./tools/notsd-move

13 files changed, 1388 insertions, 0 deletions

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 <http://www.gnu.org/licenses/>.
+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 <http://www.gnu.org/licenses/>.
+***/
+
+#include <errno.h>
+#include <getopt.h>
+#include <signal.h>
+#include <stdbool.h>
+#include <stdlib.h>
+#include <sys/mman.h>
+#include <sys/mount.h>
+#include <sys/reboot.h>
+#include <sys/stat.h>
+#include <unistd.h>
+
+#include <linux/reboot.h>
+
+#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", &param);
+                        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 @@
+<?xml version='1.0'?> <!--*-nxml-*-->
+<!DOCTYPE refentry PUBLIC "-//OASIS//DTD DocBook XML V4.2//EN"
+  "http://www.oasis-open.org/docbook/xml/4.2/docbookx.dtd">
+
+<!--
+  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 <http://www.gnu.org/licenses/>.
+-->
+
+<refentry id="systemd-halt.service">
+
+  <refentryinfo>
+    <title>systemd-halt.service</title>
+    <productname>systemd</productname>
+
+    <authorgroup>
+      <author>
+        <contrib>Developer</contrib>
+        <firstname>Lennart</firstname>
+        <surname>Poettering</surname>
+        <email>lennart@poettering.net</email>
+      </author>
+    </authorgroup>
+  </refentryinfo>
+
+  <refmeta>
+    <refentrytitle>systemd-halt.service</refentrytitle>
+    <manvolnum>8</manvolnum>
+  </refmeta>
+
+  <refnamediv>
+    <refname>systemd-halt.service</refname>
+    <refname>systemd-poweroff.service</refname>
+    <refname>systemd-reboot.service</refname>
+    <refname>systemd-kexec.service</refname>
+    <refname>systemd-shutdown</refname>
+    <refpurpose>System shutdown logic</refpurpose>
+  </refnamediv>
+
+  <refsynopsisdiv>
+    <para><filename>systemd-halt.service</filename></para>
+    <para><filename>systemd-poweroff.service</filename></para>
+    <para><filename>systemd-reboot.service</filename></para>
+    <para><filename>systemd-kexec.service</filename></para>
+    <para><filename>/usr/lib/systemd/systemd-shutdown</filename></para>
+    <para><filename>/usr/lib/systemd/system-shutdown/</filename></para>
+  </refsynopsisdiv>
+
+  <refsect1>
+    <title>Description</title>
+
+    <para><filename>systemd-halt.service</filename> is a system
+    service that is pulled in by <filename>halt.target</filename> and
+    is responsible for the actual system halt. Similarly,
+    <filename>systemd-poweroff.service</filename> is pulled in by
+    <filename>poweroff.target</filename>,
+    <filename>systemd-reboot.service</filename> by
+    <filename>reboot.target</filename> and
+    <filename>systemd-kexec.service</filename> by
+    <filename>kexec.target</filename> to execute the respective
+    actions.</para>
+
+    <para>When these services are run, they ensure that PID 1 is
+    replaced by the
+    <filename>/usr/lib/systemd/systemd-shutdown</filename> 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.</para>
+
+    <para>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.</para>
+
+    <para>Immediately before executing the actual system
+    halt/poweroff/reboot/kexec <filename>systemd-shutdown</filename>
+    will run all executables in
+    <filename>/usr/lib/systemd/system-shutdown/</filename> and pass
+    one arguments to them: either <literal>halt</literal>,
+    <literal>poweroff</literal>, <literal>reboot</literal> or
+    <literal>kexec</literal>, 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.</para>
+
+    <para>Note that <filename>systemd-halt.service</filename> (and the
+    related units) should never be executed directly. Instead, trigger
+    system shutdown with a command such as <literal>systemctl
+    halt</literal> or suchlike.</para>
+  </refsect1>
+
+  <refsect1>
+    <title>See Also</title>
+    <para>
+      <citerefentry><refentrytitle>systemd</refentrytitle><manvolnum>1</manvolnum></citerefentry>,
+      <citerefentry><refentrytitle>systemctl</refentrytitle><manvolnum>1</manvolnum></citerefentry>,
+      <citerefentry><refentrytitle>systemd.special</refentrytitle><manvolnum>7</manvolnum></citerefentry>,
+      <citerefentry><refentrytitle>reboot</refentrytitle><manvolnum>2</manvolnum></citerefentry>,
+      <citerefentry><refentrytitle>systemd-suspend.service</refentrytitle><manvolnum>8</manvolnum></citerefentry>
+    </para>
+  </refsect1>
+
+</refentry>
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 <http://www.gnu.org/licenses/>.
+***/
+
+#include <errno.h>
+#include <fcntl.h>
+#include <string.h>
+#include <sys/mount.h>
+#include <sys/swap.h>
+
+#include <linux/dm-ioctl.h>
+#include <linux/loop.h>
+
+#include <libudev.h>
+
+#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 <http://www.gnu.org/licenses/>.
+***/
+
+int umount_all(bool *changed);
+
+int swapoff_all(bool *changed);
+
+int loopback_detach_all(bool *changed);
+
+int dm_detach_all(bool *changed);