling a unit + file globally for all future logins of all users. + + + + + + + + When used with enable and + disable, do not implicitly reload daemon + configuration after executing the changes. + + + + + + + + When used with start and related + commands, disables asking for passwords. Background services + may require input of a password or passphrase string, for + example to unlock system hard disks or cryptographic + certificates. Unless this option is specified and the + command is invoked from a terminal, + systemctl will query the user on the + terminal for the necessary secrets. Use this option to + switch this behavior off. In this case, the password must be + supplied by some other means (for example graphical password + agents) or the service might fail. This also disables + querying the user for authentication for privileged + operations. + + + + + + + + When used with kill, choose which + processes to send a signal to. Must be one of + , or + to select whether to kill only the main + process, the control process or all processes of the + unit. The main process of the unit is the one that defines + the life-time of it. A control process of a unit is one that + is invoked by the manager to induce state changes of it. For + example, all processes started due to the + ExecStartPre=, + ExecStop= or + ExecReload= settings of service units are + control processes. Note that there is only one control + process per unit at a time, as only one state change is + executed at a time. For services of type + Type=forking, the initial process started + by the manager for ExecStart= is a + control process, while the process ultimately forked off by + that one is then considered the main process of the unit (if + it can be determined). This is different for service units + of other types, where the process forked off by the manager + for ExecStart= is always the main process + itself. A service unit consists of zero or one main process, + zero or one control process plus any number of additional + processes. Not all unit types manage processes of these + types however. For example, for mount units, control processes + are defined (which are the invocations of + &MOUNT_PATH; and + &UMOUNT_PATH;), but no main process + is defined. If omitted, defaults to + . + + + + + + + + + + When used with kill, choose which + signal to send to selected processes. Must be one of the + well-known signal specifiers such as SIGTERM, SIGINT or + SIGSTOP. If omitted, defaults to + . + + + + + + + + + When used with enable, overwrite + any existing conflicting symlinks. + + When used with edit, create all of the + specified units which do not already exist. + + When used with halt, poweroff, reboot or + kexec, execute the selected operation without shutting down all units. However, all + processes will be killed forcibly and all file systems are unmounted or remounted read-only. This is hence a + drastic but relatively safe option to request an immediate reboot. If is specified + twice for these operations (with the exception of kexec), they will be executed + immediately, without terminating any processes or unmounting any file systems. Warning: specifying + twice with any of these operations might result in data loss. Note that when + is specified twice the selected operation is executed by + systemctl itself, and the system manager is not contacted. This means the command should + succeed even when the system manager hangs or crashed. + + + + + + + + When used with halt, + poweroff, reboot or + kexec, set a short message explaining the reason + for the operation. The message will be logged together with the + default shutdown message. + + + + + + + + When used with enable, the units + will also be started. When used with disable or + mask, the units will also be stopped. The start + or stop operation is only carried out when the respective enable or + disable operation has been successful. + + + + + + + + When used with + enable/disable/is-enabled + (and related commands), use an alternate root path when + looking for unit files. + + + + + + + + + When used with enable, + disable, edit, + (and related commands), make changes only temporarily, so + that they are lost on the next reboot. This will have the + effect that changes are not made in subdirectories of + /etc but in /run, + with identical immediate effects, however, since the latter + is lost on reboot, the changes are lost too. + + Similarly, when used with + set-property, make changes only + temporarily, so that they are lost on the next + reboot. + + + + + + + + Takes one of full (the default), + enable-only, + disable-only. When used with the + preset or preset-all + commands, controls whether units shall be disabled and + enabled according to the preset rules, or only enabled, or + only disabled. + + + + + + + + + When used with status, controls the + number of journal lines to show, counting from the most + recent ones. Takes a positive integer argument. Defaults to + 10. + + + + + + + + + When used with status, controls the + formatting of the journal entries that are shown. For the + available choices, see + journalctl1. + Defaults to short. + + + + + + + + When used with the reboot command, + indicate to the system's firmware to boot into setup + mode. Note that this is currently only supported on some EFI + systems and only if the system was booted in EFI + mode. + + + + + + + + When used with list-dependencies, + list-units or list-machines, the + the output is printed as a list instead of a tree, and the bullet + circles are omitted. + + + + + + + + + + + + + + + Commands + + The following commands are understood: + + + Unit Commands + + + + list-units PATTERN... + + + List units that systemd has loaded. This includes units that + are either referenced directly or through a dependency, or units that were active in the + past and have failed. By default only units which are active, have pending jobs, or have + failed are shown; this can be changed with option . If one or more + PATTERNs are specified, only units matching one of them are + shown. The units that are shown are additionally filtered by + and if those options are specified. + + This is the default command. + + + + + list-sockets PATTERN... + + + List socket units ordered by listening address. + If one or more PATTERNs are + specified, only socket units matching one of them are + shown. Produces output similar to + +LISTEN UNIT ACTIVATES +/dev/initctl systemd-initctl.socket systemd-initctl.service +... +[::]:22 sshd.socket sshd.service +kobject-uevent 1 systemd-udevd-kernel.socket systemd-udevd.service + +5 sockets listed. + Note: because the addresses might contains spaces, this output + is not suitable for programmatic consumption. + + + See also the options , + , and . + + + + + list-timers PATTERN... + + + List timer units ordered by the time they elapse + next. If one or more PATTERNs + are specified, only units matching one of them are shown. + + + See also the options and + . + + + + + start PATTERN... + + + Start (activate) one or more units specified on the + command line. + + Note that glob patterns operate on the set of primary names of currently loaded units. Units which + are not active and are not in a failed state usually are not loaded, and will not be matched by any + pattern. In addition, in case of instantiated units, systemd is often unaware of the instance name until + the instance has been started. Therefore, using glob patterns with start has limited + usefulness. Also, secondary alias names of units are not considered. + + + + stop PATTERN... + + + Stop (deactivate) one or more units specified on the + command line. + + + + reload PATTERN... + + + Asks all units listed on the command line to reload + their configuration. Note that this will reload the + service-specific configuration, not the unit configuration + file of systemd. If you want systemd to reload the + configuration file of a unit, use the + daemon-reload command. In other words: + for the example case of Apache, this will reload Apache's + httpd.conf in the web server, not the + apache.service systemd unit + file. + + This command should not be confused with the + daemon-reload command. + + + + + restart PATTERN... + + + Restart one or more units specified on the command + line. If the units are not running yet, they will be + started. + + + + try-restart PATTERN... + + + Restart one or more units specified on the command + line if the units are running. This does nothing if units are not + running. + + + + + reload-or-restart PATTERN... + + + Reload one or more units if they support it. If not, + restart them instead. If the units are not running yet, they + will be started. + + + + try-reload-or-restart PATTERN... + + + Reload one or more units if they support it. If not, + restart them instead. This does nothing if the units are not + running. + + + + + isolate NAME + + + Start the unit specified on the command line and its + dependencies and stop all others. If a unit name with no + extension is given, an extension of + .target will be assumed. + + This is similar to changing the runlevel in a + traditional init system. The isolate + command will immediately stop processes that are not enabled + in the new unit, possibly including the graphical + environment or terminal you are currently using. + + Note that this is allowed only on units where + is enabled. See + systemd.unit5 + for details. + + + + kill PATTERN... + + + Send a signal to one or more processes of the + unit. Use to select which + process to kill. Use to select + the signal to send. + + + + is-active PATTERN... + + + Check whether any of the specified units are active + (i.e. running). Returns an exit code + 0 if at least one is active, or + non-zero otherwise. Unless is + specified, this will also print the current unit state to + standard output. + + + + is-failed PATTERN... + + + Check whether any of the specified units are in a + "failed" state. Returns an exit code + 0 if at least one has failed, + non-zero otherwise. Unless is + specified, this will also print the current unit state to + standard output. + + + + status PATTERN...|PID...] + + + Show terse runtime status information about one or + more units, followed by most recent log data from the + journal. If no units are specified, show system status. If + combined with , also show the status of + all units (subject to limitations specified with + ). If a PID is passed, show information + about the unit the process belongs to. + + This function is intended to generate human-readable + output. If you are looking for computer-parsable output, + use show instead. By default, this + function only shows 10 lines of output and ellipsizes + lines to fit in the terminal window. This can be changed + with and , + see above. In addition, journalctl + --unit=NAME or + journalctl + --user-unit=NAME use + a similar filter for messages and might be more + convenient. + + + + + show PATTERN...|JOB... + + + Show properties of one or more units, jobs, or the + manager itself. If no argument is specified, properties of + the manager will be shown. If a unit name is specified, + properties of the unit is shown, and if a job ID is + specified, properties of the job is shown. By default, empty + properties are suppressed. Use to + show those too. To select specific properties to show, use + . This command is intended to be + used whenever computer-parsable output is required. Use + status if you are looking for formatted + human-readable output. + + + + cat PATTERN... + + + Show backing files of one or more units. Prints the + "fragment" and "drop-ins" (source files) of units. Each + file is preceded by a comment which includes the file + name. + + + + set-property NAME ASSIGNMENT... + + + Set the specified unit properties at runtime where + this is supported. This allows changing configuration + parameter properties such as resource control settings at + runtime. Not all properties may be changed at runtime, but + many resource control settings (primarily those in + systemd.resource-control5) + may. The changes are applied instantly, and stored on disk + for future boots, unless is + passed, in which case the settings only apply until the + next reboot. The syntax of the property assignment follows + closely the syntax of assignments in unit files. + + Example: systemctl set-property foobar.service CPUShares=777 + + If the specified unit appears to be inactive, the + changes will be only stored on disk as described + previously hence they will be effective when the unit will + be started. + + Note that this command allows changing multiple + properties at the same time, which is preferable over + setting them individually. Like unit file configuration + settings, assigning the empty list to list parameters will + reset the list. + + + + + help PATTERN...|PID... + + + Show manual pages for one or more units, if + available. If a PID is given, the manual pages for the unit + the process belongs to are shown. + + + + + reset-failed [PATTERN...] + + + Reset the failed state of the + specified units, or if no unit name is passed, reset the state of all + units. When a unit fails in some way (i.e. process exiting + with non-zero error code, terminating abnormally or timing + out), it will automatically enter the + failed state and its exit code and status + is recorded for introspection by the administrator until the + service is restarted or reset with this command. + + + + + + list-dependencies + NAME + + + + Shows units required and wanted by the specified + unit. This recursively lists units following the + Requires=, + Requisite=, + ConsistsOf=, + Wants=, BindsTo= + dependencies. If no unit is specified, + default.target is implied. + + By default, only target units are recursively + expanded. When is passed, all other + units are recursively expanded as well. + + Options , + , + may be used to change what types of dependencies + are shown. + + + + + + + Unit File Commands + + + + list-unit-files PATTERN... + + + List unit files installed on the system, in combination with their enablement state (as reported by + is-enabled). If one or more PATTERNs are specified, only unit + files whose name matches one of them are shown (patterns matching unit file system paths are not + supported). + + + + + enable NAME... + enable PATH... + + + Enable one or more units or unit instances. This will create a set of symlinks, as encoded in the + [Install] sections of the indicated unit files. After the symlinks have been created, + the system manager configuration is reloaded (in a way equivalent to daemon-reload), in + order to ensure the changes are taken into account immediately. Note that this does + not have the effect of also starting any of the units being enabled. If this is + desired, combine this command with the switch, or invoke start + with appropriate arguments later. Note that in case of unit instance enablement (i.e. enablement of units of + the form foo@bar.service), symlinks named the same as instances are created in the + unit configuration diectory, however they point to the single template unit file they are instantiated + from. + + This command expects either valid unit names (in which case various unit file directories are + automatically searched for unit files with appropriate names), or absolute paths to unit files (in which + case these files are read directly). If a specified unit file is located outside of the usual unit file + directories, an additional symlink is created, linking it into the unit configuration path, thus ensuring + it is found when requested by commands such as start. + + This command will print the file system operations executed. This output may be suppressed by passing + . + + + Note that this operation creates only the symlinks suggested in the [Install] + section of the unit files. While this command is the recommended way to manipulate the unit configuration + directory, the administrator is free to make additional changes manually by placing or removing symlinks + below this directory. This is particularly useful to create configurations that deviate from the suggested + default installation. In this case, the administrator must make sure to invoke + daemon-reload manually as necessary, in order to ensure the changes are taken into + account. + + + Enabling units should not be confused with starting (activating) units, as done by the + start command. Enabling and starting units is orthogonal: units may be enabled without + being started and started without being enabled. Enabling simply hooks the unit into various suggested + places (for example, so that the unit is automatically started on boot or when a particular kind of + hardware is plugged in). Starting actually spawns the daemon process (in case of service units), or binds + the socket (in case of socket units), and so on. + + Depending on whether , , , + or is specified, this enables the unit for the system, for the calling user only, + for only this boot of the system, or for all future logins of all users, or only this boot. Note that in + the last case, no systemd daemon configuration is reloaded. + + Using enable on masked units is not supported and results in an error. + + + + + disable NAME... + + + Disables one or more units. This removes all symlinks to the unit files backing the specified units + from the unit configuration directory, and hence undoes any changes made by enable or + link. Note that this removes all symlinks to matching unit files, + including manually created symlinks, and not just those actually created by enable or + link. Note that while disable undoes the effect of + enable, the two commands are otherwise not symmetric, as disable may + remove more symlinks than a prior enable invocation of the same unit created. + + This command expects valid unit names only, it does not accept paths to unit files. + + In addition to the units specified as arguments, all units are disabled that are listed in the + Also= setting contained in the [Install] section of any of the unit + files being operated on. + + This command implicitly reloads the system manager configuration after completing the operation. Note + that this command does not implicitly stop the units that are being disabled. If this is desired, either + combine this command with the switch, or invoke the stop command + with appropriate arguments later. + + This command will print information about the file system operations (symlink removals) + executed. This output may be suppressed by passing . + + + This command honors , , + and in a similar way as enable. + + + + + reenable NAME... + + + Reenable one or more units, as specified on the command line. This is a combination of + disable and enable and is useful to reset the symlinks a unit file is + enabled with to the defaults configured in its [Install] section. This commands expects + a unit uname only, it does not accept paths to unit files. + + + + + preset NAME... + + + Reset the enable/disable status one or more unit files, as specified on + the command line, to the defaults configured in the preset policy files. This + has the same effect as disable or + enable, depending how the unit is listed in the preset + files. + + Use to control whether units shall be + enabled and disabled, or only enabled, or only disabled. + + If the unit carries no install information, it will be silently ignored + by this command. + + For more information on the preset policy format, see + systemd.preset5. + For more information on the concept of presets, please consult the + Preset + document. + + + + + preset-all + + + Resets all installed unit files to the defaults + configured in the preset policy file (see above). + + Use to control + whether units shall be enabled and disabled, or only + enabled, or only disabled. + + + + + is-enabled NAME... + + + Checks whether any of the specified unit files are + enabled (as with enable). Returns an + exit code of 0 if at least one is enabled, non-zero + otherwise. Prints the current enable status (see table). + To suppress this output, use . + + + + + <command>is-enabled</command> output + + + + + + Name + Description + Exit Code + + + + + enabled + Enabled via .wants/, .requires/ or alias symlinks (permanently in /etc/systemd/system/, or transiently in /run/systemd/system/). + 0 + + + enabled-runtime + + + linked + Made available through one or more symlinks to the unit file (permanently in /etc/systemd/system/ or transiently in /run/systemd/system/), even though the unit file might reside outside of the unit file search path. + > 0 + + + linked-runtime + + + masked + Completely disabled, so that any start operation on it fails (permanently in /etc/systemd/system/ or transiently in /run/systemd/systemd/). + > 0 + + + masked-runtime + + + static + The unit file is not enabled, and has no provisions for enabling in the [Install] unit file section. + 0 + + + indirect + The unit file itself is not enabled, but it has a non-empty Also= setting in the [Install] unit file section, listing other unit files that might be enabled. + 0 + + + disabled + The unit file is not enabled, but contains an [Install] section with installation instructions. + > 0 + + + generated + The unit file was generated dynamically via a generator tool. See systemd.generator7. Generated unit files may not be enabled, they are enabled implicitly by their generator. + 0 + + + transient + The unit file has been created dynamically with the runtime API. Transient units may not be enabled. + 0 + + + bad + The unit file is invalid or another error occurred. Note that is-enabled will not actually return this state, but print an error message instead. However the unit file listing printed by list-unit-files might show it. + > 0 + + + +
+ + + + + + mask NAME... + + + Mask one or more units, as specified on the command line. This will link these unit files to + /dev/null, making it impossible to start them. This is a stronger version of + disable, since it prohibits all kinds of activation of the unit, including enablement + and manual activation. Use this option with care. This honors the option to only + mask temporarily until the next reboot of the system. The option may be used to + ensure that the units are also stopped. This command expects valid unit names only, it does not accept unit + file paths. + + + + + unmask NAME... + + + Unmask one or more unit files, as specified on the command line. This will undo the effect of + mask. This command expects valid unit names only, it does not accept unit file + paths. + + + + + link PATH... + + + Link a unit file that is not in the unit file search paths into the unit file search path. This + command expects an absolute path to a unit file. The effect of this may be undone with + disable. The effect of this command is that a unit file is made available for commands + such as start, even though it is not installed directly in the unit search path. + + + + + revert NAME... + + + Revert one or more unit files to their vendor versions. This command removes drop-in configuration + files that modify the specified units, as well as any user-configured unit file that overrides a matching + vendor supplied unit file. Specifically, for a unit foo.service the matching directories + foo.service.d/ with all their contained files are removed, both below the persistent and + runtime configuration directories (i.e. below /etc/systemd/system and + /run/systemd/system); if the unit file has a vendor-supplied version (i.e. a unit file + located below /usr) any matching peristent or runtime unit file that overrides it is + removed, too. Note that if a unit file has no vendor-supplied version (i.e. is only defined below + /etc/systemd/system or /run/systemd/system, but not in a unit + file stored below /usr), then it is not removed. Also, if a unit is masked, it is + unmasked. + + Effectively, this command may be used to undo all changes made with systemctl + edit, systemctl set-property and systemctl mask and puts + the original unit file with its settings back in effect. + + + + + add-wants TARGET + NAME... + add-requires TARGET + NAME... + + + Adds Wants= or Requires= + dependencies, respectively, to the specified + TARGET for one or more units. + + This command honors , + , and + in a way similar to + enable. + + + + + + edit NAME... + + + Edit a drop-in snippet or a whole replacement file if + is specified, to extend or override the + specified unit. + + Depending on whether (the default), + , or is specified, + this command creates a drop-in file for each unit either for the system, + for the calling user, or for all futures logins of all users. Then, + the editor (see the "Environment" section below) is invoked on + temporary files which will be written to the real location if the + editor exits successfully. + + If is specified, this will copy the + original units instead of creating drop-in files. + + If is specified and any units do + not already exist, new unit files will be opened for editing. + + If is specified, the changes will + be made temporarily in /run and they will be + lost on the next reboot. + + If the temporary file is empty upon exit, the modification of + the related unit is canceled. + + After the units have been edited, systemd configuration is + reloaded (in a way that is equivalent to daemon-reload). + + + Note that this command cannot be used to remotely edit units + and that you cannot temporarily edit units which are in + /etc, since they take precedence over + /run. + + + + + get-default + + + Return the default target to boot into. This returns + the target unit name default.target + is aliased (symlinked) to. + + + + + set-default NAME + + + Set the default target to boot into. This sets + (symlinks) the default.target alias + to the given target unit. + + + + + + + + Machine Commands + + + + list-machines PATTERN... + + + List the host and all running local containers with + their state. If one or more + PATTERNs are specified, only + containers matching one of them are shown. + + + + + + + + Job Commands + + + + list-jobs PATTERN... + + + List jobs that are in progress. If one or more + PATTERNs are specified, only + jobs for units matching one of them are shown. + + + + cancel JOB... + + + Cancel one or more jobs specified on the command line + by their numeric job IDs. If no job ID is specified, cancel + all pending jobs. + + + + + + + Environment Commands + + + + show-environment + + + Dump the systemd manager environment block. The + environment block will be dumped in straight-forward form + suitable for sourcing into a shell script. This environment + block will be passed to all processes the manager + spawns. + + + + set-environment VARIABLE=VALUE... + + + Set one or more systemd manager environment variables, + as specified on the command line. + + + + unset-environment VARIABLE... + + + Unset one or more systemd manager environment + variables. If only a variable name is specified, it will be + removed regardless of its value. If a variable and a value + are specified, the variable is only removed if it has the + specified value. + + + + + import-environment + VARIABLE... + + + + Import all, one or more environment variables set on + the client into the systemd manager environment block. If + no arguments are passed, the entire environment block is + imported. Otherwise, a list of one or more environment + variable names should be passed, whose client-side values + are then imported into the manager's environment + block. + + + + + + + Manager Lifecycle Commands + + + + daemon-reload + + + Reload the systemd manager configuration. This will + rerun all generators (see + systemd.generator7), + reload all unit files, and recreate the entire dependency + tree. While the daemon is being reloaded, all sockets + systemd listens on behalf of user configuration will stay + accessible. + + This command should not be confused with the + reload command. + + + + daemon-reexec + + + Reexecute the systemd manager. This will serialize the + manager state, reexecute the process and deserialize the + state again. This command is of little use except for + debugging and package upgrades. Sometimes, it might be + helpful as a heavy-weight daemon-reload. + While the daemon is being reexecuted, all sockets systemd listening + on behalf of user configuration will stay accessible. + + + + + + + + System Commands + + + + is-system-running + + + Checks whether the system is operational. This + returns success (exit code 0) when the system is fully up + and running, specifically not in startup, shutdown or + maintenance mode, and with no failed services. Failure is + returned otherwise (exit code non-zero). In addition, the + current state is printed in a short string to standard + output, see the table below. Use to + suppress this output. + + + <command>is-system-running</command> output + + + + + + + Name + Description + Exit Code + + + + + initializing + Early bootup, before + basic.target is reached + or the maintenance state entered. + + > 0 + + + starting + Late bootup, before the job queue + becomes idle for the first time, or one of the + rescue targets are reached. + > 0 + + + running + The system is fully + operational. + 0 + + + degraded + The system is operational but one or more + units failed. + > 0 + + + maintenance + The rescue or emergency target is + active. + > 0 + + + stopping + The manager is shutting + down. + > 0 + + + offline + The manager is not + running. Specifically, this is the operational + state if an incompatible program is running as + system manager (PID 1). + > 0 + + + unknown + The operational state could not be + determined, due to lack of resources or another + error cause. + > 0 + + + +
+ + + + + default + + + Enter default mode. This is mostly equivalent to + isolate default.target. + + + + + rescue + + + Enter rescue mode. This is mostly equivalent to + isolate rescue.target, but also prints a + wall message to all users. + + + + emergency + + + Enter emergency mode. This is mostly equivalent to + isolate emergency.target, but also prints + a wall message to all users. + + + + halt + + + Shut down and halt the system. This is mostly equivalent to start halt.target + --job-mode=replace-irreversibly, but also prints a wall message to all users. If combined with + , shutdown of all running services is skipped, however all processes are killed and + all file systems are unmounted or mounted read-only, immediately followed by the system halt. If + is specified twice, the operation is immediately executed without terminating any + processes or unmounting any file systems. This may result in data loss. Note that when + is specified twice the halt operation is executed by + systemctl itself, and the system manager is not contacted. This means the command should + succeed even when the system manager hangs or crashed. + + + + poweroff + + + Shut down and power-off the system. This is mostly equivalent to start poweroff.target + --job-mode=replace-irreversibly, but also prints a wall message to all users. If combined with + , shutdown of all running services is skipped, however all processes are killed and + all file systems are unmounted or mounted read-only, immediately followed by the powering off. If + is specified twice, the operation is immediately executed without terminating any + processes or unmounting any file systems. This may result in data loss. Note that when + is specified twice the power-off operation is executed by + systemctl itself, and the system manager is not contacted. This means the command should + succeed even when the system manager hangs or crashed. + + + + reboot arg + + + Shut down and reboot the system. This is mostly equivalent to start reboot.target + --job-mode=replace-irreversibly, but also prints a wall message to all users. If combined with + , shutdown of all running services is skipped, however all processes are killed and + all file systems are unmounted or mounted read-only, immediately followed by the reboot. If + is specified twice, the operation is immediately executed without terminating any + processes or unmounting any file systems. This may result in data loss. Note that when + is specified twice the reboot operation is executed by + systemctl itself, and the system manager is not contacted. This means the command should + succeed even when the system manager hangs or crashed. + + If the optional argument + arg is given, it will be passed + as the optional argument to the + reboot2 + system call. The value is architecture and firmware + specific. As an example, recovery might + be used to trigger system recovery, and + fota might be used to trigger a + firmware over the air update. + + + + + kexec + + + Shut down and reboot the system via kexec. This is + mostly equivalent to start kexec.target --job-mode=replace-irreversibly, + but also prints a wall message to all users. If combined + with , shutdown of all running + services is skipped, however all processes are killed and + all file systems are unmounted or mounted read-only, + immediately followed by the reboot. + + + + + exit EXIT_CODE + + + Ask the systemd manager to quit. This is only + supported for user service managers (i.e. in conjunction + with the option) or in containers + and is equivalent to poweroff otherwise. + + The systemd manager can exit with a non-zero exit + code if the optional argument + EXIT_CODE is given. + + + + + switch-root ROOT INIT + + + Switches to a different root directory and executes a + new system manager process below it. This is intended for + usage in initial RAM disks ("initrd"), and will transition + from the initrd's system manager process (a.k.a. "init" + process) to the main system manager process. This call takes two + arguments: the directory that is to become the new root directory, and + the path to the new system manager binary below it to + execute as PID 1. If the latter is omitted or the empty + string, a systemd binary will automatically be searched for + and used as init. If the system manager path is omitted or + equal to the empty string, the state of the initrd's system + manager process is passed to the main system manager, which + allows later introspection of the state of the services + involved in the initrd boot. + + + + + suspend + + + Suspend the system. This will trigger activation of + the special suspend.target target. + + + + + + hibernate + + + Hibernate the system. This will trigger activation of + the special hibernate.target target. + + + + + + hybrid-sleep + + + Hibernate and suspend the system. This will trigger + activation of the special + hybrid-sleep.target target. + + + + + + + Parameter Syntax + + Unit commands listed above take either a single unit name (designated as NAME), + or multiple unit specifications (designated as PATTERN...). In the first case, the + unit name with or without a suffix must be given. If the suffix is not specified (unit name is "abbreviated"), + systemctl will append a suitable suffix, .service by default, and a type-specific suffix in + case of commands which operate only on specific unit types. For example, + # systemctl start sshd and + # systemctl start sshd.service + are equivalent, as are + # systemctl isolate default + and + # systemctl isolate default.target + Note that (absolute) paths to device nodes are automatically converted to device unit names, and other (absolute) + paths to mount unit names. + # systemctl status /dev/sda +# systemctl status /home + are equivalent to: + # systemctl status dev-sda.device +# systemctl status home.mount + In the second case, shell-style globs will be matched against the primary names of all currently loaded units; + literal unit names, with or without a suffix, will be treated as in the first case. This means that literal unit + names always refer to exactly one unit, but globs may match zero units and this is not considered an + error. + + Glob patterns use + fnmatch3, + so normal shell-style globbing rules are used, and + *, ?, + [] may be used. See + glob7 + for more details. The patterns are matched against the primary names of + currently loaded units, and patterns which do not match anything + are silently skipped. For example: + # systemctl stop sshd@*.service + will stop all sshd@.service instances. Note that alias names of units, and units that aren't + loaded are not considered for glob expansion. + + + For unit file commands, the specified NAME should be the name of the unit file + (possibly abbreviated, see above), or the absolute path to the unit file: + # systemctl enable foo.service + or + # systemctl link /path/to/foo.service + + + + + + + Exit status + + On success, 0 is returned, a non-zero failure + code otherwise. + + + + Environment + + + + $SYSTEMD_EDITOR + + Editor to use when editing units; overrides + $EDITOR and $VISUAL. If neither + $SYSTEMD_EDITOR nor $EDITOR nor + $VISUAL are present or if it is set to an empty + string or if their execution failed, systemctl will try to execute well + known editors in this order: + editor1, + nano1, + vim1, + vi1. + + + + + + + + + See Also + + systemd1, + journalctl1, + loginctl1, + machinectl1, + systemd.unit5, + systemd.resource-control5, + systemd.special7, + wall1, + systemd.preset5, + systemd.generator7, + glob7 + + + + diff --git a/src/grp-system/systemctl/systemd-sysv-install.SKELETON b/src/grp-system/systemctl/systemd-sysv-install.SKELETON new file mode 100755 index 0000000000..a53a3e6221 --- /dev/null +++ b/src/grp-system/systemctl/systemd-sysv-install.SKELETON @@ -0,0 +1,47 @@ +#!/bin/sh +# This script is called by "systemctl enable/disable" when the given unit is a +# SysV init.d script. It needs to call the distribution's mechanism for +# enabling/disabling those, such as chkconfig, update-rc.d, or similar. This +# can optionally take a --root argument for enabling a SysV init script +# in a chroot or similar. +set -e + +usage() { + echo "Usage: $0 [--root=path] enable|disable|is-enabled " >&2 + exit 1 +} + +# parse options +eval set -- "$(getopt -o r: --long root: -- "$@")" +while true; do + case "$1" in + -r|--root) + ROOT="$2" + shift 2 ;; + --) shift ; break ;; + *) usage ;; + esac +done + +NAME="$2" +[ -n "$NAME" ] || usage + +case "$1" in + enable) + # call the command to enable SysV init script $NAME here + # (consider optional $ROOT) + echo "IMPLEMENT ME: enabling SysV init.d script $NAME" + ;; + disable) + # call the command to disable SysV init script $NAME here + # (consider optional $ROOT) + echo "IMPLEMENT ME: disabling SysV init.d script $NAME" + ;; + is-enabled) + # exit with 0 if $NAME is enabled, non-zero if it is disabled + # (consider optional $ROOT) + echo "IMPLEMENT ME: checking SysV init.d script $NAME" + ;; + *) + usage ;; +esac diff --git a/src/grp-system/systemctl/systemd.preset.xml b/src/grp-system/systemctl/systemd.preset.xml new file mode 100644 index 0000000000..b7164014f0 --- /dev/null +++ b/src/grp-system/systemctl/systemd.preset.xml @@ -0,0 +1,189 @@ + + + + + + + + systemd.preset + systemd + + + + Developer + Lennart + Poettering + lennart@poettering.net + + + + + + systemd.preset + 5 + + + + systemd.preset + Service enablement presets + + + + /etc/systemd/system-preset/*.preset + /run/systemd/system-preset/*.preset + /usr/lib/systemd/system-preset/*.preset + /etc/systemd/user-preset/*.preset + /run/systemd/user-preset/*.preset + /usr/lib/systemd/user-preset/*.preset + + + + Description + + Preset files may be used to encode policy which units shall + be enabled by default and which ones shall be disabled. They are + read by systemctl preset (for more information + see + systemctl1) + which uses this information to enable or disable a unit according + to preset policy. systemctl preset is used by + the post install scriptlets of RPM packages (or other OS package + formats), to enable/disable specific units by default on package + installation, enforcing distribution, spin or administrator preset + policy. This allows choosing a certain set of units to be + enabled/disabled even before installing the actual package. + + For more information on the preset logic please have a look + at the Presets + document. + + It is not recommended to ship preset files within the + respective software packages implementing the units, but rather + centralize them in a distribution or spin default policy, which + can be amended by administrator policy. + + If no preset files exist, systemctl + preset will enable all units that are installed by + default. If this is not desired and all units shall rather be + disabled, it is necessary to ship a preset file with a single, + catchall "disable *" line. (See example 1, + below.) + + + + Preset File Format + + The preset files contain a list of directives consisting of + either the word enable or + disable followed by a space and a unit name + (possibly with shell style wildcards), separated by newlines. + Empty lines and lines whose first non-whitespace character is # or + ; are ignored. + + Two different directives are understood: + enable may be used to enable units by default, + disable to disable units by default. + + If multiple lines apply to a unit name, the first matching + one takes precedence over all others. + + Each preset file shall be named in the style of + <priority>-<policy-name>.preset. Files + in /etc/ override files with the same name in + /usr/lib/ and /run/. + Files in /run/ override files with the same + name in /usr/lib/. Packages should install + their preset files in /usr/lib/. Files in + /etc/ are reserved for the local + administrator, who may use this logic to override the preset files + installed by vendor packages. All preset files are sorted by their + filename in lexicographic order, regardless of which of the + directories they reside in. If multiple files specify the same + unit name, the entry in the file with the lexicographically + earliest name will be applied. It is recommended to prefix all + filenames with a two-digit number and a dash, to simplify the + ordering of the files. + + If the administrator wants to disable a preset file supplied + by the vendor, the recommended way is to place a symlink to + /dev/null in + /etc/systemd/system-preset/ bearing the same + filename. + + + + Example + + + Default off example <filename>/usr/lib/systemd/system-preset/99-default.preset</filename>: + + disable * + + + This disables all units. Due to the filename prefix + 99-, it will be read last and hence can easily + be overridden by spin or administrator preset policy or + suchlike. + + + A GNOME spin example <filename>/usr/lib/systemd/system-preset/50-gnome.preset</filename>: + + enable gdm.service +enable colord.service +enable accounts-daemon.service +enable avahi-daemon.* + + + + This enables the three mentioned units, plus all + avahi-daemon regardless of which unit type. A + file like this could be useful for inclusion in a GNOME spin of a + distribution. It will ensure that the units necessary for GNOME + are properly enabled as they are installed. It leaves all other + units untouched, and subject to other (later) preset files, for + example like the one from the first example above. + + + Administrator policy <filename>/etc/systemd/system-preset/00-lennart.preset</filename>: + + enable httpd.service +enable sshd.service +enable postfix.service +disable * + + + This enables three specific services and disables all + others. This is useful for administrators to specifically select + the units to enable, and disable all others. Due to the filename + prefix 00- it will be read early and hence + overrides all other preset policy files. + + + + See Also + + systemd1, + systemctl1, + systemd-delta1 + + + + diff --git a/src/grp-system/systemctl/telinit.xml b/src/grp-system/systemctl/telinit.xml new file mode 100644 index 0000000000..02d31fbd46 --- /dev/null +++ b/src/grp-system/systemctl/telinit.xml @@ -0,0 +1,179 @@ + + + + + + + + + telinit + systemd + + + + Developer + Lennart + Poettering + lennart@poettering.net + + + + + + telinit + 8 + + + + telinit + Change SysV runlevel + + + + + telinit OPTIONS COMMAND + + + + + Description + + telinit may be used to change the SysV + system runlevel. Since the concept of SysV runlevels is obsolete + the runlevel requests will be transparently translated into + systemd unit activation requests. + + + + + Options + + The following options are understood: + + + + + + + + + + + + Do not send wall message before + reboot/halt/power-off. + + + + The following commands are understood: + + + + 0 + + Power-off the machine. This is translated into + an activation request for poweroff.target + and is equivalent to systemctl + poweroff. + + + + 6 + + Reboot the machine. This is translated into an + activation request for reboot.target and + is equivalent to systemctl + reboot. + + + + 2 + 3 + 4 + 5 + + Change the SysV runlevel. This is translated + into an activation request for + runlevel2.target, + runlevel3.target, ... and is equivalent + to systemctl isolate runlevel2.target, + systemctl isolate runlevel3.target, + ... + + + + 1 + s + S + + Change into system rescue mode. This is + translated into an activation request for + rescue.target and is equivalent to + systemctl rescue. + + + + q + Q + + Reload daemon configuration. This is + equivalent to systemctl + daemon-reload. + + + + u + U + + Serialize state, reexecute daemon and + deserialize state again. This is equivalent to + systemctl daemon-reexec. + + + + + + + Exit status + + On success, 0 is returned, a non-zero failure + code otherwise. + + + + Notes + + This is a legacy command available for compatibility only. + It should not be used anymore, as the concept of runlevels is + obsolete. + + + + See Also + + systemd1, + systemctl1, + wall1 + + + + diff --git a/src/grp-system/systemd-shutdown/Makefile b/src/grp-system/systemd-shutdown/Makefile new file mode 100644 index 0000000000..f68758174a --- /dev/null +++ b/src/grp-system/systemd-shutdown/Makefile @@ -0,0 +1,39 @@ +# -*- Mode: makefile; indent-tabs-mode: t -*- +# +# This file is part of systemd. +# +# Copyright 2010-2012 Lennart Poettering +# Copyright 2010-2012 Kay Sievers +# Copyright 2013 Zbigniew Jędrzejewski-Szmek +# Copyright 2013 David Strauss +# Copyright 2016 Luke Shumaker +# +# systemd is free software; you can redistribute it and/or modify it +# under the terms of the GNU Lesser General Public License as published by +# the Free Software Foundation; either version 2.1 of the License, or +# (at your option) any later version. +# +# systemd is distributed in the hope that it will be useful, but +# WITHOUT ANY WARRANTY; without even the implied warranty of +# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU +# Lesser General Public License for more details. +# +# You should have received a copy of the GNU Lesser General Public License +# along with systemd; If not, see . +include $(dir $(lastword $(MAKEFILE_LIST)))/../../../config.mk +include $(topsrcdir)/build-aux/Makefile.head.mk + +rootlibexec_PROGRAMS += systemd-shutdown +systemd_shutdown_SOURCES = \ + src/core/umount.c \ + src/core/umount.h \ + src/core/shutdown.c \ + src/core/mount-setup.c \ + src/core/mount-setup.h \ + src/core/killall.h \ + src/core/killall.c + +systemd_shutdown_LDADD = \ + libsystemd-shared.la + +include $(topsrcdir)/build-aux/Makefile.tail.mk diff --git a/src/grp-system/systemd-shutdown/halt.target b/src/grp-system/systemd-shutdown/halt.target new file mode 100644 index 0000000000..a21d984b26 --- /dev/null +++ b/src/grp-system/systemd-shutdown/halt.target @@ -0,0 +1,17 @@ +# This file is part of systemd. +# +# systemd is free software; you can redistribute it and/or modify it +# under the terms of the GNU Lesser General Public License as published by +# the Free Software Foundation; either version 2.1 of the License, or +# (at your option) any later version. + +[Unit] +Description=Halt +Documentation=man:systemd.special(7) +DefaultDependencies=no +Requires=systemd-halt.service +After=systemd-halt.service +AllowIsolate=yes + +[Install] +Alias=ctrl-alt-del.target diff --git a/src/grp-system/systemd-shutdown/kexec.target b/src/grp-system/systemd-shutdown/kexec.target new file mode 100644 index 0000000000..90795d0c5a --- /dev/null +++ b/src/grp-system/systemd-shutdown/kexec.target @@ -0,0 +1,17 @@ +# This file is part of systemd. +# +# systemd is free software; you can redistribute it and/or modify it +# under the terms of the GNU Lesser General Public License as published by +# the Free Software Foundation; either version 2.1 of the License, or +# (at your option) any later version. + +[Unit] +Description=Reboot via kexec +Documentation=man:systemd.special(7) +DefaultDependencies=no +Requires=systemd-kexec.service +After=systemd-kexec.service +AllowIsolate=yes + +[Install] +Alias=ctrl-alt-del.target diff --git a/src/grp-system/systemd-shutdown/poweroff.target b/src/grp-system/systemd-shutdown/poweroff.target new file mode 100644 index 0000000000..dd92d816ca --- /dev/null +++ b/src/grp-system/systemd-shutdown/poweroff.target @@ -0,0 +1,19 @@ +# This file is part of systemd. +# +# systemd is free software; you can redistribute it and/or modify it +# under the terms of the GNU Lesser General Public License as published by +# the Free Software Foundation; either version 2.1 of the License, or +# (at your option) any later version. + +[Unit] +Description=Power-Off +Documentation=man:systemd.special(7) +DefaultDependencies=no +Requires=systemd-poweroff.service +After=systemd-poweroff.service +AllowIsolate=yes +JobTimeoutSec=30min +JobTimeoutAction=poweroff-force + +[Install] +Alias=ctrl-alt-del.target diff --git a/src/grp-system/systemd-shutdown/reboot.target b/src/grp-system/systemd-shutdown/reboot.target new file mode 100644 index 0000000000..668b98d9e4 --- /dev/null +++ b/src/grp-system/systemd-shutdown/reboot.target @@ -0,0 +1,19 @@ +# This file is part of systemd. +# +# systemd is free software; you can redistribute it and/or modify it +# under the terms of the GNU Lesser General Public License as published by +# the Free Software Foundation; either version 2.1 of the License, or +# (at your option) any later version. + +[Unit] +Description=Reboot +Documentation=man:systemd.special(7) +DefaultDependencies=no +Requires=systemd-reboot.service +After=systemd-reboot.service +AllowIsolate=yes +JobTimeoutSec=30min +JobTimeoutAction=reboot-force + +[Install] +Alias=ctrl-alt-del.target diff --git a/src/grp-system/systemd-shutdown/shutdown.c b/src/grp-system/systemd-shutdown/shutdown.c new file mode 100644 index 0000000000..11e2143089 --- /dev/null +++ b/src/grp-system/systemd-shutdown/shutdown.c @@ -0,0 +1,446 @@ +/*** + This file is part of systemd. + + Copyright 2010 ProFUSION embedded systems + + systemd is free software; you can redistribute it and/or modify it + under the terms of the GNU Lesser General Public License as published by + the Free Software Foundation; either version 2.1 of the License, or + (at your option) any later version. + + systemd is distributed in the hope that it will be useful, but + WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + Lesser General Public License for more details. + + You should have received a copy of the GNU Lesser General Public License + along with systemd; If not, see . +***/ + +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include + +#include + +#include "core/killall.h" +#include "systemd-basic/alloc-util.h" +#include "systemd-basic/cgroup-util.h" +#include "systemd-basic/def.h" +#include "systemd-basic/fileio.h" +#include "systemd-basic/log.h" +#include "systemd-basic/missing.h" +#include "systemd-basic/parse-util.h" +#include "systemd-basic/process-util.h" +#include "systemd-basic/string-util.h" +#include "systemd-basic/terminal-util.h" +#include "systemd-basic/util.h" +#include "systemd-basic/virt.h" +#include "systemd-shared/switch-root.h" +#include "systemd-shared/watchdog.h" + +#include "umount.h" + +#define FINALIZE_ATTEMPTS 50 + +static char* arg_verb; +static uint8_t arg_exit_code; + +static int parse_argv(int argc, char *argv[]) { + enum { + ARG_LOG_LEVEL = 0x100, + ARG_LOG_TARGET, + ARG_LOG_COLOR, + ARG_LOG_LOCATION, + ARG_EXIT_CODE, + }; + + static const struct option options[] = { + { "log-level", required_argument, NULL, ARG_LOG_LEVEL }, + { "log-target", required_argument, NULL, ARG_LOG_TARGET }, + { "log-color", optional_argument, NULL, ARG_LOG_COLOR }, + { "log-location", optional_argument, NULL, ARG_LOG_LOCATION }, + { "exit-code", required_argument, NULL, ARG_EXIT_CODE }, + {} + }; + + int c, r; + + assert(argc >= 1); + assert(argv); + + /* "-" prevents getopt from permuting argv[] and moving the verb away + * from argv[1]. Our interface to initrd promises it'll be there. */ + while ((c = getopt_long(argc, argv, "-", options, NULL)) >= 0) + switch (c) { + + case ARG_LOG_LEVEL: + r = log_set_max_level_from_string(optarg); + if (r < 0) + log_error("Failed to parse log level %s, ignoring.", optarg); + + break; + + case ARG_LOG_TARGET: + r = log_set_target_from_string(optarg); + if (r < 0) + log_error("Failed to parse log target %s, ignoring", optarg); + + break; + + case ARG_LOG_COLOR: + + if (optarg) { + r = log_show_color_from_string(optarg); + if (r < 0) + log_error("Failed to parse log color setting %s, ignoring", optarg); + } else + log_show_color(true); + + break; + + case ARG_LOG_LOCATION: + if (optarg) { + r = log_show_location_from_string(optarg); + if (r < 0) + log_error("Failed to parse log location setting %s, ignoring", optarg); + } else + log_show_location(true); + + break; + + case ARG_EXIT_CODE: + r = safe_atou8(optarg, &arg_exit_code); + if (r < 0) + log_error("Failed to parse exit code %s, ignoring", optarg); + + break; + + case '\001': + if (!arg_verb) + arg_verb = optarg; + else + log_error("Excess arguments, ignoring"); + break; + + case '?': + return -EINVAL; + + default: + assert_not_reached("Unhandled option code."); + } + + if (!arg_verb) { + log_error("Verb argument missing."); + return -EINVAL; + } + + return 0; +} + +static int switch_root_initramfs(void) { + if (mount("/run/initramfs", "/run/initramfs", NULL, MS_BIND, NULL) < 0) + return log_error_errno(errno, "Failed to mount bind /run/initramfs on /run/initramfs: %m"); + + if (mount(NULL, "/run/initramfs", NULL, MS_PRIVATE, NULL) < 0) + return log_error_errno(errno, "Failed to make /run/initramfs private mount: %m"); + + /* switch_root with MS_BIND, because there might still be processes lurking around, which have open file descriptors. + * /run/initramfs/shutdown will take care of these. + * Also do not detach the old root, because /run/initramfs/shutdown needs to access it. + */ + return switch_root("/run/initramfs", "/oldroot", false, MS_BIND); +} + +int main(int argc, char *argv[]) { + bool need_umount, need_swapoff, need_loop_detach, need_dm_detach; + bool in_container, use_watchdog = false; + _cleanup_free_ char *cgroup = NULL; + char *arguments[3]; + unsigned retries; + int cmd, r; + static const char* const dirs[] = {SYSTEM_SHUTDOWN_PATH, NULL}; + + log_parse_environment(); + r = parse_argv(argc, argv); + if (r < 0) + goto error; + + /* journald will die if not gone yet. The log target defaults + * to console, but may have been changed by command line options. */ + + log_close_console(); /* force reopen of /dev/console */ + log_open(); + + umask(0022); + + if (getpid() != 1) { + log_error("Not executed by init (PID 1)."); + r = -EPERM; + goto error; + } + + if (streq(arg_verb, "reboot")) + cmd = RB_AUTOBOOT; + else if (streq(arg_verb, "poweroff")) + cmd = RB_POWER_OFF; + else if (streq(arg_verb, "halt")) + cmd = RB_HALT_SYSTEM; + else if (streq(arg_verb, "kexec")) + cmd = LINUX_REBOOT_CMD_KEXEC; + else if (streq(arg_verb, "exit")) + cmd = 0; /* ignored, just checking that arg_verb is valid */ + else { + r = -EINVAL; + log_error("Unknown action '%s'.", arg_verb); + goto error; + } + + (void) cg_get_root_path(&cgroup); + in_container = detect_container() > 0; + + use_watchdog = !!getenv("WATCHDOG_USEC"); + + /* Lock us into memory */ + mlockall(MCL_CURRENT|MCL_FUTURE); + + /* Synchronize everything that is not written to disk yet at this point already. This is a good idea so that + * slow IO is processed here already and the final process killing spree is not impacted by processes + * desperately trying to sync IO to disk within their timeout. */ + if (!in_container) + sync(); + + log_info("Sending SIGTERM to remaining processes..."); + broadcast_signal(SIGTERM, true, true); + + log_info("Sending SIGKILL to remaining processes..."); + broadcast_signal(SIGKILL, true, false); + + need_umount = !in_container; + need_swapoff = !in_container; + need_loop_detach = !in_container; + need_dm_detach = !in_container; + + /* Unmount all mountpoints, swaps, and loopback devices */ + for (retries = 0; retries < FINALIZE_ATTEMPTS; retries++) { + bool changed = false; + + if (use_watchdog) + watchdog_ping(); + + /* Let's trim the cgroup tree on each iteration so + that we leave an empty cgroup tree around, so that + container managers get a nice notify event when we + are down */ + if (cgroup) + cg_trim(SYSTEMD_CGROUP_CONTROLLER, cgroup, false); + + if (need_umount) { + log_info("Unmounting file systems."); + r = umount_all(&changed); + if (r == 0) { + need_umount = false; + log_info("All filesystems unmounted."); + } else if (r > 0) + log_info("Not all file systems unmounted, %d left.", r); + else + log_error_errno(r, "Failed to unmount file systems: %m"); + } + + if (need_swapoff) { + log_info("Deactivating swaps."); + r = swapoff_all(&changed); + if (r == 0) { + need_swapoff = false; + log_info("All swaps deactivated."); + } else if (r > 0) + log_info("Not all swaps deactivated, %d left.", r); + else + log_error_errno(r, "Failed to deactivate swaps: %m"); + } + + if (need_loop_detach) { + log_info("Detaching loop devices."); + r = loopback_detach_all(&changed); + if (r == 0) { + need_loop_detach = false; + log_info("All loop devices detached."); + } else if (r > 0) + log_info("Not all loop devices detached, %d left.", r); + else + log_error_errno(r, "Failed to detach loop devices: %m"); + } + + if (need_dm_detach) { + log_info("Detaching DM devices."); + r = dm_detach_all(&changed); + if (r == 0) { + need_dm_detach = false; + log_info("All DM devices detached."); + } else if (r > 0) + log_info("Not all DM devices detached, %d left.", r); + else + log_error_errno(r, "Failed to detach DM devices: %m"); + } + + if (!need_umount && !need_swapoff && !need_loop_detach && !need_dm_detach) { + if (retries > 0) + log_info("All filesystems, swaps, loop devices, DM devices detached."); + /* Yay, done */ + goto initrd_jump; + } + + /* If in this iteration we didn't manage to + * unmount/deactivate anything, we simply give up */ + if (!changed) { + log_info("Cannot finalize remaining%s%s%s%s continuing.", + need_umount ? " file systems," : "", + need_swapoff ? " swap devices," : "", + need_loop_detach ? " loop devices," : "", + need_dm_detach ? " DM devices," : ""); + goto initrd_jump; + } + + log_debug("After %u retries, couldn't finalize remaining %s%s%s%s trying again.", + retries + 1, + need_umount ? " file systems," : "", + need_swapoff ? " swap devices," : "", + need_loop_detach ? " loop devices," : "", + need_dm_detach ? " DM devices," : ""); + } + + log_error("Too many iterations, giving up."); + + initrd_jump: + + arguments[0] = NULL; + arguments[1] = arg_verb; + arguments[2] = NULL; + execute_directories(dirs, DEFAULT_TIMEOUT_USEC, arguments); + + if (!in_container && !in_initrd() && + access("/run/initramfs/shutdown", X_OK) == 0) { + r = switch_root_initramfs(); + if (r >= 0) { + argv[0] = (char*) "/shutdown"; + + setsid(); + make_console_stdio(); + + log_info("Successfully changed into root pivot.\n" + "Returning to initrd..."); + + execv("/shutdown", argv); + log_error_errno(errno, "Failed to execute shutdown binary: %m"); + } else + log_error_errno(r, "Failed to switch root to \"/run/initramfs\": %m"); + + } + + if (need_umount || need_swapoff || need_loop_detach || need_dm_detach) + log_error("Failed to finalize %s%s%s%s ignoring", + need_umount ? " file systems," : "", + need_swapoff ? " swap devices," : "", + need_loop_detach ? " loop devices," : "", + need_dm_detach ? " DM devices," : ""); + + /* The kernel will automatically flush ATA disks and suchlike on reboot(), but the file systems need to be + * sync'ed explicitly in advance. So let's do this here, but not needlessly slow down containers. Note that we + * sync'ed things already once above, but we did some more work since then which might have caused IO, hence + * let's doit once more. */ + if (!in_container) + sync(); + + if (streq(arg_verb, "exit")) { + if (in_container) + exit(arg_exit_code); + else { + /* We cannot exit() on the host, fallback on another + * method. */ + cmd = RB_POWER_OFF; + } + } + + switch (cmd) { + + case LINUX_REBOOT_CMD_KEXEC: + + if (!in_container) { + /* We cheat and exec kexec to avoid doing all its work */ + pid_t pid; + + log_info("Rebooting with kexec."); + + pid = fork(); + if (pid < 0) + log_error_errno(errno, "Failed to fork: %m"); + else if (pid == 0) { + + const char * const args[] = { + KEXEC, "-e", NULL + }; + + /* Child */ + + execv(args[0], (char * const *) args); + _exit(EXIT_FAILURE); + } else + wait_for_terminate_and_warn("kexec", pid, true); + } + + cmd = RB_AUTOBOOT; + /* Fall through */ + + case RB_AUTOBOOT: + + if (!in_container) { + _cleanup_free_ char *param = NULL; + + r = read_one_line_file("/run/systemd/reboot-param", ¶m); + if (r < 0) + log_warning_errno(r, "Failed to read reboot parameter file: %m"); + + if (!isempty(param)) { + log_info("Rebooting with argument '%s'.", param); + syscall(SYS_reboot, LINUX_REBOOT_MAGIC1, LINUX_REBOOT_MAGIC2, LINUX_REBOOT_CMD_RESTART2, param); + log_warning_errno(errno, "Failed to reboot with parameter, retrying without: %m"); + } + } + + log_info("Rebooting."); + break; + + case RB_POWER_OFF: + log_info("Powering off."); + break; + + case RB_HALT_SYSTEM: + log_info("Halting system."); + break; + + default: + assert_not_reached("Unknown magic"); + } + + reboot(cmd); + if (errno == EPERM && in_container) { + /* If we are in a container, and we lacked + * CAP_SYS_BOOT just exit, this will kill our + * container for good. */ + log_info("Exiting container."); + exit(0); + } + + r = log_error_errno(errno, "Failed to invoke reboot(): %m"); + + error: + log_emergency_errno(r, "Critical error while doing system shutdown: %m"); + freeze(); +} diff --git a/src/grp-system/systemd-shutdown/systemd-halt.service.in b/src/grp-system/systemd-shutdown/systemd-halt.service.in new file mode 100644 index 0000000000..d55d622c1c --- /dev/null +++ b/src/grp-system/systemd-shutdown/systemd-halt.service.in @@ -0,0 +1,17 @@ +# This file is part of systemd. +# +# systemd is free software; you can redistribute it and/or modify it +# under the terms of the GNU Lesser General Public License as published by +# the Free Software Foundation; either version 2.1 of the License, or +# (at your option) any later version. + +[Unit] +Description=Halt +Documentation=man:systemd-halt.service(8) +DefaultDependencies=no +Requires=shutdown.target umount.target final.target +After=shutdown.target umount.target final.target + +[Service] +Type=oneshot +ExecStart=@SYSTEMCTL@ --force halt diff --git a/src/grp-system/systemd-shutdown/systemd-kexec.service.in b/src/grp-system/systemd-shutdown/systemd-kexec.service.in new file mode 100644 index 0000000000..61303f917f --- /dev/null +++ b/src/grp-system/systemd-shutdown/systemd-kexec.service.in @@ -0,0 +1,17 @@ +# This file is part of systemd. +# +# systemd is free software; you can redistribute it and/or modify it +# under the terms of the GNU Lesser General Public License as published by +# the Free Software Foundation; either version 2.1 of the License, or +# (at your option) any later version. + +[Unit] +Description=Reboot via kexec +Documentation=man:systemd-halt.service(8) +DefaultDependencies=no +Requires=shutdown.target umount.target final.target +After=shutdown.target umount.target final.target + +[Service] +Type=oneshot +ExecStart=@SYSTEMCTL@ --force kexec diff --git a/src/grp-system/systemd-shutdown/systemd-poweroff.service.in b/src/grp-system/systemd-shutdown/systemd-poweroff.service.in new file mode 100644 index 0000000000..3630719733 --- /dev/null +++ b/src/grp-system/systemd-shutdown/systemd-poweroff.service.in @@ -0,0 +1,17 @@ +# This file is part of systemd. +# +# systemd is free software; you can redistribute it and/or modify it +# under the terms of the GNU Lesser General Public License as published by +# the Free Software Foundation; either version 2.1 of the License, or +# (at your option) any later version. + +[Unit] +Description=Power-Off +Documentation=man:systemd-halt.service(8) +DefaultDependencies=no +Requires=shutdown.target umount.target final.target +After=shutdown.target umount.target final.target + +[Service] +Type=oneshot +ExecStart=@SYSTEMCTL@ --force poweroff diff --git a/src/grp-system/systemd-shutdown/systemd-reboot.service.in b/src/grp-system/systemd-shutdown/systemd-reboot.service.in new file mode 100644 index 0000000000..d99bd3e701 --- /dev/null +++ b/src/grp-system/systemd-shutdown/systemd-reboot.service.in @@ -0,0 +1,17 @@ +# This file is part of systemd. +# +# systemd is free software; you can redistribute it and/or modify it +# under the terms of the GNU Lesser General Public License as published by +# the Free Software Foundation; either version 2.1 of the License, or +# (at your option) any later version. + +[Unit] +Description=Reboot +Documentation=man:systemd-halt.service(8) +DefaultDependencies=no +Requires=shutdown.target umount.target final.target +After=shutdown.target umount.target final.target + +[Service] +Type=oneshot +ExecStart=@SYSTEMCTL@ --force reboot diff --git a/src/grp-system/systemd-shutdown/systemd-shutdown.xml b/src/grp-system/systemd-shutdown/systemd-shutdown.xml new file mode 100644 index 0000000000..d16e5d628f --- /dev/null +++ b/src/grp-system/systemd-shutdown/systemd-shutdown.xml @@ -0,0 +1,119 @@ + + + + + + + + + systemd-halt.service + systemd + + + + Developer + Lennart + Poettering + lennart@poettering.net + + + + + + systemd-halt.service + 8 + + + + systemd-halt.service + systemd-poweroff.service + systemd-reboot.service + systemd-kexec.service + systemd-shutdown + System shutdown logic + + + + systemd-halt.service + systemd-poweroff.service + systemd-reboot.service + systemd-kexec.service + /usr/lib/systemd/systemd-shutdown + /usr/lib/systemd/system-shutdown/ + + + + Description + + systemd-halt.service is a system + service that is pulled in by halt.target and + is responsible for the actual system halt. Similarly, + systemd-poweroff.service is pulled in by + poweroff.target, + systemd-reboot.service by + reboot.target and + systemd-kexec.service by + kexec.target to execute the respective + actions. + + When these services are run, they ensure that PID 1 is + replaced by the + /usr/lib/systemd/systemd-shutdown tool which + is then responsible for the actual shutdown. Before shutting down, + this binary will try to unmount all remaining file systems, + disable all remaining swap devices, detach all remaining storage + devices and kill all remaining processes. + + It is necessary to have this code in a separate binary + because otherwise rebooting after an upgrade might be broken — the + running PID 1 could still depend on libraries which are not + available any more, thus keeping the file system busy, which then + cannot be re-mounted read-only. + + Immediately before executing the actual system + halt/poweroff/reboot/kexec systemd-shutdown + will run all executables in + /usr/lib/systemd/system-shutdown/ and pass + one arguments to them: either halt, + poweroff, reboot or + kexec, depending on the chosen action. All + executables in this directory are executed in parallel, and + execution of the action is not continued before all executables + finished. + + Note that systemd-halt.service (and the + related units) should never be executed directly. Instead, trigger + system shutdown with a command such as systemctl + halt or suchlike. + + + + See Also + + systemd1, + systemctl1, + systemd.special7, + reboot2, + systemd-suspend.service8 + + + + diff --git a/src/grp-system/systemd-shutdown/umount.c b/src/grp-system/systemd-shutdown/umount.c new file mode 100644 index 0000000000..8d280c8c8c --- /dev/null +++ b/src/grp-system/systemd-shutdown/umount.c @@ -0,0 +1,616 @@ +/*** + This file is part of systemd. + + Copyright 2010 ProFUSION embedded systems + + systemd is free software; you can redistribute it and/or modify it + under the terms of the GNU Lesser General Public License as published by + the Free Software Foundation; either version 2.1 of the License, or + (at your option) any later version. + + systemd is distributed in the hope that it will be useful, but + WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + Lesser General Public License for more details. + + You should have received a copy of the GNU Lesser General Public License + along with systemd; If not, see . +***/ + +#include +#include +#include +#include +#include + +#include +#include + +#include + +#include "core/mount-setup.h" +#include "systemd-basic/alloc-util.h" +#include "systemd-basic/escape.h" +#include "systemd-basic/fd-util.h" +#include "systemd-basic/list.h" +#include "systemd-basic/path-util.h" +#include "systemd-basic/string-util.h" +#include "systemd-basic/util.h" +#include "systemd-basic/virt.h" +#include "systemd-shared/fstab-util.h" +#include "systemd-shared/udev-util.h" + +#include "umount.h" + +typedef struct MountPoint { + char *path; + char *options; + dev_t devnum; + LIST_FIELDS(struct MountPoint, mount_point); +} MountPoint; + +static void mount_point_free(MountPoint **head, MountPoint *m) { + assert(head); + assert(m); + + LIST_REMOVE(mount_point, *head, m); + + free(m->path); + free(m); +} + +static void mount_points_list_free(MountPoint **head) { + assert(head); + + while (*head) + mount_point_free(head, *head); +} + +static int mount_points_list_get(MountPoint **head) { + _cleanup_fclose_ FILE *proc_self_mountinfo = NULL; + unsigned int i; + int r; + + assert(head); + + proc_self_mountinfo = fopen("/proc/self/mountinfo", "re"); + if (!proc_self_mountinfo) + return -errno; + + for (i = 1;; i++) { + _cleanup_free_ char *path = NULL, *options = NULL; + char *p = NULL; + MountPoint *m; + int k; + + k = fscanf(proc_self_mountinfo, + "%*s " /* (1) mount id */ + "%*s " /* (2) parent id */ + "%*s " /* (3) major:minor */ + "%*s " /* (4) root */ + "%ms " /* (5) mount point */ + "%*s" /* (6) mount flags */ + "%*[^-]" /* (7) optional fields */ + "- " /* (8) separator */ + "%*s " /* (9) file system type */ + "%*s" /* (10) mount source */ + "%ms" /* (11) mount options */ + "%*[^\n]", /* some rubbish at the end */ + &path, &options); + if (k != 2) { + if (k == EOF) + break; + + log_warning("Failed to parse /proc/self/mountinfo:%u.", i); + continue; + } + + r = cunescape(path, UNESCAPE_RELAX, &p); + if (r < 0) + return r; + + /* Ignore mount points we can't unmount because they + * are API or because we are keeping them open (like + * /dev/console). Also, ignore all mounts below API + * file systems, since they are likely virtual too, + * and hence not worth spending time on. Also, in + * unprivileged containers we might lack the rights to + * unmount these things, hence don't bother. */ + if (mount_point_is_api(p) || + mount_point_ignore(p) || + path_startswith(p, "/dev") || + path_startswith(p, "/sys") || + path_startswith(p, "/proc")) { + free(p); + continue; + } + + m = new0(MountPoint, 1); + if (!m) { + free(p); + return -ENOMEM; + } + + m->path = p; + m->options = options; + options = NULL; + + LIST_PREPEND(mount_point, *head, m); + } + + return 0; +} + +static int swap_list_get(MountPoint **head) { + _cleanup_fclose_ FILE *proc_swaps = NULL; + unsigned int i; + int r; + + assert(head); + + proc_swaps = fopen("/proc/swaps", "re"); + if (!proc_swaps) + return (errno == ENOENT) ? 0 : -errno; + + (void) fscanf(proc_swaps, "%*s %*s %*s %*s %*s\n"); + + for (i = 2;; i++) { + MountPoint *swap; + char *dev = NULL, *d; + int k; + + k = fscanf(proc_swaps, + "%ms " /* device/file */ + "%*s " /* type of swap */ + "%*s " /* swap size */ + "%*s " /* used */ + "%*s\n", /* priority */ + &dev); + + if (k != 1) { + if (k == EOF) + break; + + log_warning("Failed to parse /proc/swaps:%u.", i); + free(dev); + continue; + } + + if (endswith(dev, " (deleted)")) { + free(dev); + continue; + } + + r = cunescape(dev, UNESCAPE_RELAX, &d); + free(dev); + if (r < 0) + return r; + + swap = new0(MountPoint, 1); + if (!swap) { + free(d); + return -ENOMEM; + } + + swap->path = d; + LIST_PREPEND(mount_point, *head, swap); + } + + return 0; +} + +static int loopback_list_get(MountPoint **head) { + _cleanup_udev_enumerate_unref_ struct udev_enumerate *e = NULL; + struct udev_list_entry *item = NULL, *first = NULL; + _cleanup_udev_unref_ struct udev *udev = NULL; + int r; + + assert(head); + + udev = udev_new(); + if (!udev) + return -ENOMEM; + + e = udev_enumerate_new(udev); + if (!e) + return -ENOMEM; + + r = udev_enumerate_add_match_subsystem(e, "block"); + if (r < 0) + return r; + + r = udev_enumerate_add_match_sysname(e, "loop*"); + if (r < 0) + return r; + + r = udev_enumerate_add_match_sysattr(e, "loop/backing_file", NULL); + if (r < 0) + return r; + + r = udev_enumerate_scan_devices(e); + if (r < 0) + return r; + + first = udev_enumerate_get_list_entry(e); + udev_list_entry_foreach(item, first) { + MountPoint *lb; + _cleanup_udev_device_unref_ struct udev_device *d; + char *loop; + const char *dn; + + d = udev_device_new_from_syspath(udev, udev_list_entry_get_name(item)); + if (!d) + return -ENOMEM; + + dn = udev_device_get_devnode(d); + if (!dn) + continue; + + loop = strdup(dn); + if (!loop) + return -ENOMEM; + + lb = new0(MountPoint, 1); + if (!lb) { + free(loop); + return -ENOMEM; + } + + lb->path = loop; + LIST_PREPEND(mount_point, *head, lb); + } + + return 0; +} + +static int dm_list_get(MountPoint **head) { + _cleanup_udev_enumerate_unref_ struct udev_enumerate *e = NULL; + struct udev_list_entry *item = NULL, *first = NULL; + _cleanup_udev_unref_ struct udev *udev = NULL; + int r; + + assert(head); + + udev = udev_new(); + if (!udev) + return -ENOMEM; + + e = udev_enumerate_new(udev); + if (!e) + return -ENOMEM; + + r = udev_enumerate_add_match_subsystem(e, "block"); + if (r < 0) + return r; + + r = udev_enumerate_add_match_sysname(e, "dm-*"); + if (r < 0) + return r; + + r = udev_enumerate_scan_devices(e); + if (r < 0) + return r; + + first = udev_enumerate_get_list_entry(e); + udev_list_entry_foreach(item, first) { + MountPoint *m; + _cleanup_udev_device_unref_ struct udev_device *d; + dev_t devnum; + char *node; + const char *dn; + + d = udev_device_new_from_syspath(udev, udev_list_entry_get_name(item)); + if (!d) + return -ENOMEM; + + devnum = udev_device_get_devnum(d); + dn = udev_device_get_devnode(d); + if (major(devnum) == 0 || !dn) + continue; + + node = strdup(dn); + if (!node) + return -ENOMEM; + + m = new(MountPoint, 1); + if (!m) { + free(node); + return -ENOMEM; + } + + m->path = node; + m->devnum = devnum; + LIST_PREPEND(mount_point, *head, m); + } + + return 0; +} + +static int delete_loopback(const char *device) { + _cleanup_close_ int fd = -1; + int r; + + fd = open(device, O_RDONLY|O_CLOEXEC); + if (fd < 0) + return errno == ENOENT ? 0 : -errno; + + r = ioctl(fd, LOOP_CLR_FD, 0); + if (r >= 0) + return 1; + + /* ENXIO: not bound, so no error */ + if (errno == ENXIO) + return 0; + + return -errno; +} + +static int delete_dm(dev_t devnum) { + _cleanup_close_ int fd = -1; + int r; + struct dm_ioctl dm = { + .version = {DM_VERSION_MAJOR, + DM_VERSION_MINOR, + DM_VERSION_PATCHLEVEL}, + .data_size = sizeof(dm), + .dev = devnum, + }; + + assert(major(devnum) != 0); + + fd = open("/dev/mapper/control", O_RDWR|O_CLOEXEC); + if (fd < 0) + return -errno; + + r = ioctl(fd, DM_DEV_REMOVE, &dm); + return r >= 0 ? 0 : -errno; +} + +static int mount_points_list_umount(MountPoint **head, bool *changed, bool log_error) { + MountPoint *m, *n; + int n_failed = 0; + + assert(head); + + LIST_FOREACH_SAFE(mount_point, m, n, *head) { + + /* If we are in a container, don't attempt to + read-only mount anything as that brings no real + benefits, but might confuse the host, as we remount + the superblock here, not the bind mound. */ + if (detect_container() <= 0) { + _cleanup_free_ char *options = NULL; + /* MS_REMOUNT requires that the data parameter + * should be the same from the original mount + * except for the desired changes. Since we want + * to remount read-only, we should filter out + * rw (and ro too, because it confuses the kernel) */ + (void) fstab_filter_options(m->options, "rw\0ro\0", NULL, NULL, &options); + + /* We always try to remount directories + * read-only first, before we go on and umount + * them. + * + * Mount points can be stacked. If a mount + * point is stacked below / or /usr, we + * cannot umount or remount it directly, + * since there is no way to refer to the + * underlying mount. There's nothing we can do + * about it for the general case, but we can + * do something about it if it is aliased + * somehwere else via a bind mount. If we + * explicitly remount the super block of that + * alias read-only we hence should be + * relatively safe regarding keeping the fs we + * can otherwise not see dirty. */ + log_info("Remounting '%s' read-only with options '%s'.", m->path, options); + (void) mount(NULL, m->path, NULL, MS_REMOUNT|MS_RDONLY, options); + } + + /* Skip / and /usr since we cannot unmount that + * anyway, since we are running from it. They have + * already been remounted ro. */ + if (path_equal(m->path, "/") +#ifndef HAVE_SPLIT_USR + || path_equal(m->path, "/usr") +#endif + || path_startswith(m->path, "/run/initramfs") + ) + continue; + + /* Trying to umount. We don't force here since we rely + * on busy NFS and FUSE file systems to return EBUSY + * until we closed everything on top of them. */ + log_info("Unmounting %s.", m->path); + if (umount2(m->path, 0) == 0) { + if (changed) + *changed = true; + + mount_point_free(head, m); + } else if (log_error) { + log_warning_errno(errno, "Could not unmount %s: %m", m->path); + n_failed++; + } + } + + return n_failed; +} + +static int swap_points_list_off(MountPoint **head, bool *changed) { + MountPoint *m, *n; + int n_failed = 0; + + assert(head); + + LIST_FOREACH_SAFE(mount_point, m, n, *head) { + log_info("Deactivating swap %s.", m->path); + if (swapoff(m->path) == 0) { + if (changed) + *changed = true; + + mount_point_free(head, m); + } else { + log_warning_errno(errno, "Could not deactivate swap %s: %m", m->path); + n_failed++; + } + } + + return n_failed; +} + +static int loopback_points_list_detach(MountPoint **head, bool *changed) { + MountPoint *m, *n; + int n_failed = 0, k; + struct stat root_st; + + assert(head); + + k = lstat("/", &root_st); + + LIST_FOREACH_SAFE(mount_point, m, n, *head) { + int r; + struct stat loopback_st; + + if (k >= 0 && + major(root_st.st_dev) != 0 && + lstat(m->path, &loopback_st) >= 0 && + root_st.st_dev == loopback_st.st_rdev) { + n_failed++; + continue; + } + + log_info("Detaching loopback %s.", m->path); + r = delete_loopback(m->path); + if (r >= 0) { + if (r > 0 && changed) + *changed = true; + + mount_point_free(head, m); + } else { + log_warning_errno(errno, "Could not detach loopback %s: %m", m->path); + n_failed++; + } + } + + return n_failed; +} + +static int dm_points_list_detach(MountPoint **head, bool *changed) { + MountPoint *m, *n; + int n_failed = 0, k; + struct stat root_st; + + assert(head); + + k = lstat("/", &root_st); + + LIST_FOREACH_SAFE(mount_point, m, n, *head) { + int r; + + if (k >= 0 && + major(root_st.st_dev) != 0 && + root_st.st_dev == m->devnum) { + n_failed++; + continue; + } + + log_info("Detaching DM %u:%u.", major(m->devnum), minor(m->devnum)); + r = delete_dm(m->devnum); + if (r >= 0) { + if (changed) + *changed = true; + + mount_point_free(head, m); + } else { + log_warning_errno(errno, "Could not detach DM %s: %m", m->path); + n_failed++; + } + } + + return n_failed; +} + +int umount_all(bool *changed) { + int r; + bool umount_changed; + LIST_HEAD(MountPoint, mp_list_head); + + LIST_HEAD_INIT(mp_list_head); + r = mount_points_list_get(&mp_list_head); + if (r < 0) + goto end; + + /* retry umount, until nothing can be umounted anymore */ + do { + umount_changed = false; + + mount_points_list_umount(&mp_list_head, &umount_changed, false); + if (umount_changed) + *changed = true; + + } while (umount_changed); + + /* umount one more time with logging enabled */ + r = mount_points_list_umount(&mp_list_head, &umount_changed, true); + if (r <= 0) + goto end; + + end: + mount_points_list_free(&mp_list_head); + + return r; +} + +int swapoff_all(bool *changed) { + int r; + LIST_HEAD(MountPoint, swap_list_head); + + LIST_HEAD_INIT(swap_list_head); + + r = swap_list_get(&swap_list_head); + if (r < 0) + goto end; + + r = swap_points_list_off(&swap_list_head, changed); + + end: + mount_points_list_free(&swap_list_head); + + return r; +} + +int loopback_detach_all(bool *changed) { + int r; + LIST_HEAD(MountPoint, loopback_list_head); + + LIST_HEAD_INIT(loopback_list_head); + + r = loopback_list_get(&loopback_list_head); + if (r < 0) + goto end; + + r = loopback_points_list_detach(&loopback_list_head, changed); + + end: + mount_points_list_free(&loopback_list_head); + + return r; +} + +int dm_detach_all(bool *changed) { + int r; + LIST_HEAD(MountPoint, dm_list_head); + + LIST_HEAD_INIT(dm_list_head); + + r = dm_list_get(&dm_list_head); + if (r < 0) + goto end; + + r = dm_points_list_detach(&dm_list_head, changed); + + end: + mount_points_list_free(&dm_list_head); + + return r; +} diff --git a/src/grp-system/systemd-shutdown/umount.h b/src/grp-system/systemd-shutdown/umount.h new file mode 100644 index 0000000000..4e2215a47d --- /dev/null +++ b/src/grp-system/systemd-shutdown/umount.h @@ -0,0 +1,28 @@ +#pragma once + +/*** + This file is part of systemd. + + Copyright 2010 ProFUSION embedded systems + + systemd is free software; you can redistribute it and/or modify it + under the terms of the GNU Lesser General Public License as published by + the Free Software Foundation; either version 2.1 of the License, or + (at your option) any later version. + + systemd is distributed in the hope that it will be useful, but + WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + Lesser General Public License for more details. + + You should have received a copy of the GNU Lesser General Public License + along with systemd; If not, see . +***/ + +int umount_all(bool *changed); + +int swapoff_all(bool *changed); + +int loopback_detach_all(bool *changed); + +int dm_detach_all(bool *changed); diff --git a/src/grp-system/systemd/50-systemd-user.xorg b/src/grp-system/systemd/50-systemd-user.xorg new file mode 100755 index 0000000000..4d49767228 --- /dev/null +++ b/src/grp-system/systemd/50-systemd-user.xorg @@ -0,0 +1,7 @@ +#!/bin/sh + +systemctl --user import-environment DISPLAY XAUTHORITY + +if which dbus-update-activation-environment >/dev/null 2>&1; then + dbus-update-activation-environment DISPLAY XAUTHORITY +fi diff --git a/src/grp-system/systemd/Makefile b/src/grp-system/systemd/Makefile new file mode 100644 index 0000000000..7f7fbb963e --- /dev/null +++ b/src/grp-system/systemd/Makefile @@ -0,0 +1,73 @@ +# -*- Mode: makefile; indent-tabs-mode: t -*- +# +# This file is part of systemd. +# +# Copyright 2010-2012 Lennart Poettering +# Copyright 2010-2012 Kay Sievers +# Copyright 2013 Zbigniew Jędrzejewski-Szmek +# Copyright 2013 David Strauss +# Copyright 2016 Luke Shumaker +# +# systemd is free software; you can redistribute it and/or modify it +# under the terms of the GNU Lesser General Public License as published by +# the Free Software Foundation; either version 2.1 of the License, or +# (at your option) any later version. +# +# systemd is distributed in the hope that it will be useful, but +# WITHOUT ANY WARRANTY; without even the implied warranty of +# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU +# Lesser General Public License for more details. +# +# You should have received a copy of the GNU Lesser General Public License +# along with systemd; If not, see . +include $(dir $(lastword $(MAKEFILE_LIST)))/../../../config.mk +include $(topsrcdir)/build-aux/Makefile.head.mk + +rootlibexec_PROGRAMS += systemd +systemd_SOURCES = \ + src/core/main.c + +systemd_CFLAGS = \ + $(SECCOMP_CFLAGS) \ + $(MOUNT_CFLAGS) + +systemd_LDADD = \ + libcore.la + +dist_pkgsysconf_DATA += \ + src/core/system.conf \ + src/core/user.conf + +dist_dbuspolicy_DATA += \ + src/core/org.freedesktop.systemd1.conf + +dist_dbussystemservice_DATA += \ + src/core/org.freedesktop.systemd1.service + +polkitpolicy_in_in_files += \ + src/core/org.freedesktop.systemd1.policy.in.in + +pkgconfigdata_DATA += \ + src/core/systemd.pc + +nodist_rpmmacros_DATA = \ + src/core/macros.systemd + +BUILT_SOURCES += \ + src/core/triggers.systemd + +EXTRA_DIST += \ + src/core/systemd.pc.in \ + src/core/macros.systemd.in \ + src/core/triggers.systemd.in + +dist_xinitrc_SCRIPTS = \ + xorg/50-systemd-user.sh + +dist_systemunit_DATA_busnames += \ + units/org.freedesktop.systemd1.busname + +BUSNAMES_TARGET_WANTS += \ + org.freedesktop.systemd1.busname + +include $(topsrcdir)/build-aux/Makefile.tail.mk diff --git a/src/grp-system/systemd/macros.systemd.in b/src/grp-system/systemd/macros.systemd.in new file mode 100644 index 0000000000..6e8a3b3e3d --- /dev/null +++ b/src/grp-system/systemd/macros.systemd.in @@ -0,0 +1,113 @@ +# -*- Mode: rpm-spec; indent-tabs-mode: nil -*- */ +# +# This file is part of systemd. +# +# Copyright 2012 Lennart Poettering +# +# systemd is free software; you can redistribute it and/or modify it +# under the terms of the GNU Lesser General Public License as published by +# the Free Software Foundation; either version 2.1 of the License, or +# (at your option) any later version. +# +# systemd is distributed in the hope that it will be useful, but +# WITHOUT ANY WARRANTY; without even the implied warranty of +# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU +# Lesser General Public License for more details. +# +# You should have received a copy of the GNU Lesser General Public License +# along with systemd; If not, see . + +# RPM macros for packages installing systemd unit files + +%_unitdir @systemunitdir@ +%_userunitdir @userunitdir@ +%_presetdir @systempresetdir@ +%_udevhwdbdir @udevhwdbdir@ +%_udevrulesdir @udevrulesdir@ +%_journalcatalogdir @catalogdir@ +%_tmpfilesdir @tmpfilesdir@ +%_sysusersdir @sysusersdir@ +%_sysctldir @sysctldir@ +%_binfmtdir @binfmtdir@ +%_systemdgeneratordir @systemgeneratordir@ +%_systemdusergeneratordir @usergeneratordir@ + +%systemd_requires \ +Requires(post): systemd \ +Requires(preun): systemd \ +Requires(postun): systemd \ +%{nil} + +%systemd_ordering \ +OrderWithRequires(post): systemd \ +OrderWithRequires(preun): systemd \ +OrderWithRequires(postun): systemd \ +%{nil} + +%systemd_post() \ +if [ $1 -eq 1 ] ; then \ + # Initial installation \ + systemctl --no-reload preset %{?*} >/dev/null 2>&1 || : \ +fi \ +%{nil} + +%systemd_user_post() %{expand:%systemd_post \\--user \\--global %%{?*}} + +%systemd_preun() \ +if [ $1 -eq 0 ] ; then \ + # Package removal, not upgrade \ + systemctl --no-reload disable --now %{?*} > /dev/null 2>&1 || : \ +fi \ +%{nil} + +%systemd_user_preun() \ +if [ $1 -eq 0 ] ; then \ + # Package removal, not upgrade \ + systemctl --no-reload --user --global disable %{?*} > /dev/null 2>&1 || : \ +fi \ +%{nil} + +%systemd_postun() %{nil} + +%systemd_user_postun() %{nil} + +%systemd_postun_with_restart() \ +if [ $1 -ge 1 ] ; then \ + # Package upgrade, not uninstall \ + systemctl try-restart %{?*} >/dev/null 2>&1 || : \ +fi \ +%{nil} + +%systemd_user_postun_with_restart() %{nil} + +%udev_hwdb_update() \ +udevadm hwdb --update >/dev/null 2>&1 || : \ +%{nil} + +%udev_rules_update() \ +udevadm control --reload >/dev/null 2>&1 || : \ +%{nil} + +%journal_catalog_update() \ +journalctl --update-catalog >/dev/null 2>&1 || : \ +%{nil} + +%tmpfiles_create() \ +systemd-tmpfiles --create %{?*} >/dev/null 2>&1 || : \ +%{nil} + +%sysusers_create() \ +systemd-sysusers %{?*} >/dev/null 2>&1 || : \ +%{nil} + +%sysusers_create_inline() \ +echo %{?*} | systemd-sysusers - >/dev/null 2>&1 || : \ +%{nil} + +%sysctl_apply() \ +@rootlibexecdir@/systemd-sysctl %{?*} >/dev/null 2>&1 || : \ +%{nil} + +%binfmt_apply() \ +@rootlibexecdir@/systemd-binfmt %{?*} >/dev/null 2>&1 || : \ +%{nil} diff --git a/src/grp-system/systemd/main.c b/src/grp-system/systemd/main.c new file mode 100644 index 0000000000..1c53f9aad3 --- /dev/null +++ b/src/grp-system/systemd/main.c @@ -0,0 +1,2195 @@ +/*** + This file is part of systemd. + + Copyright 2010 Lennart Poettering + + systemd is free software; you can redistribute it and/or modify it + under the terms of the GNU Lesser General Public License as published by + the Free Software Foundation; either version 2.1 of the License, or + (at your option) any later version. + + systemd is distributed in the hope that it will be useful, but + WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + Lesser General Public License for more details. + + You should have received a copy of the GNU Lesser General Public License + along with systemd; If not, see . +***/ + +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#ifdef HAVE_SECCOMP +#include +#endif +#ifdef HAVE_VALGRIND_VALGRIND_H +#include +#endif + +#include +#include + +#include "core/dbus-manager.h" +#include "core/hostname-setup.h" +#include "core/ima-setup.h" +#include "core/killall.h" +#include "core/kmod-setup.h" +#include "core/load-fragment.h" +#include "core/loopback-setup.h" +#include "core/machine-id-setup.h" +#include "core/manager.h" +#include "core/mount-setup.h" +#include "core/selinux-setup.h" +#include "core/smack-setup.h" +#include "sd-bus/bus-error.h" +#include "sd-bus/bus-util.h" +#include "systemd-basic/alloc-util.h" +#include "systemd-basic/architecture.h" +#include "systemd-basic/build.h" +#include "systemd-basic/capability-util.h" +#include "systemd-basic/clock-util.h" +#include "systemd-basic/cpu-set-util.h" +#include "systemd-basic/def.h" +#include "systemd-basic/env-util.h" +#include "systemd-basic/fd-util.h" +#include "systemd-basic/fileio.h" +#include "systemd-basic/formats-util.h" +#include "systemd-basic/fs-util.h" +#include "systemd-basic/log.h" +#include "systemd-basic/missing.h" +#include "systemd-basic/parse-util.h" +#include "systemd-basic/proc-cmdline.h" +#include "systemd-basic/process-util.h" +#include "systemd-basic/raw-clone.h" +#include "systemd-basic/rlimit-util.h" +#include "systemd-basic/selinux-util.h" +#include "systemd-basic/signal-util.h" +#include "systemd-basic/special.h" +#include "systemd-basic/stat-util.h" +#include "systemd-basic/stdio-util.h" +#include "systemd-basic/strv.h" +#include "systemd-basic/terminal-util.h" +#include "systemd-basic/umask-util.h" +#include "systemd-basic/user-util.h" +#include "systemd-basic/virt.h" +#include "systemd-shared/conf-parser.h" +#include "systemd-shared/fdset.h" +#include "systemd-shared/pager.h" +#include "systemd-shared/switch-root.h" +#include "systemd-shared/watchdog.h" + +static enum { + ACTION_RUN, + ACTION_HELP, + ACTION_VERSION, + ACTION_TEST, + ACTION_DUMP_CONFIGURATION_ITEMS, + ACTION_DONE +} arg_action = ACTION_RUN; +static char *arg_default_unit = NULL; +static bool arg_system = false; +static bool arg_dump_core = true; +static int arg_crash_chvt = -1; +static bool arg_crash_shell = false; +static bool arg_crash_reboot = false; +static bool arg_confirm_spawn = false; +static ShowStatus arg_show_status = _SHOW_STATUS_UNSET; +static bool arg_switched_root = false; +static bool arg_no_pager = false; +static char ***arg_join_controllers = NULL; +static ExecOutput arg_default_std_output = EXEC_OUTPUT_JOURNAL; +static ExecOutput arg_default_std_error = EXEC_OUTPUT_INHERIT; +static usec_t arg_default_restart_usec = DEFAULT_RESTART_USEC; +static usec_t arg_default_timeout_start_usec = DEFAULT_TIMEOUT_USEC; +static usec_t arg_default_timeout_stop_usec = DEFAULT_TIMEOUT_USEC; +static usec_t arg_default_start_limit_interval = DEFAULT_START_LIMIT_INTERVAL; +static unsigned arg_default_start_limit_burst = DEFAULT_START_LIMIT_BURST; +static usec_t arg_runtime_watchdog = 0; +static usec_t arg_shutdown_watchdog = 10 * USEC_PER_MINUTE; +static char **arg_default_environment = NULL; +static struct rlimit *arg_default_rlimit[_RLIMIT_MAX] = {}; +static uint64_t arg_capability_bounding_set = CAP_ALL; +static nsec_t arg_timer_slack_nsec = NSEC_INFINITY; +static usec_t arg_default_timer_accuracy_usec = 1 * USEC_PER_MINUTE; +static Set* arg_syscall_archs = NULL; +static FILE* arg_serialization = NULL; +static bool arg_default_cpu_accounting = false; +static bool arg_default_io_accounting = false; +static bool arg_default_blockio_accounting = false; +static bool arg_default_memory_accounting = false; +static bool arg_default_tasks_accounting = true; +static uint64_t arg_default_tasks_max = UINT64_MAX; +static sd_id128_t arg_machine_id = {}; + +noreturn static void freeze_or_reboot(void) { + + if (arg_crash_reboot) { + log_notice("Rebooting in 10s..."); + (void) sleep(10); + + log_notice("Rebooting now..."); + (void) reboot(RB_AUTOBOOT); + log_emergency_errno(errno, "Failed to reboot: %m"); + } + + log_emergency("Freezing execution."); + freeze(); +} + +noreturn static void crash(int sig) { + struct sigaction sa; + pid_t pid; + + if (getpid() != 1) + /* Pass this on immediately, if this is not PID 1 */ + (void) raise(sig); + else if (!arg_dump_core) + log_emergency("Caught <%s>, not dumping core.", signal_to_string(sig)); + else { + sa = (struct sigaction) { + .sa_handler = nop_signal_handler, + .sa_flags = SA_NOCLDSTOP|SA_RESTART, + }; + + /* We want to wait for the core process, hence let's enable SIGCHLD */ + (void) sigaction(SIGCHLD, &sa, NULL); + + pid = raw_clone(SIGCHLD); + if (pid < 0) + log_emergency_errno(errno, "Caught <%s>, cannot fork for core dump: %m", signal_to_string(sig)); + else if (pid == 0) { + /* Enable default signal handler for core dump */ + + sa = (struct sigaction) { + .sa_handler = SIG_DFL, + }; + (void) sigaction(sig, &sa, NULL); + + /* Don't limit the coredump size */ + (void) setrlimit(RLIMIT_CORE, &RLIMIT_MAKE_CONST(RLIM_INFINITY)); + + /* Just to be sure... */ + (void) chdir("/"); + + /* Raise the signal again */ + pid = raw_getpid(); + (void) kill(pid, sig); /* raise() would kill the parent */ + + assert_not_reached("We shouldn't be here..."); + _exit(EXIT_FAILURE); + } else { + siginfo_t status; + int r; + + /* Order things nicely. */ + r = wait_for_terminate(pid, &status); + if (r < 0) + log_emergency_errno(r, "Caught <%s>, waitpid() failed: %m", signal_to_string(sig)); + else if (status.si_code != CLD_DUMPED) + log_emergency("Caught <%s>, core dump failed (child "PID_FMT", code=%s, status=%i/%s).", + signal_to_string(sig), + pid, sigchld_code_to_string(status.si_code), + status.si_status, + strna(status.si_code == CLD_EXITED + ? exit_status_to_string(status.si_status, EXIT_STATUS_FULL) + : signal_to_string(status.si_status))); + else + log_emergency("Caught <%s>, dumped core as pid "PID_FMT".", signal_to_string(sig), pid); + } + } + + if (arg_crash_chvt >= 0) + (void) chvt(arg_crash_chvt); + + sa = (struct sigaction) { + .sa_handler = SIG_IGN, + .sa_flags = SA_NOCLDSTOP|SA_NOCLDWAIT|SA_RESTART, + }; + + /* Let the kernel reap children for us */ + (void) sigaction(SIGCHLD, &sa, NULL); + + if (arg_crash_shell) { + log_notice("Executing crash shell in 10s..."); + (void) sleep(10); + + pid = raw_clone(SIGCHLD); + if (pid < 0) + log_emergency_errno(errno, "Failed to fork off crash shell: %m"); + else if (pid == 0) { + (void) setsid(); + (void) make_console_stdio(); + (void) execle("/bin/sh", "/bin/sh", NULL, environ); + + log_emergency_errno(errno, "execle() failed: %m"); + _exit(EXIT_FAILURE); + } else { + log_info("Spawned crash shell as PID "PID_FMT".", pid); + (void) wait_for_terminate(pid, NULL); + } + } + + freeze_or_reboot(); +} + +static void install_crash_handler(void) { + static const struct sigaction sa = { + .sa_handler = crash, + .sa_flags = SA_NODEFER, /* So that we can raise the signal again from the signal handler */ + }; + int r; + + /* We ignore the return value here, since, we don't mind if we + * cannot set up a crash handler */ + r = sigaction_many(&sa, SIGNALS_CRASH_HANDLER, -1); + if (r < 0) + log_debug_errno(r, "I had trouble setting up the crash handler, ignoring: %m"); +} + +static int console_setup(void) { + _cleanup_close_ int tty_fd = -1; + int r; + + tty_fd = open_terminal("/dev/console", O_WRONLY|O_NOCTTY|O_CLOEXEC); + if (tty_fd < 0) + return log_error_errno(tty_fd, "Failed to open /dev/console: %m"); + + /* We don't want to force text mode. plymouth may be showing + * pictures already from initrd. */ + r = reset_terminal_fd(tty_fd, false); + if (r < 0) + return log_error_errno(r, "Failed to reset /dev/console: %m"); + + return 0; +} + +static int parse_crash_chvt(const char *value) { + int b; + + if (safe_atoi(value, &arg_crash_chvt) >= 0) + return 0; + + b = parse_boolean(value); + if (b < 0) + return b; + + if (b > 0) + arg_crash_chvt = 0; /* switch to where kmsg goes */ + else + arg_crash_chvt = -1; /* turn off switching */ + + return 0; +} + +static int set_machine_id(const char *m) { + sd_id128_t t; + assert(m); + + if (sd_id128_from_string(m, &t) < 0) + return -EINVAL; + + if (sd_id128_is_null(t)) + return -EINVAL; + + arg_machine_id = t; + return 0; +} + +static int parse_proc_cmdline_item(const char *key, const char *value) { + + int r; + + assert(key); + + if (streq(key, "systemd.unit") && value) { + + if (!in_initrd()) + return free_and_strdup(&arg_default_unit, value); + + } else if (streq(key, "rd.systemd.unit") && value) { + + if (in_initrd()) + return free_and_strdup(&arg_default_unit, value); + + } else if (streq(key, "systemd.dump_core") && value) { + + r = parse_boolean(value); + if (r < 0) + log_warning("Failed to parse dump core switch %s. Ignoring.", value); + else + arg_dump_core = r; + + } else if (streq(key, "systemd.crash_chvt") && value) { + + if (parse_crash_chvt(value) < 0) + log_warning("Failed to parse crash chvt switch %s. Ignoring.", value); + + } else if (streq(key, "systemd.crash_shell") && value) { + + r = parse_boolean(value); + if (r < 0) + log_warning("Failed to parse crash shell switch %s. Ignoring.", value); + else + arg_crash_shell = r; + + } else if (streq(key, "systemd.crash_reboot") && value) { + + r = parse_boolean(value); + if (r < 0) + log_warning("Failed to parse crash reboot switch %s. Ignoring.", value); + else + arg_crash_reboot = r; + + } else if (streq(key, "systemd.confirm_spawn") && value) { + + r = parse_boolean(value); + if (r < 0) + log_warning("Failed to parse confirm spawn switch %s. Ignoring.", value); + else + arg_confirm_spawn = r; + + } else if (streq(key, "systemd.show_status") && value) { + + r = parse_show_status(value, &arg_show_status); + if (r < 0) + log_warning("Failed to parse show status switch %s. Ignoring.", value); + + } else if (streq(key, "systemd.default_standard_output") && value) { + + r = exec_output_from_string(value); + if (r < 0) + log_warning("Failed to parse default standard output switch %s. Ignoring.", value); + else + arg_default_std_output = r; + + } else if (streq(key, "systemd.default_standard_error") && value) { + + r = exec_output_from_string(value); + if (r < 0) + log_warning("Failed to parse default standard error switch %s. Ignoring.", value); + else + arg_default_std_error = r; + + } else if (streq(key, "systemd.setenv") && value) { + + if (env_assignment_is_valid(value)) { + char **env; + + env = strv_env_set(arg_default_environment, value); + if (env) + arg_default_environment = env; + else + log_warning_errno(ENOMEM, "Setting environment variable '%s' failed, ignoring: %m", value); + } else + log_warning("Environment variable name '%s' is not valid. Ignoring.", value); + + } else if (streq(key, "systemd.machine_id") && value) { + + r = set_machine_id(value); + if (r < 0) + log_warning("MachineID '%s' is not valid. Ignoring.", value); + + } else if (streq(key, "quiet") && !value) { + + if (arg_show_status == _SHOW_STATUS_UNSET) + arg_show_status = SHOW_STATUS_AUTO; + + } else if (streq(key, "debug") && !value) { + + /* Note that log_parse_environment() handles 'debug' + * too, and sets the log level to LOG_DEBUG. */ + + if (detect_container() > 0) + log_set_target(LOG_TARGET_CONSOLE); + + } else if (!value) { + const char *target; + + /* SysV compatibility */ + target = runlevel_to_target(key); + if (target) + return free_and_strdup(&arg_default_unit, target); + + } else if (streq(key, "systemd.default_timeout_start_sec") && value) { + + r = parse_sec(value, &arg_default_timeout_start_usec); + if (r < 0) + log_warning_errno(r, "Failed to parse default start timeout: %s, ignoring.", value); + + if (arg_default_timeout_start_usec <= 0) + arg_default_timeout_start_usec = USEC_INFINITY; + } + + return 0; +} + +#define DEFINE_SETTER(name, func, descr) \ + static int name(const char *unit, \ + const char *filename, \ + unsigned line, \ + const char *section, \ + unsigned section_line, \ + const char *lvalue, \ + int ltype, \ + const char *rvalue, \ + void *data, \ + void *userdata) { \ + \ + int r; \ + \ + assert(filename); \ + assert(lvalue); \ + assert(rvalue); \ + \ + r = func(rvalue); \ + if (r < 0) \ + log_syntax(unit, LOG_ERR, filename, line, r, \ + "Invalid " descr "'%s': %m", \ + rvalue); \ + \ + return 0; \ + } + +DEFINE_SETTER(config_parse_level2, log_set_max_level_from_string, "log level") +DEFINE_SETTER(config_parse_target, log_set_target_from_string, "target") +DEFINE_SETTER(config_parse_color, log_show_color_from_string, "color" ) +DEFINE_SETTER(config_parse_location, log_show_location_from_string, "location") + +static int config_parse_cpu_affinity2( + const char *unit, + const char *filename, + unsigned line, + const char *section, + unsigned section_line, + const char *lvalue, + int ltype, + const char *rvalue, + void *data, + void *userdata) { + + _cleanup_cpu_free_ cpu_set_t *c = NULL; + int ncpus; + + ncpus = parse_cpu_set_and_warn(rvalue, &c, unit, filename, line, lvalue); + if (ncpus < 0) + return ncpus; + + if (sched_setaffinity(0, CPU_ALLOC_SIZE(ncpus), c) < 0) + log_warning("Failed to set CPU affinity: %m"); + + return 0; +} + +static int config_parse_show_status( + const char* unit, + const char *filename, + unsigned line, + const char *section, + unsigned section_line, + const char *lvalue, + int ltype, + const char *rvalue, + void *data, + void *userdata) { + + int k; + ShowStatus *b = data; + + assert(filename); + assert(lvalue); + assert(rvalue); + assert(data); + + k = parse_show_status(rvalue, b); + if (k < 0) { + log_syntax(unit, LOG_ERR, filename, line, k, "Failed to parse show status setting, ignoring: %s", rvalue); + return 0; + } + + return 0; +} + +static int config_parse_crash_chvt( + const char* unit, + const char *filename, + unsigned line, + const char *section, + unsigned section_line, + const char *lvalue, + int ltype, + const char *rvalue, + void *data, + void *userdata) { + + int r; + + assert(filename); + assert(lvalue); + assert(rvalue); + + r = parse_crash_chvt(rvalue); + if (r < 0) { + log_syntax(unit, LOG_ERR, filename, line, r, "Failed to parse CrashChangeVT= setting, ignoring: %s", rvalue); + return 0; + } + + return 0; +} + +static int config_parse_join_controllers(const char *unit, + const char *filename, + unsigned line, + const char *section, + unsigned section_line, + const char *lvalue, + int ltype, + const char *rvalue, + void *data, + void *userdata) { + + const char *whole_rvalue = rvalue; + unsigned n = 0; + + assert(filename); + assert(lvalue); + assert(rvalue); + + arg_join_controllers = strv_free_free(arg_join_controllers); + + for (;;) { + _cleanup_free_ char *word = NULL; + char **l; + int r; + + r = extract_first_word(&rvalue, &word, WHITESPACE, EXTRACT_QUOTES); + if (r < 0) { + log_syntax(unit, LOG_ERR, filename, line, r, "Invalid value for %s: %s", lvalue, whole_rvalue); + return r; + } + if (r == 0) + break; + + l = strv_split(word, ","); + if (!l) + return log_oom(); + strv_uniq(l); + + if (strv_length(l) <= 1) { + strv_free(l); + continue; + } + + if (!arg_join_controllers) { + arg_join_controllers = new(char**, 2); + if (!arg_join_controllers) { + strv_free(l); + return log_oom(); + } + + arg_join_controllers[0] = l; + arg_join_controllers[1] = NULL; + + n = 1; + } else { + char ***a; + char ***t; + + t = new0(char**, n+2); + if (!t) { + strv_free(l); + return log_oom(); + } + + n = 0; + + for (a = arg_join_controllers; *a; a++) { + + if (strv_overlap(*a, l)) { + if (strv_extend_strv(&l, *a, false) < 0) { + strv_free(l); + strv_free_free(t); + return log_oom(); + } + + } else { + char **c; + + c = strv_copy(*a); + if (!c) { + strv_free(l); + strv_free_free(t); + return log_oom(); + } + + t[n++] = c; + } + } + + t[n++] = strv_uniq(l); + + strv_free_free(arg_join_controllers); + arg_join_controllers = t; + } + } + if (!isempty(rvalue)) + log_syntax(unit, LOG_ERR, filename, line, 0, "Trailing garbage, ignoring."); + + return 0; +} + +static int parse_config_file(void) { + + const ConfigTableItem items[] = { + { "Manager", "LogLevel", config_parse_level2, 0, NULL }, + { "Manager", "LogTarget", config_parse_target, 0, NULL }, + { "Manager", "LogColor", config_parse_color, 0, NULL }, + { "Manager", "LogLocation", config_parse_location, 0, NULL }, + { "Manager", "DumpCore", config_parse_bool, 0, &arg_dump_core }, + { "Manager", "CrashChVT", /* legacy */ config_parse_crash_chvt, 0, NULL }, + { "Manager", "CrashChangeVT", config_parse_crash_chvt, 0, NULL }, + { "Manager", "CrashShell", config_parse_bool, 0, &arg_crash_shell }, + { "Manager", "CrashReboot", config_parse_bool, 0, &arg_crash_reboot }, + { "Manager", "ShowStatus", config_parse_show_status, 0, &arg_show_status }, + { "Manager", "CPUAffinity", config_parse_cpu_affinity2, 0, NULL }, + { "Manager", "JoinControllers", config_parse_join_controllers, 0, &arg_join_controllers }, + { "Manager", "RuntimeWatchdogSec", config_parse_sec, 0, &arg_runtime_watchdog }, + { "Manager", "ShutdownWatchdogSec", config_parse_sec, 0, &arg_shutdown_watchdog }, + { "Manager", "CapabilityBoundingSet", config_parse_capability_set, 0, &arg_capability_bounding_set }, +#ifdef HAVE_SECCOMP + { "Manager", "SystemCallArchitectures", config_parse_syscall_archs, 0, &arg_syscall_archs }, +#endif + { "Manager", "TimerSlackNSec", config_parse_nsec, 0, &arg_timer_slack_nsec }, + { "Manager", "DefaultTimerAccuracySec", config_parse_sec, 0, &arg_default_timer_accuracy_usec }, + { "Manager", "DefaultStandardOutput", config_parse_output, 0, &arg_default_std_output }, + { "Manager", "DefaultStandardError", config_parse_output, 0, &arg_default_std_error }, + { "Manager", "DefaultTimeoutStartSec", config_parse_sec, 0, &arg_default_timeout_start_usec }, + { "Manager", "DefaultTimeoutStopSec", config_parse_sec, 0, &arg_default_timeout_stop_usec }, + { "Manager", "DefaultRestartSec", config_parse_sec, 0, &arg_default_restart_usec }, + { "Manager", "DefaultStartLimitInterval", config_parse_sec, 0, &arg_default_start_limit_interval }, /* obsolete alias */ + { "Manager", "DefaultStartLimitIntervalSec",config_parse_sec, 0, &arg_default_start_limit_interval }, + { "Manager", "DefaultStartLimitBurst", config_parse_unsigned, 0, &arg_default_start_limit_burst }, + { "Manager", "DefaultEnvironment", config_parse_environ, 0, &arg_default_environment }, + { "Manager", "DefaultLimitCPU", config_parse_limit, RLIMIT_CPU, arg_default_rlimit }, + { "Manager", "DefaultLimitFSIZE", config_parse_limit, RLIMIT_FSIZE, arg_default_rlimit }, + { "Manager", "DefaultLimitDATA", config_parse_limit, RLIMIT_DATA, arg_default_rlimit }, + { "Manager", "DefaultLimitSTACK", config_parse_limit, RLIMIT_STACK, arg_default_rlimit }, + { "Manager", "DefaultLimitCORE", config_parse_limit, RLIMIT_CORE, arg_default_rlimit }, + { "Manager", "DefaultLimitRSS", config_parse_limit, RLIMIT_RSS, arg_default_rlimit }, + { "Manager", "DefaultLimitNOFILE", config_parse_limit, RLIMIT_NOFILE, arg_default_rlimit }, + { "Manager", "DefaultLimitAS", config_parse_limit, RLIMIT_AS, arg_default_rlimit }, + { "Manager", "DefaultLimitNPROC", config_parse_limit, RLIMIT_NPROC, arg_default_rlimit }, + { "Manager", "DefaultLimitMEMLOCK", config_parse_limit, RLIMIT_MEMLOCK, arg_default_rlimit }, + { "Manager", "DefaultLimitLOCKS", config_parse_limit, RLIMIT_LOCKS, arg_default_rlimit }, + { "Manager", "DefaultLimitSIGPENDING", config_parse_limit, RLIMIT_SIGPENDING, arg_default_rlimit }, + { "Manager", "DefaultLimitMSGQUEUE", config_parse_limit, RLIMIT_MSGQUEUE, arg_default_rlimit }, + { "Manager", "DefaultLimitNICE", config_parse_limit, RLIMIT_NICE, arg_default_rlimit }, + { "Manager", "DefaultLimitRTPRIO", config_parse_limit, RLIMIT_RTPRIO, arg_default_rlimit }, + { "Manager", "DefaultLimitRTTIME", config_parse_limit, RLIMIT_RTTIME, arg_default_rlimit }, + { "Manager", "DefaultCPUAccounting", config_parse_bool, 0, &arg_default_cpu_accounting }, + { "Manager", "DefaultIOAccounting", config_parse_bool, 0, &arg_default_io_accounting }, + { "Manager", "DefaultBlockIOAccounting", config_parse_bool, 0, &arg_default_blockio_accounting }, + { "Manager", "DefaultMemoryAccounting", config_parse_bool, 0, &arg_default_memory_accounting }, + { "Manager", "DefaultTasksAccounting", config_parse_bool, 0, &arg_default_tasks_accounting }, + { "Manager", "DefaultTasksMax", config_parse_tasks_max, 0, &arg_default_tasks_max }, + {} + }; + + const char *fn, *conf_dirs_nulstr; + + fn = arg_system ? + PKGSYSCONFDIR "/system.conf" : + PKGSYSCONFDIR "/user.conf"; + + conf_dirs_nulstr = arg_system ? + CONF_PATHS_NULSTR("systemd/system.conf.d") : + CONF_PATHS_NULSTR("systemd/user.conf.d"); + + config_parse_many(fn, conf_dirs_nulstr, "Manager\0", config_item_table_lookup, items, false, NULL); + + /* Traditionally "0" was used to turn off the default unit timeouts. Fix this up so that we used USEC_INFINITY + * like everywhere else. */ + if (arg_default_timeout_start_usec <= 0) + arg_default_timeout_start_usec = USEC_INFINITY; + if (arg_default_timeout_stop_usec <= 0) + arg_default_timeout_stop_usec = USEC_INFINITY; + + return 0; +} + +static void manager_set_defaults(Manager *m) { + + assert(m); + + m->default_timer_accuracy_usec = arg_default_timer_accuracy_usec; + m->default_std_output = arg_default_std_output; + m->default_std_error = arg_default_std_error; + m->default_timeout_start_usec = arg_default_timeout_start_usec; + m->default_timeout_stop_usec = arg_default_timeout_stop_usec; + m->default_restart_usec = arg_default_restart_usec; + m->default_start_limit_interval = arg_default_start_limit_interval; + m->default_start_limit_burst = arg_default_start_limit_burst; + m->default_cpu_accounting = arg_default_cpu_accounting; + m->default_io_accounting = arg_default_io_accounting; + m->default_blockio_accounting = arg_default_blockio_accounting; + m->default_memory_accounting = arg_default_memory_accounting; + m->default_tasks_accounting = arg_default_tasks_accounting; + m->default_tasks_max = arg_default_tasks_max; + + manager_set_default_rlimits(m, arg_default_rlimit); + manager_environment_add(m, NULL, arg_default_environment); +} + +static int parse_argv(int argc, char *argv[]) { + + enum { + ARG_LOG_LEVEL = 0x100, + ARG_LOG_TARGET, + ARG_LOG_COLOR, + ARG_LOG_LOCATION, + ARG_UNIT, + ARG_SYSTEM, + ARG_USER, + ARG_TEST, + ARG_NO_PAGER, + ARG_VERSION, + ARG_DUMP_CONFIGURATION_ITEMS, + ARG_DUMP_CORE, + ARG_CRASH_CHVT, + ARG_CRASH_SHELL, + ARG_CRASH_REBOOT, + ARG_CONFIRM_SPAWN, + ARG_SHOW_STATUS, + ARG_DESERIALIZE, + ARG_SWITCHED_ROOT, + ARG_DEFAULT_STD_OUTPUT, + ARG_DEFAULT_STD_ERROR, + ARG_MACHINE_ID + }; + + static const struct option options[] = { + { "log-level", required_argument, NULL, ARG_LOG_LEVEL }, + { "log-target", required_argument, NULL, ARG_LOG_TARGET }, + { "log-color", optional_argument, NULL, ARG_LOG_COLOR }, + { "log-location", optional_argument, NULL, ARG_LOG_LOCATION }, + { "unit", required_argument, NULL, ARG_UNIT }, + { "system", no_argument, NULL, ARG_SYSTEM }, + { "user", no_argument, NULL, ARG_USER }, + { "test", no_argument, NULL, ARG_TEST }, + { "no-pager", no_argument, NULL, ARG_NO_PAGER }, + { "help", no_argument, NULL, 'h' }, + { "version", no_argument, NULL, ARG_VERSION }, + { "dump-configuration-items", no_argument, NULL, ARG_DUMP_CONFIGURATION_ITEMS }, + { "dump-core", optional_argument, NULL, ARG_DUMP_CORE }, + { "crash-chvt", required_argument, NULL, ARG_CRASH_CHVT }, + { "crash-shell", optional_argument, NULL, ARG_CRASH_SHELL }, + { "crash-reboot", optional_argument, NULL, ARG_CRASH_REBOOT }, + { "confirm-spawn", optional_argument, NULL, ARG_CONFIRM_SPAWN }, + { "show-status", optional_argument, NULL, ARG_SHOW_STATUS }, + { "deserialize", required_argument, NULL, ARG_DESERIALIZE }, + { "switched-root", no_argument, NULL, ARG_SWITCHED_ROOT }, + { "default-standard-output", required_argument, NULL, ARG_DEFAULT_STD_OUTPUT, }, + { "default-standard-error", required_argument, NULL, ARG_DEFAULT_STD_ERROR, }, + { "machine-id", required_argument, NULL, ARG_MACHINE_ID }, + {} + }; + + int c, r; + + assert(argc >= 1); + assert(argv); + + if (getpid() == 1) + opterr = 0; + + while ((c = getopt_long(argc, argv, "hDbsz:", options, NULL)) >= 0) + + switch (c) { + + case ARG_LOG_LEVEL: + r = log_set_max_level_from_string(optarg); + if (r < 0) { + log_error("Failed to parse log level %s.", optarg); + return r; + } + + break; + + case ARG_LOG_TARGET: + r = log_set_target_from_string(optarg); + if (r < 0) { + log_error("Failed to parse log target %s.", optarg); + return r; + } + + break; + + case ARG_LOG_COLOR: + + if (optarg) { + r = log_show_color_from_string(optarg); + if (r < 0) { + log_error("Failed to parse log color setting %s.", optarg); + return r; + } + } else + log_show_color(true); + + break; + + case ARG_LOG_LOCATION: + if (optarg) { + r = log_show_location_from_string(optarg); + if (r < 0) { + log_error("Failed to parse log location setting %s.", optarg); + return r; + } + } else + log_show_location(true); + + break; + + case ARG_DEFAULT_STD_OUTPUT: + r = exec_output_from_string(optarg); + if (r < 0) { + log_error("Failed to parse default standard output setting %s.", optarg); + return r; + } else + arg_default_std_output = r; + break; + + case ARG_DEFAULT_STD_ERROR: + r = exec_output_from_string(optarg); + if (r < 0) { + log_error("Failed to parse default standard error output setting %s.", optarg); + return r; + } else + arg_default_std_error = r; + break; + + case ARG_UNIT: + + r = free_and_strdup(&arg_default_unit, optarg); + if (r < 0) + return log_error_errno(r, "Failed to set default unit %s: %m", optarg); + + break; + + case ARG_SYSTEM: + arg_system = true; + break; + + case ARG_USER: + arg_system = false; + break; + + case ARG_TEST: + arg_action = ACTION_TEST; + break; + + case ARG_NO_PAGER: + arg_no_pager = true; + break; + + case ARG_VERSION: + arg_action = ACTION_VERSION; + break; + + case ARG_DUMP_CONFIGURATION_ITEMS: + arg_action = ACTION_DUMP_CONFIGURATION_ITEMS; + break; + + case ARG_DUMP_CORE: + if (!optarg) + arg_dump_core = true; + else { + r = parse_boolean(optarg); + if (r < 0) + return log_error_errno(r, "Failed to parse dump core boolean: %s", optarg); + arg_dump_core = r; + } + break; + + case ARG_CRASH_CHVT: + r = parse_crash_chvt(optarg); + if (r < 0) + return log_error_errno(r, "Failed to parse crash virtual terminal index: %s", optarg); + break; + + case ARG_CRASH_SHELL: + if (!optarg) + arg_crash_shell = true; + else { + r = parse_boolean(optarg); + if (r < 0) + return log_error_errno(r, "Failed to parse crash shell boolean: %s", optarg); + arg_crash_shell = r; + } + break; + + case ARG_CRASH_REBOOT: + if (!optarg) + arg_crash_reboot = true; + else { + r = parse_boolean(optarg); + if (r < 0) + return log_error_errno(r, "Failed to parse crash shell boolean: %s", optarg); + arg_crash_reboot = r; + } + break; + + case ARG_CONFIRM_SPAWN: + r = optarg ? parse_boolean(optarg) : 1; + if (r < 0) { + log_error("Failed to parse confirm spawn boolean %s.", optarg); + return r; + } + arg_confirm_spawn = r; + break; + + case ARG_SHOW_STATUS: + if (optarg) { + r = parse_show_status(optarg, &arg_show_status); + if (r < 0) { + log_error("Failed to parse show status boolean %s.", optarg); + return r; + } + } else + arg_show_status = SHOW_STATUS_YES; + break; + + case ARG_DESERIALIZE: { + int fd; + FILE *f; + + r = safe_atoi(optarg, &fd); + if (r < 0 || fd < 0) { + log_error("Failed to parse deserialize option %s.", optarg); + return -EINVAL; + } + + (void) fd_cloexec(fd, true); + + f = fdopen(fd, "r"); + if (!f) + return log_error_errno(errno, "Failed to open serialization fd: %m"); + + safe_fclose(arg_serialization); + arg_serialization = f; + + break; + } + + case ARG_SWITCHED_ROOT: + arg_switched_root = true; + break; + + case ARG_MACHINE_ID: + r = set_machine_id(optarg); + if (r < 0) { + log_error("MachineID '%s' is not valid.", optarg); + return r; + } + break; + + case 'h': + arg_action = ACTION_HELP; + break; + + case 'D': + log_set_max_level(LOG_DEBUG); + break; + + case 'b': + case 's': + case 'z': + /* Just to eat away the sysvinit kernel + * cmdline args without getopt() error + * messages that we'll parse in + * parse_proc_cmdline_word() or ignore. */ + + case '?': + if (getpid() != 1) + return -EINVAL; + else + return 0; + + default: + assert_not_reached("Unhandled option code."); + } + + if (optind < argc && getpid() != 1) { + /* Hmm, when we aren't run as init system + * let's complain about excess arguments */ + + log_error("Excess arguments."); + return -EINVAL; + } + + return 0; +} + +static int help(void) { + + printf("%s [OPTIONS...]\n\n" + "Starts up and maintains the system or user services.\n\n" + " -h --help Show this help\n" + " --test Determine startup sequence, dump it and exit\n" + " --no-pager Do not pipe output into a pager\n" + " --dump-configuration-items Dump understood unit configuration items\n" + " --unit=UNIT Set default unit\n" + " --system Run a system instance, even if PID != 1\n" + " --user Run a user instance\n" + " --dump-core[=BOOL] Dump core on crash\n" + " --crash-vt=NR Change to specified VT on crash\n" + " --crash-reboot[=BOOL] Reboot on crash\n" + " --crash-shell[=BOOL] Run shell on crash\n" + " --confirm-spawn[=BOOL] Ask for confirmation when spawning processes\n" + " --show-status[=BOOL] Show status updates on the console during bootup\n" + " --log-target=TARGET Set log target (console, journal, kmsg, journal-or-kmsg, null)\n" + " --log-level=LEVEL Set log level (debug, info, notice, warning, err, crit, alert, emerg)\n" + " --log-color[=BOOL] Highlight important log messages\n" + " --log-location[=BOOL] Include code location in log messages\n" + " --default-standard-output= Set default standard output for services\n" + " --default-standard-error= Set default standard error output for services\n", + program_invocation_short_name); + + return 0; +} + +static int prepare_reexecute(Manager *m, FILE **_f, FDSet **_fds, bool switching_root) { + _cleanup_fdset_free_ FDSet *fds = NULL; + _cleanup_fclose_ FILE *f = NULL; + int r; + + assert(m); + assert(_f); + assert(_fds); + + r = manager_open_serialization(m, &f); + if (r < 0) + return log_error_errno(r, "Failed to create serialization file: %m"); + + /* Make sure nothing is really destructed when we shut down */ + m->n_reloading++; + bus_manager_send_reloading(m, true); + + fds = fdset_new(); + if (!fds) + return log_oom(); + + r = manager_serialize(m, f, fds, switching_root); + if (r < 0) + return log_error_errno(r, "Failed to serialize state: %m"); + + if (fseeko(f, 0, SEEK_SET) == (off_t) -1) + return log_error_errno(errno, "Failed to rewind serialization fd: %m"); + + r = fd_cloexec(fileno(f), false); + if (r < 0) + return log_error_errno(r, "Failed to disable O_CLOEXEC for serialization: %m"); + + r = fdset_cloexec(fds, false); + if (r < 0) + return log_error_errno(r, "Failed to disable O_CLOEXEC for serialization fds: %m"); + + *_f = f; + *_fds = fds; + + f = NULL; + fds = NULL; + + return 0; +} + +static int bump_rlimit_nofile(struct rlimit *saved_rlimit) { + struct rlimit nl; + int r; + + assert(saved_rlimit); + + /* Save the original RLIMIT_NOFILE so that we can reset it + * later when transitioning from the initrd to the main + * systemd or suchlike. */ + if (getrlimit(RLIMIT_NOFILE, saved_rlimit) < 0) + return log_error_errno(errno, "Reading RLIMIT_NOFILE failed: %m"); + + /* Make sure forked processes get the default kernel setting */ + if (!arg_default_rlimit[RLIMIT_NOFILE]) { + struct rlimit *rl; + + rl = newdup(struct rlimit, saved_rlimit, 1); + if (!rl) + return log_oom(); + + arg_default_rlimit[RLIMIT_NOFILE] = rl; + } + + /* Bump up the resource limit for ourselves substantially */ + nl.rlim_cur = nl.rlim_max = 64*1024; + r = setrlimit_closest(RLIMIT_NOFILE, &nl); + if (r < 0) + return log_error_errno(r, "Setting RLIMIT_NOFILE failed: %m"); + + return 0; +} + +static void test_usr(void) { + + /* Check that /usr is not a separate fs */ + + if (dir_is_empty("/usr") <= 0) + return; + + log_warning("/usr appears to be on its own filesystem and is not already mounted. This is not a supported setup. " + "Some things will probably break (sometimes even silently) in mysterious ways. " + "Consult http://freedesktop.org/wiki/Software/systemd/separate-usr-is-broken for more information."); +} + +static int initialize_join_controllers(void) { + /* By default, mount "cpu" + "cpuacct" together, and "net_cls" + * + "net_prio". We'd like to add "cpuset" to the mix, but + * "cpuset" doesn't really work for groups with no initialized + * attributes. */ + + arg_join_controllers = new(char**, 3); + if (!arg_join_controllers) + return -ENOMEM; + + arg_join_controllers[0] = strv_new("cpu", "cpuacct", NULL); + if (!arg_join_controllers[0]) + goto oom; + + arg_join_controllers[1] = strv_new("net_cls", "net_prio", NULL); + if (!arg_join_controllers[1]) + goto oom; + + arg_join_controllers[2] = NULL; + return 0; + +oom: + arg_join_controllers = strv_free_free(arg_join_controllers); + return -ENOMEM; +} + +static int enforce_syscall_archs(Set *archs) { +#ifdef HAVE_SECCOMP + scmp_filter_ctx *seccomp; + Iterator i; + void *id; + int r; + + seccomp = seccomp_init(SCMP_ACT_ALLOW); + if (!seccomp) + return log_oom(); + + SET_FOREACH(id, arg_syscall_archs, i) { + r = seccomp_arch_add(seccomp, PTR_TO_UINT32(id) - 1); + if (r == -EEXIST) + continue; + if (r < 0) { + log_error_errno(r, "Failed to add architecture to seccomp: %m"); + goto finish; + } + } + + r = seccomp_attr_set(seccomp, SCMP_FLTATR_CTL_NNP, 0); + if (r < 0) { + log_error_errno(r, "Failed to unset NO_NEW_PRIVS: %m"); + goto finish; + } + + r = seccomp_load(seccomp); + if (r < 0) + log_error_errno(r, "Failed to add install architecture seccomp: %m"); + +finish: + seccomp_release(seccomp); + return r; +#else + return 0; +#endif +} + +static int status_welcome(void) { + _cleanup_free_ char *pretty_name = NULL, *ansi_color = NULL; + int r; + + r = parse_env_file("/etc/os-release", NEWLINE, + "PRETTY_NAME", &pretty_name, + "ANSI_COLOR", &ansi_color, + NULL); + if (r == -ENOENT) + r = parse_env_file("/usr/lib/os-release", NEWLINE, + "PRETTY_NAME", &pretty_name, + "ANSI_COLOR", &ansi_color, + NULL); + + if (r < 0 && r != -ENOENT) + log_warning_errno(r, "Failed to read os-release file: %m"); + + if (log_get_show_color()) + return status_printf(NULL, false, false, + "\nWelcome to \x1B[%sm%s\x1B[0m!\n", + isempty(ansi_color) ? "1" : ansi_color, + isempty(pretty_name) ? "GNU/Linux" : pretty_name); + else + return status_printf(NULL, false, false, + "\nWelcome to %s!\n", + isempty(pretty_name) ? "GNU/Linux" : pretty_name); +} + +static int write_container_id(void) { + const char *c; + int r; + + c = getenv("container"); + if (isempty(c)) + return 0; + + RUN_WITH_UMASK(0022) + r = write_string_file("/run/systemd/container", c, WRITE_STRING_FILE_CREATE); + if (r < 0) + return log_warning_errno(r, "Failed to write /run/systemd/container, ignoring: %m"); + + return 1; +} + +static int bump_unix_max_dgram_qlen(void) { + _cleanup_free_ char *qlen = NULL; + unsigned long v; + int r; + + /* Let's bump the net.unix.max_dgram_qlen sysctl. The kernel + * default of 16 is simply too low. We set the value really + * really early during boot, so that it is actually applied to + * all our sockets, including the $NOTIFY_SOCKET one. */ + + r = read_one_line_file("/proc/sys/net/unix/max_dgram_qlen", &qlen); + if (r < 0) + return log_warning_errno(r, "Failed to read AF_UNIX datagram queue length, ignoring: %m"); + + r = safe_atolu(qlen, &v); + if (r < 0) + return log_warning_errno(r, "Failed to parse AF_UNIX datagram queue length, ignoring: %m"); + + if (v >= DEFAULT_UNIX_MAX_DGRAM_QLEN) + return 0; + + qlen = mfree(qlen); + if (asprintf(&qlen, "%lu\n", DEFAULT_UNIX_MAX_DGRAM_QLEN) < 0) + return log_oom(); + + r = write_string_file("/proc/sys/net/unix/max_dgram_qlen", qlen, 0); + if (r < 0) + return log_full_errno(IN_SET(r, -EROFS, -EPERM, -EACCES) ? LOG_DEBUG : LOG_WARNING, r, + "Failed to bump AF_UNIX datagram queue length, ignoring: %m"); + + return 1; +} + +static int fixup_environment(void) { + _cleanup_free_ char *term = NULL; + int r; + + /* We expect the environment to be set correctly + * if run inside a container. */ + if (detect_container() > 0) + return 0; + + /* When started as PID1, the kernel uses /dev/console + * for our stdios and uses TERM=linux whatever the + * backend device used by the console. We try to make + * a better guess here since some consoles might not + * have support for color mode for example. + * + * However if TERM was configured through the kernel + * command line then leave it alone. */ + + r = get_proc_cmdline_key("TERM=", &term); + if (r < 0) + return r; + + if (r == 0) { + term = strdup(default_term_for_tty("/dev/console") + 5); + if (!term) + return -ENOMEM; + } + + if (setenv("TERM", term, 1) < 0) + return -errno; + + return 0; +} + +int main(int argc, char *argv[]) { + Manager *m = NULL; + int r, retval = EXIT_FAILURE; + usec_t before_startup, after_startup; + char timespan[FORMAT_TIMESPAN_MAX]; + FDSet *fds = NULL; + bool reexecute = false; + const char *shutdown_verb = NULL; + dual_timestamp initrd_timestamp = DUAL_TIMESTAMP_NULL; + dual_timestamp userspace_timestamp = DUAL_TIMESTAMP_NULL; + dual_timestamp kernel_timestamp = DUAL_TIMESTAMP_NULL; + dual_timestamp security_start_timestamp = DUAL_TIMESTAMP_NULL; + dual_timestamp security_finish_timestamp = DUAL_TIMESTAMP_NULL; + static char systemd[] = "systemd"; + bool skip_setup = false; + unsigned j; + bool loaded_policy = false; + bool arm_reboot_watchdog = false; + bool queue_default_job = false; + bool empty_etc = false; + char *switch_root_dir = NULL, *switch_root_init = NULL; + struct rlimit saved_rlimit_nofile = RLIMIT_MAKE_CONST(0); + const char *error_message = NULL; + +#ifdef HAVE_SYSV_COMPAT + if (getpid() != 1 && strstr(program_invocation_short_name, "init")) { + /* This is compatibility support for SysV, where + * calling init as a user is identical to telinit. */ + + execv(SYSTEMCTL_BINARY_PATH, argv); + log_error_errno(errno, "Failed to exec " SYSTEMCTL_BINARY_PATH ": %m"); + return 1; + } +#endif + + dual_timestamp_from_monotonic(&kernel_timestamp, 0); + dual_timestamp_get(&userspace_timestamp); + + /* Determine if this is a reexecution or normal bootup. We do + * the full command line parsing much later, so let's just + * have a quick peek here. */ + if (strv_find(argv+1, "--deserialize")) + skip_setup = true; + + /* If we have switched root, do all the special setup + * things */ + if (strv_find(argv+1, "--switched-root")) + skip_setup = false; + + /* If we get started via the /sbin/init symlink then we are + called 'init'. After a subsequent reexecution we are then + called 'systemd'. That is confusing, hence let's call us + systemd right-away. */ + program_invocation_short_name = systemd; + prctl(PR_SET_NAME, systemd); + + saved_argv = argv; + saved_argc = argc; + + log_set_upgrade_syslog_to_journal(true); + + /* Disable the umask logic */ + if (getpid() == 1) + umask(0); + + if (getpid() == 1 && detect_container() <= 0) { + + /* Running outside of a container as PID 1 */ + arg_system = true; + log_set_target(LOG_TARGET_KMSG); + log_open(); + + if (in_initrd()) + initrd_timestamp = userspace_timestamp; + + if (!skip_setup) { + r = mount_setup_early(); + if (r < 0) { + error_message = "Failed to early mount API filesystems"; + goto finish; + } + dual_timestamp_get(&security_start_timestamp); + if (mac_selinux_setup(&loaded_policy) < 0) { + error_message = "Failed to load SELinux policy"; + goto finish; + } else if (ima_setup() < 0) { + error_message = "Failed to load IMA policy"; + goto finish; + } else if (mac_smack_setup(&loaded_policy) < 0) { + error_message = "Failed to load SMACK policy"; + goto finish; + } + dual_timestamp_get(&security_finish_timestamp); + } + + if (mac_selinux_init() < 0) { + error_message = "Failed to initialize SELinux policy"; + goto finish; + } + + if (!skip_setup) { + if (clock_is_localtime(NULL) > 0) { + int min; + + /* + * The very first call of settimeofday() also does a time warp in the kernel. + * + * In the rtc-in-local time mode, we set the kernel's timezone, and rely on + * external tools to take care of maintaining the RTC and do all adjustments. + * This matches the behavior of Windows, which leaves the RTC alone if the + * registry tells that the RTC runs in UTC. + */ + r = clock_set_timezone(&min); + if (r < 0) + log_error_errno(r, "Failed to apply local time delta, ignoring: %m"); + else + log_info("RTC configured in localtime, applying delta of %i minutes to system time.", min); + } else if (!in_initrd()) { + /* + * Do a dummy very first call to seal the kernel's time warp magic. + * + * Do not call this from inside the initrd. The initrd might not + * carry /etc/adjtime with LOCAL, but the real system could be set up + * that way. In such case, we need to delay the time-warp or the sealing + * until we reach the real system. + * + * Do no set the kernel's timezone. The concept of local time cannot + * be supported reliably, the time will jump or be incorrect at every daylight + * saving time change. All kernel local time concepts will be treated + * as UTC that way. + */ + (void) clock_reset_timewarp(); + } + + r = clock_apply_epoch(); + if (r < 0) + log_error_errno(r, "Current system time is before build time, but cannot correct: %m"); + else if (r > 0) + log_info("System time before build time, advancing clock."); + } + + /* Set the default for later on, but don't actually + * open the logs like this for now. Note that if we + * are transitioning from the initrd there might still + * be journal fd open, and we shouldn't attempt + * opening that before we parsed /proc/cmdline which + * might redirect output elsewhere. */ + log_set_target(LOG_TARGET_JOURNAL_OR_KMSG); + + } else if (getpid() == 1) { + /* Running inside a container, as PID 1 */ + arg_system = true; + log_set_target(LOG_TARGET_CONSOLE); + log_close_console(); /* force reopen of /dev/console */ + log_open(); + + /* For the later on, see above... */ + log_set_target(LOG_TARGET_JOURNAL); + + /* clear the kernel timestamp, + * because we are in a container */ + kernel_timestamp = DUAL_TIMESTAMP_NULL; + } else { + /* Running as user instance */ + arg_system = false; + log_set_target(LOG_TARGET_AUTO); + log_open(); + + /* clear the kernel timestamp, + * because we are not PID 1 */ + kernel_timestamp = DUAL_TIMESTAMP_NULL; + } + + if (getpid() == 1) { + /* Don't limit the core dump size, so that coredump handlers such as systemd-coredump (which honour the limit) + * will process core dumps for system services by default. */ + (void) setrlimit(RLIMIT_CORE, &RLIMIT_MAKE_CONST(RLIM_INFINITY)); + + /* But at the same time, turn off the core_pattern logic by default, so that no coredumps are stored + * until the systemd-coredump tool is enabled via sysctl. */ + if (!skip_setup) + (void) write_string_file("/proc/sys/kernel/core_pattern", "|/bin/false", 0); + } + + if (arg_system) { + if (fixup_environment() < 0) { + error_message = "Failed to fix up PID1 environment"; + goto finish; + } + + /* Try to figure out if we can use colors with the console. No + * need to do that for user instances since they never log + * into the console. */ + log_show_color(colors_enabled()); + make_null_stdio(); + } + + /* Initialize default unit */ + r = free_and_strdup(&arg_default_unit, SPECIAL_DEFAULT_TARGET); + if (r < 0) { + log_emergency_errno(r, "Failed to set default unit %s: %m", SPECIAL_DEFAULT_TARGET); + error_message = "Failed to set default unit"; + goto finish; + } + + r = initialize_join_controllers(); + if (r < 0) { + error_message = "Failed to initialize cgroup controllers"; + goto finish; + } + + /* Mount /proc, /sys and friends, so that /proc/cmdline and + * /proc/$PID/fd is available. */ + if (getpid() == 1) { + + /* Load the kernel modules early, so that we kdbus.ko is loaded before kdbusfs shall be mounted */ + if (!skip_setup) + kmod_setup(); + + r = mount_setup(loaded_policy); + if (r < 0) { + error_message = "Failed to mount API filesystems"; + goto finish; + } + } + + /* Reset all signal handlers. */ + (void) reset_all_signal_handlers(); + (void) ignore_signals(SIGNALS_IGNORE, -1); + + arg_default_tasks_max = system_tasks_max_scale(15U, 100U); /* 15% the system PIDs equals 4915 by default. */ + + if (parse_config_file() < 0) { + error_message = "Failed to parse config file"; + goto finish; + } + + if (arg_system) { + r = parse_proc_cmdline(parse_proc_cmdline_item); + if (r < 0) + log_warning_errno(r, "Failed to parse kernel command line, ignoring: %m"); + } + + /* Note that this also parses bits from the kernel command + * line, including "debug". */ + log_parse_environment(); + + if (parse_argv(argc, argv) < 0) { + error_message = "Failed to parse commandline arguments"; + goto finish; + } + + if (arg_action == ACTION_TEST && + geteuid() == 0) { + log_error("Don't run test mode as root."); + goto finish; + } + + if (!arg_system && + arg_action == ACTION_RUN && + sd_booted() <= 0) { + log_error("Trying to run as user instance, but the system has not been booted with systemd."); + goto finish; + } + + if (arg_system && + arg_action == ACTION_RUN && + running_in_chroot() > 0) { + log_error("Cannot be run in a chroot() environment."); + goto finish; + } + + if (arg_action == ACTION_TEST) + skip_setup = true; + + if (arg_action == ACTION_TEST || arg_action == ACTION_HELP) + pager_open(arg_no_pager, false); + + if (arg_action == ACTION_HELP) { + retval = help(); + goto finish; + } else if (arg_action == ACTION_VERSION) { + retval = version(); + goto finish; + } else if (arg_action == ACTION_DUMP_CONFIGURATION_ITEMS) { + unit_dump_config_items(stdout); + retval = EXIT_SUCCESS; + goto finish; + } else if (arg_action == ACTION_DONE) { + retval = EXIT_SUCCESS; + goto finish; + } + + if (!arg_system && + !getenv("XDG_RUNTIME_DIR")) { + log_error("Trying to run as user instance, but $XDG_RUNTIME_DIR is not set."); + goto finish; + } + + assert_se(arg_action == ACTION_RUN || arg_action == ACTION_TEST); + + /* Close logging fds, in order not to confuse fdset below */ + log_close(); + + /* Remember open file descriptors for later deserialization */ + r = fdset_new_fill(&fds); + if (r < 0) { + log_emergency_errno(r, "Failed to allocate fd set: %m"); + error_message = "Failed to allocate fd set"; + goto finish; + } else + fdset_cloexec(fds, true); + + if (arg_serialization) + assert_se(fdset_remove(fds, fileno(arg_serialization)) >= 0); + + if (arg_system) + /* Become a session leader if we aren't one yet. */ + setsid(); + + /* Move out of the way, so that we won't block unmounts */ + assert_se(chdir("/") == 0); + + /* Reset the console, but only if this is really init and we + * are freshly booted */ + if (arg_system && arg_action == ACTION_RUN) { + + /* If we are init, we connect stdin/stdout/stderr to + * /dev/null and make sure we don't have a controlling + * tty. */ + release_terminal(); + + if (getpid() == 1 && !skip_setup) + console_setup(); + } + + /* Open the logging devices, if possible and necessary */ + log_open(); + + if (arg_show_status == _SHOW_STATUS_UNSET) + arg_show_status = SHOW_STATUS_YES; + + /* Make sure we leave a core dump without panicing the + * kernel. */ + if (getpid() == 1) { + install_crash_handler(); + + r = mount_cgroup_controllers(arg_join_controllers); + if (r < 0) + goto finish; + } + + if (arg_system) { + int v; + + log_info(PACKAGE_STRING " running in %ssystem mode. (" SYSTEMD_FEATURES ")", + arg_action == ACTION_TEST ? "test " : "" ); + + v = detect_virtualization(); + if (v > 0) + log_info("Detected virtualization %s.", virtualization_to_string(v)); + + write_container_id(); + + log_info("Detected architecture %s.", architecture_to_string(uname_architecture())); + + if (in_initrd()) + log_info("Running in initial RAM disk."); + + /* Let's check whether /etc is already populated. We + * don't actually really check for that, but use + * /etc/machine-id as flag file. This allows container + * managers and installers to provision a couple of + * files already. If the container manager wants to + * provision the machine ID itself it should pass + * $container_uuid to PID 1. */ + + empty_etc = access("/etc/machine-id", F_OK) < 0; + if (empty_etc) + log_info("Running with unpopulated /etc."); + } else { + _cleanup_free_ char *t; + + t = uid_to_name(getuid()); + log_debug(PACKAGE_STRING " running in %suser mode for user "UID_FMT"/%s. (" SYSTEMD_FEATURES ")", + arg_action == ACTION_TEST ? " test" : "", getuid(), t); + } + + if (arg_system && !skip_setup) { + if (arg_show_status > 0) + status_welcome(); + + hostname_setup(); + machine_id_setup(NULL, arg_machine_id, NULL); + loopback_setup(); + bump_unix_max_dgram_qlen(); + + test_usr(); + } + + if (arg_system && arg_runtime_watchdog > 0 && arg_runtime_watchdog != USEC_INFINITY) + watchdog_set_timeout(&arg_runtime_watchdog); + + if (arg_timer_slack_nsec != NSEC_INFINITY) + if (prctl(PR_SET_TIMERSLACK, arg_timer_slack_nsec) < 0) + log_error_errno(errno, "Failed to adjust timer slack: %m"); + + if (!cap_test_all(arg_capability_bounding_set)) { + r = capability_bounding_set_drop_usermode(arg_capability_bounding_set); + if (r < 0) { + log_emergency_errno(r, "Failed to drop capability bounding set of usermode helpers: %m"); + error_message = "Failed to drop capability bounding set of usermode helpers"; + goto finish; + } + r = capability_bounding_set_drop(arg_capability_bounding_set, true); + if (r < 0) { + log_emergency_errno(r, "Failed to drop capability bounding set: %m"); + error_message = "Failed to drop capability bounding set"; + goto finish; + } + } + + if (arg_syscall_archs) { + r = enforce_syscall_archs(arg_syscall_archs); + if (r < 0) { + error_message = "Failed to set syscall architectures"; + goto finish; + } + } + + if (!arg_system) + /* Become reaper of our children */ + if (prctl(PR_SET_CHILD_SUBREAPER, 1) < 0) + log_warning_errno(errno, "Failed to make us a subreaper: %m"); + + if (arg_system) { + bump_rlimit_nofile(&saved_rlimit_nofile); + + if (empty_etc) { + r = unit_file_preset_all(UNIT_FILE_SYSTEM, false, NULL, UNIT_FILE_PRESET_ENABLE_ONLY, false, NULL, 0); + if (r < 0) + log_full_errno(r == -EEXIST ? LOG_NOTICE : LOG_WARNING, r, "Failed to populate /etc with preset unit settings, ignoring: %m"); + else + log_info("Populated /etc with preset unit settings."); + } + } + + r = manager_new(arg_system ? UNIT_FILE_SYSTEM : UNIT_FILE_USER, arg_action == ACTION_TEST, &m); + if (r < 0) { + log_emergency_errno(r, "Failed to allocate manager object: %m"); + error_message = "Failed to allocate manager object"; + goto finish; + } + + m->confirm_spawn = arg_confirm_spawn; + m->runtime_watchdog = arg_runtime_watchdog; + m->shutdown_watchdog = arg_shutdown_watchdog; + m->userspace_timestamp = userspace_timestamp; + m->kernel_timestamp = kernel_timestamp; + m->initrd_timestamp = initrd_timestamp; + m->security_start_timestamp = security_start_timestamp; + m->security_finish_timestamp = security_finish_timestamp; + + manager_set_defaults(m); + manager_set_show_status(m, arg_show_status); + manager_set_first_boot(m, empty_etc); + + /* Remember whether we should queue the default job */ + queue_default_job = !arg_serialization || arg_switched_root; + + before_startup = now(CLOCK_MONOTONIC); + + r = manager_startup(m, arg_serialization, fds); + if (r < 0) + log_error_errno(r, "Failed to fully start up daemon: %m"); + + /* This will close all file descriptors that were opened, but + * not claimed by any unit. */ + fds = fdset_free(fds); + + arg_serialization = safe_fclose(arg_serialization); + + if (queue_default_job) { + _cleanup_(sd_bus_error_free) sd_bus_error error = SD_BUS_ERROR_NULL; + Unit *target = NULL; + Job *default_unit_job; + + log_debug("Activating default unit: %s", arg_default_unit); + + r = manager_load_unit(m, arg_default_unit, NULL, &error, &target); + if (r < 0) + log_error("Failed to load default target: %s", bus_error_message(&error, r)); + else if (target->load_state == UNIT_ERROR || target->load_state == UNIT_NOT_FOUND) + log_error_errno(target->load_error, "Failed to load default target: %m"); + else if (target->load_state == UNIT_MASKED) + log_error("Default target masked."); + + if (!target || target->load_state != UNIT_LOADED) { + log_info("Trying to load rescue target..."); + + r = manager_load_unit(m, SPECIAL_RESCUE_TARGET, NULL, &error, &target); + if (r < 0) { + log_emergency("Failed to load rescue target: %s", bus_error_message(&error, r)); + error_message = "Failed to load rescue target"; + goto finish; + } else if (target->load_state == UNIT_ERROR || target->load_state == UNIT_NOT_FOUND) { + log_emergency_errno(target->load_error, "Failed to load rescue target: %m"); + error_message = "Failed to load rescue target"; + goto finish; + } else if (target->load_state == UNIT_MASKED) { + log_emergency("Rescue target masked."); + error_message = "Rescue target masked"; + goto finish; + } + } + + assert(target->load_state == UNIT_LOADED); + + if (arg_action == ACTION_TEST) { + printf("-> By units:\n"); + manager_dump_units(m, stdout, "\t"); + } + + r = manager_add_job(m, JOB_START, target, JOB_ISOLATE, &error, &default_unit_job); + if (r == -EPERM) { + log_debug("Default target could not be isolated, starting instead: %s", bus_error_message(&error, r)); + + sd_bus_error_free(&error); + + r = manager_add_job(m, JOB_START, target, JOB_REPLACE, &error, &default_unit_job); + if (r < 0) { + log_emergency("Failed to start default target: %s", bus_error_message(&error, r)); + error_message = "Failed to start default target"; + goto finish; + } + } else if (r < 0) { + log_emergency("Failed to isolate default target: %s", bus_error_message(&error, r)); + error_message = "Failed to isolate default target"; + goto finish; + } + + m->default_unit_job_id = default_unit_job->id; + + after_startup = now(CLOCK_MONOTONIC); + log_full(arg_action == ACTION_TEST ? LOG_INFO : LOG_DEBUG, + "Loaded units and determined initial transaction in %s.", + format_timespan(timespan, sizeof(timespan), after_startup - before_startup, 100 * USEC_PER_MSEC)); + + if (arg_action == ACTION_TEST) { + printf("-> By jobs:\n"); + manager_dump_jobs(m, stdout, "\t"); + retval = EXIT_SUCCESS; + goto finish; + } + } + + for (;;) { + r = manager_loop(m); + if (r < 0) { + log_emergency_errno(r, "Failed to run main loop: %m"); + error_message = "Failed to run main loop"; + goto finish; + } + + switch (m->exit_code) { + + case MANAGER_RELOAD: + log_info("Reloading."); + + r = parse_config_file(); + if (r < 0) + log_error("Failed to parse config file."); + + manager_set_defaults(m); + + r = manager_reload(m); + if (r < 0) + log_error_errno(r, "Failed to reload: %m"); + break; + + case MANAGER_REEXECUTE: + + if (prepare_reexecute(m, &arg_serialization, &fds, false) < 0) { + error_message = "Failed to prepare for reexecution"; + goto finish; + } + + reexecute = true; + log_notice("Reexecuting."); + goto finish; + + case MANAGER_SWITCH_ROOT: + /* Steal the switch root parameters */ + switch_root_dir = m->switch_root; + switch_root_init = m->switch_root_init; + m->switch_root = m->switch_root_init = NULL; + + if (!switch_root_init) + if (prepare_reexecute(m, &arg_serialization, &fds, true) < 0) { + error_message = "Failed to prepare for reexecution"; + goto finish; + } + + reexecute = true; + log_notice("Switching root."); + goto finish; + + case MANAGER_EXIT: + retval = m->return_value; + + if (MANAGER_IS_USER(m)) { + log_debug("Exit."); + goto finish; + } + + /* fallthrough */ + case MANAGER_REBOOT: + case MANAGER_POWEROFF: + case MANAGER_HALT: + case MANAGER_KEXEC: { + static const char * const table[_MANAGER_EXIT_CODE_MAX] = { + [MANAGER_EXIT] = "exit", + [MANAGER_REBOOT] = "reboot", + [MANAGER_POWEROFF] = "poweroff", + [MANAGER_HALT] = "halt", + [MANAGER_KEXEC] = "kexec" + }; + + assert_se(shutdown_verb = table[m->exit_code]); + arm_reboot_watchdog = m->exit_code == MANAGER_REBOOT; + + log_notice("Shutting down."); + goto finish; + } + + default: + assert_not_reached("Unknown exit code."); + } + } + +finish: + pager_close(); + + if (m) + arg_shutdown_watchdog = m->shutdown_watchdog; + + m = manager_free(m); + + for (j = 0; j < ELEMENTSOF(arg_default_rlimit); j++) + arg_default_rlimit[j] = mfree(arg_default_rlimit[j]); + + arg_default_unit = mfree(arg_default_unit); + arg_join_controllers = strv_free_free(arg_join_controllers); + arg_default_environment = strv_free(arg_default_environment); + arg_syscall_archs = set_free(arg_syscall_archs); + + mac_selinux_finish(); + + if (reexecute) { + const char **args; + unsigned i, args_size; + + /* Close and disarm the watchdog, so that the new + * instance can reinitialize it, but doesn't get + * rebooted while we do that */ + watchdog_close(true); + + /* Reset the RLIMIT_NOFILE to the kernel default, so + * that the new systemd can pass the kernel default to + * its child processes */ + if (saved_rlimit_nofile.rlim_cur > 0) + (void) setrlimit(RLIMIT_NOFILE, &saved_rlimit_nofile); + + if (switch_root_dir) { + /* Kill all remaining processes from the + * initrd, but don't wait for them, so that we + * can handle the SIGCHLD for them after + * deserializing. */ + broadcast_signal(SIGTERM, false, true); + + /* And switch root with MS_MOVE, because we remove the old directory afterwards and detach it. */ + r = switch_root(switch_root_dir, "/mnt", true, MS_MOVE); + if (r < 0) + log_error_errno(r, "Failed to switch root, trying to continue: %m"); + } + + args_size = MAX(6, argc+1); + args = newa(const char*, args_size); + + if (!switch_root_init) { + char sfd[DECIMAL_STR_MAX(int) + 1]; + + /* First try to spawn ourselves with the right + * path, and with full serialization. We do + * this only if the user didn't specify an + * explicit init to spawn. */ + + assert(arg_serialization); + assert(fds); + + xsprintf(sfd, "%i", fileno(arg_serialization)); + + i = 0; + args[i++] = SYSTEMD_BINARY_PATH; + if (switch_root_dir) + args[i++] = "--switched-root"; + args[i++] = arg_system ? "--system" : "--user"; + args[i++] = "--deserialize"; + args[i++] = sfd; + args[i++] = NULL; + + assert(i <= args_size); + + /* + * We want valgrind to print its memory usage summary before reexecution. + * Valgrind won't do this is on its own on exec(), but it will do it on exit(). + * Hence, to ensure we get a summary here, fork() off a child, let it exit() cleanly, + * so that it prints the summary, and wait() for it in the parent, before proceeding into the exec(). + */ + valgrind_summary_hack(); + + (void) execv(args[0], (char* const*) args); + } + + /* Try the fallback, if there is any, without any + * serialization. We pass the original argv[] and + * envp[]. (Well, modulo the ordering changes due to + * getopt() in argv[], and some cleanups in envp[], + * but let's hope that doesn't matter.) */ + + arg_serialization = safe_fclose(arg_serialization); + fds = fdset_free(fds); + + /* Reopen the console */ + (void) make_console_stdio(); + + for (j = 1, i = 1; j < (unsigned) argc; j++) + args[i++] = argv[j]; + args[i++] = NULL; + assert(i <= args_size); + + /* Reenable any blocked signals, especially important + * if we switch from initial ramdisk to init=... */ + (void) reset_all_signal_handlers(); + (void) reset_signal_mask(); + + if (switch_root_init) { + args[0] = switch_root_init; + (void) execv(args[0], (char* const*) args); + log_warning_errno(errno, "Failed to execute configured init, trying fallback: %m"); + } + + args[0] = "/sbin/init"; + (void) execv(args[0], (char* const*) args); + + if (errno == ENOENT) { + log_warning("No /sbin/init, trying fallback"); + + args[0] = "/bin/sh"; + args[1] = NULL; + (void) execv(args[0], (char* const*) args); + log_error_errno(errno, "Failed to execute /bin/sh, giving up: %m"); + } else + log_warning_errno(errno, "Failed to execute /sbin/init, giving up: %m"); + } + + arg_serialization = safe_fclose(arg_serialization); + fds = fdset_free(fds); + +#ifdef HAVE_VALGRIND_VALGRIND_H + /* If we are PID 1 and running under valgrind, then let's exit + * here explicitly. valgrind will only generate nice output on + * exit(), not on exec(), hence let's do the former not the + * latter here. */ + if (getpid() == 1 && RUNNING_ON_VALGRIND) + return 0; +#endif + + if (shutdown_verb) { + char log_level[DECIMAL_STR_MAX(int) + 1]; + char exit_code[DECIMAL_STR_MAX(uint8_t) + 1]; + const char* command_line[11] = { + SYSTEMD_SHUTDOWN_BINARY_PATH, + shutdown_verb, + "--log-level", log_level, + "--log-target", + }; + unsigned pos = 5; + _cleanup_strv_free_ char **env_block = NULL; + + assert(command_line[pos] == NULL); + env_block = strv_copy(environ); + + xsprintf(log_level, "%d", log_get_max_level()); + + switch (log_get_target()) { + + case LOG_TARGET_KMSG: + case LOG_TARGET_JOURNAL_OR_KMSG: + case LOG_TARGET_SYSLOG_OR_KMSG: + command_line[pos++] = "kmsg"; + break; + + case LOG_TARGET_NULL: + command_line[pos++] = "null"; + break; + + case LOG_TARGET_CONSOLE: + default: + command_line[pos++] = "console"; + break; + }; + + if (log_get_show_color()) + command_line[pos++] = "--log-color"; + + if (log_get_show_location()) + command_line[pos++] = "--log-location"; + + if (streq(shutdown_verb, "exit")) { + command_line[pos++] = "--exit-code"; + command_line[pos++] = exit_code; + xsprintf(exit_code, "%d", retval); + } + + assert(pos < ELEMENTSOF(command_line)); + + if (arm_reboot_watchdog && arg_shutdown_watchdog > 0 && arg_shutdown_watchdog != USEC_INFINITY) { + char *e; + + /* If we reboot let's set the shutdown + * watchdog and tell the shutdown binary to + * repeatedly ping it */ + r = watchdog_set_timeout(&arg_shutdown_watchdog); + watchdog_close(r < 0); + + /* Tell the binary how often to ping, ignore failure */ + if (asprintf(&e, "WATCHDOG_USEC="USEC_FMT, arg_shutdown_watchdog) > 0) + (void) strv_push(&env_block, e); + } else + watchdog_close(true); + + /* Avoid the creation of new processes forked by the + * kernel; at this point, we will not listen to the + * signals anyway */ + if (detect_container() <= 0) + (void) cg_uninstall_release_agent(SYSTEMD_CGROUP_CONTROLLER); + + execve(SYSTEMD_SHUTDOWN_BINARY_PATH, (char **) command_line, env_block); + log_error_errno(errno, "Failed to execute shutdown binary, %s: %m", + getpid() == 1 ? "freezing" : "quitting"); + } + + if (getpid() == 1) { + if (error_message) + manager_status_printf(NULL, STATUS_TYPE_EMERGENCY, + ANSI_HIGHLIGHT_RED "!!!!!!" ANSI_NORMAL, + "%s, freezing.", error_message); + freeze_or_reboot(); + } + + return retval; +} diff --git a/src/grp-system/systemd/org.freedesktop.systemd1.conf b/src/grp-system/systemd/org.freedesktop.systemd1.conf new file mode 100644 index 0000000000..3c64f20872 --- /dev/null +++ b/src/grp-system/systemd/org.freedesktop.systemd1.conf @@ -0,0 +1,232 @@ + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + diff --git a/src/grp-system/systemd/org.freedesktop.systemd1.policy.in.in b/src/grp-system/systemd/org.freedesktop.systemd1.policy.in.in new file mode 100644 index 0000000000..cc39a9e1c3 --- /dev/null +++ b/src/grp-system/systemd/org.freedesktop.systemd1.policy.in.in @@ -0,0 +1,70 @@ + + + + + + + + The systemd Project + http://www.freedesktop.org/wiki/Software/systemd + + + <_description>Send passphrase back to system + <_message>Authentication is required to send the entered passphrase back to the system. + + no + no + auth_admin_keep + + @rootlibexecdir@/systemd-reply-password + + + + <_description>Manage system services or other units + <_message>Authentication is required to manage system services or other units. + + auth_admin + auth_admin + auth_admin_keep + + + + + <_description>Manage system service or unit files + <_message>Authentication is required to manage system service or unit files. + + auth_admin + auth_admin + auth_admin_keep + + + + + <_description>Set or unset system and service manager environment variables + <_message>Authentication is required to set or unset system and service manager environment variables. + + auth_admin + auth_admin + auth_admin_keep + + + + + <_description>Reload the systemd state + <_message>Authentication is required to reload the systemd state. + + auth_admin + auth_admin + auth_admin_keep + + + + diff --git a/src/grp-system/systemd/org.freedesktop.systemd1.service b/src/grp-system/systemd/org.freedesktop.systemd1.service new file mode 100644 index 0000000000..d4df3e93a2 --- /dev/null +++ b/src/grp-system/systemd/org.freedesktop.systemd1.service @@ -0,0 +1,11 @@ +# This file is part of systemd. +# +# systemd is free software; you can redistribute it and/or modify it +# under the terms of the GNU Lesser General Public License as published by +# the Free Software Foundation; either version 2.1 of the License, or +# (at your option) any later version. + +[D-BUS Service] +Name=org.freedesktop.systemd1 +Exec=/bin/false +User=root diff --git a/src/grp-system/systemd/system.conf b/src/grp-system/systemd/system.conf new file mode 100644 index 0000000000..c6bb050aac --- /dev/null +++ b/src/grp-system/systemd/system.conf @@ -0,0 +1,61 @@ +# This file is part of systemd. +# +# systemd is free software; you can redistribute it and/or modify it +# under the terms of the GNU Lesser General Public License as published by +# the Free Software Foundation; either version 2.1 of the License, or +# (at your option) any later version. +# +# Entries in this file show the compile time defaults. +# You can change settings by editing this file. +# Defaults can be restored by simply deleting this file. +# +# See systemd-system.conf(5) for details. + +[Manager] +#LogLevel=info +#LogTarget=journal-or-kmsg +#LogColor=yes +#LogLocation=no +#DumpCore=yes +#ShowStatus=yes +#CrashChangeVT=no +#CrashShell=no +#CrashReboot=no +#CPUAffinity=1 2 +#JoinControllers=cpu,cpuacct net_cls,net_prio +#RuntimeWatchdogSec=0 +#ShutdownWatchdogSec=10min +#CapabilityBoundingSet= +#SystemCallArchitectures= +#TimerSlackNSec= +#DefaultTimerAccuracySec=1min +#DefaultStandardOutput=journal +#DefaultStandardError=inherit +#DefaultTimeoutStartSec=90s +#DefaultTimeoutStopSec=90s +#DefaultRestartSec=100ms +#DefaultStartLimitIntervalSec=10s +#DefaultStartLimitBurst=5 +#DefaultEnvironment= +#DefaultCPUAccounting=no +#DefaultIOAccounting=no +#DefaultBlockIOAccounting=no +#DefaultMemoryAccounting=no +#DefaultTasksAccounting=yes +#DefaultTasksMax=15% +#DefaultLimitCPU= +#DefaultLimitFSIZE= +#DefaultLimitDATA= +#DefaultLimitSTACK= +#DefaultLimitCORE= +#DefaultLimitRSS= +#DefaultLimitNOFILE= +#DefaultLimitAS= +#DefaultLimitNPROC= +#DefaultLimitMEMLOCK= +#DefaultLimitLOCKS= +#DefaultLimitSIGPENDING= +#DefaultLimitMSGQUEUE= +#DefaultLimitNICE= +#DefaultLimitRTPRIO= +#DefaultLimitRTTIME= diff --git a/src/grp-system/systemd/systemd-system.conf.xml b/src/grp-system/systemd/systemd-system.conf.xml new file mode 100644 index 0000000000..1bb40fd234 --- /dev/null +++ b/src/grp-system/systemd/systemd-system.conf.xml @@ -0,0 +1,393 @@ + + + + + + + + systemd-system.conf + systemd + + + + Developer + Lennart + Poettering + lennart@poettering.net + + + + + + systemd-system.conf + 5 + + + + systemd-system.conf + system.conf.d + systemd-user.conf + user.conf.d + System and session service manager configuration files + + + + /etc/systemd/system.conf, + /etc/systemd/system.conf.d/*.conf, + /run/systemd/system.conf.d/*.conf, + /usr/lib/systemd/system.conf.d/*.conf + /etc/systemd/user.conf, + /etc/systemd/user.conf.d/*.conf, + /run/systemd/user.conf.d/*.conf, + /usr/lib/systemd/user.conf.d/*.conf + + + + Description + + When run as a system instance, systemd interprets the + configuration file system.conf and the files + in system.conf.d directories; when run as a + user instance, systemd interprets the configuration file + user.conf and the files in + user.conf.d directories. These configuration + files contain a few settings controlling basic manager + operations. + + + + + + Options + + All options are configured in the + [Manager] section: + + + + + LogLevel= + LogTarget= + LogColor= + LogLocation= + DumpCore=yes + CrashChangeVT=no + CrashShell=no + CrashReboot=no + ShowStatus=yes + DefaultStandardOutput=journal + DefaultStandardError=inherit + + Configures various parameters of basic manager + operation. These options may be overridden by the respective + command line arguments. See + systemd1 + for details about these command line + arguments. + + + + CPUAffinity= + + Configures the initial CPU affinity for the + init process. Takes a list of CPU indices or ranges separated + by either whitespace or commas. CPU ranges are specified by + the lower and upper CPU indices separated by a + dash. + + + + JoinControllers=cpu,cpuacct net_cls,netprio + + Configures controllers that shall be mounted + in a single hierarchy. By default, systemd will mount all + controllers which are enabled in the kernel in individual + hierarchies, with the exception of those listed in this + setting. Takes a space-separated list of comma-separated + controller names, in order to allow multiple joined + hierarchies. Defaults to 'cpu,cpuacct'. Pass an empty string + to ensure that systemd mounts all controllers in separate + hierarchies. + + Note that this option is only applied once, at very + early boot. If you use an initial RAM disk (initrd) that uses + systemd, it might hence be necessary to rebuild the initrd if + this option is changed, and make sure the new configuration + file is included in it. Otherwise, the initrd might mount the + controller hierarchies in a different configuration than + intended, and the main system cannot remount them + anymore. + + + + RuntimeWatchdogSec= + ShutdownWatchdogSec= + + Configure the hardware watchdog at runtime and + at reboot. Takes a timeout value in seconds (or in other time + units if suffixed with ms, + min, h, + d, w). If + RuntimeWatchdogSec= is set to a non-zero + value, the watchdog hardware + (/dev/watchdog) will be programmed to + automatically reboot the system if it is not contacted within + the specified timeout interval. The system manager will ensure + to contact it at least once in half the specified timeout + interval. This feature requires a hardware watchdog device to + be present, as it is commonly the case in embedded and server + systems. Not all hardware watchdogs allow configuration of the + reboot timeout, in which case the closest available timeout is + picked. ShutdownWatchdogSec= may be used to + configure the hardware watchdog when the system is asked to + reboot. It works as a safety net to ensure that the reboot + takes place even if a clean reboot attempt times out. By + default RuntimeWatchdogSec= defaults to 0 + (off), and ShutdownWatchdogSec= to 10min. + These settings have no effect if a hardware watchdog is not + available. + + + + CapabilityBoundingSet= + + Controls which capabilities to include in the + capability bounding set for PID 1 and its children. See + capabilities7 + for details. Takes a whitespace-separated list of capability + names as read by + cap_from_name3. + Capabilities listed will be included in the bounding set, all + others are removed. If the list of capabilities is prefixed + with ~, all but the listed capabilities will be included, the + effect of the assignment inverted. Note that this option also + affects the respective capabilities in the effective, + permitted and inheritable capability sets. The capability + bounding set may also be individually configured for units + using the CapabilityBoundingSet= directive + for units, but note that capabilities dropped for PID 1 cannot + be regained in individual units, they are lost for + good. + + + + SystemCallArchitectures= + + Takes a space-separated list of architecture + identifiers. Selects from which architectures system calls may + be invoked on this system. This may be used as an effective + way to disable invocation of non-native binaries system-wide, + for example to prohibit execution of 32-bit x86 binaries on + 64-bit x86-64 systems. This option operates system-wide, and + acts similar to the + SystemCallArchitectures= setting of unit + files, see + systemd.exec5 + for details. This setting defaults to the empty list, in which + case no filtering of system calls based on architecture is + applied. Known architecture identifiers are + x86, x86-64, + x32, arm and the special + identifier native. The latter implicitly + maps to the native architecture of the system (or more + specifically, the architecture the system manager was compiled + for). Set this setting to native to + prohibit execution of any non-native binaries. When a binary + executes a system call of an architecture that is not listed + in this setting, it will be immediately terminated with the + SIGSYS signal. + + + + TimerSlackNSec= + + Sets the timer slack in nanoseconds for PID 1, + which is inherited by all executed processes, unless + overridden individually, for example with the + TimerSlackNSec= setting in service units + (for details see + systemd.exec5). + The timer slack controls the accuracy of wake-ups triggered by + system timers. See + prctl2 + for more information. Note that in contrast to most other time + span definitions this parameter takes an integer value in + nano-seconds if no unit is specified. The usual time units are + understood too. + + + + DefaultTimerAccuracySec= + + Sets the default accuracy of timer units. This + controls the global default for the + AccuracySec= setting of timer units, see + systemd.timer5 + for details. AccuracySec= set in individual + units override the global default for the specific unit. + Defaults to 1min. Note that the accuracy of timer units is + also affected by the configured timer slack for PID 1, see + TimerSlackNSec= above. + + + + DefaultTimeoutStartSec= + DefaultTimeoutStopSec= + DefaultRestartSec= + + Configures the default timeouts for starting + and stopping of units, as well as the default time to sleep + between automatic restarts of units, as configured per-unit in + TimeoutStartSec=, + TimeoutStopSec= and + RestartSec= (for services, see + systemd.service5 + for details on the per-unit settings). For non-service units, + DefaultTimeoutStartSec= sets the default + TimeoutSec= + value. DefaultTimeoutStartSec= and + DefaultTimeoutStopSec= default to + 90s. DefaultRestartSec= defaults to + 100ms. + + + + DefaultStartLimitIntervalSec= + DefaultStartLimitBurst= + + Configure the default unit start rate + limiting, as configured per-service by + StartLimitIntervalSec= and + StartLimitBurst=. See + systemd.service5 + for details on the per-service settings. + DefaultStartLimitIntervalSec= defaults to + 10s. DefaultStartLimitBurst= defaults to + 5. + + + + DefaultEnvironment= + + Sets manager environment variables passed to + all executed processes. Takes a space-separated list of + variable assignments. See + environ7 + for details about environment variables. + + Example: + + DefaultEnvironment="VAR1=word1 word2" VAR2=word3 "VAR3=word 5 6" + + Sets three variables + VAR1, + VAR2, + VAR3. + + + + DefaultCPUAccounting= + DefaultBlockIOAccounting= + DefaultMemoryAccounting= + DefaultTasksAccounting= + + Configure the default resource accounting + settings, as configured per-unit by + CPUAccounting=, + BlockIOAccounting=, + MemoryAccounting= and + TasksAccounting=. See + systemd.resource-control5 + for details on the per-unit + settings. DefaulTasksAccounting= defaults + to on, the other three settings to off. + + + + DefaultTasksMax= + + Configure the default value for the per-unit TasksMax= setting. See + systemd.resource-control5 + for details. This setting applies to all unit types that support resource control settings, with the exception + of slice units. Defaults to 15%, which equals 4915 with the kernel's defaults on the host, but might be smaller + in OS containers. + + + + DefaultLimitCPU= + DefaultLimitFSIZE= + DefaultLimitDATA= + DefaultLimitSTACK= + DefaultLimitCORE= + DefaultLimitRSS= + DefaultLimitNOFILE= + DefaultLimitAS= + DefaultLimitNPROC= + DefaultLimitMEMLOCK= + DefaultLimitLOCKS= + DefaultLimitSIGPENDING= + DefaultLimitMSGQUEUE= + DefaultLimitNICE= + DefaultLimitRTPRIO= + DefaultLimitRTTIME= + + These settings control various default + resource limits for units. See + setrlimit2 + for details. The resource limit is possible to specify in two formats, + to set soft and hard limits to the same value, + or to set both limits individually (e.g. DefaultLimitAS=4G:16G). + Use the string infinity to + configure no limit on a specific resource. The multiplicative + suffixes K (=1024), M (=1024*1024) and so on for G, T, P and E + may be used for resource limits measured in bytes + (e.g. DefaultLimitAS=16G). For the limits referring to time values, + the usual time units ms, s, min, h and so on may be used (see + systemd.time7 + for details). Note that if no time unit is specified for + DefaultLimitCPU= the default unit of seconds is + implied, while for DefaultLimitRTTIME= the default + unit of microseconds is implied. Also, note that the effective + granularity of the limits might influence their + enforcement. For example, time limits specified for + DefaultLimitCPU= will be rounded up implicitly to + multiples of 1s. These settings may be overridden in individual units + using the corresponding LimitXXX= directives. Note that these resource + limits are only defaults for units, they are not applied to PID 1 + itself. + + + + + + See Also + + systemd1, + systemd.directives7, + systemd.exec5, + systemd.service5, + environ7, + capabilities7 + + + + diff --git a/src/grp-system/systemd/systemd-tmpfs.tmpfiles b/src/grp-system/systemd/systemd-tmpfs.tmpfiles new file mode 100644 index 0000000000..98050d329d --- /dev/null +++ b/src/grp-system/systemd/systemd-tmpfs.tmpfiles @@ -0,0 +1,14 @@ +# This file is part of systemd. +# +# systemd is free software; you can redistribute it and/or modify it +# under the terms of the GNU Lesser General Public License as published by +# the Free Software Foundation; either version 2.1 of the License, or +# (at your option) any later version. + +# See tmpfiles.d(5) for details + +# Exclude namespace mountpoints created with PrivateTmp=yes +x /tmp/systemd-private-%b-* +X /tmp/systemd-private-%b-*/tmp +x /var/tmp/systemd-private-%b-* +X /var/tmp/systemd-private-%b-*/tmp diff --git a/src/grp-system/systemd/systemd.automount.xml b/src/grp-system/systemd/systemd.automount.xml new file mode 100644 index 0000000000..a43dc981bd --- /dev/null +++ b/src/grp-system/systemd/systemd.automount.xml @@ -0,0 +1,173 @@ + + + + + + + + systemd.automount + systemd + + + + Developer + Lennart + Poettering + lennart@poettering.net + + + + + + systemd.automount + 5 + + + + systemd.automount + Automount unit configuration + + + + automount.automount + + + + Description + + A unit configuration file whose name ends in + .automount encodes information about a file + system automount point controlled and supervised by + systemd. + + This man page lists the configuration options specific to + this unit type. See + systemd.unit5 + for the common options of all unit configuration files. The common + configuration items are configured in the generic [Unit] and + [Install] sections. The automount specific configuration options + are configured in the [Automount] section. + + Automount units must be named after the automount directories they control. Example: the automount point + /home/lennart must be configured in a unit file + home-lennart.automount. For details about the escaping logic used to convert a file system + path to a unit name see + systemd.unit5. Note that + automount units cannot be templated, nor is it possible to add multiple names to an automount unit by creating + additional symlinks to its unit file. + + For each automount unit file a matching mount unit file (see + systemd.mount5 + for details) must exist which is activated when the automount path + is accessed. Example: if an automount unit + home-lennart.automount is active and the user + accesses /home/lennart the mount unit + home-lennart.mount will be activated. + + Automount units may be used to implement on-demand mounting + as well as parallelized mounting of file systems. + + + + Automatic Dependencies + + If an automount unit is beneath another mount unit in the + file system hierarchy, both a requirement and an ordering + dependency between both units are created automatically. + + An implicit Before= dependency is created + between an automount unit and the mount unit it activates. + + Automount units acquire automatic Before= and Conflicts= on + umount.target in order to be stopped during shutdown, unless + DefaultDependencies=no is set in the [Unit] section. + + + + + <filename>fstab</filename> + + Automount units may either be configured via unit files, or + via /etc/fstab (see + fstab5 + for details). + + For details how systemd parses + /etc/fstab see + systemd.mount5. + + If an automount point is configured in both + /etc/fstab and a unit file, the configuration + in the latter takes precedence. + + + + Options + + Automount files must include an [Automount] section, which + carries information about the file system automount points it + supervises. The options specific to the [Automount] section of + automount units are the following: + + + + + Where= + Takes an absolute path of a directory of the + automount point. If the automount point does not exist at time + that the automount point is installed, it is created. This + string must be reflected in the unit filename. (See above.) + This option is mandatory. + + + + DirectoryMode= + Directories of automount points (and any + parent directories) are automatically created if needed. This + option specifies the file system access mode used when + creating these directories. Takes an access mode in octal + notation. Defaults to 0755. + + + TimeoutIdleSec= + Configures an idle timeout. Once the mount has been + idle for the specified time, systemd will attempt to unmount. Takes a + unit-less value in seconds, or a time span value such as "5min 20s". + Pass 0 to disable the timeout logic. The timeout is disabled by + default. + + + + + + See Also + + systemd1, + systemctl1, + systemd.unit5, + systemd.mount5, + mount8, + automount8, + systemd.directives7 + + + + diff --git a/src/grp-system/systemd/systemd.device.xml b/src/grp-system/systemd/systemd.device.xml new file mode 100644 index 0000000000..effed098dd --- /dev/null +++ b/src/grp-system/systemd/systemd.device.xml @@ -0,0 +1,182 @@ + + + + + + + + systemd.device + systemd + + + + Developer + Lennart + Poettering + lennart@poettering.net + + + + + + systemd.device + 5 + + + + systemd.device + Device unit configuration + + + + device.device + + + + Description + + A unit configuration file whose name ends in + .device encodes information about a device unit + as exposed in the + sysfs/udev7 + device tree. + + This unit type has no specific options. See + systemd.unit5 + for the common options of all unit configuration files. The common + configuration items are configured in the generic + [Unit] and [Install] + sections. A separate [Device] section does not + exist, since no device-specific options may be configured. + + systemd will dynamically create device units for all kernel + devices that are marked with the "systemd" udev tag (by default + all block and network devices, and a few others). This may be used + to define dependencies between devices and other units. To tag a + udev device, use TAG+="systemd" in the udev + rules file, see + udev7 + for details. + + Device units are named after the /sys + and /dev paths they control. Example: the + device /dev/sda5 is exposed in + systemd as dev-sda5.device. For details about + the escaping logic used to convert a file system path to a unit + name see + systemd.unit5. + + + + Automatic Dependencies + + Many unit types automatically acquire dependencies on device + units of devices they require. For example, + .socket unit acquire dependencies on the + device units of the network interface specified in + BindToDevice=. Similar, swap and mount units + acquire dependencies on the units encapsulating their backing + block devices. + + + + The udev Database + + The settings of device units may either be configured via + unit files, or directly from the udev database (which is + recommended). The following udev device properties are understood + by systemd: + + + + SYSTEMD_WANTS= + SYSTEMD_USER_WANTS= + Adds dependencies of type + Wants from the device unit to all listed + units. The first form is used by the system systemd instance, + the second by user systemd instances. Those settings may be + used to activate arbitrary units when a specific device + becomes available. + + Note that this and the other tags are not taken into + account unless the device is tagged with the + systemd string in the udev database, + because otherwise the device is not exposed as a systemd unit + (see above). + + Note that systemd will only act on + Wants dependencies when a device first + becomes active. It will not act on them if they are added to + devices that are already active. Use + SYSTEMD_READY= (see below) to influence on + which udev event to trigger the dependencies. + + + + + SYSTEMD_ALIAS= + Adds an additional alias name to the device + unit. This must be an absolute path that is automatically + transformed into a unit name. (See above.) + + + + SYSTEMD_READY= + If set to 0, systemd will consider this device + unplugged even if it shows up in the udev tree. If this + property is unset or set to 1, the device will be considered + plugged if it is visible in the udev tree. This property has + no influence on the behavior when a device disappears from the + udev tree. + + This option is useful to support devices that initially + show up in an uninitialized state in the tree, and for which a + changed event is generated the moment they + are fully set up. Note that SYSTEMD_WANTS= + (see above) is not acted on as long as + SYSTEMD_READY=0 is set for a + device. + + + + ID_MODEL_FROM_DATABASE= + ID_MODEL= + + If set, this property is used as description + string for the device unit. + + + + + + + + See Also + + systemd1, + systemctl1, + systemd.unit5, + udev7, + systemd.directives7 + + + + diff --git a/src/grp-system/systemd/systemd.exec.xml b/src/grp-system/systemd/systemd.exec.xml new file mode 100644 index 0000000000..41ae6e76de --- /dev/null +++ b/src/grp-system/systemd/systemd.exec.xml @@ -0,0 +1,1612 @@ + + + + + + + + systemd.exec + systemd + + + + Developer + Lennart + Poettering + lennart@poettering.net + + + + + + systemd.exec + 5 + + + + systemd.exec + Execution environment configuration + + + + service.service, + socket.socket, + mount.mount, + swap.swap + + + + Description + + Unit configuration files for services, sockets, mount + points, and swap devices share a subset of configuration options + which define the execution environment of spawned + processes. + + This man page lists the configuration options shared by + these four unit types. See + systemd.unit5 + for the common options of all unit configuration files, and + systemd.service5, + systemd.socket5, + systemd.swap5, + and + systemd.mount5 + for more information on the specific unit configuration files. The + execution specific configuration options are configured in the + [Service], [Socket], [Mount], or [Swap] sections, depending on the + unit type. + + + + Automatic Dependencies + + A few execution parameters result in additional, automatic + dependencies to be added. + + Units with WorkingDirectory= or + RootDirectory= set automatically gain + dependencies of type Requires= and + After= on all mount units required to access + the specified paths. This is equivalent to having them listed + explicitly in RequiresMountsFor=. + + Similar, units with PrivateTmp= enabled + automatically get mount unit dependencies for all mounts + required to access /tmp and + /var/tmp. + + Units whose standard output or error output is connected to , + or (or their combinations with console output, see below) automatically acquire dependencies + of type After= on systemd-journald.socket. + + + + Options + + + + + WorkingDirectory= + + Takes a directory path relative to the service's root + directory specified by RootDirectory=, or the + special value ~. Sets the working directory + for executed processes. If set to ~, the + home directory of the user specified in + User= is used. If not set, defaults to the + root directory when systemd is running as a system instance + and the respective user's home directory if run as user. If + the setting is prefixed with the - + character, a missing working directory is not considered + fatal. If RootDirectory= is not set, then + WorkingDirectory= is relative to the root of + the system running the service manager. + Note that setting this parameter might result in + additional dependencies to be added to the unit (see + above). + + + + RootDirectory= + + Takes a directory path relative to the host's root directory + (i.e. the root of the system running the service manager). Sets the + root directory for executed processes, with the chroot2 + system call. If this is used, it must be ensured that the + process binary and all its auxiliary files are available in + the chroot() jail. Note that setting this + parameter might result in additional dependencies to be added + to the unit (see above). + + + + User= + Group= + + Sets the Unix user or group that the processes + are executed as, respectively. Takes a single user or group + name or ID as argument. If no group is set, the default group + of the user is chosen. These do not affect commands prefixed with +. + + + + SupplementaryGroups= + + Sets the supplementary Unix groups the + processes are executed as. This takes a space-separated list + of group names or IDs. This option may be specified more than + once, in which case all listed groups are set as supplementary + groups. When the empty string is assigned, the list of + supplementary groups is reset, and all assignments prior to + this one will have no effect. In any way, this option does not + override, but extends the list of supplementary groups + configured in the system group database for the + user. This does not affect commands prefixed with +. + + + + Nice= + + Sets the default nice level (scheduling + priority) for executed processes. Takes an integer between -20 + (highest priority) and 19 (lowest priority). See + setpriority2 + for details. + + + + OOMScoreAdjust= + + Sets the adjustment level for the + Out-Of-Memory killer for executed processes. Takes an integer + between -1000 (to disable OOM killing for this process) and + 1000 (to make killing of this process under memory pressure + very likely). See proc.txt + for details. + + + + IOSchedulingClass= + + Sets the I/O scheduling class for executed + processes. Takes an integer between 0 and 3 or one of the + strings , , + or . See + ioprio_set2 + for details. + + + + IOSchedulingPriority= + + Sets the I/O scheduling priority for executed + processes. Takes an integer between 0 (highest priority) and 7 + (lowest priority). The available priorities depend on the + selected I/O scheduling class (see above). See + ioprio_set2 + for details. + + + + CPUSchedulingPolicy= + + Sets the CPU scheduling policy for executed + processes. Takes one of + , + , + , + or + . See + sched_setscheduler2 + for details. + + + + CPUSchedulingPriority= + + Sets the CPU scheduling priority for executed + processes. The available priority range depends on the + selected CPU scheduling policy (see above). For real-time + scheduling policies an integer between 1 (lowest priority) and + 99 (highest priority) can be used. See + sched_setscheduler2 + for details. + + + + CPUSchedulingResetOnFork= + + Takes a boolean argument. If true, elevated + CPU scheduling priorities and policies will be reset when the + executed processes fork, and can hence not leak into child + processes. See + sched_setscheduler2 + for details. Defaults to false. + + + + CPUAffinity= + + Controls the CPU affinity of the executed + processes. Takes a list of CPU indices or ranges separated by + either whitespace or commas. CPU ranges are specified by the + lower and upper CPU indices separated by a dash. + This option may be specified more than once, in which case the + specified CPU affinity masks are merged. If the empty string + is assigned, the mask is reset, all assignments prior to this + will have no effect. See + sched_setaffinity2 + for details. + + + + UMask= + + Controls the file mode creation mask. Takes an + access mode in octal notation. See + umask2 + for details. Defaults to 0022. + + + + Environment= + + Sets environment variables for executed + processes. Takes a space-separated list of variable + assignments. This option may be specified more than once, in + which case all listed variables will be set. If the same + variable is set twice, the later setting will override the + earlier setting. If the empty string is assigned to this + option, the list of environment variables is reset, all prior + assignments have no effect. Variable expansion is not + performed inside the strings, however, specifier expansion is + possible. The $ character has no special meaning. If you need + to assign a value containing spaces to a variable, use double + quotes (") for the assignment. + + Example: + Environment="VAR1=word1 word2" VAR2=word3 "VAR3=$word 5 6" + gives three variables VAR1, + VAR2, VAR3 + with the values word1 word2, + word3, $word 5 6. + + + + See + environ7 + for details about environment variables. + + + EnvironmentFile= + Similar to Environment= but + reads the environment variables from a text file. The text + file should contain new-line-separated variable assignments. + Empty lines, lines without an = separator, + or lines starting with ; or # will be ignored, + which may be used for commenting. A line ending with a + backslash will be concatenated with the following one, + allowing multiline variable definitions. The parser strips + leading and trailing whitespace from the values of + assignments, unless you use double quotes ("). + + The argument passed should be an absolute filename or + wildcard expression, optionally prefixed with + -, which indicates that if the file does + not exist, it will not be read and no error or warning message + is logged. This option may be specified more than once in + which case all specified files are read. If the empty string + is assigned to this option, the list of file to read is reset, + all prior assignments have no effect. + + The files listed with this directive will be read + shortly before the process is executed (more specifically, + after all processes from a previous unit state terminated. + This means you can generate these files in one unit state, and + read it with this option in the next). + + Settings from these + files override settings made with + Environment=. If the same variable is set + twice from these files, the files will be read in the order + they are specified and the later setting will override the + earlier setting. + + + + PassEnvironment= + + Pass environment variables from the systemd system + manager to executed processes. Takes a space-separated list of variable + names. This option may be specified more than once, in which case all + listed variables will be set. If the empty string is assigned to this + option, the list of environment variables is reset, all prior + assignments have no effect. Variables that are not set in the system + manager will not be passed and will be silently ignored. + + Variables passed from this setting are overridden by those passed + from Environment= or + EnvironmentFile=. + + Example: + PassEnvironment=VAR1 VAR2 VAR3 + passes three variables VAR1, + VAR2, VAR3 + with the values set for those variables in PID1. + + + See + environ7 + for details about environment variables. + + + + StandardInput= + Controls where file descriptor 0 (STDIN) of + the executed processes is connected to. Takes one of + , + , + , + or + . + + If is selected, standard input + will be connected to /dev/null, i.e. all + read attempts by the process will result in immediate + EOF. + + If is selected, standard input is + connected to a TTY (as configured by + TTYPath=, see below) and the executed + process becomes the controlling process of the terminal. If + the terminal is already being controlled by another process, + the executed process waits until the current controlling + process releases the terminal. + + is similar to + , but the executed process is forcefully + and immediately made the controlling process of the terminal, + potentially removing previous controlling processes from the + terminal. + + is similar to + but if the terminal already has a + controlling process start-up of the executed process + fails. + + The option is only valid in + socket-activated services, and only when the socket + configuration file (see + systemd.socket5 + for details) specifies a single socket only. If this option is + set, standard input will be connected to the socket the + service was activated from, which is primarily useful for + compatibility with daemons designed for use with the + traditional + inetd8 + daemon. + + This setting defaults to + . + + + + StandardOutput= + Controls where file descriptor 1 (STDOUT) of + the executed processes is connected to. Takes one of + , + , + , + , + , + , + , + , + or + . + + duplicates the file descriptor + of standard input for standard output. + + connects standard output to + /dev/null, i.e. everything written to it + will be lost. + + connects standard output to a tty + (as configured via TTYPath=, see below). If + the TTY is used for output only, the executed process will not + become the controlling process of the terminal, and will not + fail or wait for other processes to release the + terminal. + + connects standard output with + the journal which is accessible via + journalctl1. + Note that everything that is written to syslog or kmsg (see + below) is implicitly stored in the journal as well, the + specific two options listed below are hence supersets of this + one. + + connects standard output to the + syslog3 + system syslog service, in addition to the journal. Note that + the journal daemon is usually configured to forward everything + it receives to syslog anyway, in which case this option is no + different from . + + connects standard output with the + kernel log buffer which is accessible via + dmesg1, + in addition to the journal. The journal daemon might be + configured to send all logs to kmsg anyway, in which case this + option is no different from . + + , + and + work in a similar way as the + three options above but copy the output to the system console + as well. + + connects standard output to a + socket acquired via socket activation. The semantics are + similar to the same option of + StandardInput=. + + If the standard output (or error output, see below) of a unit is connected to the journal, syslog or the + kernel log buffer, the unit will implicitly gain a dependency of type After= on + systemd-journald.socket (also see the automatic dependencies section above). + + This setting defaults to the value set with + in + systemd-system.conf5, + which defaults to . Note that setting + this parameter might result in additional dependencies to be + added to the unit (see above). + + + + StandardError= + Controls where file descriptor 2 (STDERR) of + the executed processes is connected to. The available options + are identical to those of StandardOutput=, + with one exception: if set to the + file descriptor used for standard output is duplicated for + standard error. This setting defaults to the value set with + in + systemd-system.conf5, + which defaults to . Note that setting + this parameter might result in additional dependencies to be + added to the unit (see above). + + + + TTYPath= + Sets the terminal device node to use if + standard input, output, or error are connected to a TTY (see + above). Defaults to + /dev/console. + + + TTYReset= + Reset the terminal device specified with + TTYPath= before and after execution. + Defaults to no. + + + TTYVHangup= + Disconnect all clients which have opened the + terminal device specified with TTYPath= + before and after execution. Defaults to + no. + + + TTYVTDisallocate= + If the terminal device specified with + TTYPath= is a virtual console terminal, try + to deallocate the TTY before and after execution. This ensures + that the screen and scrollback buffer is cleared. Defaults to + no. + + + SyslogIdentifier= + Sets the process name to prefix log lines sent + to the logging system or the kernel log buffer with. If not + set, defaults to the process name of the executed process. + This option is only useful when + StandardOutput= or + StandardError= are set to + , or + (or to the same settings in combination + with ). + + + SyslogFacility= + Sets the syslog facility to use when logging + to syslog. One of , + , , + , , + , , + , , + , , + , , + , , + , , + , or + . See + syslog3 + for details. This option is only useful when + StandardOutput= or + StandardError= are set to + . Defaults to + . + + + SyslogLevel= + The default syslog level to use when logging to + syslog or the kernel log buffer. One of + , + , + , + , + , + , + , + . See + syslog3 + for details. This option is only useful when + StandardOutput= or + StandardError= are set to + or . Note that + individual lines output by the daemon might be prefixed with a + different log level which can be used to override the default + log level specified here. The interpretation of these prefixes + may be disabled with SyslogLevelPrefix=, + see below. For details, see + sd-daemon3. + + Defaults to + . + + + + SyslogLevelPrefix= + Takes a boolean argument. If true and + StandardOutput= or + StandardError= are set to + , or + , log lines written by the executed + process that are prefixed with a log level will be passed on + to syslog with this log level set but the prefix removed. If + set to false, the interpretation of these prefixes is disabled + and the logged lines are passed on as-is. For details about + this prefixing see + sd-daemon3. + Defaults to true. + + + + TimerSlackNSec= + Sets the timer slack in nanoseconds for the + executed processes. The timer slack controls the accuracy of + wake-ups triggered by timers. See + prctl2 + for more information. Note that in contrast to most other time + span definitions this parameter takes an integer value in + nano-seconds if no unit is specified. The usual time units are + understood too. + + + + LimitCPU= + LimitFSIZE= + LimitDATA= + LimitSTACK= + LimitCORE= + LimitRSS= + LimitNOFILE= + LimitAS= + LimitNPROC= + LimitMEMLOCK= + LimitLOCKS= + LimitSIGPENDING= + LimitMSGQUEUE= + LimitNICE= + LimitRTPRIO= + LimitRTTIME= + Set soft and hard limits on various resources for executed processes. See + setrlimit2 for details on + the resource limit concept. Resource limits may be specified in two formats: either as single value to set a + specific soft and hard limit to the same value, or as colon-separated pair to set + both limits individually (e.g. LimitAS=4G:16G). Use the string infinity + to configure no limit on a specific resource. The multiplicative suffixes K, M, G, T, P and E (to the base + 1024) may be used for resource limits measured in bytes (e.g. LimitAS=16G). For the limits referring to time + values, the usual time units ms, s, min, h and so on may be used (see + systemd.time7 for + details). Note that if no time unit is specified for LimitCPU= the default unit of seconds + is implied, while for LimitRTTIME= the default unit of microseconds is implied. Also, note + that the effective granularity of the limits might influence their enforcement. For example, time limits + specified for LimitCPU= will be rounded up implicitly to multiples of 1s. For + LimitNICE= the value may be specified in two syntaxes: if prefixed with + + or -, the value is understood as regular Linux nice value in the range -20..19. If not + prefixed like this the value is understood as raw resource limit parameter in the range 0..40 (with 0 being + equivalent to 1). + + Note that most process resource limits configured with + these options are per-process, and processes may fork in order + to acquire a new set of resources that are accounted + independently of the original process, and may thus escape + limits set. Also note that LimitRSS= is not + implemented on Linux, and setting it has no effect. Often it + is advisable to prefer the resource controls listed in + systemd.resource-control5 + over these per-process limits, as they apply to services as a + whole, may be altered dynamically at runtime, and are + generally more expressive. For example, + MemoryLimit= is a more powerful (and + working) replacement for LimitRSS=. + + + Limit directives and their equivalent with ulimit + + + + + + + + Directive + ulimit equivalent + Unit + + + + + LimitCPU= + ulimit -t + Seconds + + + LimitFSIZE= + ulimit -f + Bytes + + + LimitDATA= + ulimit -d + Bytes + + + LimitSTACK= + ulimit -s + Bytes + + + LimitCORE= + ulimit -c + Bytes + + + LimitRSS= + ulimit -m + Bytes + + + LimitNOFILE= + ulimit -n + Number of File Descriptors + + + LimitAS= + ulimit -v + Bytes + + + LimitNPROC= + ulimit -u + Number of Processes + + + LimitMEMLOCK= + ulimit -l + Bytes + + + LimitLOCKS= + ulimit -x + Number of Locks + + + LimitSIGPENDING= + ulimit -i + Number of Queued Signals + + + LimitMSGQUEUE= + ulimit -q + Bytes + + + LimitNICE= + ulimit -e + Nice Level + + + LimitRTPRIO= + ulimit -r + Realtime Priority + + + LimitRTTIME= + No equivalent + Microseconds + + + +
+ + + + PAMName= + Sets the PAM service name to set up a session + as. If set, the executed process will be registered as a PAM + session under the specified service name. This is only useful + in conjunction with the User= setting. If + not set, no PAM session will be opened for the executed + processes. See + pam8 + for details. + + + + CapabilityBoundingSet= + + Controls which capabilities to include in the capability bounding set for the executed + process. See capabilities7 for + details. Takes a whitespace-separated list of capability names as read by cap_from_name3, + e.g. CAP_SYS_ADMIN, CAP_DAC_OVERRIDE, + CAP_SYS_PTRACE. Capabilities listed will be included in the bounding set, all others are + removed. If the list of capabilities is prefixed with ~, all but the listed capabilities + will be included, the effect of the assignment inverted. Note that this option also affects the respective + capabilities in the effective, permitted and inheritable capability sets. If this option is not used, the + capability bounding set is not modified on process execution, hence no limits on the capabilities of the + process are enforced. This option may appear more than once, in which case the bounding sets are merged. If the + empty string is assigned to this option, the bounding set is reset to the empty capability set, and all prior + settings have no effect. If set to ~ (without any further argument), the bounding set is + reset to the full set of available capabilities, also undoing any previous settings. This does not affect + commands prefixed with +. + + + + AmbientCapabilities= + + Controls which capabilities to include in the + ambient capability set for the executed process. Takes a + whitespace-separated list of capability names as read by + cap_from_name3, + e.g. CAP_SYS_ADMIN, + CAP_DAC_OVERRIDE, + CAP_SYS_PTRACE. This option may appear more than + once in which case the ambient capability sets are merged. + If the list of capabilities is prefixed with ~, all + but the listed capabilities will be included, the effect of the + assignment inverted. If the empty string is + assigned to this option, the ambient capability set is reset to + the empty capability set, and all prior settings have no effect. + If set to ~ (without any further argument), the + ambient capability set is reset to the full set of available + capabilities, also undoing any previous settings. Note that adding + capabilities to ambient capability set adds them to the process's + inherited capability set. + + Ambient capability sets are useful if you want to execute a process + as a non-privileged user but still want to give it some capabilities. + Note that in this case option keep-caps is + automatically added to SecureBits= to retain the + capabilities over the user change. AmbientCapabilities= does not affect + commands prefixed with +. + + + + SecureBits= + Controls the secure bits set for the executed + process. Takes a space-separated combination of options from + the following list: + , + , + , + , + , and + . + This option may appear more than once, in which case the secure + bits are ORed. If the empty string is assigned to this option, + the bits are reset to 0. This does not affect commands prefixed with +. + See capabilities7 + for details. + + + + ReadWritePaths= + ReadOnlyPaths= + InaccessiblePaths= + + Sets up a new file system namespace for + executed processes. These options may be used to limit access + a process might have to the main file system hierarchy. Each + setting takes a space-separated list of paths relative to + the host's root directory (i.e. the system running the service manager). + Note that if entries contain symlinks, they are resolved from the host's root directory as well. + Entries (files or directories) listed in + ReadWritePaths= are accessible from + within the namespace with the same access rights as from + outside. Entries listed in + ReadOnlyPaths= are accessible for + reading only, writing will be refused even if the usual file + access controls would permit this. Entries listed in + InaccessiblePaths= will be made + inaccessible for processes inside the namespace, and may not + countain any other mountpoints, including those specified by + ReadWritePaths= or + ReadOnlyPaths=. + Note that restricting access with these options does not extend + to submounts of a directory that are created later on. + Non-directory paths can be specified as well. These + options may be specified more than once, in which case all + paths listed will have limited access from within the + namespace. If the empty string is assigned to this option, the + specific list is reset, and all prior assignments have no + effect. + Paths in + ReadOnlyPaths= + and + InaccessiblePaths= + may be prefixed with + -, in which case + they will be ignored when they do not + exist. Note that using this + setting will disconnect propagation of + mounts from the service to the host + (propagation in the opposite direction + continues to work). This means that + this setting may not be used for + services which shall be able to + install mount points in the main mount + namespace. + + + + PrivateTmp= + + Takes a boolean argument. If true, sets up a + new file system namespace for the executed processes and + mounts private /tmp and + /var/tmp directories inside it that is + not shared by processes outside of the namespace. This is + useful to secure access to temporary files of the process, but + makes sharing between processes via /tmp + or /var/tmp impossible. If this is + enabled, all temporary files created by a service in these + directories will be removed after the service is stopped. + Defaults to false. It is possible to run two or more units + within the same private /tmp and + /var/tmp namespace by using the + JoinsNamespaceOf= directive, see + systemd.unit5 + for details. Note that using this setting will disconnect + propagation of mounts from the service to the host + (propagation in the opposite direction continues to work). + This means that this setting may not be used for services + which shall be able to install mount points in the main mount + namespace. + + + + PrivateDevices= + + Takes a boolean argument. If true, sets up a + new /dev namespace for the executed processes and only adds + API pseudo devices such as /dev/null, + /dev/zero or + /dev/random (as well as the pseudo TTY + subsystem) to it, but no physical devices such as + /dev/sda. This is useful to securely turn + off physical device access by the executed process. Defaults + to false. Enabling this option will also remove + CAP_MKNOD from the capability bounding + set for the unit (see above), and set + DevicePolicy=closed (see + systemd.resource-control5 + for details). Note that using this setting will disconnect + propagation of mounts from the service to the host + (propagation in the opposite direction continues to work). + This means that this setting may not be used for services + which shall be able to install mount points in the main mount + namespace. The /dev namespace will be mounted read-only and 'noexec'. + The latter may break old programs which try to set up executable + memory by using mmap2 + of /dev/zero instead of using MAP_ANON. + + + + PrivateNetwork= + + Takes a boolean argument. If true, sets up a + new network namespace for the executed processes and + configures only the loopback network device + lo inside it. No other network devices will + be available to the executed process. This is useful to + securely turn off network access by the executed process. + Defaults to false. It is possible to run two or more units + within the same private network namespace by using the + JoinsNamespaceOf= directive, see + systemd.unit5 + for details. Note that this option will disconnect all socket + families from the host, this includes AF_NETLINK and AF_UNIX. + The latter has the effect that AF_UNIX sockets in the abstract + socket namespace will become unavailable to the processes + (however, those located in the file system will continue to be + accessible). + + + + ProtectSystem= + + Takes a boolean argument or + full. If true, mounts the + /usr and /boot + directories read-only for processes invoked by this unit. If + set to full, the /etc + directory is mounted read-only, too. This setting ensures that + any modification of the vendor-supplied operating system (and + optionally its configuration) is prohibited for the service. + It is recommended to enable this setting for all long-running + services, unless they are involved with system updates or need + to modify the operating system in other ways. Note however + that processes retaining the CAP_SYS_ADMIN capability can undo + the effect of this setting. This setting is hence particularly + useful for daemons which have this capability removed, for + example with CapabilityBoundingSet=. + Defaults to off. + + + + ProtectHome= + + Takes a boolean argument or + read-only. If true, the directories + /home, /root and + /run/user + are made inaccessible and empty for processes invoked by this + unit. If set to read-only, the three + directories are made read-only instead. It is recommended to + enable this setting for all long-running services (in + particular network-facing ones), to ensure they cannot get + access to private user data, unless the services actually + require access to the user's private data. Note however that + processes retaining the CAP_SYS_ADMIN capability can undo the + effect of this setting. This setting is hence particularly + useful for daemons which have this capability removed, for + example with CapabilityBoundingSet=. + Defaults to off. + + + + MountFlags= + + Takes a mount propagation flag: + , or + , which control whether mounts in the + file system namespace set up for this unit's processes will + receive or propagate mounts or unmounts. See + mount2 + for details. Defaults to . Use + to ensure that mounts and unmounts are + propagated from the host to the container and vice versa. Use + to run processes so that none of their + mounts and unmounts will propagate to the host. Use + to also ensure that no mounts and + unmounts from the host will propagate into the unit processes' + namespace. Note that means that file + systems mounted on the host might stay mounted continuously in + the unit's namespace, and thus keep the device busy. Note that + the file system namespace related options + (PrivateTmp=, + PrivateDevices=, + ProtectSystem=, + ProtectHome=, + ReadOnlyPaths=, + InaccessiblePaths= and + ReadWritePaths=) require that mount + and unmount propagation from the unit's file system namespace + is disabled, and hence downgrade to + . + + + + UtmpIdentifier= + + Takes a four character identifier string for + an utmp5 + and wtmp entry for this service. This should only be + set for services such as getty + implementations (such as agetty8) + where utmp/wtmp entries must be created and cleared before and + after execution, or for services that shall be executed as if + they were run by a getty process (see + below). If the configured string is longer than four + characters, it is truncated and the terminal four characters + are used. This setting interprets %I style string + replacements. This setting is unset by default, i.e. no + utmp/wtmp entries are created or cleaned up for this + service. + + + + UtmpMode= + + Takes one of init, + login or user. If + UtmpIdentifier= is set, controls which + type of utmp5/wtmp + entries for this service are generated. This setting has no + effect unless UtmpIdentifier= is set + too. If init is set, only an + INIT_PROCESS entry is generated and the + invoked process must implement a + getty-compatible utmp/wtmp logic. If + login is set, first an + INIT_PROCESS entry, followed by a + LOGIN_PROCESS entry is generated. In + this case, the invoked process must implement a login1-compatible + utmp/wtmp logic. If user is set, first an + INIT_PROCESS entry, then a + LOGIN_PROCESS entry and finally a + USER_PROCESS entry is generated. In this + case, the invoked process may be any process that is suitable + to be run as session leader. Defaults to + init. + + + + SELinuxContext= + + Set the SELinux security context of the + executed process. If set, this will override the automated + domain transition. However, the policy still needs to + authorize the transition. This directive is ignored if SELinux + is disabled. If prefixed by -, all errors + will be ignored. This does not affect commands prefixed with +. + See setexeccon3 + for details. + + + + AppArmorProfile= + + Takes a profile name as argument. The process + executed by the unit will switch to this profile when started. + Profiles must already be loaded in the kernel, or the unit + will fail. This result in a non operation if AppArmor is not + enabled. If prefixed by -, all errors will + be ignored. This does not affect commands prefixed with +. + + + + SmackProcessLabel= + + Takes a security + label as argument. The process executed by the unit will be + started under this label and SMACK will decide whether the + process is allowed to run or not, based on it. The process + will continue to run under the label specified here unless the + executable has its own label, in + which case the process will transition to run under that + label. When not specified, the label that systemd is running + under is used. This directive is ignored if SMACK is + disabled. + + The value may be prefixed by -, in + which case all errors will be ignored. An empty value may be + specified to unset previous assignments. This does not affect + commands prefixed with +. + + + + + IgnoreSIGPIPE= + + Takes a boolean argument. If true, causes + SIGPIPE to be ignored in the executed + process. Defaults to true because SIGPIPE + generally is useful only in shell pipelines. + + + + NoNewPrivileges= + + Takes a boolean argument. If true, ensures + that the service process and all its children can never gain + new privileges. This option is more powerful than the + respective secure bits flags (see above), as it also prohibits + UID changes of any kind. This is the simplest, most effective + way to ensure that a process and its children can never + elevate privileges again. + + + + SystemCallFilter= + + Takes a space-separated list of system call + names. If this setting is used, all system calls executed by + the unit processes except for the listed ones will result in + immediate process termination with the + SIGSYS signal (whitelisting). If the + first character of the list is ~, the + effect is inverted: only the listed system calls will result + in immediate process termination (blacklisting). If running in + user mode, or in system mode, but without the + CAP_SYS_ADMIN capability (e.g. setting + User=nobody), + NoNewPrivileges=yes is implied. This + feature makes use of the Secure Computing Mode 2 interfaces of + the kernel ('seccomp filtering') and is useful for enforcing a + minimal sandboxing environment. Note that the + execve, + rt_sigreturn, + sigreturn, + exit_group, exit + system calls are implicitly whitelisted and do not need to be + listed explicitly. This option may be specified more than once, + in which case the filter masks are merged. If the empty string + is assigned, the filter is reset, all prior assignments will + have no effect. This does not affect commands prefixed with +. + + If you specify both types of this option (i.e. + whitelisting and blacklisting), the first encountered will + take precedence and will dictate the default action + (termination or approval of a system call). Then the next + occurrences of this option will add or delete the listed + system calls from the set of the filtered system calls, + depending of its type and the default action. (For example, if + you have started with a whitelisting of + read and write, and + right after it add a blacklisting of + write, then write + will be removed from the set.) + + As the number of possible system + calls is large, predefined sets of system calls are provided. + A set starts with @ character, followed by + name of the set. + + + Currently predefined system call sets + + + + + + + Set + Description + + + + + @clock + System calls for changing the system clock (adjtimex2, settimeofday2, and related calls) + + + @cpu-emulation + System calls for CPU emulation functionality (vm862 and related calls) + + + @debug + Debugging, performance monitoring and tracing functionality (ptrace2, perf_event_open2 and related calls) + + + @io-event + Event loop system calls (poll2, select2, epoll7, eventfd2 and related calls) + + + @ipc + SysV IPC, POSIX Message Queues or other IPC (mq_overview7, svipc7) + + + @keyring + Kernel keyring access (keyctl2 and related calls) + + + @module + Kernel module control (init_module2, delete_module2 and related calls) + + + @mount + File system mounting and unmounting (mount2, chroot2, and related calls) + + + @network-io + Socket I/O (including local AF_UNIX): socket7, unix7 + + + @obsolete + Unusual, obsolete or unimplemented (create_module2, gtty2, …) + + + @privileged + All system calls which need super-user capabilities (capabilities7) + + + @process + Process control, execution, namespaces (execve2, kill2, namespaces7, … + + + @raw-io + Raw I/O port access (ioperm2, iopl2, pciconfig_read(), … + + + +
+ + Note, that as new system calls are added to the kernel, additional system calls might be added to the groups + above, so the contents of the sets may change between systemd versions. + + + + SystemCallErrorNumber= + + Takes an errno error number + name to return when the system call filter configured with + SystemCallFilter= is triggered, instead of + terminating the process immediately. Takes an error name such + as EPERM, EACCES or + EUCLEAN. When this setting is not used, + or when the empty string is assigned, the process will be + terminated immediately when the filter is + triggered. + + + + SystemCallArchitectures= + + Takes a space-separated list of architecture + identifiers to include in the system call filter. The known + architecture identifiers are x86, + x86-64, x32, + arm as well as the special identifier + native. Only system calls of the + specified architectures will be permitted to processes of this + unit. This is an effective way to disable compatibility with + non-native architectures for processes, for example to + prohibit execution of 32-bit x86 binaries on 64-bit x86-64 + systems. The special native identifier + implicitly maps to the native architecture of the system (or + more strictly: to the architecture the system manager is + compiled for). If running in user mode, or in system mode, + but without the CAP_SYS_ADMIN + capability (e.g. setting User=nobody), + NoNewPrivileges=yes is implied. Note + that setting this option to a non-empty list implies that + native is included too. By default, this + option is set to the empty list, i.e. no architecture system + call filtering is applied. + + + + RestrictAddressFamilies= + + Restricts the set of socket address families + accessible to the processes of this unit. Takes a + space-separated list of address family names to whitelist, + such as + AF_UNIX, + AF_INET or + AF_INET6. When + prefixed with ~ the listed address + families will be applied as blacklist, otherwise as whitelist. + Note that this restricts access to the + socket2 + system call only. Sockets passed into the process by other + means (for example, by using socket activation with socket + units, see + systemd.socket5) + are unaffected. Also, sockets created with + socketpair() (which creates connected + AF_UNIX sockets only) are unaffected. Note that this option + has no effect on 32-bit x86 and is ignored (but works + correctly on x86-64). If running in user mode, or in system + mode, but without the CAP_SYS_ADMIN + capability (e.g. setting User=nobody), + NoNewPrivileges=yes is implied. By + default, no restriction applies, all address families are + accessible to processes. If assigned the empty string, any + previous list changes are undone. + + Use this option to limit exposure of processes to remote + systems, in particular via exotic network protocols. Note that + in most cases, the local AF_UNIX address + family should be included in the configured whitelist as it is + frequently used for local communication, including for + syslog2 + logging. This does not affect commands prefixed with +. + + + + Personality= + + Controls which kernel architecture uname2 shall report, + when invoked by unit processes. Takes one of the architecture identifiers x86, + x86-64, ppc, ppc-le, ppc64, + ppc64-le, s390 or s390x. Which personality + architectures are supported depends on the system architecture. Usually the 64bit versions of the various + system architectures support their immediate 32bit personality architecture counterpart, but no others. For + example, x86-64 systems support the x86-64 and + x86 personalities but no others. The personality feature is useful when running 32-bit + services on a 64-bit host system. If not specified, the personality is left unmodified and thus reflects the + personality of the host system's kernel. + + + + RuntimeDirectory= + RuntimeDirectoryMode= + + Takes a list of directory names. If set, one + or more directories by the specified names will be created + below /run (for system services) or below + $XDG_RUNTIME_DIR (for user services) when + the unit is started, and removed when the unit is stopped. The + directories will have the access mode specified in + RuntimeDirectoryMode=, and will be owned by + the user and group specified in User= and + Group=. Use this to manage one or more + runtime directories of the unit and bind their lifetime to the + daemon runtime. The specified directory names must be + relative, and may not include a /, i.e. + must refer to simple directories to create or remove. This is + particularly useful for unprivileged daemons that cannot + create runtime directories in /run due to + lack of privileges, and to make sure the runtime directory is + cleaned up automatically after use. For runtime directories + that require more complex or different configuration or + lifetime guarantees, please consider using + tmpfiles.d5. + + + + MemoryDenyWriteExecute= + + Takes a boolean argument. If set, attempts to create memory mappings that are writable and + executable at the same time, or to change existing memory mappings to become executable are prohibited. + Specifically, a system call filter is added that rejects + mmap2 + system calls with both PROT_EXEC and PROT_WRITE set + and mprotect2 + system calls with PROT_EXEC set. Note that this option is incompatible with programs + that generate program code dynamically at runtime, such as JIT execution engines, or programs compiled making + use of the code "trampoline" feature of various C compilers. This option improves service security, as it makes + harder for software exploits to change running code dynamically. + + + + + RestrictRealtime= + + Takes a boolean argument. If set, any attempts to enable realtime scheduling in a process of + the unit are refused. This restricts access to realtime task scheduling policies such as + SCHED_FIFO, SCHED_RR or SCHED_DEADLINE. See + sched7 for details about + these scheduling policies. Realtime scheduling policies may be used to monopolize CPU time for longer periods + of time, and may hence be used to lock up or otherwise trigger Denial-of-Service situations on the system. It + is hence recommended to restrict access to realtime scheduling to the few programs that actually require + them. Defaults to off. + + + + + + + Environment variables in spawned processes + + Processes started by the system are executed in a clean + environment in which select variables listed below are set. System + processes started by systemd do not inherit variables from PID 1, + but processes started by user systemd instances inherit all + environment variables from the user systemd instance. + + + + + $PATH + + Colon-separated list of directories to use + when launching executables. Systemd uses a fixed value of + /usr/local/sbin:/usr/local/bin:/usr/sbin:/usr/bin:/sbin:/bin. + + + + + $LANG + + Locale. Can be set in + locale.conf5 + or on the kernel command line (see + systemd1 + and + kernel-command-line7). + + + + + $USER + $LOGNAME + $HOME + $SHELL + + User name (twice), home directory, and the + login shell. The variables are set for the units that have + User= set, which includes user + systemd instances. See + passwd5. + + + + + $XDG_RUNTIME_DIR + + The directory for volatile state. Set for the + user systemd instance, and also in user + sessions. See + pam_systemd8. + + + + + $XDG_SESSION_ID + $XDG_SEAT + $XDG_VTNR + + The identifier of the session, the seat name, + and virtual terminal of the session. Set by + pam_systemd8 + for login sessions. $XDG_SEAT and + $XDG_VTNR will only be set when attached to + a seat and a tty. + + + + $MAINPID + + The PID of the units main process if it is + known. This is only set for control processes as invoked by + ExecReload= and similar. + + + + $MANAGERPID + + The PID of the user systemd + instance, set for processes spawned by it. + + + + $LISTEN_FDS + $LISTEN_PID + $LISTEN_FDNAMES + + Information about file descriptors passed to a + service for socket activation. See + sd_listen_fds3. + + + + + $NOTIFY_SOCKET + + The socket + sd_notify() talks to. See + sd_notify3. + + + + + $WATCHDOG_PID + $WATCHDOG_USEC + + Information about watchdog keep-alive notifications. See + sd_watchdog_enabled3. + + + + + $TERM + + Terminal type, set only for units connected to + a terminal (StandardInput=tty, + StandardOutput=tty, or + StandardError=tty). See + termcap5. + + + + + $JOURNAL_STREAM + + If the standard output or standard error output of the executed processes are connected to the + journal (for example, by setting StandardError=journal) $JOURNAL_STREAM + contains the device and inode numbers of the connection file descriptor, formatted in decimal, separated by a + colon (:). This permits invoked processes to safely detect whether their standard output or + standard error output are connected to the journal. The device and inode numbers of the file descriptors should + be compared with the values set in the environment variable to determine whether the process output is still + connected to the journal. Note that it is generally not sufficient to only check whether + $JOURNAL_STREAM is set at all as services might invoke external processes replacing their + standard output or standard error output, without unsetting the environment variable. + + This environment variable is primarily useful to allow services to optionally upgrade their used log + protocol to the native journal protocol (using + sd_journal_print3 and other + functions) if their standard output or standard error output is connected to the journal anyway, thus enabling + delivery of structured metadata along with logged messages. + + + + Additional variables may be configured by the following + means: for processes spawned in specific units, use the + Environment=, EnvironmentFile= + and PassEnvironment= options above; to specify + variables globally, use DefaultEnvironment= + (see + systemd-system.conf5) + or the kernel option systemd.setenv= (see + systemd1). + Additional variables may also be set through PAM, + cf. pam_env8. + + + + See Also + + systemd1, + systemctl1, + journalctl8, + systemd.unit5, + systemd.service5, + systemd.socket5, + systemd.swap5, + systemd.mount5, + systemd.kill5, + systemd.resource-control5, + systemd.time7, + systemd.directives7, + tmpfiles.d5, + exec3 + + + + diff --git a/src/grp-system/systemd/systemd.generator.xml b/src/grp-system/systemd/systemd.generator.xml new file mode 100644 index 0000000000..b268104c9d --- /dev/null +++ b/src/grp-system/systemd/systemd.generator.xml @@ -0,0 +1,348 @@ + + +%entities; +]> + + + + + + systemd.generator + systemd + + + + Developer + Lennart + Poettering + lennart@poettering.net + + + + + + systemd.generator + 7 + + + + systemd.generator + Systemd unit generators + + + + + /path/to/generator + normal-dir + early-dir + late-dir + + + + /run/systemd/system-generators/* +/etc/systemd/system-generators/* +/usr/local/lib/systemd/system-generators/* +&systemgeneratordir;/* + + + + /run/systemd/user-generators/* +/etc/systemd/user-generators/* +/usr/local/lib/systemd/user-generators/* +&usergeneratordir;/* + + + + + Description + Generators are small binaries that live in + &usergeneratordir;/ and other directories + listed above. + systemd1 + will execute those binaries very early at bootup and at + configuration reload time — before unit files are loaded. + Generators can dynamically generate unit files or create symbolic + links to unit files to add additional dependencies, thus extending + or overriding existing definitions. Their main purpose is to + convert configuration files that are not native unit files + dynamically into native unit files. + + Generators are loaded from a set of paths determined during + compilation, as listed above. System and user generators are loaded + from directories with names ending in + system-generators/ and + user-generators/, respectively. Generators + found in directories listed earlier override the ones with the + same name in directories lower in the list. A symlink to + /dev/null or an empty file can be used to + mask a generator, thereby preventing it from running. Please note + that the order of the two directories with the highest priority is + reversed with respect to the unit load path, and generators in + /run overwrite those in + /etc. + + After installing new generators or updating the + configuration, systemctl daemon-reload may be + executed. This will delete the previous configuration created by + generators, re-run all generators, and cause + systemd to reload units from disk. See + systemctl1 + for more information. + + + + + Writing generators + + Generators are invoked with three arguments: paths to + runtime directories where generators can place their generated + unit files or symlinks. + + + + normal-dir + argv[1] may be used to override unit files in + /usr, but not those in + /etc. This means that unit files placed + in this directory take precedence over vendor unit + configuration but not over native user/administrator unit + configuration. + + + + early-dir + argv[2] may be used to override unit files in + /usr and in + /etc. This means that unit files placed + in this directory take precedence over all configuration, + both vendor and user/administrator. + + + + late-dir + argv[3] may be used to extend the unit file tree without + overriding any other unit files. Any native configuration + files supplied by the vendor or user/administrator take + precedence over the generated ones placed in this directory. + + + + + + Notes + + + + + All generators are executed in parallel. That means all + executables are started at the very same time and need to + be able to cope with this parallelism. + + + + + + Generators are run very early at boot and cannot rely on + any external services. They may not talk to any other + process. That includes simple things such as logging to + syslog3, + or systemd itself (this means: no + systemctl1)! + Non-essential file systems like + /var and /home + are mounted after generators have run. Generators + can however rely on the most basic kernel functionality to be + available, including a mounted /sys, + /proc, /dev, + /usr. + + + + + + Units written by generators are removed when the configuration + is reloaded. That means the lifetime of the generated + units is closely bound to the reload cycles of + systemd itself. + + + + + + Generators should only be used to generate unit files, not + any other kind of configuration. Due to the lifecycle + logic mentioned above, generators are not a good fit to + generate dynamic configuration for other services. If you + need to generate dynamic configuration for other services, + do so in normal services you order before the service in + question. + + + + + + Since + syslog3 + is not available (see above), log messages have to be + written to /dev/kmsg instead. + + + + + + It is a good idea to use the + SourcePath= directive in generated unit + files to specify the source configuration file you are + generating the unit from. This makes things more easily + understood by the user and also has the benefit that + systemd can warn the user about configuration files that + changed on disk but have not been read yet by systemd. + + + + + + Generators may write out dynamic unit files or just hook + unit files into other units with the usual + .wants/ or + .requires/ symlinks. Often, it is + nicer to simply instantiate a template unit file from + /usr with a generator instead of + writing out entirely dynamic unit files. Of course, this + works only if a single parameter is to be used. + + + + + + If you are careful, you can implement generators in shell + scripts. We do recommend C code however, since generators + are executed synchronously and hence delay the + entire boot if they are slow. + + + + + Regarding overriding semantics: there are two rules we + try to follow when thinking about the overriding semantics: + + + + + User configuration should override vendor + configuration. This (mostly) means that stuff from + /etc should override stuff from + /usr. + + + + Native configuration should override non-native + configuration. This (mostly) means that stuff you + generate should never override native unit files for the + same purpose. + + + + Of these two rules the first rule is probably the more + important one and breaks the second one sometimes. Hence, + when deciding whether to user argv[1], argv[2], or argv[3], + your default choice should probably be argv[1]. + + + + + Instead of heading off now and writing all kind of + generators for legacy configuration file formats, please + think twice! It is often a better idea to just deprecate + old stuff instead of keeping it artificially alive. + + + + + + + + Examples + + systemd-fstab-generator + + systemd-fstab-generator8 + converts /etc/fstab into native mount + units. It uses argv[1] as location to place the generated unit + files in order to allow the user to override + /etc/fstab with her own native unit files, + but also to ensure that /etc/fstab + overrides any vendor default from /usr. + + + After editing /etc/fstab, the user + should invoke systemctl daemon-reload. This + will re-run all generators and cause systemd + to reload units from disk. To actually mount new directories + added to fstab, systemctl start + /path/to/mountpoint or + systemctl start local-fs.target may be used. + + + + + systemd-system-update-generator + + systemd-system-update-generator8 + temporarily redirects default.target to + system-update.target, if a system update is + scheduled. Since this needs to override the default user + configuration for default.target, it uses + argv[2]. For details about this logic, see + systemd.offline-updates7. + + + + + Debugging a generator + + dir=$(mktemp -d) +SYSTEMD_LOG_LEVEL=debug &systemgeneratordir;/systemd-fstab-generator \ + "$dir" "$dir" "$dir" +find $dir + + + + + See also + + + systemd1, + systemd-cryptsetup-generator8, + systemd-debug-generator8, + systemd-fstab-generator8, + fstab5, + systemd-getty-generator8, + systemd-gpt-auto-generator8, + systemd-hibernate-resume-generator8, + systemd-system-update-generator8, + systemd-sysv-generator8, + systemd.unit5, + systemctl1 + + + diff --git a/src/grp-system/systemd/systemd.journal-fields.xml b/src/grp-system/systemd/systemd.journal-fields.xml new file mode 100644 index 0000000000..494f97aad1 --- /dev/null +++ b/src/grp-system/systemd/systemd.journal-fields.xml @@ -0,0 +1,525 @@ + + + + + + + + + systemd.journal-fields + systemd + + + + Developer + Lennart + Poettering + lennart@poettering.net + + + + + + systemd.journal-fields + 7 + + + + systemd.journal-fields + Special journal fields + + + + Description + + Entries in the journal resemble an environment block in + their syntax but with fields that can include binary data. + Primarily, fields are formatted UTF-8 text strings, and binary + formatting is used only where formatting as UTF-8 text strings + makes little sense. New fields may freely be defined by + applications, but a few fields have special meaning. All fields + with special meanings are optional. In some cases, fields may + appear more than once per entry. + + + + User Journal Fields + + User fields are fields that are directly passed from clients + and stored in the journal. + + + + MESSAGE= + + The human-readable message string for this entry. This + is supposed to be the primary text shown to the user. It is + usually not translated (but might be in some cases), and is + not supposed to be parsed for metadata. + + + + + MESSAGE_ID= + + A 128-bit message identifier ID for recognizing + certain message types, if this is desirable. This should + contain a 128-bit ID formatted as a lower-case hexadecimal + string, without any separating dashes or suchlike. This is + recommended to be a UUID-compatible ID, but this is not + enforced, and formatted differently. Developers can generate + a new ID for this purpose with journalctl + . + + + + + + PRIORITY= + + A priority value between 0 (emerg) + and 7 (debug) formatted as a decimal + string. This field is compatible with syslog's priority + concept. + + + + + CODE_FILE= + CODE_LINE= + CODE_FUNC= + + The code location generating this message, if known. + Contains the source filename, the line number and the + function name. + + + + + ERRNO= + + The low-level Unix error number causing this entry, if + any. Contains the numeric value of + errno3 + formatted as a decimal string. + + + + + SYSLOG_FACILITY= + SYSLOG_IDENTIFIER= + SYSLOG_PID= + + Syslog compatibility fields containing the facility + (formatted as decimal string), the identifier string (i.e. + "tag"), and the client PID. (Note that the tag is usually + derived from glibc's + program_invocation_short_name variable, + see + program_invocation_short_name3.) + + + + + + + + Trusted Journal Fields + + Fields prefixed with an underscore are trusted fields, i.e. + fields that are implicitly added by the journal and cannot be + altered by client code. + + + + _PID= + _UID= + _GID= + + The process, user, and group ID of the process the + journal entry originates from formatted as a decimal + string. + + + + + _COMM= + _EXE= + _CMDLINE= + + The name, the executable path, and the command line of + the process the journal entry originates from. + + + + + _CAP_EFFECTIVE= + + The effective + capabilities7 + of the process the journal entry originates from. + + + + + _AUDIT_SESSION= + _AUDIT_LOGINUID= + + The session and login UID of the process the journal + entry originates from, as maintained by the kernel audit + subsystem. + + + + + _SYSTEMD_CGROUP= + _SYSTEMD_SESSION= + _SYSTEMD_UNIT= + _SYSTEMD_USER_UNIT= + _SYSTEMD_OWNER_UID= + _SYSTEMD_SLICE= + + + The control group path in the systemd hierarchy, the + systemd session ID (if any), the systemd unit name (if any), + the systemd user session unit name (if any), the owner UID + of the systemd session (if any) and the systemd slice unit + of the process the journal entry originates from. + + + + + _SELINUX_CONTEXT= + + The SELinux security context (label) of the process + the journal entry originates from. + + + + + _SOURCE_REALTIME_TIMESTAMP= + + The earliest trusted timestamp of the message, if any + is known that is different from the reception time of the + journal. This is the time in microseconds since the epoch + UTC, formatted as a decimal string. + + + + + _BOOT_ID= + + The kernel boot ID for the boot the message was + generated in, formatted as a 128-bit hexadecimal + string. + + + + + _MACHINE_ID= + + The machine ID of the originating host, as available + in + machine-id5. + + + + + _HOSTNAME= + + The name of the originating host. + + + + + _TRANSPORT= + + How the entry was received by the journal service. + Valid transports are: + + + + + + + + for those read from the kernel audit subsystem + + + + + + + + + + for internally generated messages + + + + + + + + + + for those received via the local syslog socket + with the syslog protocol + + + + + + + + + + for those received via the native journal + protocol + + + + + + + + + + for those read from a service's standard output + or error output + + + + + + + + + + for those read from the kernel + + + + + + + + + + + Kernel Journal Fields + + Kernel fields are fields that are used by messages + originating in the kernel and stored in the journal. + + + + _KERNEL_DEVICE= + + The kernel device name. If the entry is associated to + a block device, the major and minor of the device node, + separated by : and prefixed by + b. Similar for character devices but + prefixed by c. For network devices, this + is the interface index prefixed by n. For + all other devices, this is the subsystem name prefixed by + +, followed by :, + followed by the kernel device name. + + + + _KERNEL_SUBSYSTEM= + + The kernel subsystem name. + + + + _UDEV_SYSNAME= + + The kernel device name as it shows up in the device + tree below /sys. + + + + _UDEV_DEVNODE= + + The device node path of this device in + /dev. + + + + _UDEV_DEVLINK= + + Additional symlink names pointing to the device node + in /dev. This field is frequently set + more than once per entry. + + + + + + + Fields to log on behalf of a different program + + Fields in this section are used by programs to specify that + they are logging on behalf of another program or unit. + + + Fields used by the systemd-coredump + coredump kernel helper: + + + + + COREDUMP_UNIT= + COREDUMP_USER_UNIT= + + Used to annotate messages containing coredumps from + system and session units. See + coredumpctl1. + + + + + + Privileged programs (currently UID 0) may attach + OBJECT_PID= to a message. This will instruct + systemd-journald to attach additional fields on + behalf of the caller: + + + + OBJECT_PID=PID + + PID of the program that this message pertains to. + + + + + + OBJECT_UID= + OBJECT_GID= + OBJECT_COMM= + OBJECT_EXE= + OBJECT_CMDLINE= + OBJECT_AUDIT_SESSION= + OBJECT_AUDIT_LOGINUID= + OBJECT_SYSTEMD_CGROUP= + OBJECT_SYSTEMD_SESSION= + OBJECT_SYSTEMD_OWNER_UID= + OBJECT_SYSTEMD_UNIT= + OBJECT_SYSTEMD_USER_UNIT= + + These are additional fields added automatically by + systemd-journald. Their meaning is the + same as + _UID=, + _GID=, + _COMM=, + _EXE=, + _CMDLINE=, + _AUDIT_SESSION=, + _AUDIT_LOGINUID=, + _SYSTEMD_CGROUP=, + _SYSTEMD_SESSION=, + _SYSTEMD_UNIT=, + _SYSTEMD_USER_UNIT=, and + _SYSTEMD_OWNER_UID= + as described above, except that the process identified by + PID is described, instead of the + process which logged the message. + + + + + + + + Address Fields + + During serialization into external formats, such as the + Journal + Export Format or the Journal + JSON Format, the addresses of journal entries are + serialized into fields prefixed with double underscores. Note that + these are not proper fields when stored in the journal but for + addressing metadata of entries. They cannot be written as part of + structured log entries via calls such as + sd_journal_send3. + They may also not be used as matches for + sd_journal_add_match3 + + + + __CURSOR= + + The cursor for the entry. A cursor is an opaque text + string that uniquely describes the position of an entry in + the journal and is portable across machines, platforms and + journal files. + + + + + + __REALTIME_TIMESTAMP= + + The wallclock time + (CLOCK_REALTIME) at the point in time + the entry was received by the journal, in microseconds since + the epoch UTC, formatted as a decimal string. This has + different properties from + _SOURCE_REALTIME_TIMESTAMP=, as it is + usually a bit later but more likely to be monotonic. + + + + + + __MONOTONIC_TIMESTAMP= + + The monotonic time + (CLOCK_MONOTONIC) at the point in time + the entry was received by the journal in microseconds, + formatted as a decimal string. To be useful as an address + for the entry, this should be combined with the boot ID in + _BOOT_ID=. + + + + + + + + See Also + + systemd1, + journalctl1, + journald.conf5, + sd-journal3, + coredumpctl1, + systemd.directives7 + + + + diff --git a/src/grp-system/systemd/systemd.kill.xml b/src/grp-system/systemd/systemd.kill.xml new file mode 100644 index 0000000000..13b7ab14df --- /dev/null +++ b/src/grp-system/systemd/systemd.kill.xml @@ -0,0 +1,189 @@ + + + + + + + + systemd.kill + systemd + + + + Developer + Lennart + Poettering + lennart@poettering.net + + + + + + systemd.kill + 5 + + + + systemd.kill + Process killing procedure + configuration + + + + service.service, + socket.socket, + mount.mount, + swap.swap, + scope.scope + + + + Description + + Unit configuration files for services, sockets, mount + points, swap devices and scopes share a subset of configuration + options which define the killing procedure of processes belonging + to the unit. + + This man page lists the configuration options shared by + these five unit types. See + systemd.unit5 + for the common options shared by all unit configuration files, and + systemd.service5, + systemd.socket5, + systemd.swap5, + systemd.mount5 + and + systemd.scope5 + for more information on the configuration file options specific to + each unit type. + + The kill procedure configuration options are configured in + the [Service], [Socket], [Mount] or [Swap] section, depending on + the unit type. + + + + Options + + + + + KillMode= + Specifies how processes of this unit shall be + killed. One of + , + , + , + . + + If set to , all remaining + processes in the control group of this unit will be killed on + unit stop (for services: after the stop command is executed, + as configured with ExecStop=). If set to + , only the main process itself is + killed. If set to , the + SIGTERM signal (see below) is sent to the + main process while the subsequent SIGKILL + signal (see below) is sent to all remaining processes of the + unit's control group. If set to , no + process is killed. In this case, only the stop command will be + executed on unit stop, but no process be killed otherwise. + Processes remaining alive after stop are left in their control + group and the control group continues to exist after stop + unless it is empty. + + Processes will first be terminated via + SIGTERM (unless the signal to send is + changed via KillSignal=). Optionally, this + is immediately followed by a SIGHUP (if + enabled with SendSIGHUP=). If then, after a + delay (configured via the TimeoutStopSec= + option), processes still remain, the termination request is + repeated with the SIGKILL signal (unless + this is disabled via the SendSIGKILL= + option). See + kill2 + for more information. + + Defaults to + . + + + + KillSignal= + Specifies which signal to use when killing a + service. This controls the signal that is sent as first step + of shutting down a unit (see above), and is usually followed + by SIGKILL (see above and below). For a + list of valid signals, see + signal7. + Defaults to SIGTERM. + + Note that, right after sending the signal specified in + this setting, systemd will always send + SIGCONT, to ensure that even suspended + tasks can be terminated cleanly. + + + + + SendSIGHUP= + Specifies whether to send + SIGHUP to remaining processes immediately + after sending the signal configured with + KillSignal=. This is useful to indicate to + shells and shell-like programs that their connection has been + severed. Takes a boolean value. Defaults to "no". + + + + + SendSIGKILL= + Specifies whether to send + SIGKILL to remaining processes after a + timeout, if the normal shutdown procedure left processes of + the service around. Takes a boolean value. Defaults to "yes". + + + + + + + + See Also + + systemd1, + systemctl1, + journalctl8, + systemd.unit5, + systemd.service5, + systemd.socket5, + systemd.swap5, + systemd.mount5, + systemd.exec5, + systemd.directives7, + kill2, + signal7 + + + + diff --git a/src/grp-system/systemd/systemd.link.xml b/src/grp-system/systemd/systemd.link.xml new file mode 100644 index 0000000000..d5b4d1038d --- /dev/null +++ b/src/grp-system/systemd/systemd.link.xml @@ -0,0 +1,477 @@ + + + + + + + + systemd.link + systemd + + + + Developer + Tom + Gundersen + + + + + + systemd.link + 5 + + + + systemd.link + Network device configuration + + + + link.link + + + + Description + + Network link configuration is performed by the + net_setup_link udev builtin. + + The link files are read from the files located in the system + network directory /usr/lib/systemd/network, + the volatile runtime network directory + /run/systemd/network, and the local + administration network directory + /etc/systemd/network. Link files must have + the extension .link; other extensions are + ignored. All link files are collectively sorted and processed in + lexical order, regardless of the directories in which they live. + However, files with identical filenames replace each other. Files + in /etc have the highest priority, files in + /run take precedence over files with the same + name in /usr/lib. This can be used to + override a system-supplied link file with a local file if needed. + As a special case, an empty file (file size 0) or symlink with the + same name pointing to /dev/null disables the + configuration file entirely (it is "masked"). + + The link file contains a [Match] section, + which determines if a given link file may be applied to a given + device, as well as a [Link] section specifying + how the device should be configured. The first (in lexical order) + of the link files that matches a given device is applied. Note + that a default file 99-default.link is + shipped by the system, any user-supplied + .link should hence have a lexically earlier + name to be considered at all. + + See + udevadm8 + for diagnosing problems with .link files. + + + + [Match] Section Options + + A link file is said to match a device if each of the entries + in the [Match] section matches, or if the + section is empty. The following keys are accepted: + + + + MACAddress= + + The hardware address. + + + + OriginalName= + + A whitespace-separated list of shell-style globs matching + the device name, as exposed by the udev property + "INTERFACE". This can not be used to match on names that have + already been changed from userspace. Caution is advised when matching on + kernel-assigned names, as they are known to be unstable + between reboots. + + + + Path= + + A whitespace-separated list of shell-style globs matching + the persistent path, as exposed by the udev property + ID_PATH. + + + + Driver= + + A whitespace-separated list of shell-style globs matching + the driver currently bound to the device, + as exposed by the udev property DRIVER + of its parent device, or if that is not set, the + driver as exposed by ethtool -i + of the device itself. + + + + Type= + + A whitespace-separated list of shell-style globs matching + the device type, as exposed by the udev + property DEVTYPE. + + + + Host= + + Matches against the hostname or machine + ID of the host. See ConditionHost= in + systemd.unit5 + for details. + + + + Virtualization= + + Checks whether the system is executed in + a virtualized environment and optionally test + whether it is a specific implementation. See + ConditionVirtualization= in + systemd.unit5 + for details. + + + + KernelCommandLine= + + Checks whether a specific kernel command line option + is set (or if prefixed with the exclamation mark unset). See + ConditionKernelCommandLine= in + systemd.unit5 + for details. + + + + Architecture= + + Checks whether the system is running on a specific + architecture. See ConditionArchitecture= + in + systemd.unit5 + for details. + + + + + + + + [Link] Section Options + + The [Link] section accepts the following + keys: + + + + Description= + + A description of the device. + + + + Alias= + + The ifalias is set to this + value. + + + + MACAddressPolicy= + + The policy by which the MAC address should be set. The + available policies are: + + + + + persistent + + If the hardware has a persistent MAC address, as + most hardware should, and if it is used by the kernel, + nothing is done. Otherwise, a new MAC address is + generated which is guaranteed to be the same on every + boot for the given machine and the given device, but + which is otherwise random. This feature depends on ID_NET_NAME_* + properties to exist for the link. On hardware where these + properties are not set, the generation of a persistent MAC address + will fail. + + + + random + + If the kernel is using a random MAC address, + nothing is done. Otherwise, a new address is randomly + generated each time the device appears, typically at + boot. Either way, the random address will have the + unicast and + locally administered bits set. + + + + none + + Keeps the MAC address assigned by the kernel. + + + + + + + MACAddress= + + The MAC address to use, if no + MACAddressPolicy= + is specified. + + + + NamePolicy= + + An ordered, space-separated list of policies by which + the interface name should be set. + NamePolicy may be disabled by specifying + net.ifnames=0 on the kernel command line. + Each of the policies may fail, and the first successful one + is used. The name is not set directly, but is exported to + udev as the property ID_NET_NAME, which + is, by default, used by a udev rule to set + NAME. If the name has already been set by + userspace, no renaming is performed. The available policies + are: + + + + kernel + + If the kernel claims that the name it has set + for a device is predictable, then no renaming is + performed. + + + + database + + The name is set based on entries in the udev's + Hardware Database with the key + ID_NET_NAME_FROM_DATABASE. + + + + + onboard + + The name is set based on information given by + the firmware for on-board devices, as exported by the + udev property ID_NET_NAME_ONBOARD. + + + + + slot + + The name is set based on information given by + the firmware for hot-plug devices, as exported by the + udev property ID_NET_NAME_SLOT. + + + + + path + + The name is set based on the device's physical + location, as exported by the udev property + ID_NET_NAME_PATH. + + + + mac + + The name is set based on the device's persistent + MAC address, as exported by the udev property + ID_NET_NAME_MAC. + + + + + + + Name= + + The interface name to use in case all the + policies specified in + NamePolicy= fail, or in case + NamePolicy= is missing or + disabled. + + + + MTUBytes= + + The maximum transmission unit in bytes to set for the + device. The usual suffixes K, M, G, are supported and are + understood to the base of 1024. + + + + BitsPerSecond= + + The speed to set for the device, the value is rounded + down to the nearest Mbps. The usual suffixes K, M, G, are + supported and are understood to the base of 1000. + + + + Duplex= + + The duplex mode to set for the device. The accepted + values are half and + full. + + + + WakeOnLan= + + The Wake-on-LAN policy to set for the device. The + supported values are: + + + + phy + + Wake on PHY activity. + + + + magic + + Wake on receipt of a magic packet. + + + + + off + + Never wake. + + + + + + + + + + Examples + + + /usr/lib/systemd/network/99-default.link + + The link file 99-default.link that is + shipped with systemd defines the default naming policy for + links. + + [Link] +NamePolicy=kernel database onboard slot path +MACAddressPolicy=persistent + + + + /etc/systemd/network/10-dmz.link + + This example assigns the fixed name + dmz0 to the interface with the MAC address + 00:a0:de:63:7a:e6: + + [Match] +MACAddress=00:a0:de:63:7a:e6 + +[Link] +Name=dmz0 + + + + /etc/systemd/network/10-internet.link + + This example assigns the fixed name + internet0 to the interface with the device + path pci-0000:00:1a.0-*: + + [Match] +Path=pci-0000:00:1a.0-* + +[Link] +Name=internet0 + + + + /etc/systemd/network/25-wireless.link + + Here's an overly complex example that shows the use of a large number of [Match] and [Link] settings. + + [Match] +MACAddress=12:34:56:78:9a:bc +Driver=brcmsmac +Path=pci-0000:02:00.0-* +Type=wlan +Virtualization=no +Host=my-laptop +Architecture=x86-64 + +[Link] +Name=wireless0 +MTUBytes=1450 +BitsPerSecond=10M +WakeOnLan=magic +MACAddress=cb:a9:87:65:43:21 + + + + + See Also + + + systemd-udevd.service8 + , + + udevadm8 + , + + systemd.netdev5 + , + + systemd.network5 + + + + + diff --git a/src/grp-system/systemd/systemd.mount.xml b/src/grp-system/systemd/systemd.mount.xml new file mode 100644 index 0000000000..66cddd72e0 --- /dev/null +++ b/src/grp-system/systemd/systemd.mount.xml @@ -0,0 +1,406 @@ + + + + + + + + systemd.mount + systemd + + + + Developer + Lennart + Poettering + lennart@poettering.net + + + + + + systemd.mount + 5 + + + + systemd.mount + Mount unit configuration + + + + mount.mount + + + + Description + + A unit configuration file whose name ends in + .mount encodes information about a file system + mount point controlled and supervised by systemd. + + This man page lists the configuration options specific to + this unit type. See + systemd.unit5 + for the common options of all unit configuration files. The common + configuration items are configured in the generic [Unit] and + [Install] sections. The mount specific configuration options are + configured in the [Mount] section. + + Additional options are listed in + systemd.exec5, + which define the execution environment the + mount8 + binary is executed in, and in + systemd.kill5, + which define the way the processes are terminated, and in + systemd.resource-control5, + which configure resource control settings for the processes of the + service. Note that the User= and Group= options are not + particularly useful for mount units specifying a + Type= option or using configuration not + specified in /etc/fstab; + mount8 + will refuse options that are not listed in + /etc/fstab if it is not run as UID 0. + + Mount units must be named after the mount point directories they control. Example: the mount point /home/lennart must be configured in a unit file home-lennart.mount. + For details about the escaping logic used to convert a file system path to a unit name, see + systemd.unit5. Note that mount + units cannot be templated, nor is possible to add multiple names to a mount unit by creating additional symlinks to + it. + + Optionally, a mount unit may be accompanied by an automount + unit, to allow on-demand or parallelized mounting. See + systemd.automount5. + + Mount points created at runtime (independently of unit files + or /etc/fstab) will be monitored by systemd + and appear like any other mount unit in systemd. See + /proc/self/mountinfo description in + proc5. + + + Some file systems have special semantics as API file systems + for kernel-to-userspace and userspace-to-userspace interfaces. Some + of them may not be changed via mount units, and cannot be + disabled. For a longer discussion see API + File Systems. + + + + Automatic Dependencies + + If a mount unit is beneath another mount unit in the file + system hierarchy, both a requirement dependency and an ordering + dependency between both units are created automatically. + + Block device backed file systems automatically gain + BindsTo= and After= type + dependencies on the device unit encapsulating the block + device (see below). + + If traditional file system quota is enabled for a mount + unit, automatic Wants= and + Before= dependencies on + systemd-quotacheck.service and + quotaon.service are added. + + For mount units with DefaultDependencies=yes in the [Unit] section (the + default) a couple additional dependencies are added. Mount units referring to local file systems automatically gain + an After= dependency on local-fs-pre.target. Network mount units + automatically acquire After= dependencies on remote-fs-pre.target, + network.target and network-online.target. Towards the latter a + Wants= unit is added as well. Mount units referring to local and network file systems are + distinguished by their file system type specification. In some cases this is not sufficient (for example network + block device based mounts, such as iSCSI), in which case may be added to the mount option + string of the unit, which forces systemd to consider the mount unit a network mount. Mount units (regardless if + local or network) also acquire automatic Before= and Conflicts= on + umount.target in order to be stopped during shutdown. + + Additional implicit dependencies may be added as result of + execution and resource control parameters as documented in + systemd.exec5 + and + systemd.resource-control5. + + + + <filename>fstab</filename> + + Mount units may either be configured via unit files, or via + /etc/fstab (see + fstab5 + for details). Mounts listed in /etc/fstab + will be converted into native units dynamically at boot and when + the configuration of the system manager is reloaded. In general, + configuring mount points through /etc/fstab + is the preferred approach. See + systemd-fstab-generator8 + for details about the conversion. + + The NFS mount option for NFS background mounts + as documented in nfs5 + is not supported in /etc/fstab entries. The systemd mount option + provides similar functionality and should be used instead. + + When reading /etc/fstab a few special + mount options are understood by systemd which influence how + dependencies are created for mount points. systemd will create a + dependency of type Wants= or + (see option + below), from either local-fs.target or + remote-fs.target, depending whether the file + system is local or remote. + + + + + + + Configures a Requires= and + an After= dependency between the created + mount unit and another systemd unit, such as a device or mount + unit. The argument should be a unit name, or an absolute path + to a device node or mount point. This option may be specified + more than once. This option is particularly useful for mount + point declarations that need an additional device to be around + (such as an external journal device for journal file systems) + or an additional mount to be in place (such as an overlay file + system that merges multiple mount points). See + After= and Requires= in + systemd.unit5 + for details. + + + + + + Configures a + RequiresMountsFor= dependency between the + created mount unit and other mount units. The argument must be + an absolute path. This option may be specified more than once. + See RequiresMountsFor= in + systemd.unit5 + for details. + + + + + + An automount unit will be created for the file + system. See + systemd.automount5 + for details. + + + + + + Configures the idle timeout of the + automount unit. See TimeoutIdleSec= in + systemd.automount5 + for details. + + + + + + Configure how long systemd should wait for a + device to show up before giving up on an entry from + /etc/fstab. Specify a time in seconds or + explicitly append a unit such as s, + min, h, + ms. + + Note that this option can only be used in + /etc/fstab, and will be + ignored when part of the Options= + setting in a unit file. + + + + + + + + With , this mount will + not be added as a dependency for + local-fs.target or + remote-fs.target. This means that it will + not be mounted automatically during boot, unless it is pulled + in by some other unit. The option has the + opposite meaning and is the default. + + + + + + + With , this mount will + be only wanted, not required, by + local-fs.target or + remote-fs.target. This means that the + boot will continue even if this mount point is not mounted + successfully. + + + + + + + An additional filesystem to be mounted in the + initramfs. See initrd-fs.target + description in + systemd.special7. + + + + + If a mount point is configured in both + /etc/fstab and a unit file that is stored + below /usr, the former will take precedence. + If the unit file is stored below /etc, it + will take precedence. This means: native unit files take + precedence over traditional configuration files, but this is + superseded by the rule that configuration in + /etc will always take precedence over + configuration in /usr. + + + + Options + + Mount files must include a [Mount] section, which carries + information about the file system mount points it supervises. A + number of options that may be used in this section are shared with + other unit types. These options are documented in + systemd.exec5 + and + systemd.kill5. + The options specific to the [Mount] section of mount units are the + following: + + + + + What= + Takes an absolute path of a device node, file + or other resource to mount. See + mount8 + for details. If this refers to a device node, a dependency on + the respective device unit is automatically created. (See + systemd.device5 + for more information.) This option is + mandatory. + + + + Where= + Takes an absolute path of a directory of the + mount point. If the mount point does not exist at the time of + mounting, it is created. This string must be reflected in the + unit filename. (See above.) This option is + mandatory. + + + + Type= + Takes a string for the file system type. See + mount8 + for details. This setting is optional. + + + + Options= + + Mount options to use when mounting. This takes + a comma-separated list of options. This setting is + optional. + + + + SloppyOptions= + + Takes a boolean argument. If true, parsing of + the options specified in Options= is + relaxed, and unknown mount options are tolerated. This + corresponds with + mount8's + -s switch. Defaults to + off. + + + + DirectoryMode= + Directories of mount points (and any parent + directories) are automatically created if needed. This option + specifies the file system access mode used when creating these + directories. Takes an access mode in octal notation. Defaults + to 0755. + + + + TimeoutSec= + Configures the time to wait for the mount + command to finish. If a command does not exit within the + configured time, the mount will be considered failed and be + shut down again. All commands still running will be terminated + forcibly via SIGTERM, and after another + delay of this time with SIGKILL. (See + in + systemd.kill5.) + Takes a unit-less value in seconds, or a time span value such + as "5min 20s". Pass 0 to disable the timeout logic. The + default value is set from the manager configuration file's + DefaultTimeoutStart= + variable. + + + + Check + systemd.exec5 + and + systemd.kill5 + for more settings. + + + + See Also + + systemd1, + systemctl1, + systemd.unit5, + systemd.exec5, + systemd.kill5, + systemd.resource-control5, + systemd.service5, + systemd.device5, + proc5, + mount8, + systemd-fstab-generator8, + systemd.directives7 + + + + diff --git a/src/grp-system/systemd/systemd.netdev.xml b/src/grp-system/systemd/systemd.netdev.xml new file mode 100644 index 0000000000..a5c6f0fa40 --- /dev/null +++ b/src/grp-system/systemd/systemd.netdev.xml @@ -0,0 +1,1164 @@ + + + + + + + + + systemd.network + systemd + + + + Developer + Tom + Gundersen + teg@jklm.no + + + + + + systemd.netdev + 5 + + + + systemd.netdev + Virtual Network Device configuration + + + + netdev.netdev + + + + Description + + Network setup is performed by + systemd-networkd8. + + + Virtual Network Device files must have the extension + .netdev; other extensions are ignored. + Virtual network devices are created as soon as networkd is + started. If a netdev with the specified name already exists, + networkd will use that as-is rather than create its own. Note that + the settings of the pre-existing netdev will not be changed by + networkd. + + The .netdev files are read from the + files located in the system network directory + /usr/lib/systemd/network, the volatile + runtime network directory + /run/systemd/network and the local + administration network directory + /etc/systemd/network. All configuration files + are collectively sorted and processed in lexical order, regardless + of the directories in which they live. However, files with + identical filenames replace each other. Files in + /etc have the highest priority, files in + /run take precedence over files with the same + name in /usr/lib. This can be used to + override a system-supplied configuration file with a local file if + needed. As a special case, an empty file (file size 0) or symlink + with the same name pointing to /dev/null + disables the configuration file entirely (it is "masked"). + + + + Supported netdev kinds + + The following kinds of virtual network devices may be + configured in .netdev files: + + + Supported kinds of virtual network devices + + + + + + Kind + Description + + + bond + A bond device is an aggregation of all its slave devices. See Linux Ethernet Bonding Driver HOWTO for details.Local configuration + + bridge + A bridge device is a software switch, and each of its slave devices and the bridge itself are ports of the switch. + + dummy + A dummy device drops all packets sent to it. + + gre + A Level 3 GRE tunnel over IPv4. See RFC 2784 for details. + + gretap + A Level 2 GRE tunnel over IPv4. + + ip6gre + A Level 3 GRE tunnel over IPv6. + + ip6tnl + An IPv4 or IPv6 tunnel over IPv6 + + ip6gretap + A Level 2 GRE tunnel over IPv6. + + ipip + An IPv4 over IPv4 tunnel. + + ipvlan + An ipvlan device is a stacked device which receives packets from its underlying device based on IP address filtering. + + macvlan + A macvlan device is a stacked device which receives packets from its underlying device based on MAC address filtering. + + macvtap + A macvtap device is a stacked device which receives packets from its underlying device based on MAC address filtering. + + sit + An IPv6 over IPv4 tunnel. + + tap + A persistent Level 2 tunnel between a network device and a device node. + + tun + A persistent Level 3 tunnel between a network device and a device node. + + veth + An Ethernet tunnel between a pair of network devices. + + vlan + A VLAN is a stacked device which receives packets from its underlying device based on VLAN tagging. See IEEE 802.1Q for details. + + vti + An IPv4 over IPSec tunnel. + + vti6 + An IPv6 over IPSec tunnel. + + vxlan + A virtual extensible LAN (vxlan), for connecting Cloud computing deployments. + + vrf + A Virtual Routing and Forwarding (VRF) interface to create separate routing and forwarding domains. + + + +
+ + + + + [Match] Section Options + + A virtual network device is only created if the + [Match] section matches the current + environment, or if the section is empty. The following keys are + accepted: + + + + Host= + + Matches against the hostname or machine ID of the + host. See ConditionHost= in + systemd.unit5 + for details. + + + + + Virtualization= + + Checks whether the system is executed in a virtualized + environment and optionally test whether it is a specific + implementation. See + ConditionVirtualization= in + systemd.unit5 + for details. + + + + + KernelCommandLine= + + Checks whether a specific kernel command line option + is set (or if prefixed with the exclamation mark unset). See + ConditionKernelCommandLine= in + systemd.unit5 + for details. + + + + + Architecture= + + Checks whether the system is running on a specific + architecture. See ConditionArchitecture= in + systemd.unit5 + for details. + + + + + + + + + [NetDev] Section Options + + The [NetDev] section accepts the + following keys: + + + + Description= + + A free-form description of the netdev. + + + + Name= + + The interface name used when creating the netdev. + This option is compulsory. + + + + Kind= + + The netdev kind. This option is compulsory. See the + Supported netdev kinds section for the + valid keys. + + + + MTUBytes= + + The maximum transmission unit in bytes to set for + the device. The usual suffixes K, M, G, are supported and + are understood to the base of 1024. This key is not + currently supported for tun or + tap devices. + + + + + MACAddress= + + The MAC address to use for the device. If none is + given, one is generated based on the interface name and + the + machine-id5. + This key is not currently supported for + tun or tap devices. + + + + + + + + [Bridge] Section Options + + The [Bridge] section only applies for + netdevs of kind bridge, and accepts the + following keys: + + + + HelloTimeSec= + + HelloTimeSec specifies the number of seconds between two hello packets + sent out by the root bridge and the designated bridges. Hello packets are + used to communicate information about the topology throughout the entire + bridged local area network. + + + + MaxAgeSec= + + MaxAgeSec specifies the number of seconds of maximum message age. + If the last seen (received) hello packet is more than this number of + seconds old, the bridge in question will start the takeover procedure + in attempt to become the Root Bridge itself. + + + + ForwardDelaySec= + + ForwardDelaySec specifies the number of seconds spent in each + of the Listening and Learning states before the Forwarding state is entered. + + + + MulticastQuerier= + + A boolean. This setting controls the IFLA_BR_MCAST_QUERIER option in the kernel. + If enabled, the kernel will send general ICMP queries from a zero source address. + This feature should allow faster convergence on startup, but it causes some + multicast-aware switches to misbehave and disrupt forwarding of multicast packets. + When unset, the kernel's default setting applies. + + + + + MulticastSnooping= + + A boolean. This setting controls the IFLA_BR_MCAST_SNOOPING option in the kernel. + If enabled, IGMP snooping monitors the Internet Group Management Protocol (IGMP) traffic + between hosts and multicast routers. When unset, the kernel's default setting applies. + + + + + VLANFiltering= + + A boolean. This setting controls the IFLA_BR_VLAN_FILTERING option in the kernel. + If enabled, the bridge will be started in VLAN-filtering mode. When unset, the kernel's + default setting applies. + + + + + + + + + [VLAN] Section Options + + The [VLAN] section only applies for + netdevs of kind vlan, and accepts the + following key: + + + + Id= + + The VLAN ID to use. An integer in the range 0–4094. + This option is compulsory. + + + + + + + + [MACVLAN] Section Options + + The [MACVLAN] section only applies for + netdevs of kind macvlan, and accepts the + following key: + + + + Mode= + + The MACVLAN mode to use. The supported options are + private, + vepa, + bridge, and + passthru. + + + + + + + + + [MACVTAP] Section Options + + The [MACVTAP] section applies for + netdevs of kind macvtap and accepts the + same key as [MACVLAN]. + + + + + [IPVLAN] Section Options + + The [IPVLAN] section only applies for + netdevs of kind ipvlan, and accepts the + following key: + + + + Mode= + + The IPVLAN mode to use. The supported options are + L2 and L3. + + + + + + + + + [VXLAN] Section Options + The [VXLAN] section only applies for + netdevs of kind vxlan, and accepts the + following keys: + + + + Id= + + The VXLAN ID to use. + + + + Group= + + An assigned multicast group IP address. + + + + TOS= + + The Type Of Service byte value for a vxlan interface. + + + + TTL= + + A fixed Time To Live N on Virtual eXtensible Local + Area Network packets. N is a number in the range 1–255. 0 + is a special value meaning that packets inherit the TTL + value. + + + + MacLearning= + + A boolean. When true, enables dynamic MAC learning + to discover remote MAC addresses. + + + + FDBAgeingSec= + + The lifetime of Forwarding Database entry learnt by + the kernel, in seconds. + + + + MaximumFDBEntries= + + Configures maximum number of FDB entries. + + + + ARPProxy= + + A boolean. When true, enables ARP proxying. + + + + L2MissNotification= + + A boolean. When true, enables netlink LLADDR miss + notifications. + + + + L3MissNotification= + + A boolean. When true, enables netlink IP address miss + notifications. + + + + RouteShortCircuit= + + A boolean. When true, route short circuiting is turned + on. + + + + UDPCheckSum= + + A boolean. When true, transmitting UDP checksums when doing VXLAN/IPv4 is turned on. + + + + UDP6ZeroChecksumTx= + + A boolean. When true, sending zero checksums in VXLAN/IPv6 is turned on. + + + + UDP6ZeroCheckSumRx= + + A boolean. When true, receiving zero checksums in VXLAN/IPv6 is turned on. + + + + GroupPolicyExtension= + + A boolean. When true, it enables Group Policy VXLAN extension security label mechanism + across network peers based on VXLAN. For details about the Group Policy VXLAN, see the + + VXLAN Group Policy document. Defaults to false. + + + + DestinationPort= + + Configures the default destination UDP port on a per-device basis. + If destination port is not specified then Linux kernel default will be used. + Set destination port 4789 to get the IANA assigned value, + and destination port 0 to get default values. + + + + PortRange= + + Configures VXLAN port range. VXLAN bases source + UDP port based on flow to help the receiver to be able + to load balance based on outer header flow. It + restricts the port range to the normal UDP local + ports, and allows overriding via configuration. + + + + + + [Tunnel] Section Options + + The [Tunnel] section only applies for + netdevs of kind + ipip, + sit, + gre, + gretap, + ip6gre, + ip6gretap, + vti, + vti6, and + ip6tnl and accepts + the following keys: + + + + Local= + + A static local address for tunneled packets. It must + be an address on another interface of this host. + + + + Remote= + + The remote endpoint of the tunnel. + + + + TOS= + + The Type Of Service byte value for a tunnel interface. + For details about the TOS, see the + Type of + Service in the Internet Protocol Suite document. + + + + + TTL= + + A fixed Time To Live N on tunneled packets. N is a + number in the range 1–255. 0 is a special value meaning that + packets inherit the TTL value. The default value for IPv4 + tunnels is: inherit. The default value for IPv6 tunnels is + 64. + + + + DiscoverPathMTU= + + A boolean. When true, enables Path MTU Discovery on + the tunnel. + + + + IPv6FlowLabel= + + Configures the 20-bit flow label (see + RFC 6437) field in the IPv6 header (see + RFC 2460), which is used by a node to label packets of a flow. + It is only used for IPv6 tunnels. + A flow label of zero is used to indicate packets that have + not been labeled. + It can be configured to a value in the range 0–0xFFFFF, or be + set to inherit, in which case the original flowlabel is used. + + + + CopyDSCP= + + A boolean. When true, the Differentiated Service Code + Point (DSCP) field will be copied to the inner header from + outer header during the decapsulation of an IPv6 tunnel + packet. DSCP is a field in an IP packet that enables different + levels of service to be assigned to network traffic. + Defaults to no. + + + + + EncapsulationLimit= + + The Tunnel Encapsulation Limit option specifies how many additional + levels of encapsulation are permitted to be prepended to the packet. + For example, a Tunnel Encapsulation Limit option containing a limit + value of zero means that a packet carrying that option may not enter + another tunnel before exiting the current tunnel. + (see RFC 2473). + The valid range is 0–255 and none. Defaults to 4. + + + + + Key= + + The Key= parameter specifies the same key to use in + both directions (InputKey= and OutputKey=). + The Key= is either a number or an IPv4 address-like dotted quad. + It is used as mark-configured SAD/SPD entry as part of the lookup key (both in data + and control path) in ip xfrm (framework used to implement IPsec protocol). + See + ip-xfrm — transform configuration for details. It is only used for VTI/VTI6 + tunnels. + + + + InputKey= + + The InputKey= parameter specifies the key to use for input. + The format is same as Key=. It is only used for VTI/VTI6 tunnels. + + + + OutputKey= + + The OutputKey= parameter specifies the key to use for output. + The format is same as Key=. It is only used for VTI/VTI6 tunnels. + + + + Mode= + + An ip6tnl tunnel can be in one of three + modes + ip6ip6 for IPv6 over IPv6, + ipip6 for IPv4 over IPv6 or + any for either. + + + + + + + [Peer] Section Options + + The [Peer] section only applies for + netdevs of kind veth and accepts the + following keys: + + + + Name= + + The interface name used when creating the netdev. + This option is compulsory. + + + + MACAddress= + + The peer MACAddress, if not set, it is generated in + the same way as the MAC address of the main + interface. + + + + + + [Tun] Section Options + + The [Tun] section only applies for + netdevs of kind tun, and accepts the following + keys: + + + + OneQueue= + Takes a boolean argument. Configures whether + all packets are queued at the device (enabled), or a fixed + number of packets are queued at the device and the rest at the + qdisc. Defaults to + no. + + + + MultiQueue= + Takes a boolean argument. Configures whether + to use multiple file descriptors (queues) to parallelize + packets sending and receiving. Defaults to + no. + + + + PacketInfo= + Takes a boolean argument. Configures whether + packets should be prepended with four extra bytes (two flag + bytes and two protocol bytes). If disabled, it indicates that + the packets will be pure IP packets. Defaults to + no. + + + + VNetHeader= + Takes a boolean argument. Configures + IFF_VNET_HDR flag for a tap device. It allows sending + and receiving larger Generic Segmentation Offload (GSO) + packets. This may increase throughput significantly. + Defaults to + no. + + + + User= + User to grant access to the + /dev/net/tun device. + + + + Group= + Group to grant access to the + /dev/net/tun device. + + + + + + + + + [Tap] Section Options + + The [Tap] section only applies for + netdevs of kind tap, and accepts the same keys + as the [Tun] section. + + + + [Bond] Section Options + + The [Bond] section accepts the following + key: + + + + Mode= + + Specifies one of the bonding policies. The default is + balance-rr (round robin). Possible values are + balance-rr, + active-backup, + balance-xor, + broadcast, + 802.3ad, + balance-tlb, and + balance-alb. + + + + + + TransmitHashPolicy= + + Selects the transmit hash policy to use for slave + selection in balance-xor, 802.3ad, and tlb modes. Possible + values are + layer2, + layer3+4, + layer2+3, + encap2+3, and + encap3+4. + + + + + + LACPTransmitRate= + + Specifies the rate with which link partner transmits + Link Aggregation Control Protocol Data Unit packets in + 802.3ad mode. Possible values are slow, + which requests partner to transmit LACPDUs every 30 seconds, + and fast, which requests partner to + transmit LACPDUs every second. The default value is + slow. + + + + + MIIMonitorSec= + + Specifies the frequency that Media Independent + Interface link monitoring will occur. A value of zero + disables MII link monitoring. This value is rounded down to + the nearest millisecond. The default value is 0. + + + + + UpDelaySec= + + Specifies the delay before a link is enabled after a + link up status has been detected. This value is rounded down + to a multiple of MIIMonitorSec. The default value is + 0. + + + + + DownDelaySec= + + Specifies the delay before a link is disabled after a + link down status has been detected. This value is rounded + down to a multiple of MIIMonitorSec. The default value is + 0. + + + + + LearnPacketIntervalSec= + + Specifies the number of seconds between instances where the bonding + driver sends learning packets to each slave peer switch. + The valid range is 1–0x7fffffff; the default value is 1. This option + has an effect only for the balance-tlb and balance-alb modes. + + + + + AdSelect= + + Specifies the 802.3ad aggregation selection logic to use. Possible values are + stable, + bandwidth and + count. + + + + + + FailOverMACPolicy= + + Specifies whether the active-backup mode should set all slaves to + the same MAC address at the time of enslavement or, when enabled, to perform special handling of the + bond's MAC address in accordance with the selected policy. The default policy is none. + Possible values are + none, + active and + follow. + + + + + + ARPValidate= + + Specifies whether or not ARP probes and replies should be + validated in any mode that supports ARP monitoring, or whether + non-ARP traffic should be filtered (disregarded) for link + monitoring purposes. Possible values are + none, + active, + backup and + all. + + + + + + ARPIntervalSec= + + Specifies the ARP link monitoring frequency in milliseconds. + A value of 0 disables ARP monitoring. The default value is 0. + + + + + + ARPIPTargets= + + Specifies the IP addresses to use as ARP monitoring peers when + ARPIntervalSec is greater than 0. These are the targets of the ARP request + sent to determine the health of the link to the targets. + Specify these values in IPv4 dotted decimal format. At least one IP + address must be given for ARP monitoring to function. The + maximum number of targets that can be specified is 16. The + default value is no IP addresses. + + + + + + ARPAllTargets= + + Specifies the quantity of ARPIPTargets that must be reachable + in order for the ARP monitor to consider a slave as being up. + This option affects only active-backup mode for slaves with + ARPValidate enabled. Possible values are + any and + all. + + + + + + PrimaryReselectPolicy= + + Specifies the reselection policy for the primary slave. This + affects how the primary slave is chosen to become the active slave + when failure of the active slave or recovery of the primary slave + occurs. This option is designed to prevent flip-flopping between + the primary slave and other slaves. Possible values are + always, + better and + failure. + + + + + + ResendIGMP= + + Specifies the number of IGMP membership reports to be issued after + a failover event. One membership report is issued immediately after + the failover, subsequent packets are sent in each 200ms interval. + The valid range is 0–255. Defaults to 1. A value of 0 + prevents the IGMP membership report from being issued in response + to the failover event. + + + + + + PacketsPerSlave= + + Specify the number of packets to transmit through a slave before + moving to the next one. When set to 0, then a slave is chosen at + random. The valid range is 0–65535. Defaults to 1. This option + only has effect when in balance-rr mode. + + + + + + GratuitousARP= + + Specify the number of peer notifications (gratuitous ARPs and + unsolicited IPv6 Neighbor Advertisements) to be issued after a + failover event. As soon as the link is up on the new slave, + a peer notification is sent on the bonding device and each + VLAN sub-device. This is repeated at each link monitor interval + (ARPIntervalSec or MIIMonitorSec, whichever is active) if the number is + greater than 1. The valid range is 0–255. The default value is 1. + These options affect only the active-backup mode. + + + + + + AllSlavesActive= + + A boolean. Specifies that duplicate frames (received on inactive ports) + should be dropped when false, or delivered when true. Normally, bonding will drop + duplicate frames (received on inactive ports), which is desirable for + most users. But there are some times it is nice to allow duplicate + frames to be delivered. The default value is false (drop duplicate frames + received on inactive ports). + + + + + + MinLinks= + + Specifies the minimum number of links that must be active before + asserting carrier. The default value is 0. + + + + + + + For more detail information see + + Linux Ethernet Bonding Driver HOWTO + + + + + Example + + /etc/systemd/network/25-bridge.netdev + + [NetDev] +Name=bridge0 +Kind=bridge + + + + /etc/systemd/network/25-vlan1.netdev + + [Match] +Virtualization=no + +[NetDev] +Name=vlan1 +Kind=vlan + +[VLAN] +Id=1 + + + /etc/systemd/network/25-ipip.netdev + [NetDev] +Name=ipip-tun +Kind=ipip +MTUBytes=1480 + +[Tunnel] +Local=192.168.223.238 +Remote=192.169.224.239 +TTL=64 + + + /etc/systemd/network/25-tap.netdev + [NetDev] +Name=tap-test +Kind=tap + +[Tap] +MultiQueue=true +PacketInfo=true + + + /etc/systemd/network/25-sit.netdev + [NetDev] +Name=sit-tun +Kind=sit +MTUBytes=1480 + +[Tunnel] +Local=10.65.223.238 +Remote=10.65.223.239 + + + + /etc/systemd/network/25-gre.netdev + [NetDev] +Name=gre-tun +Kind=gre +MTUBytes=1480 + +[Tunnel] +Local=10.65.223.238 +Remote=10.65.223.239 + + + + /etc/systemd/network/25-vti.netdev + + [NetDev] +Name=vti-tun +Kind=vti +MTUBytes=1480 + +[Tunnel] +Local=10.65.223.238 +Remote=10.65.223.239 + + + + /etc/systemd/network/25-veth.netdev + [NetDev] +Name=veth-test +Kind=veth + +[Peer] +Name=veth-peer + + + + /etc/systemd/network/25-bond.netdev + [NetDev] +Name=bond1 +Kind=bond + +[Bond] +Mode=802.3ad +TransmitHashPolicy=layer3+4 +MIIMonitorSec=1s +LACPTransmitRate=fast + + + + + /etc/systemd/network/25-dummy.netdev + [NetDev] +Name=dummy-test +Kind=dummy +MACAddress=12:34:56:78:9a:bc + + + /etc/systemd/network/25-vrf.netdev + Create a VRF interface with table 42. + [NetDev] +Name=vrf-test +Kind=vrf + +[VRF] +TableId=42 + + + + See Also + + systemd1, + systemd-networkd8, + systemd.link5, + systemd.network5 + + + + diff --git a/src/grp-system/systemd/systemd.network.xml b/src/grp-system/systemd/systemd.network.xml new file mode 100644 index 0000000000..4541a55490 --- /dev/null +++ b/src/grp-system/systemd/systemd.network.xml @@ -0,0 +1,1312 @@ + + + + + + + + + systemd.network + systemd + + + + Developer + Tom + Gundersen + teg@jklm.no + + + + + + systemd.network + 5 + + + + systemd.network + Network configuration + + + + network.network + + + + Description + + Network setup is performed by + systemd-networkd8. + + + Network files must have the extension + .network; other extensions are ignored. + Networks are applied to links whenever the links appear. + + The .network files are read from the + files located in the system network directory + /usr/lib/systemd/network, the volatile + runtime network directory + /run/systemd/network and the local + administration network directory + /etc/systemd/network. All configuration files + are collectively sorted and processed in lexical order, regardless + of the directories in which they live. However, files with + identical filenames replace each other. Files in + /etc have the highest priority, files in + /run take precedence over files with the same + name in /usr/lib. This can be used to + override a system-supplied configuration file with a local file if + needed. As a special case, an empty file (file size 0) or symlink + with the same name pointing to /dev/null + disables the configuration file entirely (it is "masked"). + + Note that an interface without any static IPv6 addresses configured, and neither DHCPv6 nor IPv6LL enabled, + shall be considered to have no IPv6 support. IPv6 will be automatically disabled for that interface by writing "1" + to /proc/sys/net/ipv6/conf/ifname/disable_ipv6. + + + + + [Match] Section Options + + The network file contains a [Match] + section, which determines if a given network file may be applied + to a given device; and a [Network] section + specifying how the device should be configured. The first (in + lexical order) of the network files that matches a given device + is applied, all later files are ignored, even if they match as + well. + + A network file is said to match a device if each of the + entries in the [Match] section matches, or if + the section is empty. The following keys are accepted: + + + + MACAddress= + + The hardware address of the interface (use full colon-delimited hexadecimal, e.g., + 01:23:45:67:89:ab). + + + + Path= + + A whitespace-separated list of shell-style globs + matching the persistent path, as exposed by the udev + property ID_PATH. + + + + Driver= + + A whitespace-separated list of shell-style globs + matching the driver currently bound to the device, as + exposed by the udev property DRIVER + of its parent device, or if that is not set the driver + as exposed by ethtool -i of the + device itself. + + + + Type= + + A whitespace-separated list of shell-style globs + matching the device type, as exposed by the udev property + DEVTYPE. + + + + Name= + + A whitespace-separated list of shell-style globs + matching the device name, as exposed by the udev property + INTERFACE. + + + + Host= + + Matches against the hostname or machine ID of the + host. See ConditionHost= in + systemd.unit5 + for details. + + + + + Virtualization= + + Checks whether the system is executed in a virtualized + environment and optionally test whether it is a specific + implementation. See ConditionVirtualization= in + systemd.unit5 + for details. + + + + + KernelCommandLine= + + Checks whether a specific kernel command line option is + set (or if prefixed with the exclamation mark unset). See + ConditionKernelCommandLine= in + systemd.unit5 + for details. + + + + + Architecture= + + Checks whether the system is running on a specific + architecture. See ConditionArchitecture= in + systemd.unit5 + for details. + + + + + + + + + [Link] Section Options + + The [Link] section accepts the following keys: + + + + MACAddress= + + The hardware address to set for the device. + + + + MTUBytes= + + The maximum transmission unit in bytes to set for the + device. The usual suffixes K, M, G, are supported and are + understood to the base of 1024. + Note that if IPv6 is enabled on the interface, and the MTU is chosen + below 1280 (the minimum MTU for IPv6) it will automatically be increased to this value. + + + + + + + [Network] Section Options + + The [Network] section accepts the following keys: + + + + Description= + + A description of the device. This is only used for + presentation purposes. + + + + DHCP= + + Enables DHCPv4 and/or DHCPv6 client support. Accepts + yes, no, + ipv4, or ipv6. + + Note that DHCPv6 will by default be triggered by Router + Advertisement, if that is enabled, regardless of this parameter. + By enabling DHCPv6 support explicitly, the DHCPv6 client will + be started regardless of the presence of routers on the link, + or what flags the routers pass. See + IPv6AcceptRA=. + + Furthermore, note that by default the domain name + specified through DHCP is not used for name resolution. + See option below. + + See the [DHCP] section below for further configuration options for the DHCP client + support. + + + + DHCPServer= + + A boolean. Enables DHCPv4 server support. Defaults + to no. Further settings for the DHCP + server may be set in the [DHCPServer] + section described below. + + + + LinkLocalAddressing= + + Enables link-local address autoconfiguration. Accepts + yes, no, + ipv4, or ipv6. Defaults to + ipv6. + + + + IPv4LLRoute= + + A boolean. When true, sets up the route needed for + non-IPv4LL hosts to communicate with IPv4LL-only hosts. Defaults + to false. + + + + + IPv6Token= + + An IPv6 address with the top 64 bits unset. When set, indicates the + 64-bit interface part of SLAAC IPv6 addresses for this link. Note that + the token is only ever used for SLAAC, and not for DHCPv6 addresses, even + in the case DHCP is requested by router advertisement. By default, the + token is autogenerated. + + + + LLMNR= + + A boolean or resolve. When true, + enables Link-Local + Multicast Name Resolution on the link. When set to + resolve, only resolution is enabled, + but not host registration and announcement. Defaults to + true. This setting is read by + systemd-resolved.service8. + + + + MulticastDNS= + + A boolean or resolve. When true, + enables Multicast + DNS support on the link. When set to + resolve, only resolution is enabled, + but not host or service registration and + announcement. Defaults to false. This setting is read by + systemd-resolved.service8. + + + + DNSSEC= + + A boolean or + allow-downgrade. When true, enables + DNSSEC + DNS validation support on the link. When set to + allow-downgrade, compatibility with + non-DNSSEC capable networks is increased, by automatically + turning off DNSEC in this case. This option defines a + per-interface setting for + resolved.conf5's + global DNSSEC= option. Defaults to + false. This setting is read by + systemd-resolved.service8. + + + + DNSSECNegativeTrustAnchors= + A space-separated list of DNSSEC negative + trust anchor domains. If specified and DNSSEC is enabled, + look-ups done via the interface's DNS server will be subject + to the list of negative trust anchors, and not require + authentication for the specified domains, or anything below + it. Use this to disable DNSSEC authentication for specific + private domains, that cannot be proven valid using the + Internet DNS hierarchy. Defaults to the empty list. This + setting is read by + systemd-resolved.service8. + + + + LLDP= + + Controls support for Ethernet LLDP packet reception. LLDP is a link-layer protocol commonly + implemented on professional routers and bridges which announces which physical port a system is connected + to, as well as other related data. Accepts a boolean or the special value + routers-only. When true, incoming LLDP packets are accepted and a database of all LLDP + neighbors maintained. If routers-only is set only LLDP data of various types of routers + is collected and LLDP data about other types of devices ignored (such as stations, telephones and + others). If false, LLDP reception is disabled. Defaults to routers-only. Use + networkctl1 to query the + collected neighbor data. LLDP is only available on Ethernet links. See EmitLLDP= below + for enabling LLDP packet emission from the local system. + + + + + EmitLLDP= + + Controls support for Ethernet LLDP packet emission. Accepts a boolean parameter or the special values + nearest-bridge, non-tpmr-bridge and + customer-bridge. Defaults to false, which turns off LLDP packet emission. If not false, + a short LLDP packet with information about the local system is sent out in regular intervals on the + link. The LLDP packet will contain information about the local host name, the local machine ID (as stored + in machine-id5) and the + local interface name, as well as the pretty hostname of the system (as set in + machine-info5). LLDP + emission is only available on Ethernet links. Note that this setting passes data suitable for + identification of host to the network and should thus not be enabled on untrusted networks, where such + identification data should not be made available. Use this option to permit other systems to identify on + which interfaces they are connected to this system. The three special values control propagation of the + LLDP packets. The nearest-bridge setting permits propagation only to the nearest + connected bridge, non-tpmr-bridge permits propagation across Two-Port MAC Relays, but + not any other bridges, and customer-bridge permits propagation until a customer bridge + is reached. For details about these concepts, see IEEE 802.1AB-2009. Note that + configuring this setting to true is equivalent to nearest-bridge, the recommended and + most restricted level of propagation. See LLDP= above for an option to enable LLDP + reception. + + + + BindCarrier= + + A link name or a list of link names. When set, controls the behavior of the current + link. When all links in the list are in an operational down state, the current link is brought + down. When at least one link has carrier, the current interface is brought up. + + + + + Address= + + A static IPv4 or IPv6 address and its prefix length, + separated by a / character. Specify + this key more than once to configure several addresses. + The format of the address must be as described in + inet_pton3. + This is a short-hand for an [Address] section only + containing an Address key (see below). This option may be + specified more than once. + + + If the specified address is 0.0.0.0 (for IPv4) or + [::] (for IPv6), a new address range of the requested size + is automatically allocated from a system-wide pool of + unused ranges. The allocated range is checked against all + current network interfaces and all known network + configuration files to avoid address range conflicts. The + default system-wide pool consists of 192.168.0.0/16, + 172.16.0.0/12 and 10.0.0.0/8 for IPv4, and fc00::/7 for + IPv6. This functionality is useful to manage a large + number of dynamically created network interfaces with the + same network configuration and automatic address range + assignment. + + + + + Gateway= + + The gateway address, which must be in the format + described in + inet_pton3. + This is a short-hand for a [Route] section only containing + a Gateway key. This option may be specified more than + once. + + + + DNS= + + A DNS server address, which must be in the format + described in + inet_pton3. + This option may be specified more than once. This setting is read by + systemd-resolved.service8. + + + + Domains= + + The domains used for DNS host name resolution on this link. Takes a list of DNS domain names which + are used as search suffixes for extending single-label host names (host names containing no dots) to become + fully qualified domain names (FQDNs). If a single-label host name is resolved on this interface, each of + the specified search domains are appended to it in turn, converting it into a fully qualified domain name, + until one of them may be successfully resolved. + + The specified domains are also used for routing of DNS queries: look-ups for host names ending in the + domains specified here are preferably routed to the DNS servers configured for this interface. If a domain + name is prefixed with ~, the domain name becomes a pure "routing" domain, is used for + DNS query routing purposes only and is not used in the described domain search logic. By specifying a + routing domain of ~. (the tilde indicating definition of a routing domain, the dot + referring to the DNS root domain which is the implied suffix of all valid DNS names) it is possible to + route all DNS traffic preferably to the DNS server specified for this interface. The route domain logic is + particularly useful on multi-homed hosts with DNS servers serving particular private DNS zones on each + interface. + + This setting is read by + systemd-resolved.service8. + + + + NTP= + + An NTP server address. This option may be specified more than once. This setting is read by + systemd-timesyncd.service8. + + + + IPForward= + Configures IP packet forwarding for the + system. If enabled, incoming packets on any network + interface will be forwarded to any other interfaces + according to the routing table. Takes either a boolean + argument, or the values ipv4 or + ipv6, which only enable IP packet + forwarding for the specified address family. This controls + the net.ipv4.ip_forward and + net.ipv6.conf.all.forwarding sysctl + options of the network interface (see ip-sysctl.txt + for details about sysctl options). Defaults to + no. + + Note: this setting controls a global kernel option, + and does so one way only: if a network that has this setting + enabled is set up the global setting is turned on. However, + it is never turned off again, even after all networks with + this setting enabled are shut down again. + + To allow IP packet forwarding only between specific + network interfaces use a firewall. + + + + IPMasquerade= + Configures IP masquerading for the network + interface. If enabled, packets forwarded from the network + interface will be appear as coming from the local host. + Takes a boolean argument. Implies + IPForward=ipv4. Defaults to + no. + + + IPv6PrivacyExtensions= + Configures use of stateless temporary + addresses that change over time (see RFC 4941, + Privacy Extensions for Stateless Address Autoconfiguration + in IPv6). Takes a boolean or the special values + prefer-public and + kernel. When true, enables the privacy + extensions and prefers temporary addresses over public + addresses. When prefer-public, enables the + privacy extensions, but prefers public addresses over + temporary addresses. When false, the privacy extensions + remain disabled. When kernel, the kernel's + default setting will be left in place. Defaults to + no. + + + IPv6AcceptRA= + Enable or disable IPv6 Router Advertisement (RA) reception support for the interface. Takes + a boolean parameter. If true, RAs are accepted; if false, RAs are ignored, independently of the local + forwarding state. When not set, the kernel default is used, and RAs are accepted only when local forwarding + is disabled for that interface. When RAs are accepted, they may trigger the start of the DHCPv6 client if + the relevant flags are set in the RA data, or if no routers are found on the link. + + Further settings for the IPv6 RA support may be configured in the + [IPv6AcceptRA] section, see below. + + Also see ip-sysctl.txt in the kernel + documentation regarding accept_ra, but note that systemd's setting of + 1 (i.e. true) corresponds to kernel's setting of 2. + + + + IPv6DuplicateAddressDetection= + Configures the amount of IPv6 Duplicate + Address Detection (DAD) probes to send. Defaults to unset. + + + + IPv6HopLimit= + Configures IPv6 Hop Limit. For each router that + forwards the packet, the hop limit is decremented by 1. When the + hop limit field reaches zero, the packet is discarded. + Defaults to unset. + + + + ProxyARP= + A boolean. Configures proxy ARP. Proxy ARP is the technique in which one host, + usually a router, answers ARP requests intended for another machine. By "faking" its identity, + the router accepts responsibility for routing packets to the "real" destination. (see RFC 1027. + Defaults to unset. + + + + Bridge= + + The name of the bridge to add the link to. + + + + Bond= + + The name of the bond to add the link to. + + + + VRF= + + The name of the VRF to add the link to. + + + + VLAN= + + The name of a VLAN to create on the link. This + option may be specified more than once. + + + + MACVLAN= + + The name of a MACVLAN to create on the link. This + option may be specified more than once. + + + + VXLAN= + + The name of a VXLAN to create on the link. This + option may be specified more than once. + + + + Tunnel= + + The name of a Tunnel to create on the link. This + option may be specified more than once. + + + + + + + + [Address] Section Options + + An [Address] section accepts the + following keys. Specify several [Address] + sections to configure several addresses. + + + + Address= + + As in the [Network] section. This + key is mandatory. + + + + Peer= + + The peer address in a point-to-point connection. + Accepts the same format as the Address + key. + + + + Broadcast= + + The broadcast address, which must be in the format + described in + inet_pton3. + This key only applies to IPv4 addresses. If it is not + given, it is derived from the Address + key. + + + + Label= + + An address label. + + + + PreferredLifetime= + + Allows the default "preferred lifetime" of the address to be overridden. + Only three settings are accepted: forever or infinity + which is the default and means that the address never expires, and 0 which means + that the address is considered immediately "expired" and will not be used, + unless explicitly requested. A setting of PreferredLifetime=0 is useful for + addresses which are added to be used only by a specific application, + which is then configured to use them explicitly. + + + + + + + [Route] Section Options + The [Route] section accepts the + following keys. Specify several [Route] + sections to configure several routes. + + + + Gateway= + + As in the [Network] section. + + + + Destination= + + The destination prefix of the route. Possibly + followed by a slash and the prefix length. If omitted, a + full-length host route is assumed. + + + + Source= + + The source prefix of the route. Possibly followed by + a slash and the prefix length. If omitted, a full-length + host route is assumed. + + + + Metric= + + The metric of the route (an unsigned integer). + + + + Scope= + + The scope of the route, which can be global, + link or host. Defaults to + global. + + + + PreferredSource= + + The preferred source address of the route. The address + must be in the format described in + inet_pton3. + + + + Table=num + + The table identifier for the route (a number between 1 and 4294967295, or 0 to unset). + The table can be retrieved using ip route show table num. + + + + + + + + [DHCP] Section Options + The [DHCP] section configures the + DHCPv4 and DHCP6 client, if it is enabled with the + DHCP= setting described above: + + + + UseDNS= + + When true (the default), the DNS servers received + from the DHCP server will be used and take precedence over + any statically configured ones. + + This corresponds to the + option in resolv.conf5. + + + + UseNTP= + + When true (the default), the NTP servers received + from the DHCP server will be used by systemd-timesyncd + and take precedence over any statically configured ones. + + + + UseMTU= + + When true, the interface maximum transmission unit + from the DHCP server will be used on the current link. + Defaults to false. + + + + SendHostname= + + When true (the default), the machine's hostname will + be sent to the DHCP server. + + + + UseHostname= + + When true (the default), the hostname received from + the DHCP server will be set as the transient hostname of the system + + + + + Hostname= + + Use this value for the hostname which is sent to the + DHCP server, instead of machine's hostname. + + + + UseDomains= + + Takes a boolean argument, or the special value route. When true, the domain name + received from the DHCP server will be used as DNS search domain over this link, similar to the effect of + the setting. If set to route, the domain name received from + the DHCP server will be used for routing DNS queries only, but not for searching, similar to the effect of + the setting when the argument is prefixed with ~. Defaults to + false. + + It is recommended to enable this option only on trusted networks, as setting this affects resolution + of all host names, in particular of single-label names. It is generally safer to use the supplied domain + only as routing domain, rather than as search domain, in order to not have it affect local resolution of + single-label names. + + When set to true, this setting corresponds to the option in resolv.conf5. + + + + UseRoutes= + + When true (the default), the static routes will be + requested from the DHCP server and added to the routing + table with a metric of 1024. + + + + + UseTimezone= + + When true, the timezone received from the + DHCP server will be set as timezone of the local + system. Defaults to no. + + + + CriticalConnection= + + When true, the connection will never be torn down + even if the DHCP lease expires. This is contrary to the + DHCP specification, but may be the best choice if, say, + the root filesystem relies on this connection. Defaults to + false. + + + + + ClientIdentifier= + + The DHCPv4 client identifier to use. Either mac to use the MAC address of the link + or duid (the default, see below) to use an RFC4361-compliant Client ID. + + + + + VendorClassIdentifier= + + The vendor class identifier used to identify vendor + type and configuration. + + + + + DUIDType= + + Override the global DUIDType setting for this network. See + networkd.conf5 + for a description of possible values. + + + + + DUIDRawData= + + Override the global DUIDRawData setting for this network. See + networkd.conf5 + for a description of possible values. + + + + + IAID= + + The DHCP Identity Association Identifier (IAID) for the interface, a 32-bit unsigned integer. + + + + + RequestBroadcast= + + Request the server to use broadcast messages before + the IP address has been configured. This is necessary for + devices that cannot receive RAW packets, or that cannot + receive packets at all before an IP address has been + configured. On the other hand, this must not be enabled on + networks where broadcasts are filtered out. + + + + + RouteMetric= + + Set the routing metric for routes specified by the + DHCP server. + + + + + + + [IPv6AcceptRA] Section Options + The [IPv6AcceptRA] section configures the IPv6 Router Advertisement + (RA) client, if it is enabled with the IPv6AcceptRA= setting described + above: + + + + UseDNS= + + When true (the default), the DNS servers received in the Router Advertisement will be used and take + precedence over any statically configured ones. + + This corresponds to the option in resolv.conf5. + + + + + UseDomains= + + Takes a boolean argument, or the special value route. When true, the domain name + received via IPv6 Router Advertisement (RA) will be used as DNS search domain over this link, similar to + the effect of the setting. If set to route, the domain name + received via IPv6 RA will be used for routing DNS queries only, but not for searching, similar to the + effect of the setting when the argument is prefixed with + ~. Defaults to false. + + It is recommended to enable this option only on trusted networks, as setting this affects resolution + of all host names, in particular of single-label names. It is generally safer to use the supplied domain + only as routing domain, rather than as search domain, in order to not have it affect local resolution of + single-label names. + + When set to true, this setting corresponds to the option in resolv.conf5. + + + + + + + + [DHCPServer] Section Options + The [DHCPServer] section contains + settings for the DHCP server, if enabled via the + DHCPServer= option described above: + + + + + PoolOffset= + PoolSize= + + Configures the pool of addresses to hand out. The pool + is a contiguous sequence of IP addresses in the subnet configured for + the server address, which does not include the subnet nor the broadcast + address. PoolOffset= takes the offset of the pool + from the start of subnet, or zero to use the default value. + PoolSize= takes the number of IP addresses in the + pool or zero to use the default value. By default, the pool starts at + the first address after the subnet address and takes up the rest of + the subnet, excluding the broadcast address. If the pool includes + the server address (the default), this is reserved and not handed + out to clients. + + + + DefaultLeaseTimeSec= + MaxLeaseTimeSec= + + Control the default and maximum DHCP lease + time to pass to clients. These settings take time values in seconds or + another common time unit, depending on the suffix. The default + lease time is used for clients that did not ask for a specific + lease time. If a client asks for a lease time longer than the + maximum lease time, it is automatically shortened to the + specified time. The default lease time defaults to 1h, the + maximum lease time to 12h. Shorter lease times are beneficial + if the configuration data in DHCP leases changes frequently + and clients shall learn the new settings with shorter + latencies. Longer lease times reduce the generated DHCP + network traffic. + + + + EmitDNS= + DNS= + + Configures whether the DHCP leases handed out + to clients shall contain DNS server information. The + EmitDNS= setting takes a boolean argument + and defaults to yes. The DNS servers to + pass to clients may be configured with the + DNS= option, which takes a list of IPv4 + addresses. If the EmitDNS= option is + enabled but no servers configured, the servers are + automatically propagated from an "uplink" interface that has + appropriate servers set. The "uplink" interface is determined + by the default route of the system with the highest + priority. Note that this information is acquired at the time + the lease is handed out, and does not take uplink interfaces + into account that acquire DNS or NTP server information at a + later point. DNS server propagation does not take + /etc/resolv.conf into account. Also, note + that the leases are not refreshed if the uplink network + configuration changes. To ensure clients regularly acquire the + most current uplink DNS server information, it is thus + advisable to shorten the DHCP lease time via + MaxLeaseTimeSec= described + above. + + + + EmitNTP= + NTP= + + Similar to the EmitDNS= and + DNS= settings described above, these + settings configure whether and what NTP server information + shall be emitted as part of the DHCP lease. The same syntax, + propagation semantics and defaults apply as for + EmitDNS= and + DNS=. + + + + EmitRouter= + + Similar to the EmitDNS= + setting described above, this setting configures whether the + DHCP lease should contain the router option. The same syntax, + propagation semantics and defaults apply as for + EmitDNS=. + + + + EmitTimezone= + Timezone= + + Configures whether the DHCP leases handed out + to clients shall contain timezone information. The + EmitTimezone= setting takes a boolean + argument and defaults to yes. The + Timezone= setting takes a timezone string + (such as Europe/Berlin or + UTC) to pass to clients. If no explicit + timezone is set, the system timezone of the local host is + propagated, as determined by the + /etc/localtime symlink. + + + + + + + [Bridge] Section Options + The [Bridge] section accepts the + following keys. + + + UnicastFlood= + + A boolean. Controls whether the bridge should flood + traffic for which an FDB entry is missing and the destination + is unknown through this port. Defaults to on. + + + + + HairPin= + + A boolean. Configures whether traffic may be sent back + out of the port on which it was received. By default, this + flag is false, and the bridge will not forward traffic back + out of the receiving port. + + + + UseBPDU= + + A boolean. Configures whether STP Bridge Protocol Data Units will be + processed by the bridge port. Defaults to yes. + + + + FastLeave= + + A boolean. This flag allows the bridge to immediately stop multicast + traffic on a port that receives an IGMP Leave message. It is only used with + IGMP snooping if enabled on the bridge. Defaults to off. + + + + AllowPortToBeRoot= + + A boolean. Configures whether a given port is allowed to + become a root port. Only used when STP is enabled on the bridge. + Defaults to on. + + + + Cost= + + Sets the "cost" of sending packets of this interface. + Each port in a bridge may have a different speed and the cost + is used to decide which link to use. Faster interfaces + should have lower costs. + + + + + + [BridgeFDB] Section Options + The [BridgeFDB] section manages the + forwarding database table of a port and accepts the following + keys. Specify several [BridgeFDB] sections to + configure several static MAC table entries. + + + + MACAddress= + + As in the [Network] section. This + key is mandatory. + + + + VLANId= + + The VLAN ID for the new static MAC table entry. If + omitted, no VLAN ID info is appended to the new static MAC + table entry. + + + + + + [BridgeVLAN] Section Options + The [BridgeVLAN] section manages the VLAN ID configuration of a bridge port and accepts + the following keys. Specify several [BridgeVLAN] sections to configure several VLAN entries. + The VLANFiltering= option has to be enabled, see [Bridge] section in + systemd.netdev5. + + + + VLAN= + + The VLAN ID allowed on the port. This can be either a single ID or a range M-N. VLAN IDs are valid + from 1 to 4094. + + + + EgressUntagged= + + The VLAN ID specified here will be used to untag frames on egress. Configuring + EgressUntagged= implicates the use of VLAN= above and will enable the + VLAN ID for ingress as well. This can be either a single ID or a range M-N. + + + + PVID= + + The Port VLAN ID specified here is assigned to all untagged frames at ingress. + PVID= can be used only once. Configuring PVID= implicates the use of + VLAN= above and will enable the VLAN ID for ingress as well. + + + + + + + Example + + /etc/systemd/network/50-static.network + + [Match] +Name=enp2s0 + +[Network] +Address=192.168.0.15/24 +Gateway=192.168.0.1 + + + + /etc/systemd/network/80-dhcp.network + + [Match] +Name=en* + +[Network] +DHCP=yes + + + + /etc/systemd/network/25-bridge-static.network + + [Match] +Name=bridge0 + +[Network] +Address=192.168.0.15/24 +Gateway=192.168.0.1 +DNS=192.168.0.1 + + + + /etc/systemd/network/25-bridge-slave-interface.network + + [Match] +Name=enp2s0 + +[Network] +Bridge=bridge0 + + + /etc/systemd/network/25-bridge-slave-interface-vlan.network + + [Match] +Name=enp2s0 + +[Network] +Bridge=bridge0 + +[BridgeVLAN] +VLAN=1-32 +PVID=42 +EgressUntagged=42 + +[BridgeVLAN] +VLAN=100-200 + +[BridgeVLAN] +EgressUntagged=300-400 + + + /etc/systemd/network/25-ipip.network + + [Match] +Name=em1 + +[Network] +Tunnel=ipip-tun + + + + /etc/systemd/network/25-sit.network + + [Match] +Name=em1 + +[Network] +Tunnel=sit-tun + + + + /etc/systemd/network/25-gre.network + + [Match] +Name=em1 + +[Network] +Tunnel=gre-tun + + + + /etc/systemd/network/25-vti.network + + [Match] +Name=em1 + +[Network] +Tunnel=vti-tun + + + + /etc/systemd/network/25-bond.network + + [Match] +Name=bond1 + +[Network] +DHCP=yes + + + + + /etc/systemd/network/25-vrf.network + Add the bond1 interface to the VRF master interface vrf-test. This will redirect routes generated on this interface to be within the routing table defined during VRF creation. Traffic won't be redirected towards the VRFs routing table unless specific ip-rules are added. + [Match] +Name=bond1 + +[Network] +VRF=vrf-test + + + + + + + See Also + + systemd1, + systemd-networkd.service8, + systemd.link5, + systemd.netdev5, + systemd-resolved.service8 + + + + diff --git a/src/grp-system/systemd/systemd.nspawn.xml b/src/grp-system/systemd/systemd.nspawn.xml new file mode 100644 index 0000000000..b1344d6c10 --- /dev/null +++ b/src/grp-system/systemd/systemd.nspawn.xml @@ -0,0 +1,463 @@ + + +%entities; +]> + + + + + + + systemd.nspawn + systemd + + + + Developer + Lennart + Poettering + lennart@poettering.net + + + + + + systemd.nspawn + 5 + + + + systemd.nspawn + Container settings + + + + /etc/systemd/nspawn/machine.nspawn + /run/systemd/nspawn/machine.nspawn + /var/lib/machines/machine.nspawn + + + + Description + + An nspawn container settings file (suffix + .nspawn) encodes additional runtime + information about a local container, and is searched, read and + used by + systemd-nspawn1 + when starting a container. Files of this type are named after the + containers they define settings for. They are optional, and only + required for containers whose execution environment shall differ + from the defaults. Files of this type mostly contain settings that + may also be set on the systemd-nspawn command + line, and make it easier to persistently attach specific settings + to specific containers. The syntax of these files is inspired by + .desktop files following the XDG + Desktop Entry Specification, which in turn are inspired by + Microsoft Windows .ini files. + + Boolean arguments used in these settings files can be + written in various formats. For positive settings, the strings + , , + and are equivalent. For negative settings, the + strings , , + and are + equivalent. + + Empty lines and lines starting with # or ; are + ignored. This may be used for commenting. Lines ending + in a backslash are concatenated with the following + line while reading and the backslash is replaced by a + space character. This may be used to wrap long lines. + + + + + <filename>.nspawn</filename> File Discovery + + Files are searched by appending the + .nspawn suffix to the machine name of the + container, as specified with the + switch of systemd-nspawn, or derived from the + directory or image file name. This file is first searched in + /etc/systemd/nspawn/ and + /run/systemd/nspawn/. If found in these + directories, its settings are read and all of them take full effect + (but are possibly overridden by corresponding command line + arguments). If not found, the file will then be searched next to + the image file or in the immediate parent of the root directory of + the container. If the file is found there, only a subset of the + settings will take effect however. All settings that possibly + elevate privileges or grant additional access to resources of the + host (such as files or directories) are ignored. To which options + this applies is documented below. + + Persistent settings files created and maintained by the + administrator (and thus trusted) should be placed in + /etc/systemd/nspawn/, while automatically + downloaded (and thus potentially untrusted) settings files are + placed in /var/lib/machines/ instead (next to + the container images), where their security impact is limited. In + order to add privileged settings to .nspawn + files acquired from the image vendor, it is recommended to copy the + settings files into /etc/systemd/nspawn/ and + edit them there, so that the privileged options become + available. The precise algorithm for how the files are searched and + interpreted may be configured with + systemd-nspawn's + switch, see + systemd-nspawn1 + for details. + + + + [Exec] Section Options + + Settings files may include an [Exec] + section, which carries various execution parameters: + + + + + Boot= + + Takes a boolean argument, which defaults to off. If enabled, systemd-nspawn + will automatically search for an init executable and invoke it. In this case, the + specified parameters using Parameters= are passed as additional arguments to the + init process. This setting corresponds to the switch on the + systemd-nspawn command line. This option may not be combined with + ProcessTwo=yes. This option is the default if the + systemd-nspawn@.service template unit file is used. + + + + ProcessTwo= + + Takes a boolean argument, which defaults to off. If enabled, the specified program is run as + PID 2. A stub init process is run as PID 1. This setting corresponds to the switch + on the systemd-nspawn command line. This option may not be combined with + Boot=yes. + + + + Parameters= + + Takes a space-separated list of + arguments. This is either a command line, beginning with the + binary name to execute, or – if Boot= is + enabled – the list of arguments to pass to the init + process. This setting corresponds to the command line + parameters passed on the systemd-nspawn + command line. + + + + Environment= + + Takes an environment variable assignment + consisting of key and value, separated by + =. Sets an environment variable for the + main process invoked in the container. This setting may be + used multiple times to set multiple environment variables. It + corresponds to the command line + switch. + + + + User= + + Takes a UNIX user name. Specifies the user + name to invoke the main process of the container as. This user + must be known in the container's user database. This + corresponds to the command line + switch. + + + + WorkingDirectory= + + Selects the working directory for the process invoked in the container. Expects an absolute + path in the container's file system namespace. This corresponds to the command line + switch. + + + + Capability= + DropCapability= + + Takes a space-separated list of Linux process + capabilities (see + capabilities7 + for details). The Capability= setting + specifies additional capabilities to pass on top of the + default set of capabilities. The + DropCapability= setting specifies + capabilities to drop from the default set. These settings + correspond to the and + command line + switches. Note that Capability= is a + privileged setting, and only takes effect in + .nspawn files in + /etc/systemd/nspawn/ and + /run/system/nspawn/ (see above). On the + other hand, DropCapability= takes effect in + all cases. + + + + KillSignal= + + Specify the process signal to send to the + container's PID 1 when nspawn itself receives SIGTERM, in + order to trigger an orderly shutdown of the container. + Defaults to SIGRTMIN+3 if is used + (on systemd-compatible init systems SIGRTMIN+3 triggers an + orderly shutdown). For a list of valid signals, see + signal7. + + + + Personality= + + Configures the kernel personality for the + container. This is equivalent to the + switch. + + + + MachineID= + + Configures the 128-bit machine ID (UUID) to pass to + the container. This is equivalent to the + command line switch. This option is + privileged (see above). + + + + PrivateUsers= + + Configures support for usernamespacing. This is equivalent to the + command line switch, and takes the same options. This option is privileged + (see above). This option is the default if the systemd-nspawn@.service template unit file + is used. + + + + NotifyReady= + + Configures support for notifications from the container's init process. + This is equivalent to use command line switch, + and takes the same options. See systemd-nspawn1 + for details about the specific options supported. + + + + + + [Files] Section Options + + Settings files may include a [Files] + section, which carries various parameters configuring the file + system of the container: + + + + + ReadOnly= + + Takes a boolean argument, which defaults to off. If + specified, the container will be run with a read-only file + system. This setting corresponds to the + command line + switch. + + + + Volatile= + + Takes a boolean argument, or the special value + state. This configures whether to run the + container with volatile state and/or configuration. This + option is equivalent to , see + systemd-nspawn1 + for details about the specific options + supported. + + + + Bind= + BindReadOnly= + + Adds a bind mount from the host into the + container. Takes a single path, a pair of two paths separated + by a colon, or a triplet of two paths plus an option string + separated by colons. This option may be used multiple times to + configure multiple bind mounts. This option is equivalent to + the command line switches and + , see + systemd-nspawn1 + for details about the specific options supported. This setting + is privileged (see above). + + + + TemporaryFileSystem= + + Adds a tmpfs mount to the + container. Takes a path or a pair of path and option string, + separated by a colon. This option may be used multiple times to + configure multiple tmpfs mounts. This + option is equivalent to the command line switch + , see + systemd-nspawn1 + for details about the specific options supported. This setting + is privileged (see above). + + + + PrivateUsersChown= + + Configures whether the ownership of the files and directories in the container tree shall be + adjusted to the UID/GID range used, if necessary and user namespacing is enabled. This is equivalent to the + command line switch. This option is privileged (see + above). + + + + + + + [Network] Section Options + + Settings files may include a [Network] + section, which carries various parameters configuring the network + connectivity of the container: + + + + + Private= + + Takes a boolean argument, which defaults to off. If + enabled, the container will run in its own network namespace + and not share network interfaces and configuration with the + host. This setting corresponds to the + command line + switch. + + + + VirtualEthernet= + + Takes a boolean argument. Configures whether to create a virtual Ethernet connection + (veth) between host and the container. This setting implies + Private=yes. This setting corresponds to the command line + switch. This option is privileged (see above). This option is the default if the + systemd-nspawn@.service template unit file is used. + + + + VirtualEthernetExtra= + + Takes a colon-separated pair of interface + names. Configures an additional virtual Ethernet connection + (veth) between host and the container. The + first specified name is the interface name on the host, the + second the interface name in the container. The latter may be + omitted in which case it is set to the same name as the host + side interface. This setting implies + Private=yes. This setting corresponds to + the command line + switch, and maybe be used multiple times. It is independent of + VirtualEthernet=. This option is privileged + (see above). + + + + Interface= + + Takes a space-separated list of interfaces to + add to the container. This option corresponds to the + command line switch and + implies Private=yes. This option is + privileged (see above). + + + + MACVLAN= + IPVLAN= + + Takes a space-separated list of interfaces to + add MACLVAN or IPVLAN interfaces to, which are then added to + the container. These options correspond to the + and + command line switches and + imply Private=yes. These options are + privileged (see above). + + + + Bridge= + + Takes an interface name. This setting implies + VirtualEthernet=yes and + Private=yes and has the effect that the + host side of the created virtual Ethernet link is connected to + the specified bridge interface. This option corresponds to the + command line switch. This + option is privileged (see above). + + + + Zone= + + Takes a network zone name. This setting implies VirtualEthernet=yes and + Private=yes and has the effect that the host side of the created virtual Ethernet link is + connected to an automatically managed bridge interface named after the passed argument, prefixed with + vz-. This option corresponds to the command line + switch. This option is privileged (see above). + + + + Port= + + Exposes a TCP or UDP port of the container on + the host. This option corresponds to the + command line switch, see + systemd-nspawn1 + for the precise syntax of the argument this option takes. This + option is privileged (see above). + + + + + + See Also + + systemd1, + systemd-nspawn1, + systemd.directives7 + + + + diff --git a/src/grp-system/systemd/systemd.offline-updates.xml b/src/grp-system/systemd/systemd.offline-updates.xml new file mode 100644 index 0000000000..ae53b8552d --- /dev/null +++ b/src/grp-system/systemd/systemd.offline-updates.xml @@ -0,0 +1,169 @@ + + + + + + + + systemd.offline-updates + systemd + + + + Developer + Lennart + Poettering + lennart@poettering.net + + + + + + systemd.offline-updates + 7 + + + + systemd.offline-updates + Implementation of offline updates in systemd + + + + Implementing Offline System Updates + + This man page describes how to implement "offline" system updates with systemd. By "offline" + OS updates we mean package installations and updates that are run with the system booted into a + special system update mode, in order to avoid problems related to conflicts of libraries and + services that are currently running with those on disk. This document is inspired by this + GNOME design whiteboard. + + + The logic: + + + + The package manager prepares system updates by downloading all (RPM or DEB or + whatever) packages to update off-line in a special directory + /var/lib/system-update (or + another directory of the package/upgrade manager's choice). + + + + When the user OK'ed the update, the symlink /system-update is + created that points to /var/lib/system-update (or + wherever the directory with the upgrade files is located) and the system is rebooted. This + symlink is in the root directory, since we need to check for it very early at boot, at a + time where /var is not available yet. + + + + Very early in the new boot + systemd-update-generator8 + checks whether /system-update exists. If so, it (temporarily and for + this boot only) redirects (i.e. symlinks) default.target to + system-update.target, a special target that is pulls in the base system + (i.e. sysinit.target, so that all file systems are mounted but little + else) and the system update units. + + + + The system now continues to boot into default.target, and thus + into system-update.target. This target pulls in the system update unit, + which starts the system update script after all file systems have been mounted. + + + + As the first step, the update script should check if the + /system-update symlink points to the location used by that update + script. In case it does not exists or points to a different location, the script must exit + without error. It is possible for multiple update services to be installed, and for multiple + update scripts to be launched in parallel, and only the one that corresponds to the tool + that created the symlink before reboot should perform any actions. It + is unsafe to run multiple updates in parallel. + + + + The update script should now do its job. If applicable and possible, it should + create a file system snapshot, then install all packages. + After completion (regardless whether the update succeeded or failed) the machine + must be rebooted, for example by calling systemctl reboot. + In addition, on failure the script should revert to the old file system snapshot + (without the symlink). + + + + The system is rebooted. Since the /system-update symlink is gone, + the generator won't redirect default.target after reboot and the + system now boots into the default target again. + + + + + + Recommendations + + + + To make things a bit more robust we recommend hooking the update script into + system-update.target via a .wants/ + symlink in the distribution package, rather than depending on systemctl + enable in the postinst scriptlets of your package. More specifically, for your + update script create a .service file, without [Install] section, and then add a symlink like + /usr/lib/systemd/system-update.target.wants/foobar.service + → ../foobar.service to your package. + + + + Make sure to remove the /system-update symlink as early as + possible in the update script to avoid reboot loops in case the update fails. + + + + Use FailureAction=reboot in the service file for your update script + to ensure that a reboot is automatically triggered if the update fails. + FailureAction= makes sure that the specified unit is activated if your + script exits uncleanly (by non-zero error code, or signal/coredump). If your script succeeds + you should trigger the reboot in your own code, for example by invoking logind's + Reboot() call or calling systemct reboot. See + logind dbus API + for details. + + + + The update service should declare DefaultDependencies=false, + and pull in any services it requires explicitly. + + + + + + See also + + + Implementing Offline System Updates, + systemd1, + systemd.generator7, + systemd-update-generator8, + dnf.plugin.system-upgrade8 + + + diff --git a/src/grp-system/systemd/systemd.path.xml b/src/grp-system/systemd/systemd.path.xml new file mode 100644 index 0000000000..7200c8fe27 --- /dev/null +++ b/src/grp-system/systemd/systemd.path.xml @@ -0,0 +1,202 @@ + + + + + + + + systemd.path + systemd + + + + Developer + Lennart + Poettering + lennart@poettering.net + + + + + + systemd.path + 5 + + + + systemd.path + Path unit configuration + + + + path.path + + + + Description + + A unit configuration file whose name ends in + .path encodes information about a path + monitored by systemd, for path-based activation. + + This man page lists the configuration options specific to + this unit type. See + systemd.unit5 + for the common options of all unit configuration files. The common + configuration items are configured in the generic [Unit] and + [Install] sections. The path specific configuration options are + configured in the [Path] section. + + For each path file, a matching unit file must exist, + describing the unit to activate when the path changes. By default, + a service by the same name as the path (except for the suffix) is + activated. Example: a path file foo.path + activates a matching service foo.service. The + unit to activate may be controlled by Unit= + (see below). + + Internally, path units use the + inotify7 + API to monitor file systems. Due to that, it suffers by the same + limitations as inotify, and for example cannot be used to monitor + files or directories changed by other machines on remote NFS file + systems. + + + + Automatic Dependencies + + If a path unit is beneath another mount unit in the file + system hierarchy, both a requirement and an ordering dependency + between both units are created automatically. + + An implicit Before= dependency is added + between a path unit and the unit it is supposed to activate. + + Unless DefaultDependencies=false in the [Unit] section is used, path + units will implicitly have dependencies of type Before= on paths.target, + dependencies of type After= and Requires= on + sysinit.target, and have dependencies of type Conflicts= and + Before= on shutdown.target. These ensure that path units are terminated + cleanly prior to system shutdown. Only path units involved with early boot or late system shutdown should disable + this option. + + + + + Options + + Path files must include a [Path] section, which carries + information about the path(s) it monitors. The options specific to + the [Path] section of path units are the following: + + + + PathExists= + PathExistsGlob= + PathChanged= + PathModified= + DirectoryNotEmpty= + + Defines paths to monitor for certain changes: + PathExists= may be used to watch the mere + existence of a file or directory. If the file specified + exists, the configured unit is activated. + PathExistsGlob= works similar, but checks + for the existence of at least one file matching the globbing + pattern specified. PathChanged= may be used + to watch a file or directory and activate the configured unit + whenever it changes. It is not activated on every write to the + watched file but it is activated if the file which was open + for writing gets closed. PathModified= is + similar, but additionally it is activated also on simple + writes to the watched file. + DirectoryNotEmpty= may be used to watch a + directory and activate the configured unit whenever it + contains at least one file. + + The arguments of these directives must be absolute file + system paths. + + Multiple directives may be combined, of the same and of + different types, to watch multiple paths. If the empty string + is assigned to any of these options, the list of paths to + watch is reset, and any prior assignments of these options + will not have any effect. + + If a path already exists (in case of + PathExists= and + PathExistsGlob=) or a directory already is + not empty (in case of DirectoryNotEmpty=) + at the time the path unit is activated, then the configured + unit is immediately activated as well. Something similar does + not apply to PathChanged= and + PathModified=. + + If the path itself or any of the containing directories + are not accessible, systemd will watch for + permission changes and notice that conditions are satisfied + when permissions allow that. + + + Unit= + + The unit to activate when any of the + configured paths changes. The argument is a unit name, whose + suffix is not .path. If not specified, this + value defaults to a service that has the same name as the path + unit, except for the suffix. (See above.) It is recommended + that the unit name that is activated and the unit name of the + path unit are named identical, except for the + suffix. + + + MakeDirectory= + + Takes a boolean argument. If true, the + directories to watch are created before watching. This option + is ignored for PathExists= settings. + Defaults to . + + + DirectoryMode= + + If MakeDirectory= is + enabled, use the mode specified here to create the directories + in question. Takes an access mode in octal notation. Defaults + to . + + + + + + See Also + + systemd1, + systemctl1, + systemd.unit5, + systemd.service5, + inotify7, + systemd.directives7 + + + + diff --git a/src/grp-system/systemd/systemd.pc.in b/src/grp-system/systemd/systemd.pc.in new file mode 100644 index 0000000000..ac52b30dd3 --- /dev/null +++ b/src/grp-system/systemd/systemd.pc.in @@ -0,0 +1,34 @@ +# This file is part of systemd. +# +# systemd is free software; you can redistribute it and/or modify it +# under the terms of the GNU Lesser General Public License as published by +# the Free Software Foundation; either version 2.1 of the License, or +# (at your option) any later version. + +prefix=@prefix@ +systemdutildir=@rootlibexecdir@ +systemdsystemunitdir=@systemunitdir@ +systemdsystempresetdir=@systempresetdir@ +systemduserunitdir=@userunitdir@ +systemduserpresetdir=@userpresetdir@ +systemdsystemconfdir=@pkgsysconfdir@/system +systemduserconfdir=@pkgsysconfdir@/user +systemdsystemunitpath=${systemdsystemconfdir}:/etc/systemd/system:/run/systemd/system:/usr/local/lib/systemd/system:${systemdsystemunitdir}:/usr/lib/systemd/system:/lib/systemd/system +systemduserunitpath=${systemduserconfdir}:/etc/systemd/user:/run/systemd/user:/usr/local/lib/systemd/user:/usr/local/share/systemd/user:${systemduserunitdir}:/usr/lib/systemd/user:/usr/share/systemd/user +systemdsystemgeneratordir=@systemgeneratordir@ +systemdusergeneratordir=@usergeneratordir@ +systemdsleepdir=@systemsleepdir@ +systemdshutdowndir=@systemshutdowndir@ +tmpfilesdir=@tmpfilesdir@ +sysusersdir=@sysusersdir@ +sysctldir=@sysctldir@ +binfmtdir=@binfmtdir@ +modulesloaddir=@modulesloaddir@ +catalogdir=@catalogdir@ +systemuidmax=@systemuidmax@ +systemgidmax=@systemgidmax@ + +Name: systemd +Description: systemd System and Service Manager +URL: @PACKAGE_URL@ +Version: @PACKAGE_VERSION@ diff --git a/src/grp-system/systemd/systemd.resource-control.xml b/src/grp-system/systemd/systemd.resource-control.xml new file mode 100644 index 0000000000..bf44a68345 --- /dev/null +++ b/src/grp-system/systemd/systemd.resource-control.xml @@ -0,0 +1,696 @@ + + + + + + + + systemd.resource-control + systemd + + + + Developer + Lennart + Poettering + lennart@poettering.net + + + + + + systemd.resource-control + 5 + + + + systemd.resource-control + Resource control unit settings + + + + + slice.slice, + scope.scope, + service.service, + socket.socket, + mount.mount, + swap.swap + + + + + Description + + Unit configuration files for services, slices, scopes, + sockets, mount points, and swap devices share a subset of + configuration options for resource control of spawned + processes. Internally, this relies on the Control Groups + kernel concept for organizing processes in a hierarchical tree of + named groups for the purpose of resource management. + + This man page lists the configuration options shared by + those six unit types. See + systemd.unit5 + for the common options of all unit configuration files, and + systemd.slice5, + systemd.scope5, + systemd.service5, + systemd.socket5, + systemd.mount5, + and + systemd.swap5 + for more information on the specific unit configuration files. The + resource control configuration options are configured in the + [Slice], [Scope], [Service], [Socket], [Mount], or [Swap] + sections, depending on the unit type. + + See the New + Control Group Interfaces for an introduction on how to make + use of resource control APIs from programs. + + + + Automatic Dependencies + + Units with the Slice= setting set automatically acquire Requires= and + After= dependencies on the specified slice unit. + + + + Unified and Legacy Control Group Hierarchies + + The unified control group hierarchy is the new version of kernel control group interface. Depending on the + resource type, there are differences in resource control capabilities. Also, because of interface changes, some + resource types have a separate set of options on the unified hierarchy. + + + + + + + IO prefixed settings are superset of and replace BlockIO + prefixed ones. On unified hierarchy, IO resource control also applies to buffered writes. + + + + + + MemoryMax= replaces MemoryLimit=. MemoryLow= + and MemoryHigh= are effective only on unified hierarchy. + + + + + + To ease the transition, there is best-effort translation between the two versions of settings. If all + settings of a unit for a given resource type are for the other hierarchy type, the settings are translated and + applied. If there are any valid settings for the hierarchy in use, all translations are disabled for the resource + type. Mixing the two types of settings on a unit can lead to confusing results. + + + + Options + + Units of the types listed above can have settings + for resource control configuration: + + + + + CPUAccounting= + + + Turn on CPU usage accounting for this unit. Takes a + boolean argument. Note that turning on CPU accounting for + one unit will also implicitly turn it on for all units + contained in the same slice and for all its parent slices + and the units contained therein. The system default for this + setting may be controlled with + DefaultCPUAccounting= in + systemd-system.conf5. + + + + + CPUShares=weight + StartupCPUShares=weight + + + Assign the specified CPU time share weight to the + processes executed. These options take an integer value and + control the cpu.shares control group + attribute. The allowed range is 2 to 262144. Defaults to + 1024. For details about this control group attribute, see + sched-design-CFS.txt. + The available CPU time is split up among all units within + one slice relative to their CPU time share weight. + + While StartupCPUShares= only + applies to the startup phase of the system, + CPUShares= applies to normal runtime of + the system, and if the former is not set also to the startup + phase. Using StartupCPUShares= allows + prioritizing specific services at boot-up differently than + during normal runtime. + + These options imply + CPUAccounting=true. + + + + + CPUQuota= + + + Assign the specified CPU time quota to the processes + executed. Takes a percentage value, suffixed with "%". The + percentage specifies how much CPU time the unit shall get at + maximum, relative to the total CPU time available on one + CPU. Use values > 100% for allotting CPU time on more than + one CPU. This controls the + cpu.cfs_quota_us control group + attribute. For details about this control group attribute, + see sched-design-CFS.txt. + + Example: CPUQuota=20% ensures that + the executed processes will never get more than 20% CPU time + on one CPU. + + Implies CPUAccounting=true. + + + + + MemoryAccounting= + + + Turn on process and kernel memory accounting for this + unit. Takes a boolean argument. Note that turning on memory + accounting for one unit will also implicitly turn it on for + all units contained in the same slice and for all its parent + slices and the units contained therein. The system default + for this setting may be controlled with + DefaultMemoryAccounting= in + systemd-system.conf5. + + + + + MemoryLow=bytes + + + Specify the best-effort memory usage protection of the executed processes in this unit. If the memory + usages of this unit and all its ancestors are below their low boundaries, this unit's memory won't be + reclaimed as long as memory can be reclaimed from unprotected units. + + Takes a memory size in bytes. If the value is suffixed with K, M, G or T, the specified memory size is + parsed as Kilobytes, Megabytes, Gigabytes, or Terabytes (with the base 1024), respectively. Alternatively, a + percentage value may be specified, which is taken relative to the installed physical memory on the + system. This controls the memory.low control group attribute. For details about this + control group attribute, see cgroup-v2.txt. + + Implies MemoryAccounting=true. + + This setting is supported only if the unified control group hierarchy is used. + + + + + MemoryHigh=bytes + + + Specify the high limit on memory usage of the executed processes in this unit. Memory usage may go + above the limit if unavoidable, but the processes are heavily slowed down and memory is taken away + aggressively in such cases. This is the main mechanism to control memory usage of a unit. + + Takes a memory size in bytes. If the value is suffixed with K, M, G or T, the specified memory size is + parsed as Kilobytes, Megabytes, Gigabytes, or Terabytes (with the base 1024), respectively. Alternatively, a + percentage value may be specified, which is taken relative to the installed physical memory on the + system. If assigned the + special value infinity, no memory limit is applied. This controls the + memory.high control group attribute. For details about this control group attribute, see + cgroup-v2.txt. + + Implies MemoryAccounting=true. + + This setting is supported only if the unified control group hierarchy is used. + + + + + MemoryMax=bytes + + + Specify the absolute limit on memory usage of the executed processes in this unit. If memory usage + cannot be contained under the limit, out-of-memory killer is invoked inside the unit. It is recommended to + use MemoryHigh= as the main control mechanism and use MemoryMax= as the + last line of defense. + + Takes a memory size in bytes. If the value is suffixed with K, M, G or T, the specified memory size is + parsed as Kilobytes, Megabytes, Gigabytes, or Terabytes (with the base 1024), respectively. Alternatively, a + percentage value may be specified, which is taken relative to the installed physical memory on the system. If + assigned the special value infinity, no memory limit is applied. This controls the + memory.max control group attribute. For details about this control group attribute, see + cgroup-v2.txt. + + Implies MemoryAccounting=true. + + This setting is supported only if the unified control group hierarchy is used. Use + MemoryLimit= on systems using the legacy control group hierarchy. + + + + + MemoryLimit=bytes + + + Specify the limit on maximum memory usage of the executed processes. The limit specifies how much + process and kernel memory can be used by tasks in this unit. Takes a memory size in bytes. If the value is + suffixed with K, M, G or T, the specified memory size is parsed as Kilobytes, Megabytes, Gigabytes, or + Terabytes (with the base 1024), respectively. Alternatively, a percentage value may be specified, which is + taken relative to the installed physical memory on the system. If assigned the special value + infinity, no memory limit is applied. This controls the + memory.limit_in_bytes control group attribute. For details about this control group + attribute, see memory.txt. + + Implies MemoryAccounting=true. + + This setting is supported only if the legacy control group hierarchy is used. Use + MemoryMax= on systems using the unified control group hierarchy. + + + + + TasksAccounting= + + + Turn on task accounting for this unit. Takes a + boolean argument. If enabled, the system manager will keep + track of the number of tasks in the unit. The number of + tasks accounted this way includes both kernel threads and + userspace processes, with each thread counting + individually. Note that turning on tasks accounting for one + unit will also implicitly turn it on for all units contained + in the same slice and for all its parent slices and the + units contained therein. The system default for this setting + may be controlled with + DefaultTasksAccounting= in + systemd-system.conf5. + + + + + TasksMax=N + + + Specify the maximum number of tasks that may be created in the unit. This ensures that the number of + tasks accounted for the unit (see above) stays below a specific limit. This either takes an absolute number + of tasks or a percentage value that is taken relative to the configured maximum number of tasks on the + system. If assigned the special value infinity, no tasks limit is applied. This controls + the pids.max control group attribute. For details about this control group attribute, see + pids.txt. + + Implies TasksAccounting=true. The + system default for this setting may be controlled with + DefaultTasksMax= in + systemd-system.conf5. + + + + + IOAccounting= + + + Turn on Block I/O accounting for this unit, if the unified control group hierarchy is used on the + system. Takes a boolean argument. Note that turning on block I/O accounting for one unit will also implicitly + turn it on for all units contained in the same slice and all for its parent slices and the units contained + therein. The system default for this setting may be controlled with DefaultIOAccounting= + in + systemd-system.conf5. + + This setting is supported only if the unified control group hierarchy is used. Use + BlockIOAccounting= on systems using the legacy control group hierarchy. + + + + + IOWeight=weight + StartupIOWeight=weight + + + Set the default overall block I/O weight for the executed processes, if the unified control group + hierarchy is used on the system. Takes a single weight value (between 1 and 10000) to set the default block + I/O weight. This controls the io.weight control group attribute, which defaults to + 100. For details about this control group attribute, see cgroup-v2.txt. The available I/O + bandwidth is split up among all units within one slice relative to their block I/O weight. + + While StartupIOWeight= only applies + to the startup phase of the system, + IOWeight= applies to the later runtime of + the system, and if the former is not set also to the startup + phase. This allows prioritizing specific services at boot-up + differently than during runtime. + + Implies IOAccounting=true. + + This setting is supported only if the unified control group hierarchy is used. Use + BlockIOWeight= and StartupBlockIOWeight= on systems using the legacy + control group hierarchy. + + + + + IODeviceWeight=device weight + + + Set the per-device overall block I/O weight for the executed processes, if the unified control group + hierarchy is used on the system. Takes a space-separated pair of a file path and a weight value to specify + the device specific weight value, between 1 and 10000. (Example: "/dev/sda 1000"). The file path may be + specified as path to a block device node or as any other file, in which case the backing block device of the + file system of the file is determined. This controls the io.weight control group + attribute, which defaults to 100. Use this option multiple times to set weights for multiple devices. For + details about this control group attribute, see cgroup-v2.txt. + + Implies IOAccounting=true. + + This setting is supported only if the unified control group hierarchy is used. Use + BlockIODeviceWeight= on systems using the legacy control group hierarchy. + + + + + IOReadBandwidthMax=device bytes + IOWriteBandwidthMax=device bytes + + + Set the per-device overall block I/O bandwidth maximum limit for the executed processes, if the unified + control group hierarchy is used on the system. This limit is not work-conserving and the executed processes + are not allowed to use more even if the device has idle capacity. Takes a space-separated pair of a file + path and a bandwidth value (in bytes per second) to specify the device specific bandwidth. The file path may + be a path to a block device node, or as any other file in which case the backing block device of the file + system of the file is used. If the bandwidth is suffixed with K, M, G, or T, the specified bandwidth is + parsed as Kilobytes, Megabytes, Gigabytes, or Terabytes, respectively, to the base of 1000. (Example: + "/dev/disk/by-path/pci-0000:00:1f.2-scsi-0:0:0:0 5M"). This controls the io.max control + group attributes. Use this option multiple times to set bandwidth limits for multiple devices. For details + about this control group attribute, see cgroup-v2.txt. + + + Implies IOAccounting=true. + + This setting is supported only if the unified control group hierarchy is used. Use + BlockIOAccounting= on systems using the legacy control group hierarchy. + + + + + IOReadIOPSMax=device IOPS + IOWriteIOPSMax=device IOPS + + + Set the per-device overall block I/O IOs-Per-Second maximum limit for the executed processes, if the + unified control group hierarchy is used on the system. This limit is not work-conserving and the executed + processes are not allowed to use more even if the device has idle capacity. Takes a space-separated pair of + a file path and an IOPS value to specify the device specific IOPS. The file path may be a path to a block + device node, or as any other file in which case the backing block device of the file system of the file is + used. If the IOPS is suffixed with K, M, G, or T, the specified IOPS is parsed as KiloIOPS, MegaIOPS, + GigaIOPS, or TeraIOPS, respectively, to the base of 1000. (Example: + "/dev/disk/by-path/pci-0000:00:1f.2-scsi-0:0:0:0 1K"). This controls the io.max control + group attributes. Use this option multiple times to set IOPS limits for multiple devices. For details about + this control group attribute, see cgroup-v2.txt. + + + Implies IOAccounting=true. + + This setting is supported only if the unified control group hierarchy is used. + + + + + BlockIOAccounting= + + + Turn on Block I/O accounting for this unit, if the legacy control group hierarchy is used on the + system. Takes a boolean argument. Note that turning on block I/O accounting for one unit will also implicitly + turn it on for all units contained in the same slice and all for its parent slices and the units contained + therein. The system default for this setting may be controlled with + DefaultBlockIOAccounting= in + systemd-system.conf5. + + This setting is supported only if the legacy control group hierarchy is used. Use + IOAccounting= on systems using the unified control group hierarchy. + + + + + BlockIOWeight=weight + StartupBlockIOWeight=weight + + Set the default overall block I/O weight for the executed processes, if the legacy control + group hierarchy is used on the system. Takes a single weight value (between 10 and 1000) to set the default + block I/O weight. This controls the blkio.weight control group attribute, which defaults to + 500. For details about this control group attribute, see blkio-controller.txt. + The available I/O bandwidth is split up among all units within one slice relative to their block I/O + weight. + + While StartupBlockIOWeight= only + applies to the startup phase of the system, + BlockIOWeight= applies to the later runtime + of the system, and if the former is not set also to the + startup phase. This allows prioritizing specific services at + boot-up differently than during runtime. + + Implies + BlockIOAccounting=true. + + This setting is supported only if the legacy control group hierarchy is used. Use + IOWeight= and StartupIOWeight= on systems using the unified control group + hierarchy. + + + + + + BlockIODeviceWeight=device weight + + + Set the per-device overall block I/O weight for the executed processes, if the legacy control group + hierarchy is used on the system. Takes a space-separated pair of a file path and a weight value to specify + the device specific weight value, between 10 and 1000. (Example: "/dev/sda 500"). The file path may be + specified as path to a block device node or as any other file, in which case the backing block device of the + file system of the file is determined. This controls the blkio.weight_device control group + attribute, which defaults to 1000. Use this option multiple times to set weights for multiple devices. For + details about this control group attribute, see blkio-controller.txt. + + Implies + BlockIOAccounting=true. + + This setting is supported only if the legacy control group hierarchy is used. Use + IODeviceWeight= on systems using the unified control group hierarchy. + + + + + BlockIOReadBandwidth=device bytes + BlockIOWriteBandwidth=device bytes + + + Set the per-device overall block I/O bandwidth limit for the executed processes, if the legacy control + group hierarchy is used on the system. Takes a space-separated pair of a file path and a bandwidth value (in + bytes per second) to specify the device specific bandwidth. The file path may be a path to a block device + node, or as any other file in which case the backing block device of the file system of the file is used. If + the bandwidth is suffixed with K, M, G, or T, the specified bandwidth is parsed as Kilobytes, Megabytes, + Gigabytes, or Terabytes, respectively, to the base of 1000. (Example: + "/dev/disk/by-path/pci-0000:00:1f.2-scsi-0:0:0:0 5M"). This controls the + blkio.throttle.read_bps_device and blkio.throttle.write_bps_device + control group attributes. Use this option multiple times to set bandwidth limits for multiple devices. For + details about these control group attributes, see blkio-controller.txt. + + + Implies + BlockIOAccounting=true. + + This setting is supported only if the legacy control group hierarchy is used. Use + IOReadBandwidthMax= and IOWriteBandwidthMax= on systems using the + unified control group hierarchy. + + + + + DeviceAllow= + + + Control access to specific device nodes by the + executed processes. Takes two space-separated strings: a + device node specifier followed by a combination of + r, w, + m to control + reading, writing, + or creation of the specific device node(s) by the unit + (mknod), respectively. This controls + the devices.allow and + devices.deny control group + attributes. For details about these control group + attributes, see devices.txt. + + The device node specifier is either a path to a device + node in the file system, starting with + /dev/, or a string starting with either + char- or block- + followed by a device group name, as listed in + /proc/devices. The latter is useful to + whitelist all current and future devices belonging to a + specific device group at once. The device group is matched + according to file name globbing rules, you may hence use the + * and ? + wildcards. Examples: /dev/sda5 is a + path to a device node, referring to an ATA or SCSI block + device. char-pts and + char-alsa are specifiers for all pseudo + TTYs and all ALSA sound devices, + respectively. char-cpu/* is a specifier + matching all CPU related device groups. + + + + + DevicePolicy=auto|closed|strict + + + + Control the policy for allowing device access: + + + + + + means to only allow types of access that are + explicitly specified. + + + + + + + in addition, allows access to standard pseudo + devices including + /dev/null, + /dev/zero, + /dev/full, + /dev/random, and + /dev/urandom. + + + + + + + + + in addition, allows access to all devices if no + explicit DeviceAllow= is present. + This is the default. + + + + + + + + + Slice= + + + The name of the slice unit to place the unit + in. Defaults to system.slice for all + non-instantiated units of all unit types (except for slice + units themselves see below). Instance units are by default + placed in a subslice of system.slice + that is named after the template name. + + This option may be used to arrange systemd units in a + hierarchy of slices each of which might have resource + settings applied. + + For units of type slice, the only accepted value for + this setting is the parent slice. Since the name of a slice + unit implies the parent slice, it is hence redundant to ever + set this parameter directly for slice units. + + Special care should be taken when relying on the default slice assignment in templated service units + that have DefaultDependencies=no set, see + systemd.service5, section + "Automatic Dependencies" for details. + + + + + + Delegate= + + + Turns on delegation of further resource control + partitioning to processes of the unit. For unprivileged + services (i.e. those using the User= + setting), this allows processes to create a subhierarchy + beneath its control group path. For privileged services and + scopes, this ensures the processes will have all control + group controllers enabled. + + + + + + + + See Also + + systemd1, + systemd.unit5, + systemd.service5, + systemd.slice5, + systemd.scope5, + systemd.socket5, + systemd.mount5, + systemd.swap5, + systemd.directives7, + systemd.special7, + The documentation for control groups and specific controllers in the Linux kernel: + cgroups.txt, + cpuacct.txt, + memory.txt, + blkio-controller.txt. + + + diff --git a/src/grp-system/systemd/systemd.scope.xml b/src/grp-system/systemd/systemd.scope.xml new file mode 100644 index 0000000000..f69b2ef635 --- /dev/null +++ b/src/grp-system/systemd/systemd.scope.xml @@ -0,0 +1,108 @@ + + + + + + + + systemd.scope + systemd + + + + Developer + Lennart + Poettering + lennart@poettering.net + + + + + + systemd.scope + 5 + + + + systemd.scope + Scope unit configuration + + + + scope.scope + + + + Description + + Scope units are not configured via unit configuration files, + but are only created programmatically using the bus interfaces of + systemd. They are named similar to filenames. A unit whose name + ends in .scope refers to a scope unit. Scopes + units manage a set of system processes. Unlike service units, scope + units manage externally created processes, and do not fork off + processes on its own. + + The main purpose of scope units is grouping worker processes + of a system service for organization and for managing resources. + + systemd-run may + be used to easily launch a command in a new scope unit from the + command line. + + See the New + Control Group Interfaces for an introduction on how to make + use of scope units from programs. + + + + Automatic Dependencies + + Unless DefaultDependencies=false + is used, scope units will implicitly have dependencies of + type Conflicts= and + Before= on + shutdown.target. These ensure + that scope units are removed prior to system + shutdown. Only scope units involved with early boot or + late system shutdown should disable this option. + + + Additional implicit dependencies may be added as result of + resource control parameters as documented in + systemd.resource-control5. + + + + + See Also + + systemd1, + systemd-run1, + systemd.unit5, + systemd.resource-control5, + systemd.service5, + systemd.directives7. + + + + diff --git a/src/grp-system/systemd/systemd.service.xml b/src/grp-system/systemd/systemd.service.xml new file mode 100644 index 0000000000..875d368fcf --- /dev/null +++ b/src/grp-system/systemd/systemd.service.xml @@ -0,0 +1,1350 @@ + + + + + + + + systemd.service + systemd + + + + Developer + Lennart + Poettering + lennart@poettering.net + + + + + + systemd.service + 5 + + + + systemd.service + Service unit configuration + + + + service.service + + + + Description + + A unit configuration file whose name ends in + .service encodes information about a process + controlled and supervised by systemd. + + This man page lists the configuration options specific to + this unit type. See + systemd.unit5 + for the common options of all unit configuration files. The common + configuration items are configured in the generic + [Unit] and [Install] + sections. The service specific configuration options are + configured in the [Service] section. + + Additional options are listed in + systemd.exec5, + which define the execution environment the commands are executed + in, and in + systemd.kill5, + which define the way the processes of the service are terminated, + and in + systemd.resource-control5, + which configure resource control settings for the processes of the + service. + + If a service is requested under a certain name but no unit + configuration file is found, systemd looks for a SysV init script + by the same name (with the .service suffix + removed) and dynamically creates a service unit from that script. + This is useful for compatibility with SysV. Note that this + compatibility is quite comprehensive but not 100%. For details + about the incompatibilities, see the Incompatibilities + with SysV document. + + + + Automatic Dependencies + + Services with Type=dbus set automatically + acquire dependencies of type Requires= and + After= on + dbus.socket. + + Socket activated service are automatically ordered after + their activated .socket units via an + automatic After= dependency. + + Unless DefaultDependencies= in the [Unit] is set to + , service units will implicitly have dependencies of type Requires= and + After= on sysinit.target, a dependency of type After= on + basic.target as well as dependencies of type Conflicts= and + Before= on shutdown.target. These ensure that normal service units pull in + basic system initialization, and are terminated cleanly prior to system shutdown. Only services involved with early + boot or late system shutdown should disable this option. + + Instanced service units (i.e. service units with an @ in their name) are assigned by + default a per-template slice unit (see + systemd.slice5), named after the + template unit, containing all instances of the specific template. This slice is normally stopped at shutdown, + together with all template instances. If that is not desired, set DefaultDependencies=no in the + template unit, and either define your own per-template slice unit file that also sets + DefaultDependencies=no, or set Slice=system.slice (or another suitable slice) + in the template unit. Also see + systemd.resource-control5. + + Additional implicit dependencies may be added as result of + execution and resource control parameters as documented in + systemd.exec5 + and + systemd.resource-control5. + + + + Options + + Service files must include a [Service] + section, which carries information about the service and the + process it supervises. A number of options that may be used in + this section are shared with other unit types. These options are + documented in + systemd.exec5 + and + systemd.kill5. + The options specific to the [Service] section + of service units are the following: + + + + Type= + + Configures the process start-up type for this + service unit. One of + , + , + , + , + or + . + + If set to (the default if + neither Type= nor + BusName=, but ExecStart= + are specified), it is expected that the process configured + with ExecStart= is the main process of the + service. In this mode, if the process offers functionality to + other processes on the system, its communication channels + should be installed before the daemon is started up (e.g. + sockets set up by systemd, via socket activation), as systemd + will immediately proceed starting follow-up units. + + If set to , it is expected that + the process configured with ExecStart= will + call fork() as part of its start-up. The + parent process is expected to exit when start-up is complete + and all communication channels are set up. The child continues + to run as the main daemon process. This is the behavior of + traditional UNIX daemons. If this setting is used, it is + recommended to also use the PIDFile= + option, so that systemd can identify the main process of the + daemon. systemd will proceed with starting follow-up units as + soon as the parent process exits. + + Behavior of is similar to + ; however, it is expected that the + process has to exit before systemd starts follow-up units. + RemainAfterExit= is particularly useful for + this type of service. This is the implied default if neither + Type= or ExecStart= are + specified. + + Behavior of is similar to + ; however, it is expected that the + daemon acquires a name on the D-Bus bus, as configured by + BusName=. systemd will proceed with + starting follow-up units after the D-Bus bus name has been + acquired. Service units with this option configured implicitly + gain dependencies on the dbus.socket + unit. This type is the default if BusName= + is specified. + + Behavior of is similar to + ; however, it is expected that the + daemon sends a notification message via + sd_notify3 + or an equivalent call when it has finished starting up. + systemd will proceed with starting follow-up units after this + notification message has been sent. If this option is used, + NotifyAccess= (see below) should be set to + open access to the notification socket provided by systemd. If + NotifyAccess= is missing or set to + , it will be forcibly set to + . Note that currently + Type= will not work + if used in combination with + PrivateNetwork=. + + Behavior of is very similar to + ; however, actual execution of the + service binary is delayed until all jobs are dispatched. This + may be used to avoid interleaving of output of shell services + with the status output on the console. + + + + + RemainAfterExit= + + Takes a boolean value that specifies whether + the service shall be considered active even when all its + processes exited. Defaults to . + + + + + GuessMainPID= + + Takes a boolean value that specifies whether + systemd should try to guess the main PID of a service if it + cannot be determined reliably. This option is ignored unless + is set and + is unset because for the other types + or with an explicitly configured PID file, the main PID is + always known. The guessing algorithm might come to incorrect + conclusions if a daemon consists of more than one process. If + the main PID cannot be determined, failure detection and + automatic restarting of a service will not work reliably. + Defaults to . + + + + + PIDFile= + + Takes an absolute file name pointing to the + PID file of this daemon. Use of this option is recommended for + services where Type= is set to + . systemd will read the PID of the + main process of the daemon after start-up of the service. + systemd will not write to the file configured here, although + it will remove the file after the service has shut down if it + still exists. + + + + + + BusName= + + Takes a D-Bus bus name that this service is + reachable as. This option is mandatory for services where + Type= is set to + . + + + + + ExecStart= + Commands with their arguments that are + executed when this service is started. The value is split into + zero or more command lines according to the rules described + below (see section "Command Lines" below). + + + When Type= is not + , only one command may and must be + given. When Type=oneshot is used, zero or + more commands may be specified. This can be specified by + providing multiple command lines in the same directive, or + alternatively, this directive may be specified more than once + with the same effect. If the empty string is assigned to this + option, the list of commands to start is reset, prior + assignments of this option will have no effect. If no + ExecStart= is specified, then the service + must have RemainAfterExit=yes set. + + For each of the specified commands, the first argument must be an absolute path to an + executable. Optionally, if this file name is prefixed with @, the second token will be + passed as argv[0] to the executed process, followed by the further arguments specified. If + the absolute filename is prefixed with -, an exit code of the command normally considered a + failure (i.e. non-zero exit status or abnormal exit due to signal) is ignored and considered success. If the + absolute path is prefixed with + then it is executed with full + privileges. -, @, and + may be used together and they + can appear in any order. + + If more than one command is specified, the commands are + invoked sequentially in the order they appear in the unit + file. If one of the commands fails (and is not prefixed with + -), other lines are not executed, and the + unit is considered failed. + + Unless Type=forking is set, the + process started via this command line will be considered the + main process of the daemon. + + + + + ExecStartPre= + ExecStartPost= + Additional commands that are executed before + or after the command in ExecStart=, + respectively. Syntax is the same as for + ExecStart=, except that multiple command + lines are allowed and the commands are executed one after the + other, serially. + + If any of those commands (not prefixed with + -) fail, the rest are not executed and the + unit is considered failed. + + ExecStart= commands are only run after + all ExecStartPre= commands that were not prefixed + with a - exit successfully. + + ExecStartPost= commands are only run after + the service has started successfully, as determined by Type= + (i.e. the process has been started for Type=simple + or Type=idle, the process exits successfully for + Type=oneshot, the initial process exits successfully + for Type=forking, READY=1 is sent + for Type=notify, or the BusName= + has been taken for Type=dbus). + + Note that ExecStartPre= may not be + used to start long-running processes. All processes forked + off by processes invoked via ExecStartPre= will + be killed before the next service process is run. + + Note that if any of the commands specified in ExecStartPre=, + ExecStart=, or ExecStartPost= fail (and are not prefixed with + -, see above) or time out before the service is fully up, execution continues with commands + specified in ExecStopPost=, the commands in ExecStop= are skipped. + + + + + ExecReload= + Commands to execute to trigger a configuration + reload in the service. This argument takes multiple command + lines, following the same scheme as described for + ExecStart= above. Use of this setting is + optional. Specifier and environment variable substitution is + supported here following the same scheme as for + ExecStart=. + + One additional, special environment variable is set: if + known, $MAINPID is set to the main process + of the daemon, and may be used for command lines like the + following: + + /bin/kill -HUP $MAINPID + + Note however that reloading a daemon by sending a signal + (as with the example line above) is usually not a good choice, + because this is an asynchronous operation and hence not + suitable to order reloads of multiple services against each + other. It is strongly recommended to set + ExecReload= to a command that not only + triggers a configuration reload of the daemon, but also + synchronously waits for it to complete. + + + + + ExecStop= + Commands to execute to stop the service + started via ExecStart=. This argument takes + multiple command lines, following the same scheme as described + for ExecStart= above. Use of this setting + is optional. After the commands configured in this option are + run, all processes remaining for a service are terminated + according to the KillMode= setting (see + systemd.kill5). + If this option is not specified, the process is terminated by + sending the signal specified in KillSignal= + when service stop is requested. Specifier and environment + variable substitution is supported (including + $MAINPID, see above). + + Note that it is usually not sufficient to specify a + command for this setting that only asks the service to + terminate (for example, by queuing some form of termination + signal for it), but does not wait for it to do so. Since the + remaining processes of the services are killed using + SIGKILL immediately after the command + exited, this would not result in a clean stop. The specified + command should hence be a synchronous operation, not an + asynchronous one. + + Note that the commands specified in ExecStop= are only executed when the service + started successfully first. They are not invoked if the service was never started at all, or in case its + start-up failed, for example because any of the commands specified in ExecStart=, + ExecStartPre= or ExecStartPost= failed (and weren't prefixed with + -, see above) or timed out. Use ExecStopPost= to invoke commands when a + service failed to start up correctly and is shut down again. + + It is recommended to use this setting for commands that communicate with the service requesting clean + termination. When the commands specified with this option are executed it should be assumed that the service is + still fully up and is able to react correctly to all commands. For post-mortem clean-up steps use + ExecStopPost= instead. + + + + ExecStopPost= + Additional commands that are executed after the service is stopped. This includes cases where + the commands configured in ExecStop= were used, where the service does not have any + ExecStop= defined, or where the service exited unexpectedly. This argument takes multiple + command lines, following the same scheme as described for ExecStart=. Use of these settings + is optional. Specifier and environment variable substitution is supported. Note that – unlike + ExecStop= – commands specified with this setting are invoked when a service failed to start + up correctly and is shut down again. + + It is recommended to use this setting for clean-up operations that shall be executed even when the + service failed to start up correctly. Commands configured with this setting need to be able to operate even if + the service failed starting up half-way and left incompletely initialized data around. As the service's + processes have been terminated already when the commands specified with this setting are executed they should + not attempt to communicate with them. + + + + RestartSec= + Configures the time to sleep before restarting + a service (as configured with Restart=). + Takes a unit-less value in seconds, or a time span value such + as "5min 20s". Defaults to 100ms. + + + + TimeoutStartSec= + Configures the time to wait for start-up. If a + daemon service does not signal start-up completion within the + configured time, the service will be considered failed and + will be shut down again. Takes a unit-less value in seconds, + or a time span value such as "5min 20s". Pass + infinity to disable the timeout logic. Defaults to + DefaultTimeoutStartSec= from the manager + configuration file, except when + Type=oneshot is used, in which case the + timeout is disabled by default (see + systemd-system.conf5). + + + + + TimeoutStopSec= + Configures the time to wait for stop. If a + service is asked to stop, but does not terminate in the + specified time, it will be terminated forcibly via + SIGTERM, and after another timeout of + equal duration with SIGKILL (see + KillMode= in + systemd.kill5). + Takes a unit-less value in seconds, or a time span value such + as "5min 20s". Pass infinity to disable the + timeout logic. Defaults to + DefaultTimeoutStopSec= from the manager + configuration file (see + systemd-system.conf5). + + + + + TimeoutSec= + A shorthand for configuring both + TimeoutStartSec= and + TimeoutStopSec= to the specified value. + + + + + RuntimeMaxSec= + + Configures a maximum time for the service to run. If this is used and the service has been + active for longer than the specified time it is terminated and put into a failure state. Note that this setting + does not have any effect on Type=oneshot services, as they terminate immediately after + activation completed. Pass infinity (the default) to configure no runtime + limit. + + + + WatchdogSec= + Configures the watchdog timeout for a service. + The watchdog is activated when the start-up is completed. The + service must call + sd_notify3 + regularly with WATCHDOG=1 (i.e. the + "keep-alive ping"). If the time between two such calls is + larger than the configured time, then the service is placed in + a failed state and it will be terminated with + SIGABRT. By setting + Restart= to , + , or + , the service will be automatically + restarted. The time configured here will be passed to the + executed service process in the + WATCHDOG_USEC= environment variable. This + allows daemons to automatically enable the keep-alive pinging + logic if watchdog support is enabled for the service. If this + option is used, NotifyAccess= (see below) + should be set to open access to the notification socket + provided by systemd. If NotifyAccess= is + not set, it will be implicitly set to . + Defaults to 0, which disables this feature. The service can + check whether the service manager expects watchdog keep-alive + notifications. See + sd_watchdog_enabled3 + for details. + sd_event_set_watchdog3 + may be used to enable automatic watchdog notification support. + + + + + Restart= + Configures whether the service shall be + restarted when the service process exits, is killed, or a + timeout is reached. The service process may be the main + service process, but it may also be one of the processes + specified with ExecStartPre=, + ExecStartPost=, + ExecStop=, + ExecStopPost=, or + ExecReload=. When the death of the process + is a result of systemd operation (e.g. service stop or + restart), the service will not be restarted. Timeouts include + missing the watchdog "keep-alive ping" deadline and a service + start, reload, and stop operation timeouts. + + Takes one of + , + , + , + , + , + , or + . + If set to (the default), the service will + not be restarted. If set to , it + will be restarted only when the service process exits cleanly. + In this context, a clean exit means an exit code of 0, or one + of the signals + SIGHUP, + SIGINT, + SIGTERM or + SIGPIPE, and + additionally, exit statuses and signals specified in + SuccessExitStatus=. If set to + , the service will be restarted + when the process exits with a non-zero exit code, is + terminated by a signal (including on core dump, but excluding + the aforementioned four signals), when an operation (such as + service reload) times out, and when the configured watchdog + timeout is triggered. If set to , + the service will be restarted when the process is terminated + by a signal (including on core dump, excluding the + aforementioned four signals), when an operation times out, or + when the watchdog timeout is triggered. If set to + , the service will be restarted only + if the service process exits due to an uncaught signal not + specified as a clean exit status. If set to + , the service will be restarted + only if the watchdog timeout for the service expires. If set + to , the service will be restarted + regardless of whether it exited cleanly or not, got terminated + abnormally by a signal, or hit a timeout. + + + Exit causes and the effect of the <varname>Restart=</varname> settings on them + + + + + + + Restart settings/Exit causes + + + + + + + + + + + + Clean exit code or signal + + X + X + + + + + + + Unclean exit code + + X + + X + + + + + + Unclean signal + + X + + X + X + X + + + + Timeout + + X + + X + X + + + + + Watchdog + + X + + X + X + + X + + + +
+ + As exceptions to the setting above, the service will not + be restarted if the exit code or signal is specified in + RestartPreventExitStatus= (see below). + Also, the services will always be restarted if the exit code + or signal is specified in + RestartForceExitStatus= (see below). + + Setting this to is the + recommended choice for long-running services, in order to + increase reliability by attempting automatic recovery from + errors. For services that shall be able to terminate on their + own choice (and avoid immediate restarting), + is an alternative choice. + + + + + SuccessExitStatus= + Takes a list of exit status definitions that, + when returned by the main service process, will be considered + successful termination, in addition to the normal successful + exit code 0 and the signals SIGHUP, + SIGINT, SIGTERM, and + SIGPIPE. Exit status definitions can + either be numeric exit codes or termination signal names, + separated by spaces. For example: + + SuccessExitStatus=1 2 8 SIGKILL + + ensures that exit codes 1, 2, 8 and + the termination signal SIGKILL are + considered clean service terminations. + + + Note that if a process has a signal handler installed + and exits by calling + _exit2 + in response to a signal, the information about the signal is + lost. Programs should instead perform cleanup and kill + themselves with the same signal instead. See + Proper + handling of SIGINT/SIGQUIT — How to be a proper + program. + + This option may appear more than once, in which case the + list of successful exit statuses is merged. If the empty + string is assigned to this option, the list is reset, all + prior assignments of this option will have no + effect. + + + + RestartPreventExitStatus= + Takes a list of exit status definitions that, + when returned by the main service process, will prevent + automatic service restarts, regardless of the restart setting + configured with Restart=. Exit status + definitions can either be numeric exit codes or termination + signal names, and are separated by spaces. Defaults to the + empty list, so that, by default, no exit status is excluded + from the configured restart logic. For example: + + RestartPreventExitStatus=1 6 SIGABRT + + ensures that exit codes 1 and 6 and the termination signal + SIGABRT will not result in automatic + service restarting. This option may appear more than once, in + which case the list of restart-preventing statuses is + merged. If the empty string is assigned to this option, the + list is reset and all prior assignments of this option will + have no effect. + + + + RestartForceExitStatus= + Takes a list of exit status definitions that, + when returned by the main service process, will force automatic + service restarts, regardless of the restart setting configured + with Restart=. The argument format is + similar to + RestartPreventExitStatus=. + + + + PermissionsStartOnly= + Takes a boolean argument. If true, the + permission-related execution options, as configured with + User= and similar options (see + systemd.exec5 + for more information), are only applied to the process started + with + ExecStart=, and not to the various other + ExecStartPre=, + ExecStartPost=, + ExecReload=, + ExecStop=, and + ExecStopPost= + commands. If false, the setting is applied to all configured + commands the same way. Defaults to false. + + + + RootDirectoryStartOnly= + Takes a boolean argument. If true, the root + directory, as configured with the + RootDirectory= option (see + systemd.exec5 + for more information), is only applied to the process started + with ExecStart=, and not to the various + other ExecStartPre=, + ExecStartPost=, + ExecReload=, ExecStop=, + and ExecStopPost= commands. If false, the + setting is applied to all configured commands the same way. + Defaults to false. + + + + NonBlocking= + Set the O_NONBLOCK flag + for all file descriptors passed via socket-based activation. + If true, all file descriptors >= 3 (i.e. all except stdin, + stdout, and stderr) will have the + O_NONBLOCK flag set and hence are in + non-blocking mode. This option is only useful in conjunction + with a socket unit, as described in + systemd.socket5. + Defaults to false. + + + + NotifyAccess= + Controls access to the service status + notification socket, as accessible via the + sd_notify3 + call. Takes one of (the default), + or . If + , no daemon status updates are accepted + from the service processes, all status update messages are + ignored. If , only service updates sent + from the main process of the service are accepted. If + , all services updates from all members of + the service's control group are accepted. This option should + be set to open access to the notification socket when using + Type=notify or + WatchdogSec= (see above). If those options + are used but NotifyAccess= is not + configured, it will be implicitly set to + . + + + + Sockets= + Specifies the name of the socket units this + service shall inherit socket file descriptors from when the + service is started. Normally, it should not be necessary to use + this setting, as all socket file descriptors whose unit shares + the same name as the service (subject to the different unit + name suffix of course) are passed to the spawned + process. + + Note that the same socket file descriptors may be passed + to multiple processes simultaneously. Also note that a + different service may be activated on incoming socket traffic + than the one which is ultimately configured to inherit the + socket file descriptors. Or, in other words: the + Service= setting of + .socket units does not have to match the + inverse of the Sockets= setting of the + .service it refers to. + + This option may appear more than once, in which case the + list of socket units is merged. If the empty string is + assigned to this option, the list of sockets is reset, and all + prior uses of this setting will have no + effect. + + + + FailureAction= + Configure the action to take when the service enters a failed state. Takes the same values as + the unit setting StartLimitAction= and executes the same actions (see + systemd.unit5). Defaults to + . + + + + FileDescriptorStoreMax= + Configure how many file descriptors may be + stored in the service manager for the service using + sd_pid_notify_with_fds3's + FDSTORE=1 messages. This is useful for + implementing service restart schemes where the state is + serialized to /run and the file + descriptors passed to the service manager, to allow restarts + without losing state. Defaults to 0, i.e. no file descriptors + may be stored in the service manager by default. All file + descriptors passed to the service manager from a specific + service are passed back to the service's main process on the + next service restart. Any file descriptors passed to the + service manager are automatically closed when POLLHUP or + POLLERR is seen on them, or when the service is fully stopped + and no job queued or being executed for it. + + + + USBFunctionDescriptors= + Configure the location of a file containing + USB + FunctionFS descriptors, for implementation of USB + gadget functions. This is used only in conjunction with a + socket unit with ListenUSBFunction= + configured. The contents of this file are written to the + ep0 file after it is + opened. + + + + USBFunctionStrings= + Configure the location of a file containing + USB FunctionFS strings. Behavior is similar to + USBFunctionDescriptors= + above. + + + + + Check + systemd.exec5 + and + systemd.kill5 + for more settings. + + + + + Command lines + + This section describes command line parsing and + variable and specifier substitutions for + ExecStart=, + ExecStartPre=, + ExecStartPost=, + ExecReload=, + ExecStop=, and + ExecStopPost= options. + + Multiple command lines may be concatenated in a single + directive by separating them with semicolons (these semicolons + must be passed as separate words). Lone semicolons may be escaped + as \;. + + Each command line is split on whitespace, with the first + item being the command to execute, and the subsequent items being + the arguments. Double quotes ("...") and single quotes ('...') may + be used, in which case everything until the next matching quote + becomes part of the same argument. C-style escapes are also + supported. The table below contains the list of allowed escape + patterns. Only patterns which match the syntax in the table are + allowed; others will result in an error, and must be escaped by + doubling the backslash. Quotes themselves are removed after + parsing and escape sequences substituted. In addition, a trailing + backslash (\) may be used to merge lines. + + + This syntax is intended to be very similar to shell syntax, + but only the meta-characters and expansions described in the + following paragraphs are understood. Specifically, redirection + using + <, + <<, + >, and + >>, pipes using + |, running programs in the background using + &, and other elements of shell + syntax are not supported. + + The command to execute must be an absolute path name. It may + contain spaces, but control characters are not allowed. + + The command line accepts % specifiers as + described in + systemd.unit5. + Note that the first argument of the command line (i.e. the program + to execute) may not include specifiers. + + Basic environment variable substitution is supported. Use + ${FOO} as part of a word, or as a word of its + own, on the command line, in which case it will be replaced by the + value of the environment variable including all whitespace it + contains, resulting in a single argument. Use + $FOO as a separate word on the command line, in + which case it will be replaced by the value of the environment + variable split at whitespace, resulting in zero or more arguments. + For this type of expansion, quotes are respected when splitting + into words, and afterwards removed. + + Example: + + Environment="ONE=one" 'TWO=two two' +ExecStart=/bin/echo $ONE $TWO ${TWO} + + This will execute /bin/echo with four + arguments: one, two, + two, and two two. + + Example: + Environment=ONE='one' "TWO='two two' too" THREE= +ExecStart=/bin/echo ${ONE} ${TWO} ${THREE} +ExecStart=/bin/echo $ONE $TWO $THREE + This results in echo being + called twice, the first time with arguments + 'one', + 'two two' too, , + and the second time with arguments + one, two two, + too. + + + To pass a literal dollar sign, use $$. + Variables whose value is not known at expansion time are treated + as empty strings. Note that the first argument (i.e. the program + to execute) may not be a variable. + + Variables to be used in this fashion may be defined through + Environment= and + EnvironmentFile=. In addition, variables listed + in the section "Environment variables in spawned processes" in + systemd.exec5, + which are considered "static configuration", may be used (this + includes e.g. $USER, but not + $TERM). + + Note that shell command lines are not directly supported. If + shell command lines are to be used, they need to be passed + explicitly to a shell implementation of some kind. Example: + ExecStart=/bin/sh -c 'dmesg | tac' + + Example: + + ExecStart=/bin/echo one ; /bin/echo "two two" + + This will execute /bin/echo two times, + each time with one argument: one and + two two, respectively. Because two commands are + specified, Type=oneshot must be used. + + Example: + + ExecStart=/bin/echo / >/dev/null & \; \ +/bin/ls + + This will execute /bin/echo + with five arguments: /, + >/dev/null, + &, ;, and + /bin/ls. + + + C escapes supported in command lines and environment variables + + + + + + Literal + Actual value + + + + + \a + bell + + + \b + backspace + + + \f + form feed + + + \n + newline + + + \r + carriage return + + + \t + tab + + + \v + vertical tab + + + \\ + backslash + + + \" + double quotation mark + + + \' + single quotation mark + + + \s + space + + + \xxx + character number xx in hexadecimal encoding + + + \nnn + character number nnn in octal encoding + + + +
+ + + + Examples + + + Simple service + + The following unit file creates a service that will + execute /usr/sbin/foo-daemon. Since no + Type= is specified, the default + Type= will be assumed. + systemd will assume the unit to be started immediately after the + program has begun executing. + + [Unit] +Description=Foo + +[Service] +ExecStart=/usr/sbin/foo-daemon + +[Install] +WantedBy=multi-user.target + + Note that systemd assumes here that the process started by + systemd will continue running until the service terminates. If + the program daemonizes itself (i.e. forks), please use + Type= instead. + + Since no ExecStop= was specified, + systemd will send SIGTERM to all processes started from this + service, and after a timeout also SIGKILL. This behavior can be + modified, see + systemd.kill5 + for details. + + Note that this unit type does not include any type of + notification when a service has completed initialization. For + this, you should use other unit types, such as + Type= if the service + understands systemd's notification protocol, + Type= if the service + can background itself or + Type= if the unit + acquires a DBus name once initialization is complete. See + below. + + + + Oneshot service + + Sometimes, units should just execute an action without + keeping active processes, such as a filesystem check or a + cleanup action on boot. For this, + Type= exists. Units + of this type will wait until the process specified terminates + and then fall back to being inactive. The following unit will + perform a cleanup action: + + [Unit] +Description=Cleanup old Foo data + +[Service] +Type=oneshot +ExecStart=/usr/sbin/foo-cleanup + +[Install] +WantedBy=multi-user.target + + Note that systemd will consider the unit to be in the + state "starting" until the program has terminated, so ordered + dependencies will wait for the program to finish before starting + themselves. The unit will revert to the "inactive" state after + the execution is done, never reaching the "active" state. That + means another request to start the unit will perform the action + again. + + Type= are the + only service units that may have more than one + ExecStart= specified. They will be executed + in order until either they are all successful or one of them + fails. + + + + Stoppable oneshot service + + Similarly to the oneshot services, there are sometimes + units that need to execute a program to set up something and + then execute another to shut it down, but no process remains + active while they are considered "started". Network + configuration can sometimes fall into this category. Another use + case is if a oneshot service shall not be executed each time + when they are pulled in as a dependency, but only the first + time. + + For this, systemd knows the setting + RemainAfterExit=, which + causes systemd to consider the unit to be active if the start + action exited successfully. This directive can be used with all + types, but is most useful with + Type= and + Type=. With + Type=, systemd waits + until the start action has completed before it considers the + unit to be active, so dependencies start only after the start + action has succeeded. With + Type=, dependencies + will start immediately after the start action has been + dispatched. The following unit provides an example for a simple + static firewall. + + [Unit] +Description=Simple firewall + +[Service] +Type=oneshot +RemainAfterExit=yes +ExecStart=/usr/local/sbin/simple-firewall-start +ExecStop=/usr/local/sbin/simple-firewall-stop + +[Install] +WantedBy=multi-user.target + + Since the unit is considered to be running after the start + action has exited, invoking systemctl start + on that unit again will cause no action to be taken. + + + + Traditional forking services + + Many traditional daemons/services background (i.e. fork, + daemonize) themselves when starting. Set + Type= in the + service's unit file to support this mode of operation. systemd + will consider the service to be in the process of initialization + while the original program is still running. Once it exits + successfully and at least a process remains (and + RemainAfterExit=), the + service is considered started. + + Often, a traditional daemon only consists of one process. + Therefore, if only one process is left after the original + process terminates, systemd will consider that process the main + process of the service. In that case, the + $MAINPID variable will be available in + ExecReload=, ExecStop=, + etc. + + In case more than one process remains, systemd will be + unable to determine the main process, so it will not assume + there is one. In that case, $MAINPID will not + expand to anything. However, if the process decides to write a + traditional PID file, systemd will be able to read the main PID + from there. Please set PIDFile= accordingly. + Note that the daemon should write that file before finishing + with its initialization. Otherwise, systemd might try to read the + file before it exists. + + The following example shows a simple daemon that forks and + just starts one process in the background: + + [Unit] +Description=Some simple daemon + +[Service] +Type=forking +ExecStart=/usr/sbin/my-simple-daemon -d + +[Install] +WantedBy=multi-user.target + + Please see + systemd.kill5 + for details on how you can influence the way systemd terminates + the service. + + + + DBus services + + For services that acquire a name on the DBus system bus, + use Type= and set + BusName= accordingly. The service should not + fork (daemonize). systemd will consider the service to be + initialized once the name has been acquired on the system bus. + The following example shows a typical DBus service: + + [Unit] +Description=Simple DBus service + +[Service] +Type=dbus +BusName=org.example.simple-dbus-service +ExecStart=/usr/sbin/simple-dbus-service + +[Install] +WantedBy=multi-user.target + + For bus-activatable services, do not + include a [Install] section in the systemd + service file, but use the SystemdService= + option in the corresponding DBus service file, for example + (/usr/share/dbus-1/system-services/org.example.simple-dbus-service.service): + + [D-BUS Service] +Name=org.example.simple-dbus-service +Exec=/usr/sbin/simple-dbus-service +User=root +SystemdService=simple-dbus-service.service + + Please see + systemd.kill5 + for details on how you can influence the way systemd terminates + the service. + + + + Services that notify systemd about their initialization + + Type= services + are really easy to write, but have the major disadvantage of + systemd not being able to tell when initialization of the given + service is complete. For this reason, systemd supports a simple + notification protocol that allows daemons to make systemd aware + that they are done initializing. Use + Type= for this. A + typical service file for such a daemon would look like + this: + + [Unit] +Description=Simple notifying service + +[Service] +Type=notify +ExecStart=/usr/sbin/simple-notifying-service + +[Install] +WantedBy=multi-user.target + + Note that the daemon has to support systemd's notification + protocol, else systemd will think the service has not started yet + and kill it after a timeout. For an example of how to update + daemons to support this protocol transparently, take a look at + sd_notify3. + systemd will consider the unit to be in the 'starting' state + until a readiness notification has arrived. + + Please see + systemd.kill5 + for details on how you can influence the way systemd terminates + the service. + + + + + See Also + + systemd1, + systemctl1, + systemd.unit5, + systemd.exec5, + systemd.resource-control5, + systemd.kill5, + systemd.directives7 + + + + diff --git a/src/grp-system/systemd/systemd.slice.xml b/src/grp-system/systemd/systemd.slice.xml new file mode 100644 index 0000000000..eee98d99ee --- /dev/null +++ b/src/grp-system/systemd/systemd.slice.xml @@ -0,0 +1,132 @@ + + + + + + + + systemd.slice + systemd + + + + Developer + Lennart + Poettering + lennart@poettering.net + + + + + + systemd.slice + 5 + + + + systemd.slice + Slice unit configuration + + + + slice.slice + + + + Description + + A unit configuration file whose name ends in + .slice encodes information about a slice which + is a concept for hierarchically managing resources of a group of + processes. This management is performed by creating a node in the + Linux Control Group (cgroup) tree. Units that manage processes + (primarily scope and service units) may be assigned to a specific + slice. For each slice, certain resource limits may be set that + apply to all processes of all units contained in that + slice. Slices are organized hierarchically in a tree. The name of + the slice encodes the location in the tree. The name consists of a + dash-separated series of names, which describes the path to the + slice from the root slice. The root slice is named, + -.slice. Example: + foo-bar.slice is a slice that is located + within foo.slice, which in turn is located in + the root slice -.slice. + + + Note that slice units cannot be templated, nor is possible to add multiple names to a slice unit by creating + additional symlinks to it. + + By default, service and scope units are placed in + system.slice, virtual machines and containers + registered with + systemd-machined1 + are found in machine.slice, and user sessions + handled by + systemd-logind1 + in user.slice. See + systemd.special5 + for more information. + + See + systemd.unit5 + for the common options of all unit configuration + files. The common configuration items are configured + in the generic [Unit] and [Install] sections. The + slice specific configuration options are configured in + the [Slice] section. Currently, only generic resource control settings + as described in + systemd.resource-control5 are allowed. + + + See the New + Control Group Interfaces for an introduction on how to make + use of slice units from programs. + + + + Automatic Dependencies + + Slice units automatically gain dependencies of type + After= and Requires= on + their immediate parent slice unit. + + Unless DefaultDependencies=false is used in the [Unit] section, slice + units will implicitly have dependencies of type Conflicts= and Before= on + shutdown.target. These ensure that slice units are removed prior to system shutdown. Only + slice units involved with early boot or late system shutdown should disable this option. + + + + + See Also + + systemd1, + systemd.unit5, + systemd.resource-control5, + systemd.service5, + systemd.scope5, + systemd.special7, + systemd.directives7 + + + + diff --git a/src/grp-system/systemd/systemd.socket.xml b/src/grp-system/systemd/systemd.socket.xml new file mode 100644 index 0000000000..5bf54d8ef3 --- /dev/null +++ b/src/grp-system/systemd/systemd.socket.xml @@ -0,0 +1,860 @@ + + + + + + + + systemd.socket + systemd + + + + Developer + Lennart + Poettering + lennart@poettering.net + + + + + + systemd.socket + 5 + + + + systemd.socket + Socket unit configuration + + + + socket.socket + + + + Description + + A unit configuration file whose name ends in + .socket encodes information about an IPC or + network socket or a file system FIFO controlled and supervised by + systemd, for socket-based activation. + + This man page lists the configuration options specific to + this unit type. See + systemd.unit5 + for the common options of all unit configuration files. The common + configuration items are configured in the generic [Unit] and + [Install] sections. The socket specific configuration options are + configured in the [Socket] section. + + Additional options are listed in + systemd.exec5, + which define the execution environment the + , , + and + commands are executed in, and in + systemd.kill5, + which define the way the processes are terminated, and in + systemd.resource-control5, + which configure resource control settings for the processes of the + socket. + + For each socket file, a matching service file must exist, + describing the service to start on incoming traffic on the socket + (see + systemd.service5 + for more information about .service files). The name of the + .service unit is by default the same as the name of the .socket + unit, but can be altered with the option + described below. Depending on the setting of the + option described below, this .service + unit must either be named like the .socket unit, but with the + suffix replaced, unless overridden with ; + or it must be a template unit named the same way. Example: a + socket file foo.socket needs a matching + service foo.service if + is set. If + is set, a service template file + foo@.service must exist from which services + are instantiated for each incoming connection. + + Unless DefaultDependencies= in the [Unit] section is set to + , socket units will implicitly have dependencies of type Requires= and + After= on sysinit.target as well as dependencies of type + Conflicts= and Before= on shutdown.target. These ensure + that socket units pull in basic system initialization, and are terminated cleanly prior to system shutdown. Only + sockets involved with early boot or late system shutdown should disable this option. + + Socket units will have a Before= + dependency on the service which they trigger added implicitly. No + implicit WantedBy= or + RequiredBy= dependency from the socket to the + service is added. This means that the service may be started + without the socket, in which case it must be able to open sockets + by itself. To prevent this, an explicit + Requires= dependency may be added. + + Socket units may be used to implement on-demand starting of + services, as well as parallelized starting of services. See the + blog stories linked at the end for an introduction. + + Note that the daemon software configured for socket + activation with socket units needs to be able to accept sockets + from systemd, either via systemd's native socket passing interface + (see + sd_listen_fds3 + for details) or via the traditional + inetd8-style + socket passing (i.e. sockets passed in via standard input and + output, using StandardInput=socket in the + service file). + + + + Automatic Dependencies + + Socket units automatically gain a Before= + dependency on the service units they activate. + + Socket units referring to file system paths (such as AF_UNIX + sockets or FIFOs) implicitly gain Requires= and + After= dependencies on all mount units + necessary to access those paths. + + Socket units using the BindToDevice= + setting automatically gain a BindsTo= and + After= dependency on the device unit + encapsulating the specified network interface. + + If DefaultDependencies=yes is set (the + default), socket units automatically gain a + Before= dependency on + sockets.target. They also gain a pair of + After= and Requires= + dependency on sysinit.target, and a pair of + Before= and Conflicts= + dependencies on shutdown.target. These + dependencies ensure that the socket unit is started before normal + services at boot, and is stopped on shutdown. + + Additional implicit dependencies may be added as result of + execution and resource control parameters as documented in + systemd.exec5 + and + systemd.resource-control5. + + + + Options + + Socket files must include a [Socket] section, which carries + information about the socket or FIFO it supervises. A number of + options that may be used in this section are shared with other + unit types. These options are documented in + systemd.exec5 + and + systemd.kill5. + The options specific to the [Socket] section of socket units are + the following: + + + + ListenStream= + ListenDatagram= + ListenSequentialPacket= + Specifies an address to listen on for a stream + (SOCK_STREAM), datagram + (SOCK_DGRAM), or sequential packet + (SOCK_SEQPACKET) socket, respectively. + The address can be written in various formats: + + If the address starts with a slash + (/), it is read as file system socket in + the AF_UNIX socket family. + + If the address starts with an at symbol + (@), it is read as abstract namespace + socket in the AF_UNIX family. The + @ is replaced with a + NUL character before binding. For + details, see + unix7. + + If the address string is a single number, it is read as + port number to listen on via IPv6. Depending on the value of + BindIPv6Only= (see below) this might result + in the service being available via both IPv6 and IPv4 + (default) or just via IPv6. + + + If the address string is a string in the format + v.w.x.y:z, it is read as IPv4 specifier for listening on an + address v.w.x.y on a port z. + + If the address string is a string in the format [x]:y, + it is read as IPv6 address x on a port y. Note that this might + make the service available via IPv4, too, depending on the + BindIPv6Only= setting (see below). + + + Note that SOCK_SEQPACKET (i.e. + ListenSequentialPacket=) is only available + for AF_UNIX sockets. + SOCK_STREAM (i.e. + ListenStream=) when used for IP sockets + refers to TCP sockets, SOCK_DGRAM (i.e. + ListenDatagram=) to UDP. + + These options may be specified more than once, in which + case incoming traffic on any of the sockets will trigger + service activation, and all listed sockets will be passed to + the service, regardless of whether there is incoming traffic + on them or not. If the empty string is assigned to any of + these options, the list of addresses to listen on is reset, + all prior uses of any of these options will have no + effect. + + It is also possible to have more than one socket unit + for the same service when using Service=, + and the service will receive all the sockets configured in all + the socket units. Sockets configured in one unit are passed in + the order of configuration, but no ordering between socket + units is specified. + + If an IP address is used here, it is often desirable to + listen on it before the interface it is configured on is up + and running, and even regardless of whether it will be up and + running at any point. To deal with this, it is recommended to + set the FreeBind= option described + below. + + + + ListenFIFO= + Specifies a file system FIFO to listen on. + This expects an absolute file system path as argument. + Behavior otherwise is very similar to the + ListenDatagram= directive + above. + + + + ListenSpecial= + Specifies a special file in the file system to + listen on. This expects an absolute file system path as + argument. Behavior otherwise is very similar to the + ListenFIFO= directive above. Use this to + open character device nodes as well as special files in + /proc and + /sys. + + + + ListenNetlink= + Specifies a Netlink family to create a socket + for to listen on. This expects a short string referring to the + AF_NETLINK family name (such as + audit or kobject-uevent) + as argument, optionally suffixed by a whitespace followed by a + multicast group integer. Behavior otherwise is very similar to + the ListenDatagram= directive + above. + + + + ListenMessageQueue= + Specifies a POSIX message queue name to listen + on. This expects a valid message queue name (i.e. beginning + with /). Behavior otherwise is very similar to the + ListenFIFO= directive above. On Linux + message queue descriptors are actually file descriptors and + can be inherited between processes. + + + + ListenUSBFunction= + Specifies a USB + FunctionFS endpoint location to listen on, for + implementation of USB gadget functions. This expects an + absolute file system path as the argument. Behavior otherwise + is very similar to the ListenFIFO= + directive above. Use this to open the FunctionFS endpoint + ep0. When using this option, the + activated service has to have the + USBFunctionDescriptors= and + USBFunctionStrings= options set. + + + + + SocketProtocol= + Takes a one of + or . Specifies a socket protocol + (IPPROTO_UDPLITE) UDP-Lite + (IPPROTO_SCTP) SCTP socket respectively. + + + + + BindIPv6Only= + Takes a one of , + or . Controls + the IPV6_V6ONLY socket option (see + ipv67 + for details). If , IPv6 sockets bound + will be accessible via both IPv4 and IPv6. If + , they will be accessible via IPv6 + only. If (which is the default, + surprise!), the system wide default setting is used, as + controlled by + /proc/sys/net/ipv6/bindv6only, which in + turn defaults to the equivalent of + . + + + + + Backlog= + Takes an unsigned integer argument. Specifies + the number of connections to queue that have not been accepted + yet. This setting matters only for stream and sequential + packet sockets. See + listen2 + for details. Defaults to SOMAXCONN (128). + + + + BindToDevice= + Specifies a network interface name to bind + this socket to. If set, traffic will only be accepted from the + specified network interfaces. This controls the + SO_BINDTODEVICE socket option (see socket7 + for details). If this option is used, an automatic dependency + from this socket unit on the network interface device unit + (systemd.device5 + is created. Note that setting this parameter might result in + additional dependencies to be added to the unit (see + above). + + + + SocketUser= + SocketGroup= + + Takes a UNIX user/group name. When specified, + all AF_UNIX sockets and FIFO nodes in the file system are + owned by the specified user and group. If unset (the default), + the nodes are owned by the root user/group (if run in system + context) or the invoking user/group (if run in user context). + If only a user is specified but no group, then the group is + derived from the user's default group. + + + + SocketMode= + If listening on a file system socket or FIFO, + this option specifies the file system access mode used when + creating the file node. Takes an access mode in octal + notation. Defaults to 0666. + + + + DirectoryMode= + If listening on a file system socket or FIFO, + the parent directories are automatically created if needed. + This option specifies the file system access mode used when + creating these directories. Takes an access mode in octal + notation. Defaults to 0755. + + + + Accept= + Takes a boolean argument. If true, a service + instance is spawned for each incoming connection and only the + connection socket is passed to it. If false, all listening + sockets themselves are passed to the started service unit, and + only one service unit is spawned for all connections (also see + above). This value is ignored for datagram sockets and FIFOs + where a single service unit unconditionally handles all + incoming traffic. Defaults to . For + performance reasons, it is recommended to write new daemons + only in a way that is suitable for + . A daemon listening on an + AF_UNIX socket may, but does not need to, + call + close2 + on the received socket before exiting. However, it must not + unlink the socket from a file system. It should not invoke + shutdown2 + on sockets it got with Accept=false, but it + may do so for sockets it got with + Accept=true set. Setting + Accept=true is mostly useful to allow + daemons designed for usage with + inetd8 + to work unmodified with systemd socket + activation. + + For IPv4 and IPv6 connections, the REMOTE_ADDR + environment variable will contain the remote IP address, and REMOTE_PORT + will contain the remote port. This is the same as the format used by CGI. + For SOCK_RAW, the port is the IP protocol. + + + + Writable= + Takes a boolean argument. May only be used in + conjunction with ListenSpecial=. If true, + the specified special file is opened in read-write mode, if + false, in read-only mode. Defaults to false. + + + + MaxConnections= + The maximum number of connections to + simultaneously run services instances for, when + is set. If more concurrent + connections are coming in, they will be refused until at least + one existing connection is terminated. This setting has no + effect on sockets configured with + or datagram sockets. Defaults to + 64. + + + + KeepAlive= + Takes a boolean argument. If true, the TCP/IP + stack will send a keep alive message after 2h (depending on + the configuration of + /proc/sys/net/ipv4/tcp_keepalive_time) + for all TCP streams accepted on this socket. This controls the + SO_KEEPALIVE socket option (see + socket7 + and the TCP + Keepalive HOWTO for details.) Defaults to + . + + + + KeepAliveTimeSec= + Takes time (in seconds) as argument. The connection needs to remain + idle before TCP starts sending keepalive probes. This controls the TCP_KEEPIDLE + socket option (see + socket7 + and the TCP + Keepalive HOWTO for details.) + Defaults value is 7200 seconds (2 hours). + + + + KeepAliveIntervalSec= + Takes time (in seconds) as argument between + individual keepalive probes, if the socket option SO_KEEPALIVE + has been set on this socket. This controls + the TCP_KEEPINTVL socket option (see + socket7 + and the TCP + Keepalive HOWTO for details.) Defaults value is 75 + seconds. + + + + KeepAliveProbes= + Takes an integer as argument. It is the number of + unacknowledged probes to send before considering the + connection dead and notifying the application layer. This + controls the TCP_KEEPCNT socket option (see + socket7 + and the TCP + Keepalive HOWTO for details.) Defaults value is + 9. + + + + NoDelay= + Takes a boolean argument. TCP Nagle's + algorithm works by combining a number of small outgoing + messages, and sending them all at once. This controls the + TCP_NODELAY socket option (see + tcp7 + Defaults to . + + + + Priority= + Takes an integer argument controlling the + priority for all traffic sent from this socket. This controls + the SO_PRIORITY socket option (see + socket7 + for details.). + + + + DeferAcceptSec= + + Takes time (in seconds) as argument. If set, + the listening process will be awakened only when data arrives + on the socket, and not immediately when connection is + established. When this option is set, the + TCP_DEFER_ACCEPT socket option will be + used (see + tcp7), + and the kernel will ignore initial ACK packets without any + data. The argument specifies the approximate amount of time + the kernel should wait for incoming data before falling back + to the normal behavior of honouring empty ACK packets. This + option is beneficial for protocols where the client sends the + data first (e.g. HTTP, in contrast to SMTP), because the + server process will not be woken up unnecessarily before it + can take any action. + + + If the client also uses the + TCP_DEFER_ACCEPT option, the latency of + the initial connection may be reduced, because the kernel will + send data in the final packet establishing the connection (the + third packet in the "three-way handshake"). + + Disabled by default. + + + + + ReceiveBuffer= + SendBuffer= + Takes an integer argument controlling the + receive or send buffer sizes of this socket, respectively. + This controls the SO_RCVBUF and SO_SNDBUF socket options (see + socket7 + for details.). The usual suffixes K, M, G are supported and + are understood to the base of 1024. + + + + IPTOS= + Takes an integer argument controlling the IP + Type-Of-Service field for packets generated from this socket. + This controls the IP_TOS socket option (see + ip7 + for details.). Either a numeric string or one of + , , + or may + be specified. + + + + IPTTL= + Takes an integer argument controlling the IPv4 + Time-To-Live/IPv6 Hop-Count field for packets generated from + this socket. This sets the IP_TTL/IPV6_UNICAST_HOPS socket + options (see + ip7 + and + ipv67 + for details.) + + + + Mark= + Takes an integer value. Controls the firewall + mark of packets generated by this socket. This can be used in + the firewall logic to filter packets from this socket. This + sets the SO_MARK socket option. See + iptables8 + for details. + + + + ReusePort= + Takes a boolean value. If true, allows + multiple + bind2s + to this TCP or UDP port. This controls the SO_REUSEPORT socket + option. See + socket7 + for details. + + + + SmackLabel= + SmackLabelIPIn= + SmackLabelIPOut= + Takes a string value. Controls the extended + attributes security.SMACK64, + security.SMACK64IPIN and + security.SMACK64IPOUT, respectively, i.e. + the security label of the FIFO, or the security label for the + incoming or outgoing connections of the socket, respectively. + See Smack.txt + for details. + + + + SELinuxContextFromNet= + Takes a boolean argument. When true, systemd + will attempt to figure out the SELinux label used for the + instantiated service from the information handed by the peer + over the network. Note that only the security level is used + from the information provided by the peer. Other parts of the + resulting SELinux context originate from either the target + binary that is effectively triggered by socket unit or from + the value of the SELinuxContext= option. + This configuration option only affects sockets with + Accept= mode set to + true. Also note that this option is useful + only when MLS/MCS SELinux policy is deployed. Defaults to + false. + + + + PipeSize= + Takes a size in bytes. Controls the pipe + buffer size of FIFOs configured in this socket unit. See + fcntl2 + for details. The usual suffixes K, M, G are supported and are + understood to the base of 1024. + + + + MessageQueueMaxMessages=, + MessageQueueMessageSize= + These two settings take integer values and + control the mq_maxmsg field or the mq_msgsize field, + respectively, when creating the message queue. Note that + either none or both of these variables need to be set. See + mq_setattr3 + for details. + + + + FreeBind= + Takes a boolean value. Controls whether the + socket can be bound to non-local IP addresses. This is useful + to configure sockets listening on specific IP addresses before + those IP addresses are successfully configured on a network + interface. This sets the IP_FREEBIND socket option. For + robustness reasons it is recommended to use this option + whenever you bind a socket to a specific IP address. Defaults + to . + + + + Transparent= + Takes a boolean value. Controls the + IP_TRANSPARENT socket option. Defaults to + . + + + + Broadcast= + Takes a boolean value. This controls the + SO_BROADCAST socket option, which allows broadcast datagrams + to be sent from this socket. Defaults to + . + + + + PassCredentials= + Takes a boolean value. This controls the + SO_PASSCRED socket option, which allows + AF_UNIX sockets to receive the + credentials of the sending process in an ancillary message. + Defaults to . + + + + PassSecurity= + Takes a boolean value. This controls the + SO_PASSSEC socket option, which allows + AF_UNIX sockets to receive the security + context of the sending process in an ancillary message. + Defaults to . + + + + TCPCongestion= + Takes a string value. Controls the TCP + congestion algorithm used by this socket. Should be one of + "westwood", "veno", "cubic", "lp" or any other available + algorithm supported by the IP stack. This setting applies only + to stream sockets. + + + + ExecStartPre= + ExecStartPost= + Takes one or more command lines, which are + executed before or after the listening sockets/FIFOs are + created and bound, respectively. The first token of the + command line must be an absolute filename, then followed by + arguments for the process. Multiple command lines may be + specified following the same scheme as used for + ExecStartPre= of service unit + files. + + + + ExecStopPre= + ExecStopPost= + Additional commands that are executed before + or after the listening sockets/FIFOs are closed and removed, + respectively. Multiple command lines may be specified + following the same scheme as used for + ExecStartPre= of service unit + files. + + + + TimeoutSec= + Configures the time to wait for the commands + specified in ExecStartPre=, + ExecStartPost=, + ExecStopPre= and + ExecStopPost= to finish. If a command does + not exit within the configured time, the socket will be + considered failed and be shut down again. All commands still + running will be terminated forcibly via + SIGTERM, and after another delay of this + time with SIGKILL. (See + in + systemd.kill5.) + Takes a unit-less value in seconds, or a time span value such + as "5min 20s". Pass 0 to disable the + timeout logic. Defaults to + DefaultTimeoutStartSec= from the manager + configuration file (see + systemd-system.conf5). + + + + + Service= + Specifies the service unit name to activate on + incoming traffic. This setting is only allowed for sockets + with Accept=no. It defaults to the service + that bears the same name as the socket (with the suffix + replaced). In most cases, it should not be necessary to use + this option. Note that setting this parameter might result in + additional dependencies to be added to the unit (see + above). + + + + RemoveOnStop= + Takes a boolean argument. If enabled, any file + nodes created by this socket unit are removed when it is + stopped. This applies to AF_UNIX sockets in the file system, + POSIX message queues, FIFOs, as well as any symlinks to them + configured with Symlinks=. Normally, it + should not be necessary to use this option, and is not + recommended as services might continue to run after the socket + unit has been terminated and it should still be possible to + communicate with them via their file system node. Defaults to + off. + + + + Symlinks= + Takes a list of file system paths. The + specified paths will be created as symlinks to the AF_UNIX + socket path or FIFO path of this socket unit. If this setting + is used, only one AF_UNIX socket in the file system or one + FIFO may be configured for the socket unit. Use this option to + manage one or more symlinked alias names for a socket, binding + their lifecycle together. Defaults to the empty + list. + + + + FileDescriptorName= + Assigns a name to all file descriptors this + socket unit encapsulates. This is useful to help activated + services identify specific file descriptors, if multiple fds + are passed. Services may use the + sd_listen_fds_with_names3 + call to acquire the names configured for the received file + descriptors. Names may contain any ASCII character, but must + exclude control characters and :, and must + be at most 255 characters in length. If this setting is not + used, the file descriptor name defaults to the name of the + socket unit, including its .socket + suffix. + + + + TriggerLimitIntervalSec= + TriggerLimitBurst= + + Configures a limit on how often this socket unit my be activated within a specific time + interval. The TriggerLimitIntervalSec= may be used to configure the length of the time + interval in the usual time units us, ms, s, + min, h, … and defaults to 2s (See + systemd.time7 for details on + the various time units understood). The TriggerLimitBurst= setting takes a positive integer + value and specifies the number of permitted activations per time interval, and defaults to 200 for + Accept=yes sockets (thus by default permitting 200 activations per 2s), and 20 otherwise (20 + activations per 2s). Set either to 0 to disable any form of trigger rate limiting. If the limit is hit, the + socket unit is placed into a failure mode, and will not be connectible anymore until restarted. Note that this + limit is enforced before the service activation is enqueued. + + + + + Check + systemd.exec5 + and + systemd.kill5 + for more settings. + + + + + See Also + + systemd1, + systemctl1, + systemd.unit5, + systemd.exec5, + systemd.kill5, + systemd.resource-control5, + systemd.service5, + systemd.directives7, + sd_listen_fds3, + sd_listen_fds_with_names3 + + + For more extensive descriptions see the "systemd for Developers" series: + Socket Activation, + Socket Activation, part II, + Converting inetd Services, + Socket Activated Internet Services and OS Containers. + + + + diff --git a/src/grp-system/systemd/systemd.special.xml b/src/grp-system/systemd/systemd.special.xml new file mode 100644 index 0000000000..18ad8f92e5 --- /dev/null +++ b/src/grp-system/systemd/systemd.special.xml @@ -0,0 +1,941 @@ + + + + + + + + + systemd.special + systemd + + + + Developer + Lennart + Poettering + lennart@poettering.net + + + + + + systemd.special + 7 + + + + systemd.special + Special systemd units + + + + basic.target, + bluetooth.target, + ctrl-alt-del.target, + cryptsetup.target, + cryptsetup-pre.target, + dbus.service, + dbus.socket, + default.target, + display-manager.service, + emergency.target, + exit.target, + final.target, + getty.target, + graphical.target, + halt.target, + hibernate.target, + hybrid-sleep.target, + initrd-fs.target, + kbrequest.target, + kexec.target, + local-fs.target, + local-fs-pre.target, + multi-user.target, + network.target, + network-online.target, + network-pre.target, + nss-lookup.target, + nss-user-lookup.target, + paths.target, + poweroff.target, + printer.target, + reboot.target, + remote-fs.target, + remote-fs-pre.target, + rescue.target, + initrd-root-device.target, + initrd-root-fs.target, + rpcbind.target, + runlevel2.target, + runlevel3.target, + runlevel4.target, + runlevel5.target, + shutdown.target, + sigpwr.target, + sleep.target, + slices.target, + smartcard.target, + sockets.target, + sound.target, + suspend.target, + swap.target, + sysinit.target, + syslog.socket, + system-update.target, + time-sync.target, + timers.target, + umount.target, + -.slice, + system.slice, + user.slice, + machine.slice + + + + Description + + A few units are treated specially by systemd. They have + special internal semantics and cannot be renamed. + + + + Special System Units + + + + basic.target + + A special target unit covering basic boot-up. + + systemd automatically adds dependency of the type + After= for this target unit to all + services (except for those with + DefaultDependencies=no). + + Usually, this should pull-in all local mount points plus + /var, /tmp and + /var/tmp, swap devices, sockets, timers, + path units and other basic initialization necessary for general + purpose daemons. The mentioned mount points are special cased + to allow them to be remote. + + + This target usually does not pull in any non-target units + directly, but rather does so indirectly via other early boot targets. + It is instead meant as a synchronization point for late boot + services. Refer to + bootup7 + for details on the targets involved. + + + + + + ctrl-alt-del.target + + systemd starts this target whenever Control+Alt+Del is + pressed on the console. Usually, this should be aliased + (symlinked) to reboot.target. + + + + cryptsetup.target + + A target that pulls in setup services for all + encrypted block devices. + + + + dbus.service + + A special unit for the D-Bus bus daemon. As soon as + this service is fully started up systemd will connect to it + and register its service. + + + + dbus.socket + + A special unit for the D-Bus system bus socket. All + units with Type=dbus automatically gain a + dependency on this unit. + + + + default.target + + The default unit systemd starts at bootup. Usually, + this should be aliased (symlinked) to + multi-user.target or + graphical.target. + + The default unit systemd starts at bootup can be + overridden with the systemd.unit= kernel + command line option. + + + + display-manager.service + + The display manager service. Usually, this should be + aliased (symlinked) to gdm.service or a + similar display manager service. + + + + emergency.target + + A special target unit that starts an emergency shell on the main console. This target does not pull in + any services or mounts. It is the most minimal version of starting the system in order to acquire an + interactive shell; the only processes running are usually just the system manager (PID 1) and the shell + process. This unit is supposed to be used with the kernel command line option + systemd.unit=; it is also used when a file system check on a required file system fails, + and boot-up cannot continue. Compare with rescue.target, which serves a similar purpose, + but also starts the most basic services and mounts all file systems. + + Use the systemd.unit=emergency.target kernel command line option to boot into this + mode. A short alias for this kernel command line option is emergency, for compatibility + with SysV. + + In many ways booting into emergency.target is similar to the effect of booting + with init=/bin/sh on the kernel command line, except that emergency mode provides you with + the full system and service manager, and allows starting individual units in order to continue the boot + process in steps. + + + + exit.target + + A special service unit for shutting down the system or + user service manager. It is equivalent to + poweroff.target on non-container + systems, and also works in containers. + + systemd will start this unit when it receives a + request to shut down over D-Bus or a + SIGTERM or SIGINT + signal when running as user service daemon. + + Normally, this (indirectly) pulls in + shutdown.target, which in turn should be + conflicted by all units that want to be scheduled for + shutdown when the service manager starts to exit. + + + + final.target + + A special target unit that is used during the shutdown + logic and may be used to pull in late services after all + normal services are already terminated and all mounts + unmounted. + + + + + getty.target + + A special target unit that pulls in statically + configured local TTY getty instances. + + + + + graphical.target + + A special target unit for setting up a graphical login + screen. This pulls in + multi-user.target. + + Units that are needed for graphical logins shall add + Wants= dependencies for their unit to + this unit (or multi-user.target) during + installation. This is best configured via + WantedBy=graphical.target in the unit's + [Install] section. + + + + hibernate.target + + A special target unit for hibernating the system. This + pulls in sleep.target. + + + + hybrid-sleep.target + + A special target unit for hibernating and suspending + the system at the same time. This pulls in + sleep.target. + + + + halt.target + + A special target unit for shutting down and halting + the system. Note that this target is distinct from + poweroff.target in that it generally + really just halts the system rather than powering it + down. + + Applications wanting to halt the system should start + this unit. + + + + initrd-fs.target + + systemd-fstab-generator3 + automatically adds dependencies of type + Before= to + sysroot-usr.mount and all mount points + found in /etc/fstab that have + and not have + mount options set. + + + + kbrequest.target + + systemd starts this target whenever Alt+ArrowUp is + pressed on the console. This is a good candidate to be + aliased (symlinked) to + rescue.target. + + + + kexec.target + + A special target unit for shutting down and rebooting + the system via kexec. + + Applications wanting to reboot the system with kexec + should start this unit. + + + + local-fs.target + + systemd-fstab-generator3 + automatically adds dependencies of type + Before= to all mount units that refer to + local mount points for this target unit. In addition, it + adds dependencies of type Wants= to this + target unit for those mounts listed in + /etc/fstab that have the + mount option set. + + + + multi-user.target + + A special target unit for setting up a multi-user + system (non-graphical). This is pulled in by + graphical.target. + + Units that are needed for a multi-user system shall + add Wants= dependencies for their unit to + this unit during installation. This is best configured via + WantedBy=multi-user.target in the unit's + [Install] section. + + + + network-online.target + + Units that strictly require a configured network + connection should pull in + network-online.target (via a + Wants= type dependency) and order + themselves after it. This target unit is intended to pull in + a service that delays further execution until the network is + sufficiently set up. What precisely this requires is left to + the implementation of the network managing service. + + Note the distinction between this unit and + network.target. This unit is an active + unit (i.e. pulled in by the consumer rather than the + provider of this functionality) and pulls in a service which + possibly adds substantial delays to further execution. In + contrast, network.target is a passive + unit (i.e. pulled in by the provider of the functionality, + rather than the consumer) that usually does not delay + execution much. Usually, network.target + is part of the boot of most systems, while + network-online.target is not, except + when at least one unit requires it. Also see Running + Services After the Network is up for more + information. + + All mount units for remote network file systems + automatically pull in this unit, and order themselves after + it. Note that networking daemons that simply provide + functionality to other hosts generally do not need to pull + this in. + + + + paths.target + + A special target unit that sets up all path units (see + systemd.path5 + for details) that shall be active after boot. + + It is recommended that path units installed by + applications get pulled in via Wants= + dependencies from this unit. This is best configured via a + WantedBy=paths.target in the path unit's + [Install] section. + + + + poweroff.target + + A special target unit for shutting down and powering + off the system. + + Applications wanting to power off the system should + start this unit. + + runlevel0.target is an alias for + this target unit, for compatibility with SysV. + + + + reboot.target + + A special target unit for shutting down and rebooting + the system. + + Applications wanting to reboot the system should start + this unit. + + runlevel6.target is an alias for + this target unit, for compatibility with SysV. + + + + remote-fs.target + + Similar to local-fs.target, but + for remote mount points. + + systemd automatically adds dependencies of type + After= for this target unit to all SysV + init script service units with an LSB header referring to + the $remote_fs facility. + + + + rescue.target + + A special target unit that pulls in the base system (including system mounts) and spawns a rescue + shell. Isolate to this target in order to administer the system in single-user mode with all file systems + mounted but with no services running, except for the most basic. Compare with + emergency.target, which is much more reduced and does not provide the file systems or + most basic services. + + runlevel1.target is an alias for this target unit, for compatibility with + SysV. + + Use the systemd.unit=rescue.target kernel command line option to boot into this + mode. A short alias for this kernel command line option is 1, for compatibility with + SysV. + + + + initrd-root-device.target + + A special initrd target unit that is reached when the root filesystem device is available, but before + it has been mounted. + systemd-fstab-generator3 + and + systemd-gpt-auto-generator3 + automatically setup the appropriate dependencies to make this happen. + + + + + initrd-root-fs.target + + systemd-fstab-generator3 + automatically adds dependencies of type + Before= to the + sysroot.mount unit, which is generated + from the kernel command line. + + + + + runlevel2.target + runlevel3.target + runlevel4.target + runlevel5.target + + These are targets that are called whenever the SysV + compatibility code asks for runlevel 2, 3, 4, 5, + respectively. It is a good idea to make this an alias for + (i.e. symlink to) graphical.target + (for runlevel 5) or multi-user.target + (the others). + + + + shutdown.target + + A special target unit that terminates the services on + system shutdown. + + Services that shall be terminated on system shutdown + shall add Conflicts= and + Before= dependencies to this unit for + their service unit, which is implicitly done when + DefaultDependencies=yes is set (the + default). + + + + sigpwr.target + + A special target that is started when systemd receives + the SIGPWR process signal, which is normally sent by the + kernel or UPS daemons when power fails. + + + + sleep.target + + A special target unit that is pulled in by + suspend.target, + hibernate.target and + hybrid-sleep.target and may be used to + hook units into the sleep state logic. + + + + slices.target + + A special target unit that sets up all slice units (see + systemd.slice5 for + details) that shall be active after boot. By default the generic user.slice, + system.slice, machines.slice slice units, as well as the root + slice unit -.slice are pulled in and ordered before this unit (see below). + + It's a good idea to add WantedBy=slices.target lines to the [Install] + section of all slices units that may be installed dynamically. + + + + sockets.target + + A special target unit that sets up all socket + units (see + systemd.socket5 + for details) that shall be active after boot. + + Services that can be socket-activated shall add + Wants= dependencies to this unit for + their socket unit during installation. This is best + configured via a WantedBy=sockets.target + in the socket unit's [Install] + section. + + + + suspend.target + + A special target unit for suspending the system. This + pulls in sleep.target. + + + + swap.target + + Similar to local-fs.target, but + for swap partitions and swap files. + + + + sysinit.target + + systemd automatically adds dependencies of the types + Requires= and After= + for this target unit to all services (except for those with + DefaultDependencies=no). + + This target pulls in the services required for system + initialization. System services pulled in by this target should + declare DefaultDependencies=no and specify + all their dependencies manually, including access to anything + more than a read only root filesystem. For details on the + dependencies of this target, refer to + bootup7. + + + + + syslog.socket + + The socket unit syslog implementations should listen + on. All userspace log messages will be made available on + this socket. For more information about syslog integration, + please consult the Syslog + Interface document. + + + + system-update.target + + A special target unit that is used for off-line system + updates. + systemd-system-update-generator8 + will redirect the boot process to this target if + /system-update exists. For more + information see the System + Updates Specification. + + + + timers.target + + A special target unit that sets up all timer units + (see + systemd.timer5 + for details) that shall be active after boot. + + It is recommended that timer units installed by + applications get pulled in via Wants= + dependencies from this unit. This is best configured via + WantedBy=timers.target in the timer + unit's [Install] section. + + + + umount.target + + A special target unit that unmounts all mount and + automount points on system shutdown. + + Mounts that shall be unmounted on system shutdown + shall add Conflicts dependencies to this unit for their + mount unit, which is implicitly done when + DefaultDependencies=yes is set (the + default). + + + + + + + + Special System Units for Devices + + Some target units are automatically pulled in as devices of + certain kinds show up in the system. These may be used to + automatically activate various services based on the specific type + of the available hardware. + + + + bluetooth.target + + This target is started automatically as soon as a + Bluetooth controller is plugged in or becomes available at + boot. + + This may be used to pull in Bluetooth management + daemons dynamically when Bluetooth hardware is found. + + + + printer.target + + This target is started automatically as soon as a + printer is plugged in or becomes available at boot. + + This may be used to pull in printer management daemons + dynamically when printer hardware is found. + + + + smartcard.target + + This target is started automatically as soon as a + smartcard controller is plugged in or becomes available at + boot. + + This may be used to pull in smartcard management + daemons dynamically when smartcard hardware is found. + + + + sound.target + + This target is started automatically as soon as a + sound card is plugged in or becomes available at + boot. + + This may be used to pull in audio management daemons + dynamically when audio hardware is found. + + + + + + + Special Passive System Units + + A number of special system targets are defined that can be + used to properly order boot-up of optional services. These targets + are generally not part of the initial boot transaction, unless + they are explicitly pulled in by one of the implementing services. + Note specifically that these passive target + units are generally not pulled in by the consumer of a service, + but by the provider of the service. This means: a consuming + service should order itself after these targets (as appropriate), + but not pull it in. A providing service should order itself before + these targets (as appropriate) and pull it in (via a + Wants= type dependency). + + Note that these passive units cannot be started manually, + i.e. systemctl start time-sync.target will fail + with an error. They can only be pulled in by dependency. This is + enforced since they exist for ordering purposes only and thus are + not useful as only unit within a transaction. + + + + cryptsetup-pre.target + + This passive target unit may be pulled in by services + that want to run before any encrypted block device is set + up. All encrypted block devices are set up after this target + has been reached. Since the shutdown order is implicitly the + reverse start-up order between units, this target is + particularly useful to ensure that a service is shut down + only after all encrypted block devices are fully + stopped. + + + + local-fs-pre.target + + This target unit is + automatically ordered before + all local mount points marked + with + (see above). It can be used to + execute certain units before + all local mounts. + + + + network.target + + This unit is supposed to indicate when network + functionality is available, but it is only very weakly + defined what that is supposed to mean, with one exception: + at shutdown, a unit that is ordered after + network.target will be stopped before + the network — to whatever level it might be set up then — + is shut down. It is hence useful when writing service files + that require network access on shutdown, which should order + themselves after this target, but not pull it in. Also see + Running + Services After the Network is up for more + information. Also see + network-online.target described + above. + + systemd automatically adds dependencies of type + After= for this target unit to all SysV + init script service units with an LSB header referring to + the $network facility. + + + + network-pre.target + + This passive target unit may be pulled in by services + that want to run before any network is set up, for example + for the purpose of setting up a firewall. All network + management software orders itself after this target, but + does not pull it in. + + + + nss-lookup.target + + A target that should be used as synchronization point + for all host/network name service lookups. Note that this is + independent of user/group name lookups for which + nss-user-lookup.target should be used. + All services for which the availability of full host/network + name resolution is essential should be ordered after this + target, but not pull it in. systemd automatically adds + dependencies of type After= for this + target unit to all SysV init script service units with an + LSB header referring to the $named + facility. + + + + nss-user-lookup.target + + A target that should be used as synchronization point + for all user/group name service lookups. Note that this is + independent of host/network name lookups for which + nss-lookup.target should be used. All + services for which the availability of the full user/group + database is essential should be ordered after this target, + but not pull it in. Note that system users are always + resolvable, and hence do not require any special ordering + against this target. + + + + remote-fs-pre.target + + This target unit is automatically ordered before all + remote mount point units (see above). It can be used to run + certain units before the remote mounts are established. Note + that this unit is generally not part of the initial + transaction, unless the unit that wants to be ordered before + all remote mounts pulls it in via a + Wants= type dependency. If the unit wants + to be pulled in by the first remote mount showing up, it + should use network-online.target (see + above). + + + + rpcbind.target + + The portmapper/rpcbind pulls in this target and orders + itself before it, to indicate its availability. systemd + automatically adds dependencies of type + After= for this target unit to all SysV + init script service units with an LSB header referring to + the $portmap facility. + + + + time-sync.target + + Services responsible for synchronizing the system + clock from a remote source (such as NTP client + implementations) should pull in this target and order + themselves before it. All services where correct time is + essential should be ordered after this unit, but not pull it + in. systemd automatically adds dependencies of type + After= for this target unit to all SysV + init script service units with an LSB header referring to + the $time facility. + + + + + + + Special User Units + + When systemd runs as a user instance, the following special + units are available, which have similar definitions as their + system counterparts: + exit.target, + default.target, + shutdown.target, + sockets.target, + timers.target, + paths.target, + bluetooth.target, + printer.target, + smartcard.target, + sound.target. + + + + Special Slice Units + + There are four .slice units which form + the basis of the hierarchy for assignment of resources for + services, users, and virtual machines or containers. + + + + -.slice + + The root slice is the root of the hierarchy. It + usually does not contain units directly, but may be used to + set defaults for the whole tree. + + + + + system.slice + + By default, all system services started by + systemd are found in this slice. + + + + + user.slice + + By default, all user processes and services started on + behalf of the user, including the per-user systemd instance + are found in this slice. + + + + + machine.slice + + By default, all virtual machines and containers + registered with systemd-machined are + found in this slice. + + + + + + + + See Also + + systemd1, + systemd.unit5, + systemd.service5, + systemd.socket5, + systemd.target5, + systemd.slice5, + bootup7, + systemd-fstab-generator8 + + + + diff --git a/src/grp-system/systemd/systemd.swap.xml b/src/grp-system/systemd/systemd.swap.xml new file mode 100644 index 0000000000..cf4e1ba839 --- /dev/null +++ b/src/grp-system/systemd/systemd.swap.xml @@ -0,0 +1,250 @@ + + + + + + + + systemd.swap + systemd + + + + Developer + Lennart + Poettering + lennart@poettering.net + + + + + + systemd.swap + 5 + + + + systemd.swap + Swap unit configuration + + + + swap.swap + + + + Description + + A unit configuration file whose name ends in + .swap encodes information about a swap device + or file for memory paging controlled and supervised by + systemd. + + This man page lists the configuration options specific to + this unit type. See + systemd.unit5 + for the common options of all unit configuration files. The common + configuration items are configured in the generic [Unit] and + [Install] sections. The swap specific configuration options are + configured in the [Swap] section. + + Additional options are listed in + systemd.exec5, + which define the execution environment the swapon8 + binary is executed in, in + systemd.kill5, + which define the way these processes are + terminated, and in + systemd.resource-control5, + which configure resource control settings for these processes of the + unit. + + Swap units must be named after the devices or files they control. Example: the swap device /dev/sda5 must be configured in a unit file dev-sda5.swap. For + details about the escaping logic used to convert a file system path to a unit name, see + systemd.unit5. Note that swap + units cannot be templated, nor is possible to add multiple names to a swap unit by creating additional symlinks to + it. + + + + Automatic Dependencies + + All swap units automatically get the + BindsTo= and After= + dependencies on the device units or the mount units of the files + they are activated from. + + Swap units with DefaultDependencies= in the [Unit] section enabled + implicitly acquire a Conflicts= and an After= dependency on + umount.target so that they are deactivated at shutdown, unless + DefaultDependencies=no is specified. + + Additional implicit dependencies may be added as result of + execution and resource control parameters as documented in + systemd.exec5 + and + systemd.resource-control5. + + + + <filename>fstab</filename> + + Swap units may either be configured via unit files, or via + /etc/fstab (see + fstab5 + for details). Swaps listed in /etc/fstab will + be converted into native units dynamically at boot and when the + configuration of the system manager is reloaded. See + systemd-fstab-generator8 + for details about the conversion. + + If a swap device or file is configured in both + /etc/fstab and a unit file, the configuration + in the latter takes precedence. + + When reading /etc/fstab, a few special + options are understood by systemd which influence how dependencies + are created for swap units. + + + + + + + With , the swap unit + will not be added as a dependency for + swap.target. This means that it will not + be activated automatically during boot, unless it is pulled in + by some other unit. The option has the + opposite meaning and is the default. + + + + + + + With , the swap unit + will be only wanted, not required by + swap.target. This means that the boot + will continue even if this swap device is not activated + successfully. + + + + + + + Options + + Swap files must include a [Swap] section, which carries + information about the swap device it supervises. A number of + options that may be used in this section are shared with other + unit types. These options are documented in + systemd.exec5 + and + systemd.kill5. + The options specific to the [Swap] section of swap units are the + following: + + + + + What= + Takes an absolute path of a device node or + file to use for paging. See + swapon8 + for details. If this refers to a device node, a dependency on + the respective device unit is automatically created. (See + systemd.device5 + for more information.) If this refers to a file, a dependency + on the respective mount unit is automatically created. (See + systemd.mount5 + for more information.) This option is + mandatory. + + + + Priority= + + Swap priority to use when activating the swap + device or file. This takes an integer. This setting is + optional and ignored when the priority is set by in the + Options= key. + + + + Options= + + May contain an option string for the swap + device. This may be used for controlling discard options among + other functionality, if the swap backing device supports the + discard or trim operation. (See + swapon8 + for more information.) + + + + TimeoutSec= + Configures the time to wait for the swapon + command to finish. If a command does not exit within the + configured time, the swap will be considered failed and be + shut down again. All commands still running will be terminated + forcibly via SIGTERM, and after another + delay of this time with SIGKILL. (See + in + systemd.kill5.) + Takes a unit-less value in seconds, or a time span value such + as "5min 20s". Pass 0 to disable the + timeout logic. Defaults to + DefaultTimeoutStartSec= from the manager + configuration file (see + systemd-system.conf5). + + + + + Check + systemd.exec5 + and + systemd.kill5 + for more settings. + + + + See Also + + systemd1, + systemctl1, + systemd.unit5, + systemd.exec5, + systemd.kill5, + systemd.resource-control5, + systemd.device5, + systemd.mount5, + swapon8, + systemd-fstab-generator8, + systemd.directives7 + + + + diff --git a/src/grp-system/systemd/systemd.target.xml b/src/grp-system/systemd/systemd.target.xml new file mode 100644 index 0000000000..2e35e54fc4 --- /dev/null +++ b/src/grp-system/systemd/systemd.target.xml @@ -0,0 +1,112 @@ + + + + + + + + systemd.target + systemd + + + + Developer + Lennart + Poettering + lennart@poettering.net + + + + + + systemd.target + 5 + + + + systemd.target + Target unit configuration + + + + target.target + + + + Description + + A unit configuration file whose name ends in + .target encodes information about a target unit + of systemd, which is used for grouping units and as well-known + synchronization points during start-up. + + This unit type has no specific options. See + systemd.unit5 + for the common options of all unit configuration files. The common + configuration items are configured in the generic [Unit] and + [Install] sections. A separate [Target] section does not exist, + since no target-specific options may be configured. + + Target units do not offer any additional functionality on + top of the generic functionality provided by units. They exist + merely to group units via dependencies (useful as boot targets), + and to establish standardized names for synchronization points + used in dependencies between units. Among other things, target + units are a more flexible replacement for SysV runlevels in the + classic SysV init system. (And for compatibility reasons special + target units such as runlevel3.target exist + which are used by the SysV runlevel compatibility code in systemd. + See + systemd.special7 + for details). + + + + Automatic Dependencies + + Unless DefaultDependencies= is set to + in either of releated units or an explicit ordering + dependency is already defined, target units will implicitly complement all + configured dependencies of type Wants= or + Requires= with dependencies of type + After=. Note that Wants= or + Requires= must be defined in the target unit itself — if + you for example define Wants=some.target in + some.service, the implicit ordering will not be added. + + All target units automatically gain Conflicts= + dependency against shutdown.target unless DefaultDependencies= + is set to . + + + + + See Also + + systemd1, + systemctl1, + systemd.unit5, + systemd.special7, + systemd.directives7 + + + + diff --git a/src/grp-system/systemd/systemd.time.xml b/src/grp-system/systemd/systemd.time.xml new file mode 100644 index 0000000000..aae3accb6c --- /dev/null +++ b/src/grp-system/systemd/systemd.time.xml @@ -0,0 +1,316 @@ + + + + + + + + + systemd.time + systemd + + + + Developer + Lennart + Poettering + lennart@poettering.net + + + + + + systemd.time + 7 + + + + systemd.time + Time and date specifications + + + + Description + + In systemd, timestamps, time spans, and calendar events are + displayed and may be specified in closely related syntaxes. + + + + Displaying Time Spans + + Time spans refer to time durations. On display, systemd will + present time spans as a space-separated series of time values each + suffixed by a time unit. + + 2h 30min + + All specified time values are meant to be added up. The + above hence refers to 150 minutes. + + + + Parsing Time Spans + + When parsing, systemd will accept the same time span syntax. + Separating spaces may be omitted. The following time units are + understood: + + + usec, us + msec, ms + seconds, second, sec, s + minutes, minute, min, m + hours, hour, hr, h + days, day, d + weeks, week, w + months, month, M (defined as 30.44 days) + years, year, y (define as 365.25 days) + + + If no time unit is specified, generally seconds are assumed, + but some exceptions exist and are marked as such. In a few cases + ns, nsec is accepted too, + where the granularity of the time span allows for this. + + Examples for valid time span specifications: + + 2 h +2hours +48hr +1y 12month +55s500ms +300ms20s 5day + + + + Displaying Timestamps + + Timestamps refer to specific, unique points in time. On + display, systemd will format these in the local timezone as + follows: + + Fri 2012-11-23 23:02:15 CET + + The weekday is printed according to the locale choice of the + user. + + + + Parsing Timestamps + + When parsing, systemd will accept a similar syntax, but + expects no timezone specification, unless it is given as the + literal string "UTC". In this case, the time is considered in UTC, + otherwise in the local timezone. The weekday specification is + optional, but when the weekday is specified, it must either be in + the abbreviated (Wed) or non-abbreviated + (Wednesday) English language form (case does + not matter), and is not subject to the locale choice of the user. + Either the date, or the time part may be omitted, in which case + the current date or 00:00:00, respectively, is assumed. The seconds + component of the time may also be omitted, in which case ":00" is + assumed. Year numbers may be specified in full or may be + abbreviated (omitting the century). + + A timestamp is considered invalid if a weekday is specified + and the date does not actually match the specified day of the + week. + + When parsing, systemd will also accept a few special + placeholders instead of timestamps: now may be + used to refer to the current time (or of the invocation of the + command that is currently executed). today, + yesterday, and tomorrow refer to + 00:00:00 of the current day, the day before, or the next day, + respectively. + + When parsing, systemd will also accept relative time + specifications. A time span (see above) that is prefixed with + + is evaluated to the current time plus the + specified time span. Correspondingly, a time span that is prefixed + with - is evaluated to the current time minus + the specified time span. Instead of prefixing the time span with + + or -, it may also be + suffixed with a space and the word left or + ago. + + Finally, a timespan prefixed with @ is + evaluated relative to the UNIX time epoch 1st Jan, 1970, + 00:00. + + Examples for valid timestamps and their normalized form + (assuming the current time was 2012-11-23 18:15:22 and the timezone + was UTC+8, for example TZ=Asia/Shanghai): + + Fri 2012-11-23 11:12:13 → Fri 2012-11-23 11:12:13 + 2012-11-23 11:12:13 → Fri 2012-11-23 11:12:13 +2012-11-23 11:12:13 UTC → Fri 2012-11-23 19:12:13 + 2012-11-23 → Fri 2012-11-23 00:00:00 + 12-11-23 → Fri 2012-11-23 00:00:00 + 11:12:13 → Fri 2012-11-23 11:12:13 + 11:12:13.9900009 → Fri 2012-11-23 11:12:13 + format_timestamp_us: Fri 2012-11-23 11:12:13.990000 + 11:12 → Fri 2012-11-23 11:12:00 + now → Fri 2012-11-23 18:15:22 + today → Fri 2012-11-23 00:00:00 + today UTC → Fri 2012-11-23 16:00:00 + yesterday → Fri 2012-11-22 00:00:00 + tomorrow → Fri 2012-11-24 00:00:00 + +3h30min → Fri 2012-11-23 21:45:22 + +3h30min UTC → -EINVAL + -5s → Fri 2012-11-23 18:15:17 + 11min ago → Fri 2012-11-23 18:04:22 + 11min ago UTC → -EINVAL + @1395716396 → Tue 2014-03-25 03:59:56 + + Note that timestamps printed by systemd will not be parsed + correctly by systemd, as the timezone specification is not + accepted, and printing timestamps is subject to locale settings + for the weekday, while parsing only accepts English weekday + names. + + In some cases, systemd will display a relative timestamp + (relative to the current time, or the time of invocation of the + command) instead or in addition to an absolute timestamp as + described above. A relative timestamp is formatted as + follows: + + 2 months 5 days ago + + Note that any relative timestamp will also parse correctly + where a timestamp is expected. (see above) + + + + Calendar Events + + Calendar events may be used to refer to one or more points + in time in a single expression. They form a superset of the + absolute timestamps explained above: + + Thu,Fri 2012-*-1,5 11:12:13 + + The above refers to 11:12:13 of the first or fifth day of + any month of the year 2012, but only if that day is a Thursday or + Friday. + + The weekday specification is optional. If specified, it + should consist of one or more English language weekday names, + either in the abbreviated (Wed) or non-abbreviated (Wednesday) + form (case does not matter), separated by commas. Specifying two + weekdays separated by .. refers to a range of + continuous weekdays. , and .. + may be combined freely. + + In the date and time specifications, any component may be + specified as * in which case any value will + match. Alternatively, each component can be specified as a list of + values separated by commas. Values may also be suffixed with + / and a repetition value, which indicates that + the value itself and the value plus all multiples of the repetition value + are matched. Each component may also contain a range of values + separated by ... + + The seconds component may contain decimal fractions both in + the value and the repetition. All fractions are rounded to 6 + decimal places. + + Either time or date specification may be omitted, in which + case the current day and 00:00:00 is implied, respectively. If the + second component is not specified, :00 is + assumed. + + A timezone specification is not expected, unless it is given + as the literal string "UTC", similarly to timestamps. + + The special expressions + minutely, + hourly, daily, + monthly, weekly, + yearly, + quarterly, + semiannually may be used as + calendar events which refer to + *-*-* *:*:00, + *-*-* *:00:00, + *-*-* 00:00:00, + *-*-01 00:00:00, + Mon *-*-* 00:00:00, + *-01-01 00:00:00, + *-01,04,07,10-01 00:00:00 and + *-01,07-01 00:00:00, respectively. + + + Examples for valid timestamps and their + normalized form: + + Sat,Thu,Mon..Wed,Sat..Sun → Mon..Thu,Sat,Sun *-*-* 00:00:00 + Mon,Sun 12-*-* 2,1:23 → Mon,Sun 2012-*-* 01,02:23:00 + Wed *-1 → Wed *-*-01 00:00:00 + Wed..Wed,Wed *-1 → Wed *-*-01 00:00:00 + Wed, 17:48 → Wed *-*-* 17:48:00 +Wed..Sat,Tue 12-10-15 1:2:3 → Tue..Sat 2012-10-15 01:02:03 + *-*-7 0:0:0 → *-*-07 00:00:00 + 10-15 → *-10-15 00:00:00 + monday *-12-* 17:00 → Mon *-12-* 17:00:00 + Mon,Fri *-*-3,1,2 *:30:45 → Mon,Fri *-*-01,02,03 *:30:45 + 12,14,13,12:20,10,30 → *-*-* 12,13,14:10,20,30:00 + 12..14:10,20,30 → *-*-* 12,13,14:10,20,30:00 + mon,fri *-1/2-1,3 *:30:45 → Mon,Fri *-01/2-01,03 *:30:45 + 03-05 08:05:40 → *-03-05 08:05:40 + 08:05:40 → *-*-* 08:05:40 + 05:40 → *-*-* 05:40:00 + Sat,Sun 12-05 08:05:40 → Sat,Sun *-12-05 08:05:40 + Sat,Sun 08:05:40 → Sat,Sun *-*-* 08:05:40 + 2003-03-05 05:40 → 2003-03-05 05:40:00 +05:40:23.4200004/3.1700005 → 05:40:23.420000/3.170001 + 2003-02..04-05 → 2003-02,03,04-05 00:00:00 + 2003-03-05 05:40 UTC → 2003-03-05 05:40:00 UTC + 2003-03-05 → 2003-03-05 00:00:00 + 03-05 → *-03-05 00:00:00 + hourly → *-*-* *:00:00 + daily → *-*-* 00:00:00 + daily UTC → *-*-* 00:00:00 UTC + monthly → *-*-01 00:00:00 + weekly → Mon *-*-* 00:00:00 + yearly → *-01-01 00:00:00 + annually → *-01-01 00:00:00 + *:2/3 → *-*-* *:02/3:00 + + Calendar events are used by timer units, see + systemd.timer5 + for details. + + + + + See Also + + systemd1, + journalctl1, + systemd.timer5, + systemd.unit5, + systemd.directives7 + + + + diff --git a/src/grp-system/systemd/systemd.timer.xml b/src/grp-system/systemd/systemd.timer.xml new file mode 100644 index 0000000000..4fe140e4bc --- /dev/null +++ b/src/grp-system/systemd/systemd.timer.xml @@ -0,0 +1,313 @@ + + + + + + + + systemd.timer + systemd + + + + Developer + Lennart + Poettering + lennart@poettering.net + + + + + + systemd.timer + 5 + + + + systemd.timer + Timer unit configuration + + + + timer.timer + + + + Description + + A unit configuration file whose name ends in + .timer encodes information about a timer + controlled and supervised by systemd, for timer-based + activation. + + This man page lists the configuration options specific to + this unit type. See + systemd.unit5 + for the common options of all unit configuration files. The common + configuration items are configured in the generic [Unit] and + [Install] sections. The timer specific configuration options are + configured in the [Timer] section. + + For each timer file, a matching unit file must exist, + describing the unit to activate when the timer elapses. By + default, a service by the same name as the timer (except for the + suffix) is activated. Example: a timer file + foo.timer activates a matching service + foo.service. The unit to activate may be + controlled by Unit= (see below). + + Note that in case the unit to activate is already active at the time the timer elapses it is not restarted, + but simply left running. There is no concept of spawning new service instances in this case. Due to this, services + with RemainAfterExit= set (which stay around continuously even after the service's main process + exited) are usually not suitable for activation via repetitive timers, as they will only be activated once, and + then stay around forever. + + + + Automatic Dependencies + + Timer units automatically gain a Before= + dependency on the service they are supposed to activate. + + Unless DefaultDependencies= in the [Unit] section is set to + , all timer units will implicitly have dependencies of type Requires= and + After= on sysinit.target, a dependency of type Before= + on timers.target, as well as Conflicts= and Before= on + shutdown.target to ensure that they are stopped cleanly prior to system shutdown. Timer units + with at least one OnCalendar= directive will have an additional After= + dependency on timer-sync.target to avoid being started before the system clock has been + correctly set. Only timer units involved with early boot or late system shutdown should disable the + DefaultDependencies= option. + + + + Options + + Timer files must include a [Timer] section, which carries + information about the timer it defines. The options specific to + the [Timer] section of timer units are the following: + + + + OnActiveSec= + OnBootSec= + OnStartupSec= + OnUnitActiveSec= + OnUnitInactiveSec= + + Defines monotonic timers relative to different + starting points: OnActiveSec= defines a + timer relative to the moment the timer itself is activated. + OnBootSec= defines a timer relative to when + the machine was booted up. OnStartupSec= + defines a timer relative to when systemd was first started. + OnUnitActiveSec= defines a timer relative + to when the unit the timer is activating was last activated. + OnUnitInactiveSec= defines a timer relative + to when the unit the timer is activating was last + deactivated. + + Multiple directives may be combined of the same and of + different types. For example, by combining + OnBootSec= and + OnUnitActiveSec=, it is possible to define + a timer that elapses in regular intervals and activates a + specific service each time. + + The arguments to the directives are time spans + configured in seconds. Example: "OnBootSec=50" means 50s after + boot-up. The argument may also include time units. Example: + "OnBootSec=5h 30min" means 5 hours and 30 minutes after + boot-up. For details about the syntax of time spans, see + systemd.time7. + + If a timer configured with OnBootSec= + or OnStartupSec= is already in the past + when the timer unit is activated, it will immediately elapse + and the configured unit is started. This is not the case for + timers defined in the other directives. + + These are monotonic timers, independent of wall-clock + time and timezones. If the computer is temporarily suspended, + the monotonic clock stops too. + + If the empty string is assigned to any of these options, + the list of timers is reset, and all prior assignments will + have no effect. + + Note that timers do not necessarily expire at the + precise time configured with these settings, as they are + subject to the AccuracySec= setting + below. + + + + + OnCalendar= + + Defines realtime (i.e. wallclock) timers with + calendar event expressions. See + systemd.time7 + for more information on the syntax of calendar event + expressions. Otherwise, the semantics are similar to + OnActiveSec= and related settings. + + Note that timers do not necessarily expire at the + precise time configured with this setting, as it is subject to + the AccuracySec= setting + below. + + + + AccuracySec= + + Specify the accuracy the timer shall elapse + with. Defaults to 1min. The timer is scheduled to elapse + within a time window starting with the time specified in + OnCalendar=, + OnActiveSec=, + OnBootSec=, + OnStartupSec=, + OnUnitActiveSec= or + OnUnitInactiveSec= and ending the time + configured with AccuracySec= later. Within + this time window, the expiry time will be placed at a + host-specific, randomized, but stable position that is + synchronized between all local timer units. This is done in + order to optimize power consumption to suppress unnecessary + CPU wake-ups. To get best accuracy, set this option to + 1us. Note that the timer is still subject to the timer slack + configured via + systemd-system.conf5's + TimerSlackNSec= setting. See + prctl2 + for details. To optimize power consumption, make sure to set + this value as high as possible and as low as + necessary. + + + + RandomizedDelaySec= + + Delay the timer by a randomly selected, evenly + distributed amount of time between 0 and the specified time + value. Defaults to 0, indicating that no randomized delay + shall be applied. Each timer unit will determine this delay + randomly each time it is started, and the delay will simply be + added on top of the next determined elapsing time. This is + useful to stretch dispatching of similarly configured timer + events over a certain amount time, to avoid that they all fire + at the same time, possibly resulting in resource + congestion. Note the relation to + AccuracySec= above: the latter allows the + service manager to coalesce timer events within a specified + time range in order to minimize wakeups, the former does the + opposite: it stretches timer events over a time range, to make + it unlikely that they fire simultaneously. If + RandomizedDelaySec= and + AccuracySec= are used in conjunction, first + the randomized delay is added, and then the result is + possibly further shifted to coalesce it with other timer + events happening on the system. As mentioned above + AccuracySec= defaults to 1min and + RandomizedDelaySec= to 0, thus encouraging + coalescing of timer events. In order to optimally stretch + timer events over a certain range of time, make sure to set + RandomizedDelaySec= to a higher value, and + AccuracySec=1us. + + + + Unit= + + The unit to activate when this timer elapses. + The argument is a unit name, whose suffix is not + .timer. If not specified, this value + defaults to a service that has the same name as the timer + unit, except for the suffix. (See above.) It is recommended + that the unit name that is activated and the unit name of the + timer unit are named identically, except for the + suffix. + + + + + Persistent= + + Takes a boolean argument. If true, the time + when the service unit was last triggered is stored on disk. + When the timer is activated, the service unit is triggered + immediately if it would have been triggered at least once + during the time when the timer was inactive. This is useful to + catch up on missed runs of the service when the machine was + off. Note that this setting only has an effect on timers + configured with OnCalendar=. Defaults + to false. + + + + + WakeSystem= + + Takes a boolean argument. If true, an elapsing + timer will cause the system to resume from suspend, should it + be suspended and if the system supports this. Note that this + option will only make sure the system resumes on the + appropriate times, it will not take care of suspending it + again after any work that is to be done is finished. Defaults + to false. + + + + RemainAfterElapse= + + Takes a boolean argument. If true, an elapsed + timer will stay loaded, and its state remains queriable. If + false, an elapsed timer unit that cannot elapse anymore is + unloaded. Turning this off is particularly useful for + transient timer units that shall disappear after they first + elapse. Note that this setting has an effect on repeatedly + starting a timer unit that only elapses once: if + RemainAfterElapse= is on, it will not be + started again, and is guaranteed to elapse only once. However, + if RemainAfterElapse= is off, it might be + started again if it is already elapsed, and thus be triggered + multiple times. Defaults to + yes. + + + + + + See Also + + systemd1, + systemctl1, + systemd.unit5, + systemd.service5, + systemd.time7, + systemd.directives7, + systemd-system.conf5, + prctl2 + + + + diff --git a/src/grp-system/systemd/systemd.tmpfiles b/src/grp-system/systemd/systemd.tmpfiles new file mode 100644 index 0000000000..00951c92c9 --- /dev/null +++ b/src/grp-system/systemd/systemd.tmpfiles @@ -0,0 +1,20 @@ +# This file is part of systemd. +# +# systemd is free software; you can redistribute it and/or modify it +# under the terms of the GNU Lesser General Public License as published by +# the Free Software Foundation; either version 2.1 of the License, or +# (at your option) any later version. + +# See tmpfiles.d(5) for details + +d /run/user 0755 root root - +F! /run/utmp 0664 root utmp - + +d /run/systemd/ask-password 0755 root root - +d /run/systemd/seats 0755 root root - +d /run/systemd/sessions 0755 root root - +d /run/systemd/users 0755 root root - +d /run/systemd/machines 0755 root root - +d /run/systemd/shutdown 0755 root root - + +d /var/lib/systemd 0755 root root - diff --git a/src/grp-system/systemd/systemd.unit.xml b/src/grp-system/systemd/systemd.unit.xml new file mode 100644 index 0000000000..85a7b12d76 --- /dev/null +++ b/src/grp-system/systemd/systemd.unit.xml @@ -0,0 +1,1484 @@ + + +%entities; +]> + + + + + + + systemd.unit + systemd + + + + Developer + Lennart + Poettering + lennart@poettering.net + + + + + + systemd.unit + 5 + + + + systemd.unit + Unit configuration + + + + service.service, + socket.socket, + device.device, + mount.mount, + automount.automount, + swap.swap, + target.target, + path.path, + timer.timer, + slice.slice, + scope.scope + + /etc/systemd/system/* +/run/systemd/system/* +/usr/lib/systemd/system/* +… + + + ~/.config/systemd/user/* +/etc/systemd/user/* +$XDG_RUNTIME_DIR/systemd/user/* +/run/systemd/user/* +~/.local/share/systemd/user/* +/usr/lib/systemd/user/* +… + + + + + Description + + A unit configuration file encodes information about a + service, a socket, a device, a mount point, an automount point, a + swap file or partition, a start-up target, a watched file system + path, a timer controlled and supervised by + systemd1, + a resource management slice or + a group of externally created processes. The syntax is inspired by + XDG + Desktop Entry Specification .desktop + files, which are in turn inspired by Microsoft Windows + .ini files. + + This man page lists the common configuration options of all + the unit types. These options need to be configured in the [Unit] + or [Install] sections of the unit files. + + In addition to the generic [Unit] and [Install] sections + described here, each unit may have a type-specific section, e.g. + [Service] for a service unit. See the respective man pages for + more information: + systemd.service5, + systemd.socket5, + systemd.device5, + systemd.mount5, + systemd.automount5, + systemd.swap5, + systemd.target5, + systemd.path5, + systemd.timer5, + systemd.slice5, + systemd.scope5. + + + Various settings are allowed to be specified more than once, + in which case the interpretation depends on the setting. Often, + multiple settings form a list, and setting to an empty value + "resets", which means that previous assignments are ignored. When + this is allowed, it is mentioned in the description of the + setting. Note that using multiple assignments to the same value + makes the unit file incompatible with parsers for the XDG + .desktop file format. + + Unit files are loaded from a set of paths determined during + compilation, described in the next section. + + Unit files may contain additional options on top of those + listed here. If systemd encounters an unknown option, it will + write a warning log message but continue loading the unit. If an + option or section name is prefixed with , it is + ignored completely by systemd. Options within an ignored section + do not need the prefix. Applications may use this to include + additional information in the unit files. + + Boolean arguments used in unit files can be written in + various formats. For positive settings the strings + , , + and are equivalent. For negative settings, the + strings , , + and are + equivalent. + + Time span values encoded in unit files can be written in + various formats. A stand-alone number specifies a time in seconds. + If suffixed with a time unit, the unit is honored. A concatenation + of multiple values with units is supported, in which case the + values are added up. Example: "50" refers to 50 seconds; "2min + 200ms" refers to 2 minutes plus 200 milliseconds, i.e. 120200ms. + The following time units are understood: s, min, h, d, w, ms, us. + For details see + systemd.time7. + + Empty lines and lines starting with # or ; are + ignored. This may be used for commenting. Lines ending + in a backslash are concatenated with the following + line while reading and the backslash is replaced by a + space character. This may be used to wrap long lines. + + Along with a unit file foo.service, the + directory foo.service.wants/ may exist. All + unit files symlinked from such a directory are implicitly added as + dependencies of type Wants= to the unit. This + is useful to hook units into the start-up of other units, without + having to modify their unit files. For details about the semantics + of Wants=, see below. The preferred way to + create symlinks in the .wants/ directory of a + unit file is with the enable command of the + systemctl1 + tool which reads information from the [Install] section of unit + files (see below). A similar functionality exists for + Requires= type dependencies as well, the + directory suffix is .requires/ in this + case. + + Along with a unit file foo.service, a "drop-in" directory + foo.service.d/ may exist. All files with the suffix .conf from this + directory will be parsed after the file itself is parsed. This is useful to alter or add configuration settings for + a unit, without having to modify unit files. Each drop-in file must have appropriate section headers. Note that for + instantiated units, this logic will first look for the instance .d/ subdirectory and read its + .conf files, followed by the template .d/ subdirectory and the + .conf files there. Also note that settings from the [Install] section are not + honoured in drop-in unit files, and have no effect. + + In addition to /etc/systemd/system, + the drop-in .conf files for system services + can be placed in /usr/lib/systemd/system or + /run/systemd/system directories. Drop-in + files in /etc take precedence over those in + /run which in turn take precedence over + those in /usr/lib. Drop-in files under any of + these directories take precedence over unit files wherever located. + (Of course, since /run is temporary and + /usr/lib is for vendors, it is unlikely + drop-ins should be used in either of those places.) + + + Some unit names reflect paths existing in the file system + namespace. Example: a device unit + dev-sda.device refers to a device with the + device node /dev/sda in the + file system namespace. If this applies, a special way to escape + the path name is used, so that the result is usable as part of a + filename. Basically, given a path, "/" is replaced by "-", and all + other characters which are not ASCII alphanumerics are replaced by + C-style "\x2d" escapes (except that "_" is never replaced and "." + is only replaced when it would be the first character in the + escaped path). The root directory "/" is encoded as single dash, + while otherwise the initial and ending "/" are removed from all + paths during transformation. This escaping is reversible. Properly + escaped paths can be generated using the + systemd-escape1 + command. + + Optionally, units may be instantiated from a + template file at runtime. This allows creation of + multiple units from a single configuration file. If + systemd looks for a unit configuration file, it will + first search for the literal unit name in the + file system. If that yields no success and the unit + name contains an @ character, systemd will look for a + unit template that shares the same name but with the + instance string (i.e. the part between the @ character + and the suffix) removed. Example: if a service + getty@tty3.service is requested + and no file by that name is found, systemd will look + for getty@.service and + instantiate a service from that configuration file if + it is found. + + To refer to the instance string from within the + configuration file you may use the special %i + specifier in many of the configuration options. See below for + details. + + If a unit file is empty (i.e. has the file size 0) or is + symlinked to /dev/null, its configuration + will not be loaded and it appears with a load state of + masked, and cannot be activated. Use this as an + effective way to fully disable a unit, making it impossible to + start it even manually. + + The unit file format is covered by the + Interface + Stability Promise. + + + + + Automatic Dependencies + + Note that while systemd offers a flexible dependency system + between units it is recommended to use this functionality only + sparingly and instead rely on techniques such as bus-based or + socket-based activation which make dependencies implicit, + resulting in a both simpler and more flexible system. + + A number of unit dependencies are automatically established, + depending on unit configuration. On top of that, for units with + DefaultDependencies=yes (the default) a couple + of additional dependencies are added. The precise effect of + DefaultDependencies=yes depends on the unit + type (see below). + + If DefaultDependencies=yes is set, units + that are referenced by other units of type + .target via a Wants= or + Requires= dependency might automatically gain + an Before= dependency too. See + systemd.target5 + for details. + + + + Unit File Load Path + + Unit files are loaded from a set of paths determined during + compilation, described in the two tables below. Unit files found + in directories listed earlier override files with the same name in + directories lower in the list. + + When the variable $SYSTEMD_UNIT_PATH is set, + the contents of this variable overrides the unit load path. If + $SYSTEMD_UNIT_PATH ends with an empty component + (:), the usual unit load path will be appended + to the contents of the variable. + + + + Load path when running in system mode (<option>--system</option>). + + + + + + + + Path + Description + + + + + /etc/systemd/system + Local configuration + + + /run/systemd/system + Runtime units + + + /usr/lib/systemd/system + Units of installed packages + + + +
+ + + + Load path when running in user mode (<option>--user</option>). + + + + + + + + Path + Description + + + + + $XDG_CONFIG_HOME/systemd/user + User configuration (only used when $XDG_CONFIG_HOME is set) + + + $HOME/.config/systemd/user + User configuration (only used when $XDG_CONFIG_HOME is not set) + + + /etc/systemd/user + Local configuration + + + $XDG_RUNTIME_DIR/systemd/user + Runtime units (only used when $XDG_RUNTIME_DIR is set) + + + /run/systemd/user + Runtime units + + + $XDG_DATA_HOME/systemd/user + Units of packages that have been installed in the home directory (only used when $XDG_DATA_HOME is set) + + + $HOME/.local/share/systemd/user + Units of packages that have been installed in the home directory (only used when $XDG_DATA_HOME is not set) + + + /usr/lib/systemd/user + Units of packages that have been installed system-wide + + + +
+ + Additional units might be loaded into systemd ("linked") + from directories not on the unit load path. See the + link command for + systemctl1. + Also, some units are dynamically created via a + systemd.generator7. + + + + + [Unit] Section Options + + The unit file may include a [Unit] section, which carries + generic information about the unit that is not dependent on the + type of unit: + + + + + Description= + A free-form string describing the unit. This + is intended for use in UIs to show descriptive information + along with the unit name. The description should contain a + name that means something to the end user. Apache2 + Web Server is a good example. Bad examples are + high-performance light-weight HTTP server + (too generic) or Apache2 (too specific and + meaningless for people who do not know + Apache). + + + + Documentation= + A space-separated list of URIs referencing + documentation for this unit or its configuration. Accepted are + only URIs of the types http://, + https://, file:, + info:, man:. For more + information about the syntax of these URIs, see uri7. + The URIs should be listed in order of relevance, starting with + the most relevant. It is a good idea to first reference + documentation that explains what the unit's purpose is, + followed by how it is configured, followed by any other + related documentation. This option may be specified more than + once, in which case the specified list of URIs is merged. If + the empty string is assigned to this option, the list is reset + and all prior assignments will have no + effect. + + + + Requires= + + Configures requirement dependencies on other + units. If this unit gets activated, the units listed here will + be activated as well. If one of the other units gets + deactivated or its activation fails, this unit will be + deactivated. This option may be specified more than once or + multiple space-separated units may be specified in one option + in which case requirement dependencies for all listed names + will be created. Note that requirement dependencies do not + influence the order in which services are started or stopped. + This has to be configured independently with the + After= or Before= + options. If a unit foo.service requires a + unit bar.service as configured with + Requires= and no ordering is configured + with After= or Before=, + then both units will be started simultaneously and without any + delay between them if foo.service is + activated. Often, it is a better choice to use + Wants= instead of + Requires= in order to achieve a system that + is more robust when dealing with failing services. + + Note that dependencies of this type may also be + configured outside of the unit configuration file by adding a + symlink to a .requires/ directory + accompanying the unit file. For details, see + above. + + + + Requisite= + + Similar to Requires=. + However, if the units listed here are not started already, + they will not be started and the transaction will fail + immediately. + + + + Wants= + + A weaker version of + Requires=. Units listed in this option will + be started if the configuring unit is. However, if the listed + units fail to start or cannot be added to the transaction, + this has no impact on the validity of the transaction as a + whole. This is the recommended way to hook start-up of one + unit to the start-up of another unit. + + Note that dependencies of this type may also be + configured outside of the unit configuration file by adding + symlinks to a .wants/ directory + accompanying the unit file. For details, see + above. + + + + BindsTo= + + Configures requirement dependencies, very + similar in style to Requires=, however in + addition to this behavior, it also declares that this unit is + stopped when any of the units listed suddenly disappears. + Units can suddenly, unexpectedly disappear if a service + terminates on its own choice, a device is unplugged or a mount + point unmounted without involvement of + systemd. + + + + PartOf= + + Configures dependencies similar to + Requires=, but limited to stopping and + restarting of units. When systemd stops or restarts the units + listed here, the action is propagated to this unit. Note that + this is a one-way dependency — changes to this unit do not + affect the listed units. + + + + Conflicts= + + A space-separated list of unit names. + Configures negative requirement dependencies. If a unit has a + Conflicts= setting on another unit, + starting the former will stop the latter and vice versa. Note + that this setting is independent of and orthogonal to the + After= and Before= + ordering dependencies. + + If a unit A that conflicts with a unit B is scheduled to + be started at the same time as B, the transaction will either + fail (in case both are required part of the transaction) or be + modified to be fixed (in case one or both jobs are not a + required part of the transaction). In the latter case, the job + that is not the required will be removed, or in case both are + not required, the unit that conflicts will be started and the + unit that is conflicted is stopped. + + + + Before= + After= + + A space-separated list of unit names. + Configures ordering dependencies between units. If a unit + foo.service contains a setting + and both units are being + started, bar.service's start-up is + delayed until foo.service is started up. + Note that this setting is independent of and orthogonal to the + requirement dependencies as configured by + Requires=. It is a common pattern to + include a unit name in both the After= and + Requires= option, in which case the unit + listed will be started before the unit that is configured with + these options. This option may be specified more than once, in + which case ordering dependencies for all listed names are + created. After= is the inverse of + Before=, i.e. while + After= ensures that the configured unit is + started after the listed unit finished starting up, + Before= ensures the opposite, i.e. that the + configured unit is fully started up before the listed unit is + started. Note that when two units with an ordering dependency + between them are shut down, the inverse of the start-up order + is applied. i.e. if a unit is configured with + After= on another unit, the former is + stopped before the latter if both are shut down. Given two units + with any ordering dependency between them, if one unit is shut + down and the other is started up, the shutdown is ordered + before the start-up. It doesn't matter if the ordering + dependency is After= or + Before=. It also doesn't matter which of the + two is shut down, as long as one is shut down and the other is + started up. The shutdown is ordered before the start-up in all + cases. If two units have no ordering dependencies between them, + they are shut down or started up simultaneously, and no ordering + takes place. + + + + + OnFailure= + + A space-separated list of one or more units + that are activated when this unit enters the + failed state. + + + + PropagatesReloadTo= + ReloadPropagatedFrom= + + A space-separated list of one or more units + where reload requests on this unit will be propagated to, or + reload requests on the other unit will be propagated to this + unit, respectively. Issuing a reload request on a unit will + automatically also enqueue a reload request on all units that + the reload request shall be propagated to via these two + settings. + + + + JoinsNamespaceOf= + + For units that start processes (such as + service units), lists one or more other units whose network + and/or temporary file namespace to join. This only applies to + unit types which support the + PrivateNetwork= and + PrivateTmp= directives (see + systemd.exec5 + for details). If a unit that has this setting set is started, + its processes will see the same /tmp, + /var/tmp and network namespace as one + listed unit that is started. If multiple listed units are + already started, it is not defined which namespace is joined. + Note that this setting only has an effect if + PrivateNetwork= and/or + PrivateTmp= is enabled for both the unit + that joins the namespace and the unit whose namespace is + joined. + + + + RequiresMountsFor= + + Takes a space-separated list of absolute + paths. Automatically adds dependencies of type + Requires= and After= for + all mount units required to access the specified path. + + Mount points marked with are not + mounted automatically and will be ignored for the purposes of + this option. If such a mount should be a requirement for this + unit, direct dependencies on the mount units may be added + (Requires= and After= or + some other combination). + + + + OnFailureJobMode= + + Takes a value of + fail, + replace, + replace-irreversibly, + isolate, + flush, + ignore-dependencies or + ignore-requirements. Defaults to + replace. Specifies how the units listed in + OnFailure= will be enqueued. See + systemctl1's + option for details on the + possible values. If this is set to isolate, + only a single unit may be listed in + OnFailure=.. + + + + IgnoreOnIsolate= + + Takes a boolean argument. If + , this unit will not be stopped when + isolating another unit. Defaults to + . + + + + StopWhenUnneeded= + + Takes a boolean argument. If + , this unit will be stopped when it is no + longer used. Note that, in order to minimize the work to be + executed, systemd will not stop units by default unless they + are conflicting with other units, or the user explicitly + requested their shut down. If this option is set, a unit will + be automatically cleaned up if no other active unit requires + it. Defaults to . + + + + RefuseManualStart= + RefuseManualStop= + + Takes a boolean argument. If + , this unit can only be activated or + deactivated indirectly. In this case, explicit start-up or + termination requested by the user is denied, however if it is + started or stopped as a dependency of another unit, start-up + or termination will succeed. This is mostly a safety feature + to ensure that the user does not accidentally activate units + that are not intended to be activated explicitly, and not + accidentally deactivate units that are not intended to be + deactivated. These options default to + . + + + + AllowIsolate= + + Takes a boolean argument. If + , this unit may be used with the + systemctl isolate command. Otherwise, this + will be refused. It probably is a good idea to leave this + disabled except for target units that shall be used similar to + runlevels in SysV init systems, just as a precaution to avoid + unusable system states. This option defaults to + . + + + + DefaultDependencies= + + Takes a boolean argument. If + , (the default), a few default + dependencies will implicitly be created for the unit. The + actual dependencies created depend on the unit type. For + example, for service units, these dependencies ensure that the + service is started only after basic system initialization is + completed and is properly terminated on system shutdown. See + the respective man pages for details. Generally, only services + involved with early boot or late shutdown should set this + option to . It is highly recommended to + leave this option enabled for the majority of common units. If + set to , this option does not disable + all implicit dependencies, just non-essential + ones. + + + + JobTimeoutSec= + JobTimeoutAction= + JobTimeoutRebootArgument= + + When a job for this unit is queued, a time-out may be configured. If this time limit is + reached, the job will be cancelled, the unit however will not change state or even enter the + failed mode. This value defaults to infinity (job timeouts disabled), + except for device units. NB: this timeout is independent from any unit-specific timeout (for example, the + timeout set with TimeoutStartSec= in service units) as the job timeout has no effect on the + unit itself, only on the job that might be pending for it. Or in other words: unit-specific timeouts are useful + to abort unit state changes, and revert them. The job timeout set with this option however is useful to abort + only the job waiting for the unit state to change. + + JobTimeoutAction= + optionally configures an additional + action to take when the time-out is + hit. It takes the same values as the + per-service + StartLimitAction= + setting, see + systemd.service5 + for details. Defaults to + . JobTimeoutRebootArgument= + configures an optional reboot string + to pass to the + reboot2 + system call. + + + + StartLimitIntervalSec= + StartLimitBurst= + + Configure unit start rate limiting. By default, units which are started more than 5 times + within 10 seconds are not permitted to start any more times until the 10 second interval ends. With these two + options, this rate limiting may be modified. Use StartLimitIntervalSec= to configure the + checking interval (defaults to DefaultStartLimitIntervalSec= in manager configuration file, + set to 0 to disable any kind of rate limiting). Use StartLimitBurst= to configure how many + starts per interval are allowed (defaults to DefaultStartLimitBurst= in manager + configuration file). These configuration options are particularly useful in conjunction with the service + setting Restart= (see + systemd.service5); however, + they apply to all kinds of starts (including manual), not just those triggered by the + Restart= logic. Note that units which are configured for Restart= and + which reach the start limit are not attempted to be restarted anymore; however, they may still be restarted + manually at a later point, from which point on, the restart logic is again activated. Note that + systemctl reset-failed will cause the restart rate counter for a service to be flushed, + which is useful if the administrator wants to manually start a unit and the start limit interferes with + that. Note that this rate-limiting is enforced after any unit condition checks are executed, and hence unit + activations with failing conditions are not counted by this rate limiting. Slice, target, device and scope + units do not enforce this setting, as they are unit types whose activation may either never fail, or may + succeed only a single time. + + + + StartLimitAction= + + Configure the action to take if the rate limit configured with + StartLimitIntervalSec= and StartLimitBurst= is hit. Takes one of + , , , + , , or + . If is set, hitting the rate limit will trigger no + action besides that the start will not be permitted. causes a reboot following the + normal shutdown procedure (i.e. equivalent to systemctl reboot). + causes a forced reboot which will terminate all processes forcibly but should + cause no dirty file systems on reboot (i.e. equivalent to systemctl reboot -f) and + causes immediate execution of the + reboot2 system call, which + might result in data loss. Similarly, , , + have the effect of powering down the system with similar + semantics. Defaults to . + + + + RebootArgument= + Configure the optional argument for the + reboot2 system call if + StartLimitAction= or a service's FailureAction= is a reboot action. This + works just like the optional argument to systemctl reboot command. + + + + ConditionArchitecture= + ConditionVirtualization= + ConditionHost= + ConditionKernelCommandLine= + ConditionSecurity= + ConditionCapability= + ConditionACPower= + ConditionNeedsUpdate= + ConditionFirstBoot= + ConditionPathExists= + ConditionPathExistsGlob= + ConditionPathIsDirectory= + ConditionPathIsSymbolicLink= + ConditionPathIsMountPoint= + ConditionPathIsReadWrite= + ConditionDirectoryNotEmpty= + ConditionFileNotEmpty= + ConditionFileIsExecutable= + + + + Before starting a unit, verify that the specified condition is true. If it is not true, the + starting of the unit will be (mostly silently) skipped, however all ordering dependencies of it are still + respected. A failing condition will not result in the unit being moved into a failure state. The condition is + checked at the time the queued start job is to be executed. Use condition expressions in order to silently skip + units that do not apply to the local running system, for example because the kernel or runtime environment + doesn't require its functionality. Use the various AssertArchitecture=, + AssertVirtualization=, … options for a similar mechanism that puts the unit in a failure + state and logs about the failed check (see below). + + ConditionArchitecture= may be used to + check whether the system is running on a specific + architecture. Takes one of + x86, + x86-64, + ppc, + ppc-le, + ppc64, + ppc64-le, + ia64, + parisc, + parisc64, + s390, + s390x, + sparc, + sparc64, + mips, + mips-le, + mips64, + mips64-le, + alpha, + arm, + arm-be, + arm64, + arm64-be, + sh, + sh64, + m86k, + tilegx, + cris to test + against a specific architecture. The architecture is + determined from the information returned by + uname2 + and is thus subject to + personality2. + Note that a Personality= setting in the + same unit file has no effect on this condition. A special + architecture name native is mapped to the + architecture the system manager itself is compiled for. The + test may be negated by prepending an exclamation mark. + + ConditionVirtualization= may be used + to check whether the system is executed in a virtualized + environment and optionally test whether it is a specific + implementation. Takes either boolean value to check if being + executed in any virtualized environment, or one of + vm and + container to test against a generic type of + virtualization solution, or one of + qemu, + kvm, + zvm, + vmware, + microsoft, + oracle, + xen, + bochs, + uml, + openvz, + lxc, + lxc-libvirt, + systemd-nspawn, + docker, + rkt to test + against a specific implementation. See + systemd-detect-virt1 + for a full list of known virtualization technologies and their + identifiers. If multiple virtualization technologies are + nested, only the innermost is considered. The test may be + negated by prepending an exclamation mark. + + ConditionHost= may be used to match + against the hostname or machine ID of the host. This either + takes a hostname string (optionally with shell style globs) + which is tested against the locally set hostname as returned + by + gethostname2, + or a machine ID formatted as string (see + machine-id5). + The test may be negated by prepending an exclamation + mark. + + ConditionKernelCommandLine= may be + used to check whether a specific kernel command line option is + set (or if prefixed with the exclamation mark unset). The + argument must either be a single word, or an assignment (i.e. + two words, separated =). In the former case + the kernel command line is searched for the word appearing as + is, or as left hand side of an assignment. In the latter case, + the exact assignment is looked for with right and left hand + side matching. + + ConditionSecurity= may be used to + check whether the given security module is enabled on the + system. Currently, the recognized values are + selinux, + apparmor, + ima, + smack and + audit. The test may be negated by + prepending an exclamation mark. + + ConditionCapability= may be used to + check whether the given capability exists in the capability + bounding set of the service manager (i.e. this does not check + whether capability is actually available in the permitted or + effective sets, see + capabilities7 + for details). Pass a capability name such as + CAP_MKNOD, possibly prefixed with an + exclamation mark to negate the check. + + ConditionACPower= may be used to + check whether the system has AC power, or is exclusively + battery powered at the time of activation of the unit. This + takes a boolean argument. If set to true, + the condition will hold only if at least one AC connector of + the system is connected to a power source, or if no AC + connectors are known. Conversely, if set to + false, the condition will hold only if + there is at least one AC connector known and all AC connectors + are disconnected from a power source. + + ConditionNeedsUpdate= takes one of + /var or /etc as + argument, possibly prefixed with a ! (for + inverting the condition). This condition may be used to + conditionalize units on whether the specified directory + requires an update because /usr's + modification time is newer than the stamp file + .updated in the specified directory. This + is useful to implement offline updates of the vendor operating + system resources in /usr that require + updating of /etc or + /var on the next following boot. Units + making use of this condition should order themselves before + systemd-update-done.service8, + to make sure they run before the stamp file's modification + time gets reset indicating a completed update. + + ConditionFirstBoot= takes a boolean + argument. This condition may be used to conditionalize units + on whether the system is booting up with an unpopulated + /etc directory. This may be used to + populate /etc on the first boot after + factory reset, or when a new system instances boots up for the + first time. + + With ConditionPathExists= a file + existence condition is checked before a unit is started. If + the specified absolute path name does not exist, the condition + will fail. If the absolute path name passed to + ConditionPathExists= is prefixed with an + exclamation mark (!), the test is negated, + and the unit is only started if the path does not + exist. + + ConditionPathExistsGlob= is similar + to ConditionPathExists=, but checks for the + existence of at least one file or directory matching the + specified globbing pattern. + + ConditionPathIsDirectory= is similar + to ConditionPathExists= but verifies + whether a certain path exists and is a directory. + + ConditionPathIsSymbolicLink= is + similar to ConditionPathExists= but + verifies whether a certain path exists and is a symbolic + link. + + ConditionPathIsMountPoint= is similar + to ConditionPathExists= but verifies + whether a certain path exists and is a mount point. + + ConditionPathIsReadWrite= is similar + to ConditionPathExists= but verifies + whether the underlying file system is readable and writable + (i.e. not mounted read-only). + + ConditionDirectoryNotEmpty= is + similar to ConditionPathExists= but + verifies whether a certain path exists and is a non-empty + directory. + + ConditionFileNotEmpty= is similar to + ConditionPathExists= but verifies whether a + certain path exists and refers to a regular file with a + non-zero size. + + ConditionFileIsExecutable= is similar + to ConditionPathExists= but verifies + whether a certain path exists, is a regular file and marked + executable. + + If multiple conditions are specified, the unit will be + executed if all of them apply (i.e. a logical AND is applied). + Condition checks can be prefixed with a pipe symbol (|) in + which case a condition becomes a triggering condition. If at + least one triggering condition is defined for a unit, then the + unit will be executed if at least one of the triggering + conditions apply and all of the non-triggering conditions. If + you prefix an argument with the pipe symbol and an exclamation + mark, the pipe symbol must be passed first, the exclamation + second. Except for + ConditionPathIsSymbolicLink=, all path + checks follow symlinks. If any of these options is assigned + the empty string, the list of conditions is reset completely, + all previous condition settings (of any kind) will have no + effect. + + + + AssertArchitecture= + AssertVirtualization= + AssertHost= + AssertKernelCommandLine= + AssertSecurity= + AssertCapability= + AssertACPower= + AssertNeedsUpdate= + AssertFirstBoot= + AssertPathExists= + AssertPathExistsGlob= + AssertPathIsDirectory= + AssertPathIsSymbolicLink= + AssertPathIsMountPoint= + AssertPathIsReadWrite= + AssertDirectoryNotEmpty= + AssertFileNotEmpty= + AssertFileIsExecutable= + + Similar to the ConditionArchitecture=, + ConditionVirtualization=, …, condition settings described above, these settings add + assertion checks to the start-up of the unit. However, unlike the conditions settings, any assertion setting + that is not met results in failure of the start job (which means this is logged loudly). Use assertion + expressions for units that cannot operate when specific requirements are not met, and when this is something + the administrator or user should look into. + + + + SourcePath= + A path to a configuration file this unit has + been generated from. This is primarily useful for + implementation of generator tools that convert configuration + from an external configuration file format into native unit + files. This functionality should not be used in normal + units. + + + + + + + + [Install] Section Options + + Unit files may include an [Install] section, which carries installation information for + the unit. This section is not interpreted by + systemd1 during runtime; it is + used by the enable and disable commands of the + systemctl1 tool during + installation of a unit. Note that settings in the [Install] section may not appear in + .d/*.conf unit file drop-ins (see above). + + + + Alias= + + A space-separated list of additional names this unit shall be installed under. The names listed + here must have the same suffix (i.e. type) as the unit file name. This option may be specified more than once, + in which case all listed names are used. At installation time, systemctl enable will create + symlinks from these names to the unit filename. Note that not all unit types support such alias names, and this + setting is not supported for them. Specifically, mount, slice, swap, and automount units do not support + aliasing. + + + + WantedBy= + RequiredBy= + + This option may be used more than once, or a + space-separated list of unit names may be given. A symbolic + link is created in the .wants/ or + .requires/ directory of each of the + listed units when this unit is installed by systemctl + enable. This has the effect that a dependency of + type Wants= or Requires= + is added from the listed unit to the current unit. The primary + result is that the current unit will be started when the + listed unit is started. See the description of + Wants= and Requires= in + the [Unit] section for details. + + WantedBy=foo.service in a service + bar.service is mostly equivalent to + Alias=foo.service.wants/bar.service in the + same file. In case of template units, systemctl + enable must be called with an instance name, and + this instance will be added to the + .wants/ or + .requires/ list of the listed unit. E.g. + WantedBy=getty.target in a service + getty@.service will result in + systemctl enable getty@tty2.service + creating a + getty.target.wants/getty@tty2.service + link to getty@.service. + + + + + Also= + + Additional units to install/deinstall when + this unit is installed/deinstalled. If the user requests + installation/deinstallation of a unit with this option + configured, systemctl enable and + systemctl disable will automatically + install/uninstall units listed in this option as well. + + This option may be used more than once, or a + space-separated list of unit names may be + given. + + + + DefaultInstance= + + In template unit files, this specifies for + which instance the unit shall be enabled if the template is + enabled without any explicitly set instance. This option has + no effect in non-template unit files. The specified string + must be usable as instance identifier. + + + + The following specifiers are interpreted in the Install + section: %n, %N, %p, %i, %U, %u, %m, %H, %b, %v. For their meaning + see the next section. + + + + + Specifiers + + Many settings resolve specifiers which may be used to write + generic unit files referring to runtime or unit parameters that + are replaced when the unit files are loaded. The following + specifiers are understood: + + + Specifiers available in unit files + + + + + + + Specifier + Meaning + Details + + + + + %n + Full unit name + + + + %N + Unescaped full unit name + Same as %n, but with escaping undone + + + %p + Prefix name + For instantiated units, this refers to the string before the @ character of the unit name. For non-instantiated units, this refers to the name of the unit with the type suffix removed. + + + %P + Unescaped prefix name + Same as %p, but with escaping undone + + + %i + Instance name + For instantiated units: this is the string between the @ character and the suffix of the unit name. + + + %I + Unescaped instance name + Same as %i, but with escaping undone + + + %f + Unescaped filename + This is either the unescaped instance name (if applicable) with / prepended (if applicable), or the unescaped prefix name prepended with /. + + + %c + Control group path of the unit + This path does not include the /sys/fs/cgroup/systemd/ prefix. + + + %r + Control group path of the slice the unit is placed in + This usually maps to the parent cgroup path of %c. + + + %R + Root control group path below which slices and units are placed + For system instances, this resolves to /, except in containers, where this maps to the container's root control group path. + + + %t + Runtime directory + This is either /run (for the system manager) or the path $XDG_RUNTIME_DIR resolves to (for user managers). + + + %u + User name + This is the name of the user running the service manager instance. In case of the system manager this resolves to root. + + + %U + User UID + This is the numeric UID of the user running the service manager instance. In case of the system manager this resolves to 0. + + + %h + User home directory + This is the home directory of the user running the service manager instance. In case of the system manager this resolves to /root. + + + %s + User shell + This is the shell of the user running the service manager instance. In case of the system manager this resolves to /bin/sh. + + + %m + Machine ID + The machine ID of the running system, formatted as string. See machine-id5 for more information. + + + %b + Boot ID + The boot ID of the running system, formatted as string. See random4 for more information. + + + %H + Host name + The hostname of the running system at the point in time the unit configuration is loaded. + + + %v + Kernel release + Identical to uname -r output + + + %% + Single percent sign + Use %% in place of % to specify a single percent sign. + + + +
+ + Please note that specifiers %U, + %h, %s are mostly useless + when systemd is running in system mode. PID 1 cannot query the + user account database for information, so the specifiers only work + as shortcuts for things which are already specified in a different + way in the unit file. They are fully functional when systemd is + running in mode. + + + + Examples + + + Allowing units to be enabled + + The following snippet (highlighted) allows a unit (e.g. + foo.service) to be enabled via + systemctl enable: + + [Unit] +Description=Foo + +[Service] +ExecStart=/usr/sbin/foo-daemon + +[Install] +WantedBy=multi-user.target + + After running systemctl enable, a + symlink + /etc/systemd/system/multi-user.target.wants/foo.service + linking to the actual unit will be created. It tells systemd to + pull in the unit when starting + multi-user.target. The inverse + systemctl disable will remove that symlink + again. + + + + Overriding vendor settings + + There are two methods of overriding vendor settings in + unit files: copying the unit file from + /usr/lib/systemd/system to + /etc/systemd/system and modifying the + chosen settings. Alternatively, one can create a directory named + unit.d/ within + /etc/systemd/system and place a drop-in + file name.conf + there that only changes the specific settings one is interested + in. Note that multiple such drop-in files are read if + present. + + The advantage of the first method is that one easily + overrides the complete unit, the vendor unit is not parsed at + all anymore. It has the disadvantage that improvements to the + unit file by the vendor are not automatically incorporated on + updates. + + The advantage of the second method is that one only + overrides the settings one specifically wants, where updates to + the unit by the vendor automatically apply. This has the + disadvantage that some future updates by the vendor might be + incompatible with the local changes. + + Note that for drop-in files, if one wants to remove + entries from a setting that is parsed as a list (and is not a + dependency), such as ConditionPathExists= (or + e.g. ExecStart= in service units), one needs + to first clear the list before re-adding all entries except the + one that is to be removed. See below for an example. + + This also applies for user instances of systemd, but with + different locations for the unit files. See the section on unit + load paths for further details. + + Suppose there is a vendor-supplied unit + /usr/lib/systemd/system/httpd.service with + the following contents: + + [Unit] +Description=Some HTTP server +After=remote-fs.target sqldb.service +Requires=sqldb.service +AssertPathExists=/srv/webserver + +[Service] +Type=notify +ExecStart=/usr/sbin/some-fancy-httpd-server +Nice=5 + +[Install] +WantedBy=multi-user.target + + Now one wants to change some settings as an administrator: + firstly, in the local setup, /srv/webserver + might not exist, because the HTTP server is configured to use + /srv/www instead. Secondly, the local + configuration makes the HTTP server also depend on a memory + cache service, memcached.service, that + should be pulled in (Requires=) and also be + ordered appropriately (After=). Thirdly, in + order to harden the service a bit more, the administrator would + like to set the PrivateTmp= setting (see + systemd.service5 + for details). And lastly, the administrator would like to reset + the niceness of the service to its default value of 0. + + The first possibility is to copy the unit file to + /etc/systemd/system/httpd.service and + change the chosen settings: + + [Unit] +Description=Some HTTP server +After=remote-fs.target sqldb.service memcached.service +Requires=sqldb.service memcached.service +AssertPathExists=/srv/www + +[Service] +Type=notify +ExecStart=/usr/sbin/some-fancy-httpd-server +Nice=0 +PrivateTmp=yes + +[Install] +WantedBy=multi-user.target + + Alternatively, the administrator could create a drop-in + file + /etc/systemd/system/httpd.service.d/local.conf + with the following contents: + + [Unit] +After=memcached.service +Requires=memcached.service +# Reset all assertions and then re-add the condition we want +AssertPathExists= +AssertPathExists=/srv/www + +[Service] +Nice=0 +PrivateTmp=yes + + Note that dependencies (After=, etc.) + cannot be reset to an empty list, so dependencies can only be + added in drop-ins. If you want to remove dependencies, you have + to override the entire unit. + + + + + + See Also + + systemd1, + systemctl1, + systemd.special7, + systemd.service5, + systemd.socket5, + systemd.device5, + systemd.mount5, + systemd.automount5, + systemd.swap5, + systemd.target5, + systemd.path5, + systemd.timer5, + systemd.scope5, + systemd.slice5, + systemd.time7, + systemd-analyze1, + capabilities7, + systemd.directives7, + uname1 + + + + diff --git a/src/grp-system/systemd/systemd.xml b/src/grp-system/systemd/systemd.xml new file mode 100644 index 0000000000..4f0201fc76 --- /dev/null +++ b/src/grp-system/systemd/systemd.xml @@ -0,0 +1,1155 @@ + + + + + + + + + systemd + systemd + + + + Developer + Lennart + Poettering + lennart@poettering.net + + + + + + systemd + 1 + + + + systemd + init + systemd system and service manager + + + + + systemd OPTIONS + + + init OPTIONS COMMAND + + + + + Description + + systemd is a system and service manager for GNU/Linux operating + systems. When run as first process on boot (as PID 1), it acts as + init system that brings up and maintains userspace + services. + + For compatibility with SysV, if systemd is called as + init and a PID that is not 1, it will execute + telinit and pass all command line arguments + unmodified. That means init and + telinit are mostly equivalent when invoked from + normal login sessions. See + telinit8 + for more information. + + When run as a system instance, systemd interprets the + configuration file system.conf and the files + in system.conf.d directories; when run as a + user instance, systemd interprets the configuration file + user.conf and the files in + user.conf.d directories. See + systemd-system.conf5 + for more information. + + + + Options + + The following options are understood: + + + + + + Determine startup sequence, dump it and exit. + This is an option useful for debugging only. + + + + + Dump understood unit configuration items. This + outputs a terse but complete list of configuration items + understood in unit definition files. + + + + + Set default unit to activate on startup. If + not specified, defaults to + default.target. + + + + + + For , tell systemd to + run a system instance, even if the process ID is not 1, i.e. + systemd is not run as init process. + does the opposite, running a user instance even if the process + ID is 1. Normally, it should not be necessary to pass these + options, as systemd automatically detects the mode it is + started in. These options are hence of little use except for + debugging. Note that it is not supported booting and + maintaining a full system with systemd running in + mode, but PID not 1. In practice, + passing explicitly is only useful in + conjunction with . + + + + + Enable core dumping on crash. This switch has + no effect when running as user instance. This setting may also + be enabled during boot on the kernel command line via the + systemd.dump_core= option, see + below. + + + + VT + + Switch to a specific virtual console (VT) on + crash. Takes a positive integer in the range 1–63, or a + boolean argument. If an integer is passed, selects which VT to + switch to. If yes, the VT kernel messages + are written to is selected. If no, no VT + switch is attempted. This switch has no effect when running as + user instance. This setting may also be enabled during boot, + on the kernel command line via the + systemd.crash_vt= option, see + below. + + + + + + Run a shell on crash. This switch has no + effect when running as user instance. This setting may also be + enabled during boot, on the kernel command line via the + systemd.crash_shell= option, see + below. + + + + + + Automatically reboot the system on crash. This + switch has no effect when running as user instance. This + setting may also be enabled during boot, on the kernel command + line via the systemd.crash_reboot= option, + see below. + + + + + + Ask for confirmation when spawning processes. + This switch has no effect when run as user + instance. + + + + + Show terse service status information while + booting. This switch has no effect when run as user instance. + Takes a boolean argument which may be omitted which is + interpreted as . + + + + + Set log target. Argument must be one of + , + , + , + , + . + + + + + Set log level. As + argument this accepts a numerical log + level or the well-known syslog3 + symbolic names (lowercase): + , + , + , + , + , + , + , + . + + + + + Highlight important log messages. Argument is + a boolean value. If the argument is omitted, it defaults to + . + + + + + Include code location in log messages. This is + mostly relevant for debugging purposes. Argument is a boolean + value. If the argument is omitted it defaults to + . + + + + + + Sets the default output or error output for + all services and sockets, respectively. That is, controls the + default for and + (see + systemd.exec5 + for details). Takes one of + , + , + , + , + , + , + , + , + . If the + argument is omitted + defaults to + and + to + . + + + + + + Override the machine-id set on the hard drive, + useful for network booting or for containers. May not be set + to all zeros. + + + + + + + + + Concepts + + systemd provides a dependency system between various + entities called "units" of 12 different types. Units encapsulate + various objects that are relevant for system boot-up and + maintenance. The majority of units are configured in unit + configuration files, whose syntax and basic set of options is + described in + systemd.unit5, + however some are created automatically from other configuration, + dynamically from system state or programmatically at runtime. + Units may be "active" (meaning started, bound, plugged in, ..., + depending on the unit type, see below), or "inactive" (meaning + stopped, unbound, unplugged, ...), as well as in the process of + being activated or deactivated, i.e. between the two states (these + states are called "activating", "deactivating"). A special + "failed" state is available as well, which is very similar to + "inactive" and is entered when the service failed in some way + (process returned error code on exit, or crashed, or an operation + timed out). If this state is entered, the cause will be logged, + for later reference. Note that the various unit types may have a + number of additional substates, which are mapped to the five + generalized unit states described here. + + The following unit types are available: + + + Service units, which start and control daemons + and the processes they consist of. For details, see + systemd.service5. + + Socket units, which encapsulate local IPC or + network sockets in the system, useful for socket-based + activation. For details about socket units, see + systemd.socket5, + for details on socket-based activation and other forms of + activation, see + daemon7. + + Target units are useful to group units, or + provide well-known synchronization points during boot-up, see + systemd.target5. + + Device units expose kernel devices in systemd + and may be used to implement device-based activation. For + details, see + systemd.device5. + + Mount units control mount points in the file + system, for details see + systemd.mount5. + + Automount units provide automount capabilities, + for on-demand mounting of file systems as well as parallelized + boot-up. See + systemd.automount5. + + Timer units are useful for triggering activation + of other units based on timers. You may find details in + systemd.timer5. + + Swap units are very similar to mount units and + encapsulate memory swap partitions or files of the operating + system. They are described in + systemd.swap5. + + Path units may be used to activate other + services when file system objects change or are modified. See + systemd.path5. + + Slice units may be used to group units which + manage system processes (such as service and scope units) in a + hierarchical tree for resource management purposes. See + systemd.slice5. + + Scope units are similar to service units, but + manage foreign processes instead of starting them as well. See + systemd.scope5. + + + + Units are named as their configuration files. Some units + have special semantics. A detailed list is available in + systemd.special7. + + systemd knows various kinds of dependencies, including + positive and negative requirement dependencies (i.e. + Requires= and Conflicts=) as + well as ordering dependencies (After= and + Before=). NB: ordering and requirement + dependencies are orthogonal. If only a requirement dependency + exists between two units (e.g. foo.service + requires bar.service), but no ordering + dependency (e.g. foo.service after + bar.service) and both are requested to start, + they will be started in parallel. It is a common pattern that both + requirement and ordering dependencies are placed between two + units. Also note that the majority of dependencies are implicitly + created and maintained by systemd. In most cases, it should be + unnecessary to declare additional dependencies manually, however + it is possible to do this. + + Application programs and units (via dependencies) may + request state changes of units. In systemd, these requests are + encapsulated as 'jobs' and maintained in a job queue. Jobs may + succeed or can fail, their execution is ordered based on the + ordering dependencies of the units they have been scheduled + for. + + On boot systemd activates the target unit + default.target whose job is to activate + on-boot services and other on-boot units by pulling them in via + dependencies. Usually, the unit name is just an alias (symlink) for + either graphical.target (for fully-featured + boots into the UI) or multi-user.target (for + limited console-only boots for use in embedded or server + environments, or similar; a subset of graphical.target). However, + it is at the discretion of the administrator to configure it as an + alias to any other target unit. See + systemd.special7 + for details about these target units. + + Processes systemd spawns are placed in individual Linux + control groups named after the unit which they belong to in the + private systemd hierarchy. (see cgroups.txt + for more information about control groups, or short "cgroups"). + systemd uses this to effectively keep track of processes. Control + group information is maintained in the kernel, and is accessible + via the file system hierarchy (beneath + /sys/fs/cgroup/systemd/), or in tools such as + systemd-cgls1 + or + ps1 + (ps xawf -eo pid,user,cgroup,args is + particularly useful to list all processes and the systemd units + they belong to.). + + systemd is compatible with the SysV init system to a large + degree: SysV init scripts are supported and simply read as an + alternative (though limited) configuration file format. The SysV + /dev/initctl interface is provided, and + compatibility implementations of the various SysV client tools are + available. In addition to that, various established Unix + functionality such as /etc/fstab or the + utmp database are supported. + + systemd has a minimal transaction system: if a unit is + requested to start up or shut down it will add it and all its + dependencies to a temporary transaction. Then, it will verify if + the transaction is consistent (i.e. whether the ordering of all + units is cycle-free). If it is not, systemd will try to fix it up, + and removes non-essential jobs from the transaction that might + remove the loop. Also, systemd tries to suppress non-essential + jobs in the transaction that would stop a running service. Finally + it is checked whether the jobs of the transaction contradict jobs + that have already been queued, and optionally the transaction is + aborted then. If all worked out and the transaction is consistent + and minimized in its impact it is merged with all already + outstanding jobs and added to the run queue. Effectively this + means that before executing a requested operation, systemd will + verify that it makes sense, fixing it if possible, and only + failing if it really cannot work. + + Systemd contains native implementations of various tasks + that need to be executed as part of the boot process. For example, + it sets the hostname or configures the loopback network device. It + also sets up and mounts various API file systems, such as + /sys or /proc. + + For more information about the concepts and + ideas behind systemd, please refer to the + Original Design Document. + + Note that some but not all interfaces provided + by systemd are covered by the + Interface + Stability Promise. + + Units may be generated dynamically at boot and system + manager reload time, for example based on other configuration + files or parameters passed on the kernel command line. For details, see + systemd.generator7. + + Systems which invoke systemd in a container or initrd + environment should implement the + Container Interface or + initrd Interface + specifications, respectively. + + + + Directories + + + + System unit directories + + The systemd system manager reads unit + configuration from various directories. Packages that want to + install unit files shall place them in the directory returned + by pkg-config systemd + --variable=systemdsystemunitdir. Other directories + checked are /usr/local/lib/systemd/system + and /usr/lib/systemd/system. User + configuration always takes precedence. pkg-config + systemd --variable=systemdsystemconfdir returns the + path of the system configuration directory. Packages should + alter the content of these directories only with the + enable and disable + commands of the + systemctl1 + tool. Full list of directories is provided in + systemd.unit5. + + + + + + + User unit directories + + Similar rules apply for the user unit + directories. However, here the + XDG + Base Directory specification is followed to find + units. Applications should place their unit files in the + directory returned by pkg-config systemd + --variable=systemduserunitdir. Global configuration + is done in the directory reported by pkg-config + systemd --variable=systemduserconfdir. The + enable and disable + commands of the + systemctl1 + tool can handle both global (i.e. for all users) and private + (for one user) enabling/disabling of units. Full list of + directories is provided in + systemd.unit5. + + + + + + + SysV init scripts directory + + The location of the SysV init script directory + varies between distributions. If systemd cannot find a native + unit file for a requested service, it will look for a SysV + init script of the same name (with the + .service suffix + removed). + + + + + + SysV runlevel link farm directory + + The location of the SysV runlevel link farm + directory varies between distributions. systemd will take the + link farm into account when figuring out whether a service + shall be enabled. Note that a service unit with a native unit + configuration file cannot be started by activating it in the + SysV runlevel link farm. + + + + + + Signals + + + + SIGTERM + + Upon receiving this signal the systemd system + manager serializes its state, reexecutes itself and + deserializes the saved state again. This is mostly equivalent + to systemctl daemon-reexec. + + systemd user managers will start the + exit.target unit when this signal is + received. This is mostly equivalent to systemctl + --user start exit.target. + + + + SIGINT + + Upon receiving this signal the systemd system + manager will start the + ctrl-alt-del.target unit. This is mostly + equivalent to systemctl start + ctl-alt-del.target. If this signal is received more + than 7 times per 2s, an immediate reboot is triggered. + Note that pressing Ctrl-Alt-Del on the console will trigger + this signal. Hence, if a reboot is hanging, pressing + Ctrl-Alt-Del more than 7 times in 2s is a relatively safe way + to trigger an immediate reboot. + + systemd user managers treat this signal the same way as + SIGTERM. + + + + SIGWINCH + + When this signal is received the systemd + system manager will start the + kbrequest.target unit. This is mostly + equivalent to systemctl start + kbrequest.target. + + This signal is ignored by systemd user + managers. + + + + SIGPWR + + When this signal is received the systemd + manager will start the sigpwr.target + unit. This is mostly equivalent to systemctl start + sigpwr.target. + + + + SIGUSR1 + + When this signal is received the systemd + manager will try to reconnect to the D-Bus + bus. + + + + SIGUSR2 + + When this signal is received the systemd + manager will log its complete state in human-readable form. + The data logged is the same as printed by + systemd-analyze dump. + + + + SIGHUP + + Reloads the complete daemon configuration. + This is mostly equivalent to systemctl + daemon-reload. + + + + SIGRTMIN+0 + + Enters default mode, starts the + default.target unit. This is mostly + equivalent to systemctl start + default.target. + + + + SIGRTMIN+1 + + Enters rescue mode, starts the + rescue.target unit. This is mostly + equivalent to systemctl isolate + rescue.target. + + + + SIGRTMIN+2 + + Enters emergency mode, starts the + emergency.service unit. This is mostly + equivalent to systemctl isolate + emergency.service. + + + + SIGRTMIN+3 + + Halts the machine, starts the + halt.target unit. This is mostly + equivalent to systemctl start + halt.target. + + + + SIGRTMIN+4 + + Powers off the machine, starts the + poweroff.target unit. This is mostly + equivalent to systemctl start + poweroff.target. + + + + SIGRTMIN+5 + + Reboots the machine, starts the + reboot.target unit. This is mostly + equivalent to systemctl start + reboot.target. + + + + SIGRTMIN+6 + + Reboots the machine via kexec, starts the + kexec.target unit. This is mostly + equivalent to systemctl start + kexec.target. + + + + SIGRTMIN+13 + + Immediately halts the machine. + + + + SIGRTMIN+14 + + Immediately powers off the machine. + + + + SIGRTMIN+15 + + Immediately reboots the machine. + + + + SIGRTMIN+16 + + Immediately reboots the machine with kexec. + + + + SIGRTMIN+20 + + Enables display of status messages on the + console, as controlled via + systemd.show_status=1 on the kernel command + line. + + + + SIGRTMIN+21 + + Disables display of + status messages on the console, as + controlled via + systemd.show_status=0 + on the kernel command + line. + + + + SIGRTMIN+22 + SIGRTMIN+23 + + Sets the log level to debug + (or info on + SIGRTMIN+23), as controlled via + systemd.log_level=debug (or + systemd.log_level=info on + SIGRTMIN+23) on the kernel command + line. + + + + SIGRTMIN+24 + + Immediately exits the manager (only available + for --user instances). + + + + SIGRTMIN+26 + SIGRTMIN+27 + SIGRTMIN+28 + + Sets the log level to + journal-or-kmsg (or + console on + SIGRTMIN+27, kmsg on + SIGRTMIN+28), as controlled via + systemd.log_target=journal-or-kmsg (or + systemd.log_target=console on + SIGRTMIN+27 or + systemd.log_target=kmsg on + SIGRTMIN+28) on the kernel command + line. + + + + + + Environment + + + + $SYSTEMD_LOG_LEVEL + systemd reads the log level from this + environment variable. This can be overridden with + . + + + + $SYSTEMD_LOG_TARGET + systemd reads the log target from this + environment variable. This can be overridden with + . + + + + $SYSTEMD_LOG_COLOR + Controls whether systemd highlights important + log messages. This can be overridden with + . + + + + $SYSTEMD_LOG_LOCATION + Controls whether systemd prints the code + location along with log messages. This can be overridden with + . + + + + $XDG_CONFIG_HOME + $XDG_CONFIG_DIRS + $XDG_DATA_HOME + $XDG_DATA_DIRS + + The systemd user manager uses these variables + in accordance to the XDG + Base Directory specification to find its + configuration. + + + + $SYSTEMD_UNIT_PATH + + Controls where systemd looks for unit + files. + + + + $SYSTEMD_SYSVINIT_PATH + + Controls where systemd looks for SysV init + scripts. + + + + $SYSTEMD_SYSVRCND_PATH + + Controls where systemd looks for SysV init + script runlevel link farms. + + + + $SYSTEMD_COLORS + + Controls whether colorized output should be generated. + + + + + $LISTEN_PID + $LISTEN_FDS + $LISTEN_FDNAMES + + Set by systemd for supervised processes during + socket-based activation. See + sd_listen_fds3 + for more information. + + + + $NOTIFY_SOCKET + + Set by systemd for supervised processes for + status and start-up completion notification. See + sd_notify3 + for more information. + + + + + + Kernel Command Line + + When run as system instance systemd parses a number of + kernel command line argumentsIf run inside a Linux + container these arguments may be passed as command line arguments + to systemd itself, next to any of the command line options listed + in the Options section above. If run outside of Linux containers, + these arguments are parsed from /proc/cmdline + instead.: + + + + systemd.unit= + rd.systemd.unit= + + Overrides the unit to activate on boot. + Defaults to default.target. This may be + used to temporarily boot into a different boot unit, for + example rescue.target or + emergency.service. See + systemd.special7 + for details about these units. The option prefixed with + rd. is honored only in the initial RAM disk + (initrd), while the one that is not prefixed only in the main + system. + + + + systemd.dump_core= + + Takes a boolean argument. If + , the systemd manager (PID 1) dumps core + when it crashes. Otherwise, no core dump is created. Defaults + to . + + + + systemd.crash_chvt= + + Takes a positive integer, or a boolean + argument. If a positive integer (in the range 1–63) is + specified, the system manager (PID 1) will activate the specified + virtual terminal (VT) when it crashes. Defaults to + no, meaning that no such switch is + attempted. If set to yes, the VT the + kernel messages are written to is selected. + + + + systemd.crash_shell= + + Takes a boolean argument. If + , the system manager (PID 1) spawns a + shell when it crashes, after a 10s delay. Otherwise, no shell + is spawned. Defaults to , for security + reasons, as the shell is not protected by password + authentication. + + + + systemd.crash_reboot= + + Takes a boolean argument. If + , the system manager (PID 1) will reboot + the machine automatically when it crashes, after a 10s delay. + Otherwise, the system will hang indefinitely. Defaults to + , in order to avoid a reboot loop. If + combined with systemd.crash_shell=, the + system is rebooted after the shell exits. + + + + systemd.confirm_spawn= + + Takes a boolean argument. If + , the system manager (PID 1) asks for + confirmation when spawning processes. Defaults to + . + + + + systemd.show_status= + + Takes a boolean argument or the constant + auto. If , the + systemd manager (PID 1) shows terse service status updates on + the console during bootup. auto behaves + like until a service fails or there is + a significant delay in boot. Defaults to + , unless is passed + as kernel command line option, in which case it defaults to + auto. + + + + systemd.log_target= + systemd.log_level= + systemd.log_color= + systemd.log_location= + + Controls log output, with the same effect as + the $SYSTEMD_LOG_TARGET, + $SYSTEMD_LOG_LEVEL, + $SYSTEMD_LOG_COLOR, + $SYSTEMD_LOG_LOCATION environment variables + described above. + + + + systemd.default_standard_output= + systemd.default_standard_error= + Controls default standard output and error + output for services, with the same effect as the + and + command line + arguments described above, respectively. + + + + systemd.setenv= + + Takes a string argument in the form + VARIABLE=VALUE. May be used to set default environment + variables to add to forked child processes. May be used more + than once to set multiple variables. + + + + systemd.machine_id= + + Takes a 32 character hex value to be + used for setting the machine-id. Intended mostly for + network booting where the same machine-id is desired + for every boot. + + + + quiet + + Turn off status output at boot, much like + systemd.show_status=false would. Note that + this option is also read by the kernel itself and disables + kernel log output. Passing this option hence turns off the + usual output from both the system manager and the kernel. + + + + + debug + + Turn on debugging output. This is equivalent + to systemd.log_level=debug. Note that this + option is also read by the kernel itself and enables kernel + debug output. Passing this option hence turns on the debug + output from both the system manager and the + kernel. + + + + emergency + rd.emergency + -b + + Boot into emergency mode. This is equivalent + to systemd.unit=emergency.target or + rd.systemd.unit=emergency.target, respectively, and + provided for compatibility reasons and to be easier to type. + + + + rescue + rd.rescue + single + s + S + 1 + + Boot into rescue mode. This is equivalent to + systemd.unit=rescue.target or + rd.systemd.unit=rescue.target, respectively, and + provided for compatibility reasons and to be easier to type. + + + + 2 + 3 + 4 + 5 + + Boot into the specified legacy SysV runlevel. + These are equivalent to + systemd.unit=runlevel2.target, + systemd.unit=runlevel3.target, + systemd.unit=runlevel4.target, and + systemd.unit=runlevel5.target, + respectively, and provided for compatibility reasons and to be + easier to type. + + + + locale.LANG= + locale.LANGUAGE= + locale.LC_CTYPE= + locale.LC_NUMERIC= + locale.LC_TIME= + locale.LC_COLLATE= + locale.LC_MONETARY= + locale.LC_MESSAGES= + locale.LC_PAPER= + locale.LC_NAME= + locale.LC_ADDRESS= + locale.LC_TELEPHONE= + locale.LC_MEASUREMENT= + locale.LC_IDENTIFICATION= + + Set the system locale to use. This overrides + the settings in /etc/locale.conf. For + more information, see + locale.conf5 + and + locale7. + + + + + For other kernel command line parameters understood by + components of the core OS, please refer to + kernel-command-line7. + + + + Sockets and FIFOs + + + + /run/systemd/notify + + Daemon status notification socket. This is an + AF_UNIX datagram socket and is used to + implement the daemon notification logic as implemented by + sd_notify3. + + + + + /run/systemd/private + + Used internally as communication channel + between + systemctl1 + and the systemd process. This is an + AF_UNIX stream socket. This interface is + private to systemd and should not be used in external + projects. + + + + /dev/initctl + + Limited compatibility support for the SysV + client interface, as implemented by the + systemd-initctl.service unit. This is a + named pipe in the file system. This interface is obsolete and + should not be used in new applications. + + + + + + See Also + + The systemd Homepage, + systemd-system.conf5, + locale.conf5, + systemctl1, + journalctl1, + systemd-notify1, + daemon7, + sd-daemon3, + systemd.unit5, + systemd.special5, + pkg-config1, + kernel-command-line7, + bootup7, + systemd.directives7 + + + + diff --git a/src/grp-system/systemd/triggers.systemd.in b/src/grp-system/systemd/triggers.systemd.in new file mode 100644 index 0000000000..0d8c303136 --- /dev/null +++ b/src/grp-system/systemd/triggers.systemd.in @@ -0,0 +1,66 @@ +# -*- Mode: rpm-spec; indent-tabs-mode: nil -*- */ +# +# This file is part of systemd. +# +# Copyright 2015 Zbigniew Jędrzejewski-Szmek +# +# systemd is free software; you can redistribute it and/or modify it +# under the terms of the GNU Lesser General Public License as published by +# the Free Software Foundation; either version 2.1 of the License, or +# (at your option) any later version. +# +# systemd is distributed in the hope that it will be useful, but +# WITHOUT ANY WARRANTY; without even the implied warranty of +# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU +# Lesser General Public License for more details. +# +# You should have received a copy of the GNU Lesser General Public License +# along with systemd; If not, see . + +# The contents of this are an example to be copied into systemd.spec. +# +# Minimum rpm version supported: 4.13.0 + +%transfiletriggerin -P 900900 -p -- @systemunitdir@ /etc/systemd/system +-- This script will run after any package is initially installed or +-- upgraded. We care about the case where a package is initially +-- installed, because other cases are covered by the *un scriptlets, +-- so sometimes we will reload needlessly. + +pid = posix.fork() +if pid == 0 then + assert(posix.exec("%{_bindir}/systemctl", "daemon-reload")) +elseif pid > 0 then + posix.wait(pid) +end + +%transfiletriggerun -p -- @systemunitdir@ /etc/systemd/system +-- On removal, we need to run daemon-reload after any units have been +-- removed. %transfiletriggerpostun would be ideal, but it does not get +-- executed for some reason. +-- On upgrade, we need to run daemon-reload after any new unit files +-- have been installed, but before %postun scripts in packages get +-- executed. %transfiletriggerun gets the right list of files +-- but it is invoked too early (before changes happen). +-- %filetriggerpostun happens at the right time, but it fires for +-- every package. +-- To execute the reload at the right time, we create a state +-- file in %transfiletriggerun and execute the daemon-reload in +-- the first %filetriggerpostun. + +posix.mkdir("%{_localstatedir}/lib") +posix.mkdir("%{_localstatedir}/lib/rpm-state") +posix.mkdir("%{_localstatedir}/lib/rpm-state/systemd") +io.open("%{_localstatedir}/lib/rpm-state/systemd/needs-reload", "w") + +%filetriggerpostun -P 1000100 -p -- @systemunitdir@ /etc/systemd/system +if posix.access("%{_localstatedir}/lib/rpm-state/systemd/needs-reload") then + posix.unlink("%{_localstatedir}/lib/rpm-state/systemd/needs-reload") + posix.rmdir("%{_localstatedir}/lib/rpm-state/systemd") + pid = posix.fork() + if pid == 0 then + assert(posix.exec("%{_bindir}/systemctl", "daemon-reload")) + elseif pid > 0 then + posix.wait(pid) + end +end diff --git a/src/grp-system/systemd/user.conf b/src/grp-system/systemd/user.conf new file mode 100644 index 0000000000..b427f1ef6d --- /dev/null +++ b/src/grp-system/systemd/user.conf @@ -0,0 +1,44 @@ +# This file is part of systemd. +# +# systemd is free software; you can redistribute it and/or modify it +# under the terms of the GNU Lesser General Public License as published by +# the Free Software Foundation; either version 2.1 of the License, or +# (at your option) any later version. +# +# You can override the directives in this file by creating files in +# /etc/systemd/user.conf.d/*.conf. +# +# See systemd-user.conf(5) for details + +[Manager] +#LogLevel=info +#LogTarget=console +#LogColor=yes +#LogLocation=no +#SystemCallArchitectures= +#TimerSlackNSec= +#DefaultTimerAccuracySec=1min +#DefaultStandardOutput=inherit +#DefaultStandardError=inherit +#DefaultTimeoutStartSec=90s +#DefaultTimeoutStopSec=90s +#DefaultRestartSec=100ms +#DefaultStartLimitIntervalSec=10s +#DefaultStartLimitBurst=5 +#DefaultEnvironment= +#DefaultLimitCPU= +#DefaultLimitFSIZE= +#DefaultLimitDATA= +#DefaultLimitSTACK= +#DefaultLimitCORE= +#DefaultLimitRSS= +#DefaultLimitNOFILE= +#DefaultLimitAS= +#DefaultLimitNPROC= +#DefaultLimitMEMLOCK= +#DefaultLimitLOCKS= +#DefaultLimitSIGPENDING= +#DefaultLimitMSGQUEUE= +#DefaultLimitNICE= +#DefaultLimitRTPRIO= +#DefaultLimitRTTIME= -- cgit v1.2.3-54-g00ecf