diff options
40 files changed, 699 insertions, 344 deletions
@@ -74,3 +74,18 @@ Thomas H. P. Andersen <phomes@gmail.com> Michael Olbrich <m.olbrich@pengutronix.de> Douglas Christman <DouglasChristman@gmail.com> Alexander Kuleshov <kuleshovmail@gmail.com> <0xAX@users.noreply.github.com> +Andreas Henriksson <andreas@fatal.se> +Daniel Rusek <mail@asciiwolf.com> +Dennis Wassenberg <dennis.wassenberg@secunet.com> +Reid Price <reid.price@gmail.com> +Stefan Schweter <stefan@schweter.it> +Seraphime Kirkovski <kirkseraph@gmail.com> +Bart Rulon <barron@lexmark.com> +Richard W.M. Jones <rjones@redhat.com> +Roman Stingler <coolx67@gmx.at> +Michael Hoy <rimmington@gmail.com> +Tiago Levit <liamgliam@gmail.com> +Eric Cook <llua@users.noreply.github.com> +Lukáš Nykrýn <lnykryn@redhat.com> +Heikki Kemppainen <heikki.kemppainen@nokia.com> +Hendrik Brueckner <hbrueckner@users.noreply.github.com> diff --git a/Makefile.am b/Makefile.am index a707d4a899..34c2f60330 100644 --- a/Makefile.am +++ b/Makefile.am @@ -42,9 +42,9 @@ LIBUDEV_CURRENT=7 LIBUDEV_REVISION=5 LIBUDEV_AGE=6 -LIBSYSTEMD_CURRENT=16 +LIBSYSTEMD_CURRENT=17 LIBSYSTEMD_REVISION=0 -LIBSYSTEMD_AGE=16 +LIBSYSTEMD_AGE=17 # Dirs of external packages dbuspolicydir=@dbuspolicydir@ @@ -311,6 +311,44 @@ CHANGES WITH 232 in spe * "systemctl is-enabled --full" will now show by which symlinks a unit file is enabled in the unit dependency tree. + * Support for VeraCrypt encrypted partitions has been added to the + "cryptsetup" logic and /etc/crypttab. + + * systemd-detect-virt gained support for a new --private-users switch + that checks whether the invoking processes are running inside a user + namespace. Similar, a new special value "private-users" for the + existing ConditionVirtualization= setting has been added, permitting + skipping of specific units in user namespace environments. + + Contributions from: Alban Crequy, Alexander Kuleshov, Alfie John, + Andreas Henriksson, Andrew Jeddeloh, Balázs Úr, Bart Rulon, Benjamin + Richter, Ben Gamari, Ben Harris, Brian J. Murrell, Christian Brauner, + Christian Rebischke, Clinton Roy, Colin Walters, Cristian Rodríguez, + Daniel Hahler, Daniel Mack, Daniel Maixner, Daniel Rusek, Dan Dedrick, + Davide Cavalca, David Herrmann, David Michael, Dennis Wassenberg, + Djalal Harouni, Dongsu Park, Douglas Christman, Elias Probst, Eric + Cook, Erik Karlsson, Evgeny Vereshchagin, Felipe Sateler, Felix Zhang, + Franck Bui, George Hilliard, Giuseppe Scrivano, HATAYAMA Daisuke, + Heikki Kemppainen, Hendrik Brueckner, hi117, Ismo Puustinen, Ivan + Shapovalov, Jakub Filak, Jakub Wilk, Jan Synacek, Jason Kölker, + Jean-Sébastien Bour, Jiří Pírko, Jonathan Boulle, Jorge Niedbalski, + Keith Busch, kristbaum, Kyle Russell, Lans Zhang, Lennart Poettering, + Leonardo Brondani Schenkel, Lucas Werkmeister, Luca Bruno, Lukáš + Nykrýn, Maciek Borzecki, Mantas Mikulėnas, Marc-Antoine Perennou, + Marcel Holtmann, Marcos Mello, Martin Ejdestig, Martin Pitt, Matej + Habrnal, Maxime de Roucy, Michael Biebl, Michael Chapman, Michael Hoy, + Michael Olbrich, Michael Pope, Michal Sekletar, Michal Soltys, Mike + Gilbert, Nick Owens, Patrik Flykt, Paweł Szewczyk, Peter Hutterer, + Piotr Drąg, Reid Price, Richard W.M. Jones, Roman Stingler, Ronny + Chevalier, Seraphime Kirkovski, Stefan Schweter, Steve Muir, Susant + Sahani, Tejun Heo, Thomas Blume, Thomas H. P. Andersen, Tiago Levit, + Tobias Jungel, Tomáš Janoušek, Topi Miettinen, Torstein Husebø, Umut + Tezduyar Lindskog, Vito Caputo, WaLyong Cho, Wilhelm Schuster, Yann + E. MORIN, Yi EungJun, Yuki Inoguchi, Yu Watanabe, Zbigniew + Jędrzejewski-Szmek, Zeal Jagannatha + + — Santa Fe, 2016-11-XX + CHANGES WITH 231: * In service units the various ExecXYZ= settings have been extended @@ -32,6 +32,11 @@ Janitorial Clean-ups: Features: +* drop nss-myhostname in favour of nss-resolve? + +* drop internal dlopen() based nss-dns fallback in nss-resolve, and rely on the + external nsswitch.conf based one + * add a percentage syntax for TimeoutStopSec=, e.g. TimeoutStopSec=150%, and then use that for the setting used in user@.service. It should be understood relative to the configured default value. diff --git a/configure.ac b/configure.ac index 0e87adc38f..0b10fc7de7 100644 --- a/configure.ac +++ b/configure.ac @@ -20,7 +20,7 @@ AC_PREREQ([2.64]) AC_INIT([systemd], - [231], + [232], [http://github.com/systemd/systemd/issues], [systemd], [http://www.freedesktop.org/wiki/Software/systemd]) diff --git a/man/sd_notify.xml b/man/sd_notify.xml index 025fbec6c1..94542b80b8 100644 --- a/man/sd_notify.xml +++ b/man/sd_notify.xml @@ -205,28 +205,25 @@ <varlistentry> <term>FDSTORE=1</term> - <listitem><para>Stores additional file descriptors in the - service manager. File descriptors sent this way will be - maintained per-service by the service manager and be passed - again using the usual file descriptor passing logic on the - next invocation of the service (see - <citerefentry><refentrytitle>sd_listen_fds</refentrytitle><manvolnum>3</manvolnum></citerefentry>). - This is useful for implementing service restart schemes where - services serialize their state to <filename>/run</filename>, - push their file descriptors to the system manager, and are - then restarted, retrieving their state again via socket - passing and <filename>/run</filename>. Note that the service - manager will accept messages for a service only if - <varname>FileDescriptorStoreMax=</varname> is set to non-zero - for it (defaults to zero). See - <citerefentry><refentrytitle>systemd.service</refentrytitle><manvolnum>5</manvolnum></citerefentry> - for details. Multiple arrays of file descriptors may be sent - in separate messages, in which case the arrays are combined. - Note that the service manager removes duplicate file - descriptors before passing them to the service. Use - <function>sd_pid_notify_with_fds()</function> to send messages - with <literal>FDSTORE=1</literal>, see - below.</para></listitem> + <listitem><para>Stores additional file descriptors in the service manager. File + descriptors sent this way will be maintained per-service by the service manager + and will be passed again using the usual file descriptor passing logic on the next + invocation of the service, see + <citerefentry><refentrytitle>sd_listen_fds</refentrytitle><manvolnum>3</manvolnum></citerefentry>. + This is useful for implementing service restart schemes where services serialize + their state to <filename>/run</filename>, push their file descriptors to the + system manager, and are then restarted, retrieving their state again via socket + passing and <filename>/run</filename>. Note that the service manager will accept + messages for a service only if <varname>FileDescriptorStoreMax=</varname> is set + to non-zero for it (defaults to zero, see + <citerefentry><refentrytitle>systemd.service</refentrytitle><manvolnum>5</manvolnum></citerefentry>). + File descriptors must be pollable, see + <citerefentry><refentrytitle>epoll_ctl</refentrytitle><manvolnum>2</manvolnum></citerefentry>. + Multiple arrays of file descriptors may be sent in separate messages, in which + case the arrays are combined. Note that the service manager removes duplicate + file descriptors before passing them to the service. Use + <function>sd_pid_notify_with_fds()</function> to send messages with + <literal>FDSTORE=1</literal>, see below.</para></listitem> </varlistentry> <varlistentry> diff --git a/man/systemctl.xml b/man/systemctl.xml index ae3296d132..dfa00e0c03 100644 --- a/man/systemctl.xml +++ b/man/systemctl.xml @@ -877,8 +877,8 @@ kobject-uevent 1 systemd-udevd-kernel.socket systemd-udevd.service <para>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 of the unit are shown, and if a job ID is + specified, properties of the job are shown. By default, empty properties are suppressed. Use <option>--all</option> to show those too. To select specific properties to show, use <option>--property=</option>. This command is intended to be diff --git a/man/systemd.exec.xml b/man/systemd.exec.xml index d3a19c505d..3c350df11f 100644 --- a/man/systemd.exec.xml +++ b/man/systemd.exec.xml @@ -1255,30 +1255,28 @@ <varlistentry> <term><varname>SystemCallFilter=</varname></term> - <listitem><para>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 - <constant>SIGSYS</constant> signal (whitelisting). If the - first character of the list is <literal>~</literal>, 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 - <constant>CAP_SYS_ADMIN</constant> capability (e.g. setting - <varname>User=nobody</varname>), - <varname>NoNewPrivileges=yes</varname> 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 - <function>execve</function>, - <function>rt_sigreturn</function>, - <function>sigreturn</function>, - <function>exit_group</function>, <function>exit</function> - 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 <literal>+</literal>.</para> + <listitem><para>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 + <constant>SIGSYS</constant> signal (whitelisting). If the first character of the list is <literal>~</literal>, + 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 <constant>CAP_SYS_ADMIN</constant> + capability (e.g. setting <varname>User=nobody</varname>), <varname>NoNewPrivileges=yes</varname> 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 <function>execve</function>, + <function>exit</function>, <function>exit_group</function>, <function>getrlimit</function>, + <function>rt_sigreturn</function>, <function>sigreturn</function> system calls and the system calls for + querying time and sleeping 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 + <literal>+</literal>.</para> + + <para>Note that strict system call filters may impact execution and error handling code paths of the service + invocation. Specifically, access to the <function>execve</function> system call is required for the execution + of the service binary — if it is blocked service invocation will necessarily fail. Also, if execution of the + service binary fails for some reason (for example: missing service executable), the error handling logic might + require access to an additional set of system calls in order to process and log this failure correctly. It + might be necessary to temporarily disable system call filters in order to simplify debugging of such + failures.</para> <para>If you specify both types of this option (i.e. whitelisting and blacklisting), the first encountered will @@ -1312,6 +1310,10 @@ </thead> <tbody> <row> + <entry>@basic-io</entry> + <entry>System calls for basic I/O: reading, writing, seeking, file descriptor duplication and closing (<citerefentry project='man-pages'><refentrytitle>read</refentrytitle><manvolnum>2</manvolnum></citerefentry>, <citerefentry project='man-pages'><refentrytitle>write</refentrytitle><manvolnum>2</manvolnum></citerefentry>, and related calls)</entry> + </row> + <row> <entry>@clock</entry> <entry>System calls for changing the system clock (<citerefentry project='man-pages'><refentrytitle>adjtimex</refentrytitle><manvolnum>2</manvolnum></citerefentry>, <citerefentry project='man-pages'><refentrytitle>settimeofday</refentrytitle><manvolnum>2</manvolnum></citerefentry>, and related calls)</entry> </row> @@ -1329,7 +1331,7 @@ </row> <row> <entry>@ipc</entry> - <entry>SysV IPC, POSIX Message Queues or other IPC (<citerefentry project='man-pages'><refentrytitle>mq_overview</refentrytitle><manvolnum>7</manvolnum></citerefentry>, <citerefentry project='man-pages'><refentrytitle>svipc</refentrytitle><manvolnum>7</manvolnum></citerefentry>)</entry> + <entry>Pipes, SysV IPC, POSIX Message Queues and other IPC (<citerefentry project='man-pages'><refentrytitle>mq_overview</refentrytitle><manvolnum>7</manvolnum></citerefentry>, <citerefentry project='man-pages'><refentrytitle>svipc</refentrytitle><manvolnum>7</manvolnum></citerefentry>)</entry> </row> <row> <entry>@keyring</entry> @@ -1357,17 +1359,21 @@ </row> <row> <entry>@process</entry> - <entry>Process control, execution, namespaces (<citerefentry project='man-pages'><refentrytitle>execve</refentrytitle><manvolnum>2</manvolnum></citerefentry>, <citerefentry project='man-pages'><refentrytitle>kill</refentrytitle><manvolnum>2</manvolnum></citerefentry>, <citerefentry project='man-pages'><refentrytitle>namespaces</refentrytitle><manvolnum>7</manvolnum></citerefentry>, …</entry> + <entry>Process control, execution, namespaces (<citerefentry project='man-pages'><refentrytitle>clone</refentrytitle><manvolnum>2</manvolnum></citerefentry>, <citerefentry project='man-pages'><refentrytitle>kill</refentrytitle><manvolnum>2</manvolnum></citerefentry>, <citerefentry project='man-pages'><refentrytitle>namespaces</refentrytitle><manvolnum>7</manvolnum></citerefentry>, …</entry> </row> <row> <entry>@raw-io</entry> - <entry>Raw I/O port access (<citerefentry project='man-pages'><refentrytitle>ioperm</refentrytitle><manvolnum>2</manvolnum></citerefentry>, <citerefentry project='man-pages'><refentrytitle>iopl</refentrytitle><manvolnum>2</manvolnum></citerefentry>, <function>pciconfig_read()</function>, …</entry> + <entry>Raw I/O port access (<citerefentry project='man-pages'><refentrytitle>ioperm</refentrytitle><manvolnum>2</manvolnum></citerefentry>, <citerefentry project='man-pages'><refentrytitle>iopl</refentrytitle><manvolnum>2</manvolnum></citerefentry>, <function>pciconfig_read()</function>, …)</entry> + </row> + <row> + <entry>@resources</entry> + <entry>System calls for changing resource limits, memory and scheduling parameters (<citerefentry project='man-pages'><refentrytitle>setrlimit</refentrytitle><manvolnum>2</manvolnum></citerefentry>, <citerefentry project='man-pages'><refentrytitle>setpriority</refentrytitle><manvolnum>2</manvolnum></citerefentry>, …)</entry> </row> </tbody> </tgroup> </table> - Note, that as new system calls are added to the kernel, additional system calls might be added to the groups + 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.</para> <para>It is recommended to combine the file system namespacing related options with @@ -1396,28 +1402,25 @@ <varlistentry> <term><varname>SystemCallArchitectures=</varname></term> - <listitem><para>Takes a space-separated list of architecture - identifiers to include in the system call filter. The known - architecture identifiers are <constant>x86</constant>, - <constant>x86-64</constant>, <constant>x32</constant>, - <constant>arm</constant>, <constant>s390</constant>, - <constant>s390x</constant> as well as the special identifier - <constant>native</constant>. 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 <constant>native</constant> 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 <constant>CAP_SYS_ADMIN</constant> - capability (e.g. setting <varname>User=nobody</varname>), - <varname>NoNewPrivileges=yes</varname> is implied. Note - that setting this option to a non-empty list implies that - <constant>native</constant> is included too. By default, this - option is set to the empty list, i.e. no architecture system - call filtering is applied.</para></listitem> + <listitem><para>Takes a space-separated list of architecture identifiers to + include in the system call filter. The known architecture identifiers are the same + as for <varname>ConditionArchitecture=</varname> described in + <citerefentry><refentrytitle>systemd.unit</refentrytitle><manvolnum>5</manvolnum></citerefentry>, + as well as <constant>x32</constant>, <constant>mips64-n32</constant>, + <constant>mips64-le-n32</constant>, and the special identifier + <constant>native</constant>. 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 + <constant>native</constant> 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 + <constant>CAP_SYS_ADMIN</constant> capability (e.g. setting + <varname>User=nobody</varname>), <varname>NoNewPrivileges=yes</varname> is + implied. Note that setting this option to a non-empty list implies that + <constant>native</constant> is included too. By default, this option is set to the + empty list, i.e. no architecture system call filtering is applied. + </para></listitem> </varlistentry> <varlistentry> diff --git a/man/systemd.service.xml b/man/systemd.service.xml index 90b1312603..5c65957bda 100644 --- a/man/systemd.service.xml +++ b/man/systemd.service.xml @@ -852,13 +852,13 @@ serialized to <filename>/run</filename> 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 + may be stored in the service manager. 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.</para></listitem> + and no job is queued or being executed for it.</para></listitem> </varlistentry> <varlistentry> diff --git a/src/basic/log.c b/src/basic/log.c index 2ff70be255..4919d175da 100644 --- a/src/basic/log.c +++ b/src/basic/log.c @@ -782,7 +782,7 @@ static void log_assert( return; DISABLE_WARNING_FORMAT_NONLITERAL; - xsprintf(buffer, format, text, file, line, func); + snprintf(buffer, sizeof buffer, format, text, file, line, func); REENABLE_WARNING; log_abort_msg = buffer; diff --git a/src/basic/mount-util.c b/src/basic/mount-util.c index 2985cc475a..c8f8022578 100644 --- a/src/basic/mount-util.c +++ b/src/basic/mount-util.c @@ -525,6 +525,7 @@ bool fstype_is_network(const char *fstype) { "glusterfs\0" "pvfs2\0" /* OrangeFS */ "ocfs2\0" + "lustre\0" ; const char *x; diff --git a/src/basic/special.h b/src/basic/special.h index 084d3dfa23..5276bcf598 100644 --- a/src/basic/special.h +++ b/src/basic/special.h @@ -117,3 +117,6 @@ /* The scope unit systemd itself lives in. */ #define SPECIAL_INIT_SCOPE "init.scope" + +/* The root directory. */ +#define SPECIAL_ROOT_MOUNT "-.mount" diff --git a/src/core/dbus-unit.c b/src/core/dbus-unit.c index 8f34fa1a52..69e249c844 100644 --- a/src/core/dbus-unit.c +++ b/src/core/dbus-unit.c @@ -263,10 +263,7 @@ static int property_get_can_stop( 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); + return sd_bus_message_append(reply, "b", unit_can_stop(u) && !u->refuse_manual_stop); } static int property_get_can_reload( @@ -760,6 +757,7 @@ const sd_bus_vtable bus_unit_vtable[] = { 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("Perpetual", "b", bus_property_get_bool, offsetof(Unit, perpetual), 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("StartLimitBurst", "u", bus_property_get_unsigned, offsetof(Unit, start_limit.burst), SD_BUS_VTABLE_PROPERTY_CONST), SD_BUS_PROPERTY("StartLimitAction", "s", property_get_emergency_action, offsetof(Unit, start_limit_action), SD_BUS_VTABLE_PROPERTY_CONST), diff --git a/src/core/device.c b/src/core/device.c index 8a3e888e5e..bd87a447cd 100644 --- a/src/core/device.c +++ b/src/core/device.c @@ -331,11 +331,7 @@ static int device_setup_unit(Manager *m, struct udev_device *dev, const char *pa if (!u) { delete = true; - u = unit_new(m, sizeof(Device)); - if (!u) - return log_oom(); - - r = unit_add_name(u, e); + r = unit_new_for_name(m, sizeof(Device), e, &u); if (r < 0) goto fail; diff --git a/src/core/execute.c b/src/core/execute.c index ae9df41b99..3f053602b5 100644 --- a/src/core/execute.c +++ b/src/core/execute.c @@ -787,6 +787,20 @@ static int get_fixed_supplementary_groups(const ExecContext *c, assert(c); + /* + * If user is given, then lookup GID and supplementary groups list. + * We avoid NSS lookups for gid=0. Also we have to initialize groups + * as early as possible so we keep the list of supplementary groups + * of the caller. + */ + if (user && gid_is_valid(gid) && gid != 0) { + /* First step, initialize groups from /etc/groups */ + if (initgroups(user, gid) < 0) + return -errno; + + keep_groups = true; + } + if (!c->supplementary_groups) return 0; @@ -803,18 +817,6 @@ static int get_fixed_supplementary_groups(const ExecContext *c, return -EOPNOTSUPP; /* For all other values */ } - /* - * If user is given, then lookup GID and supplementary group list. - * We avoid NSS lookups for gid=0. - */ - if (user && gid_is_valid(gid) && gid != 0) { - /* First step, initialize groups from /etc/groups */ - if (initgroups(user, gid) < 0) - return -errno; - - keep_groups = true; - } - l_gids = new(gid_t, ngroups_max); if (!l_gids) return -ENOMEM; @@ -2538,12 +2540,6 @@ static int exec_child( (void) umask(context->umask); if ((params->flags & EXEC_APPLY_PERMISSIONS) && !command->privileged) { - r = setup_smack(context, command); - if (r < 0) { - *exit_status = EXIT_SMACK_PROCESS_LABEL; - return r; - } - if (context->pam_name && username) { r = setup_pam(context->pam_name, username, uid, gid, context->tty_path, &accum_env, fds, n_fds); if (r < 0) { @@ -2577,7 +2573,7 @@ static int exec_child( return r; } - /* Drop group as early as possbile */ + /* Drop groups as early as possbile */ if ((params->flags & EXEC_APPLY_PERMISSIONS) && !command->privileged) { r = enforce_groups(context, gid, supplementary_gids, ngids); if (r < 0) { @@ -2693,6 +2689,41 @@ static int exec_child( } } + /* Apply the MAC contexts late, but before seccomp syscall filtering, as those should really be last to + * influence our own codepaths as little as possible. Moreover, applying MAC contexts usually requires + * syscalls that are subject to seccomp filtering, hence should probably be applied before the syscalls + * are restricted. */ + +#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 + + r = setup_smack(context, command); + if (r < 0) { + *exit_status = EXIT_SMACK_PROCESS_LABEL; + return r; + } + +#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 + /* PR_GET_SECUREBITS is not privileged, while * PR_SET_SECUREBITS is. So to suppress * potential EPERMs we'll try not to call @@ -2758,6 +2789,8 @@ static int exec_child( } } + /* This really should remain the last step before the execve(), to make sure our own code is unaffected + * by the filter as little as possible. */ if (context_has_syscall_filters(context)) { r = apply_seccomp(unit, context); if (r < 0) { @@ -2766,30 +2799,6 @@ static int exec_child( } } #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); @@ -3611,7 +3620,8 @@ char *exec_command_line(char **argv) { STRV_FOREACH(a, argv) k += strlen(*a)+3; - if (!(n = new(char, k))) + n = new(char, k); + if (!n) return NULL; p = n; diff --git a/src/core/job.c b/src/core/job.c index 3ecc8a1a73..ac6910a906 100644 --- a/src/core/job.c +++ b/src/core/job.c @@ -690,16 +690,16 @@ _pure_ static const char *job_get_status_message_format(Unit *u, JobType t, JobR } static void job_print_status_message(Unit *u, JobType t, JobResult result) { - static struct { + static const 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"}, + [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; @@ -767,8 +767,9 @@ static void job_log_status_message(Unit *u, JobType t, JobResult result) { if (!format) return; + /* The description might be longer than the buffer, but that's OK, we'll just truncate it here */ DISABLE_WARNING_FORMAT_NONLITERAL; - xsprintf(buf, format, unit_description(u)); + snprintf(buf, sizeof(buf), format, unit_description(u)); REENABLE_WARNING; switch (t) { diff --git a/src/core/mount.c b/src/core/mount.c index da480001e1..d749e49df5 100644 --- a/src/core/mount.c +++ b/src/core/mount.c @@ -159,17 +159,6 @@ static void mount_init(Unit *u) { 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 @@ -577,6 +566,25 @@ static int mount_add_extras(Mount *m) { return 0; } +static int mount_load_root_mount(Unit *u) { + assert(u); + + if (!unit_has_name(u, SPECIAL_ROOT_MOUNT)) + return 0; + + u->perpetual = true; + u->default_dependencies = false; + + /* The stdio/kmsg bridge socket is on /, in order to avoid a dep loop, don't use kmsg logging for -.mount */ + MOUNT(u)->exec_context.std_output = EXEC_OUTPUT_NULL; + MOUNT(u)->exec_context.std_input = EXEC_INPUT_NULL; + + if (!u->description) + u->description = strdup("Root Mount"); + + return 1; +} + static int mount_load(Unit *u) { Mount *m = MOUNT(u); int r; @@ -584,11 +592,14 @@ static int mount_load(Unit *u) { assert(u); assert(u->load_state == UNIT_STUB); - if (m->from_proc_self_mountinfo) + r = mount_load_root_mount(u); + if (r < 0) + return r; + + if (m->from_proc_self_mountinfo || u->perpetual) r = unit_load_fragment_and_dropin_optional(u); else r = unit_load_fragment_and_dropin(u); - if (r < 0) return r; @@ -1393,11 +1404,7 @@ static int mount_setup_unit( if (!u) { delete = true; - u = unit_new(m, sizeof(Mount)); - if (!u) - return log_oom(); - - r = unit_add_name(u, e); + r = unit_new_for_name(m, sizeof(Mount), e, &u); if (r < 0) goto fail; @@ -1592,11 +1599,46 @@ static int mount_get_timeout(Unit *u, usec_t *timeout) { return 1; } +static int synthesize_root_mount(Manager *m) { + Unit *u; + int r; + + assert(m); + + /* Whatever happens, we know for sure that the root directory is around, and cannot go away. Let's + * unconditionally synthesize it here and mark it as perpetual. */ + + u = manager_get_unit(m, SPECIAL_ROOT_MOUNT); + if (!u) { + r = unit_new_for_name(m, sizeof(Mount), SPECIAL_ROOT_MOUNT, &u); + if (r < 0) + return log_error_errno(r, "Failed to allocate the special " SPECIAL_ROOT_MOUNT " unit: %m"); + } + + u->perpetual = true; + MOUNT(u)->deserialized_state = MOUNT_MOUNTED; + + unit_add_to_load_queue(u); + unit_add_to_dbus_queue(u); + + return 0; +} + +static bool mount_is_mounted(Mount *m) { + assert(m); + + return UNIT(m)->perpetual || m->is_mounted; +} + static void mount_enumerate(Manager *m) { int r; assert(m); + r = synthesize_root_mount(m); + if (r < 0) + goto fail; + mnt_init_debug(0); if (!m->mount_monitor) { @@ -1703,7 +1745,7 @@ static int mount_dispatch_io(sd_event_source *source, int fd, uint32_t revents, LIST_FOREACH(units_by_type, u, m->units_by_type[UNIT_MOUNT]) { Mount *mount = MOUNT(u); - if (!mount->is_mounted) { + if (!mount_is_mounted(mount)) { /* A mount point is not around right now. It * might be gone, or might never have @@ -1764,7 +1806,7 @@ static int mount_dispatch_io(sd_event_source *source, int fd, uint32_t revents, } } - if (mount->is_mounted && + if (mount_is_mounted(mount) && mount->from_proc_self_mountinfo && mount->parameters_proc_self_mountinfo.what) { diff --git a/src/core/scope.c b/src/core/scope.c index af0c43c7da..d6e1f8e392 100644 --- a/src/core/scope.c +++ b/src/core/scope.c @@ -154,15 +154,13 @@ static int scope_load_init_scope(Unit *u) { return 0; u->transient = true; - u->no_gc = true; + u->perpetual = true; /* init.scope is a bit special, as it has to stick around forever. Because of its special semantics we * synthesize it here, instead of relying on the unit file on disk. */ u->default_dependencies = false; u->ignore_on_isolate = true; - u->refuse_manual_start = true; - u->refuse_manual_stop = true; SCOPE(u)->kill_context.kill_signal = SIGRTMIN+14; @@ -565,22 +563,15 @@ static void scope_enumerate(Manager *m) { 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); + r = unit_new_for_name(m, sizeof(Scope), SPECIAL_INIT_SCOPE, &u); if (r < 0) { - unit_free(u); - log_error_errno(r, "Failed to add the " SPECIAL_INIT_SCOPE " name: %m"); + log_error_errno(r, "Failed to allocate the special " SPECIAL_INIT_SCOPE " unit: %m"); return; } } u->transient = true; - u->no_gc = true; + u->perpetual = true; SCOPE(u)->deserialized_state = SCOPE_RUNNING; unit_add_to_load_queue(u); diff --git a/src/core/service.c b/src/core/service.c index ee4f4983fc..a7274a758f 100644 --- a/src/core/service.c +++ b/src/core/service.c @@ -289,7 +289,17 @@ static void service_fd_store_unlink(ServiceFDStore *fs) { free(fs); } -static void service_release_resources(Unit *u) { +static void service_release_fd_store(Service *s) { + assert(s); + + log_unit_debug(UNIT(s), "Releasing all stored fds"); + while (s->fd_store) + service_fd_store_unlink(s->fd_store); + + assert(s->n_fd_store == 0); +} + +static void service_release_resources(Unit *u, bool inactive) { Service *s = SERVICE(u); assert(s); @@ -297,16 +307,14 @@ static void service_release_resources(Unit *u) { if (!s->fd_store && s->stdin_fd < 0 && s->stdout_fd < 0 && s->stderr_fd < 0) return; - log_unit_debug(u, "Releasing all resources."); + log_unit_debug(u, "Releasing 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); + if (inactive) + service_release_fd_store(s); } static void service_done(Unit *u) { @@ -350,7 +358,7 @@ static void service_done(Unit *u) { s->timer_event_source = sd_event_source_unref(s->timer_event_source); - service_release_resources(u); + service_release_resources(u, true); } static int on_fd_store_io(sd_event_source *e, int fd, uint32_t revents, void *userdata) { @@ -360,6 +368,10 @@ static int on_fd_store_io(sd_event_source *e, int fd, uint32_t revents, void *us assert(fs); /* If we get either EPOLLHUP or EPOLLERR, it's time to remove this entry from the fd store */ + log_unit_debug(UNIT(fs->service), + "Received %s on stored fd %d (%s), closing.", + revents & EPOLLERR ? "EPOLLERR" : "EPOLLHUP", + fs->fd, strna(fs->fdname)); service_fd_store_unlink(fs); return 0; } @@ -368,20 +380,23 @@ static int service_add_fd_store(Service *s, int fd, const char *name) { ServiceFDStore *fs; int r; + /* fd is always consumed if we return >= 0 */ + assert(s); assert(fd >= 0); if (s->n_fd_store >= s->n_fd_store_max) - return 0; + return -EXFULL; /* Our store is full. + * Use this errno rather than E[NM]FILE to distinguish from + * the case where systemd itself hits the file limit. */ 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; + return 0; /* fd already included */ } } @@ -409,7 +424,7 @@ static int service_add_fd_store(Service *s, int fd, const char *name) { LIST_PREPEND(fd_store, s->fd_store, fs); s->n_fd_store++; - return 1; + return 1; /* fd newly stored */ } static int service_add_fd_store_set(Service *s, FDSet *fds, const char *name) { @@ -417,10 +432,7 @@ static int service_add_fd_store_set(Service *s, FDSet *fds, const char *name) { assert(s); - if (fdset_size(fds) <= 0) - return 0; - - while (s->n_fd_store < s->n_fd_store_max) { + while (fdset_size(fds) > 0) { _cleanup_close_ int fd = -1; fd = fdset_steal_first(fds); @@ -428,17 +440,17 @@ static int service_add_fd_store_set(Service *s, FDSet *fds, const char *name) { break; r = service_add_fd_store(s, fd, name); + if (r == -EXFULL) + return log_unit_warning_errno(UNIT(s), r, + "Cannot store more fds than FileDescriptorStoreMax=%u, closing remaining.", + s->n_fd_store_max); 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; - } + return log_unit_error_errno(UNIT(s), r, "Failed to add fd to store: %m"); + if (r > 0) + log_unit_debug(UNIT(s), "Added fd %u (%s) to fd store.", fd, strna(name)); + 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; } @@ -1225,6 +1237,7 @@ static int service_spawn( return r; n_fds = r; + log_unit_debug(UNIT(s), "Passing %i fds to service", n_fds); } r = service_arm_timer(s, usec_add(now(CLOCK_MONOTONIC), timeout)); @@ -2336,7 +2349,7 @@ static int service_deserialize_item(Unit *u, const char *key, const char *value, 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) + else fdset_remove(fds, fd); } diff --git a/src/core/slice.c b/src/core/slice.c index 0fef29661f..ed5d3fd701 100644 --- a/src/core/slice.c +++ b/src/core/slice.c @@ -136,15 +136,13 @@ static int slice_load_root_slice(Unit *u) { if (!unit_has_name(u, SPECIAL_ROOT_SLICE)) return 0; - u->no_gc = true; + u->perpetual = true; /* The root slice is a bit special. For example it is always running and cannot be terminated. Because of its * special semantics we synthesize it here, instead of relying on the unit file on disk. */ u->default_dependencies = false; u->ignore_on_isolate = true; - u->refuse_manual_start = true; - u->refuse_manual_stop = true; if (!u->description) u->description = strdup("Root Slice"); @@ -301,21 +299,14 @@ static void slice_enumerate(Manager *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); + r = unit_new_for_name(m, sizeof(Slice), SPECIAL_ROOT_SLICE, &u); if (r < 0) { - unit_free(u); - log_error_errno(r, "Failed to add the "SPECIAL_ROOT_SLICE " name: %m"); + log_error_errno(r, "Failed to allocate the special " SPECIAL_ROOT_SLICE " unit: %m"); return; } } - u->no_gc = true; + u->perpetual = true; SLICE(u)->deserialized_state = SLICE_ACTIVE; unit_add_to_load_queue(u); diff --git a/src/core/swap.c b/src/core/swap.c index b592abb9fb..2228a254bb 100644 --- a/src/core/swap.c +++ b/src/core/swap.c @@ -381,11 +381,7 @@ static int swap_setup_unit( if (!u) { delete = true; - u = unit_new(m, sizeof(Swap)); - if (!u) - return log_oom(); - - r = unit_add_name(u, e); + r = unit_new_for_name(m, sizeof(Swap), e, &u); if (r < 0) goto fail; diff --git a/src/core/unit.c b/src/core/unit.c index cabb1050a8..e664e23892 100644 --- a/src/core/unit.c +++ b/src/core/unit.c @@ -109,6 +109,24 @@ Unit *unit_new(Manager *m, size_t size) { return u; } +int unit_new_for_name(Manager *m, size_t size, const char *name, Unit **ret) { + Unit *u; + int r; + + u = unit_new(m, size); + if (!u) + return -ENOMEM; + + r = unit_add_name(u, name); + if (r < 0) { + unit_free(u); + return r; + } + + *ret = u; + return r; +} + bool unit_has_name(Unit *u, const char *name) { assert(u); assert(name); @@ -302,6 +320,7 @@ int unit_set_description(Unit *u, const char *description) { bool unit_check_gc(Unit *u) { UnitActiveState state; + bool inactive; assert(u); if (u->job) @@ -311,19 +330,20 @@ bool unit_check_gc(Unit *u) { return true; state = unit_active_state(u); + inactive = state == UNIT_INACTIVE; /* 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); + UNIT_VTABLE(u)->release_resources(u, inactive); /* But we keep the unit object around for longer when it is * referenced or configured to not be gc'ed */ - if (state != UNIT_INACTIVE) + if (!inactive) return true; - if (u->no_gc) + if (u->perpetual) return true; if (u->refs) @@ -924,6 +944,7 @@ void unit_dump(Unit *u, FILE *f, const char *prefix) { "%s\tGC Check Good: %s\n" "%s\tNeed Daemon Reload: %s\n" "%s\tTransient: %s\n" + "%s\tPerpetual: %s\n" "%s\tSlice: %s\n" "%s\tCGroup: %s\n" "%s\tCGroup realized: %s\n" @@ -942,6 +963,7 @@ void unit_dump(Unit *u, FILE *f, const char *prefix) { prefix, yes_no(unit_check_gc(u)), prefix, yes_no(unit_need_daemon_reload(u)), prefix, yes_no(u->transient), + prefix, yes_no(u->perpetual), prefix, strna(unit_slice_name(u)), prefix, strna(u->cgroup_path), prefix, yes_no(u->cgroup_realized), @@ -1450,7 +1472,7 @@ static void unit_status_log_starting_stopping_reloading(Unit *u, JobType t) { format = unit_get_status_message_format(u, t); DISABLE_WARNING_FORMAT_NONLITERAL; - xsprintf(buf, format, unit_description(u)); + snprintf(buf, sizeof buf, format, unit_description(u)); REENABLE_WARNING; mid = t == JOB_START ? SD_MESSAGE_UNIT_STARTING : @@ -1616,6 +1638,18 @@ int unit_stop(Unit *u) { return UNIT_VTABLE(u)->stop(u); } +bool unit_can_stop(Unit *u) { + assert(u); + + if (!unit_supported(u)) + return false; + + if (u->perpetual) + return false; + + return !!UNIT_VTABLE(u)->stop; +} + /* Errors: * -EBADR: This unit type does not support reloading. * -ENOEXEC: Unit is not started. @@ -2150,13 +2184,20 @@ bool unit_job_is_applicable(Unit *u, JobType j) { case JOB_VERIFY_ACTIVE: case JOB_START: - case JOB_STOP: case JOB_NOP: + /* Note that we don't check unit_can_start() here. That's because .device units and suchlike are not + * startable by us but may appear due to external events, and it thus makes sense to permit enqueing + * jobs for it. */ return true; + case JOB_STOP: + /* Similar as above. However, perpetual units can never be stopped (neither explicitly nor due to + * external events), hence it makes no sense to permit enqueing such a request either. */ + return !u->perpetual; + case JOB_RESTART: case JOB_TRY_RESTART: - return unit_can_start(u); + return unit_can_stop(u) && unit_can_start(u); case JOB_RELOAD: case JOB_TRY_RELOAD: diff --git a/src/core/unit.h b/src/core/unit.h index adcdee6db6..991543664b 100644 --- a/src/core/unit.h +++ b/src/core/unit.h @@ -236,6 +236,9 @@ struct Unit { /* Is this a transient unit? */ bool transient; + /* Is this a unit that is always running and cannot be stopped? */ + bool perpetual; + bool in_load_queue:1; bool in_dbus_queue:1; bool in_cleanup_queue:1; @@ -244,8 +247,6 @@ struct Unit { bool sent_dbus_new_signal:1; - bool no_gc:1; - bool in_audit:1; bool cgroup_realized:1; @@ -372,7 +373,7 @@ struct UnitVTable { /* When the unit is not running and no job for it queued we * shall release its runtime resources */ - void (*release_resources)(Unit *u); + void (*release_resources)(Unit *u, bool inactive); /* Invoked on every child that died */ void (*sigchld_event)(Unit *u, pid_t pid, int code, int status); @@ -480,6 +481,7 @@ DEFINE_CAST(SCOPE, Scope); Unit *unit_new(Manager *m, size_t size); void unit_free(Unit *u); +int unit_new_for_name(Manager *m, size_t size, const char *name, Unit **ret); int unit_add_name(Unit *u, const char *name); int unit_add_dependency(Unit *u, UnitDependency d, Unit *other, bool add_reference); @@ -524,6 +526,7 @@ 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_stop(Unit *u) _pure_; bool unit_can_isolate(Unit *u) _pure_; int unit_start(Unit *u); diff --git a/src/libsystemd/sd-bus/bus-kernel.c b/src/libsystemd/sd-bus/bus-kernel.c index 59398b841d..ad468572f3 100644 --- a/src/libsystemd/sd-bus/bus-kernel.c +++ b/src/libsystemd/sd-bus/bus-kernel.c @@ -848,8 +848,7 @@ static int bus_kernel_make_message(sd_bus *bus, struct kdbus_msg *k) { if (k->src_id == KDBUS_SRC_ID_KERNEL) bus_message_set_sender_driver(bus, m); else { - xsprintf(m->sender_buffer, ":1.%llu", - (unsigned long long)k->src_id); + xsprintf(m->sender_buffer, ":1.%"PRIu64, k->src_id); m->sender = m->creds.unique_name = m->sender_buffer; } @@ -860,8 +859,7 @@ static int bus_kernel_make_message(sd_bus *bus, struct kdbus_msg *k) { else if (k->dst_id == KDBUS_DST_ID_NAME) m->destination = bus->unique_name; /* fill in unique name if the well-known name is missing */ else { - xsprintf(m->destination_buffer, ":1.%llu", - (unsigned long long)k->dst_id); + xsprintf(m->destination_buffer, ":1.%"PRIu64, k->dst_id); m->destination = m->destination_buffer; } diff --git a/src/network/networkd-link.c b/src/network/networkd-link.c index d9e060b6cf..aefe7335b9 100644 --- a/src/network/networkd-link.c +++ b/src/network/networkd-link.c @@ -514,13 +514,12 @@ static void link_free(Link *link) { sd_lldp_unref(link->lldp); free(link->lldp_file); + ndisc_flush(link); + sd_ipv4ll_unref(link->ipv4ll); sd_dhcp6_client_unref(link->dhcp6_client); sd_ndisc_unref(link->ndisc); - set_free_free(link->ndisc_rdnss); - set_free_free(link->ndisc_dnssl); - if (link->manager) hashmap_remove(link->manager->links, INT_TO_PTR(link->ifindex)); @@ -2427,6 +2426,8 @@ static int link_drop_config(Link *link) { return r; } + ndisc_flush(link); + return 0; } diff --git a/src/network/networkd-ndisc.c b/src/network/networkd-ndisc.c index b282634e4b..4853791aa5 100644 --- a/src/network/networkd-ndisc.c +++ b/src/network/networkd-ndisc.c @@ -680,13 +680,22 @@ void ndisc_vacuum(Link *link) { SET_FOREACH(r, link->ndisc_rdnss, i) if (r->valid_until < time_now) { - (void) set_remove(link->ndisc_rdnss, r); + free(set_remove(link->ndisc_rdnss, r)); link_dirty(link); } SET_FOREACH(d, link->ndisc_dnssl, i) if (d->valid_until < time_now) { - (void) set_remove(link->ndisc_dnssl, d); + free(set_remove(link->ndisc_dnssl, d)); link_dirty(link); } } + +void ndisc_flush(Link *link) { + assert(link); + + /* Removes all RDNSS and DNSSL entries, without exception */ + + link->ndisc_rdnss = set_free_free(link->ndisc_rdnss); + link->ndisc_dnssl = set_free_free(link->ndisc_dnssl); +} diff --git a/src/network/networkd-ndisc.h b/src/network/networkd-ndisc.h index 2002f55107..127126190e 100644 --- a/src/network/networkd-ndisc.h +++ b/src/network/networkd-ndisc.h @@ -37,3 +37,4 @@ static inline char* NDISC_DNSSL_DOMAIN(const NDiscDNSSL *n) { int ndisc_configure(Link *link); void ndisc_vacuum(Link *link); +void ndisc_flush(Link *link); diff --git a/src/nspawn/nspawn.c b/src/nspawn/nspawn.c index 295293858e..c56af6e6f4 100644 --- a/src/nspawn/nspawn.c +++ b/src/nspawn/nspawn.c @@ -2260,7 +2260,7 @@ static int dissect_image( static int mount_device(const char *what, const char *where, const char *directory, bool rw) { #ifdef HAVE_BLKID _cleanup_blkid_free_probe_ blkid_probe b = NULL; - const char *fstype, *p; + const char *fstype, *p, *options; int r; assert(what); @@ -2309,7 +2309,17 @@ static int mount_device(const char *what, const char *where, const char *directo return -EOPNOTSUPP; } - return mount_verbose(LOG_ERR, what, p, fstype, MS_NODEV|(rw ? 0 : MS_RDONLY), NULL); + /* If this is a loopback device then let's mount the image with discard, so that the underlying file remains + * sparse when possible. */ + if (STR_IN_SET(fstype, "btrfs", "ext4", "vfat", "xfs")) { + const char *l; + + l = path_startswith(what, "/dev"); + if (l && startswith(l, "loop")) + options = "discard"; + } + + return mount_verbose(LOG_ERR, what, p, fstype, MS_NODEV|(rw ? 0 : MS_RDONLY), options); #else log_error("--image= is not supported, compiled without blkid support."); return -EOPNOTSUPP; diff --git a/src/shared/seccomp-util.c b/src/shared/seccomp-util.c index 6252cd16a6..c9b24f1065 100644 --- a/src/shared/seccomp-util.c +++ b/src/shared/seccomp-util.c @@ -29,23 +29,49 @@ #include "util.h" const char* seccomp_arch_to_string(uint32_t c) { + /* Maintain order used in <seccomp.h>. + * + * Names used here should be the same as those used for ConditionArchitecture=, + * except for "subarchitectures" like x32. */ - if (c == SCMP_ARCH_NATIVE) + switch(c) { + case SCMP_ARCH_NATIVE: return "native"; - if (c == SCMP_ARCH_X86) + case SCMP_ARCH_X86: return "x86"; - if (c == SCMP_ARCH_X86_64) + case SCMP_ARCH_X86_64: return "x86-64"; - if (c == SCMP_ARCH_X32) + case SCMP_ARCH_X32: return "x32"; - if (c == SCMP_ARCH_ARM) + case SCMP_ARCH_ARM: return "arm"; - if (c == SCMP_ARCH_S390) + case SCMP_ARCH_AARCH64: + return "arm64"; + case SCMP_ARCH_MIPS: + return "mips"; + case SCMP_ARCH_MIPS64: + return "mips64"; + case SCMP_ARCH_MIPS64N32: + return "mips64-n32"; + case SCMP_ARCH_MIPSEL: + return "mips-le"; + case SCMP_ARCH_MIPSEL64: + return "mips64-le"; + case SCMP_ARCH_MIPSEL64N32: + return "mips64-le-n32"; + case SCMP_ARCH_PPC: + return "ppc"; + case SCMP_ARCH_PPC64: + return "ppc64"; + case SCMP_ARCH_PPC64LE: + return "ppc64-le"; + case SCMP_ARCH_S390: return "s390"; - if (c == SCMP_ARCH_S390X) + case SCMP_ARCH_S390X: return "s390x"; - - return NULL; + default: + return NULL; + } } int seccomp_arch_from_string(const char *n, uint32_t *ret) { @@ -64,6 +90,26 @@ int seccomp_arch_from_string(const char *n, uint32_t *ret) { *ret = SCMP_ARCH_X32; else if (streq(n, "arm")) *ret = SCMP_ARCH_ARM; + else if (streq(n, "arm64")) + *ret = SCMP_ARCH_AARCH64; + else if (streq(n, "mips")) + *ret = SCMP_ARCH_MIPS; + else if (streq(n, "mips64")) + *ret = SCMP_ARCH_MIPS64; + else if (streq(n, "mips64-n32")) + *ret = SCMP_ARCH_MIPS64N32; + else if (streq(n, "mips-le")) + *ret = SCMP_ARCH_MIPSEL; + else if (streq(n, "mips64-le")) + *ret = SCMP_ARCH_MIPSEL64; + else if (streq(n, "mips64-le-n32")) + *ret = SCMP_ARCH_MIPSEL64N32; + else if (streq(n, "ppc")) + *ret = SCMP_ARCH_PPC; + else if (streq(n, "ppc64")) + *ret = SCMP_ARCH_PPC64; + else if (streq(n, "ppc64-le")) + *ret = SCMP_ARCH_PPC64LE; else if (streq(n, "s390")) *ret = SCMP_ARCH_S390; else if (streq(n, "s390x")) @@ -101,41 +147,52 @@ finish: return r; } -int seccomp_add_secondary_archs(scmp_filter_ctx c) { - -#if defined(__i386__) || defined(__x86_64__) - int r; +int seccomp_add_secondary_archs(scmp_filter_ctx ctx) { /* Add in all possible secondary archs we are aware of that * this kernel might support. */ - r = seccomp_arch_add(c, SCMP_ARCH_X86); - if (r < 0 && r != -EEXIST) - return r; - - r = seccomp_arch_add(c, SCMP_ARCH_X86_64); - if (r < 0 && r != -EEXIST) - return r; - - r = seccomp_arch_add(c, SCMP_ARCH_X32); - if (r < 0 && r != -EEXIST) - return r; + static const int seccomp_arches[] = { +#if defined(__i386__) || defined(__x86_64__) + SCMP_ARCH_X86, + SCMP_ARCH_X86_64, + SCMP_ARCH_X32, + +#elif defined(__arm__) || defined(__aarch64__) + SCMP_ARCH_ARM, + SCMP_ARCH_AARCH64, + +#elif defined(__arm__) || defined(__aarch64__) + SCMP_ARCH_ARM, + SCMP_ARCH_AARCH64, + +#elif defined(__mips__) || defined(__mips64__) + SCMP_ARCH_MIPS, + SCMP_ARCH_MIPS64, + SCMP_ARCH_MIPS64N32, + SCMP_ARCH_MIPSEL, + SCMP_ARCH_MIPSEL64, + SCMP_ARCH_MIPSEL64N32, + +#elif defined(__powerpc__) || defined(__powerpc64__) + SCMP_ARCH_PPC, + SCMP_ARCH_PPC64, + SCMP_ARCH_PPC64LE, #elif defined(__s390__) || defined(__s390x__) - int r; - - /* Add in all possible secondary archs we are aware of that - * this kernel might support. */ - - r = seccomp_arch_add(c, SCMP_ARCH_S390); - if (r < 0 && r != -EEXIST) - return r; + SCMP_ARCH_S390, + SCMP_ARCH_S390X, +#endif + }; - r = seccomp_arch_add(c, SCMP_ARCH_S390X); - if (r < 0 && r != -EEXIST) - return r; + unsigned i; + int r; -#endif + for (i = 0; i < ELEMENTSOF(seccomp_arches); i++) { + r = seccomp_arch_add(ctx, seccomp_arches[i]); + if (r < 0 && r != -EEXIST) + return r; + } return 0; } @@ -160,6 +217,24 @@ bool is_seccomp_available(void) { } const SyscallFilterSet syscall_filter_sets[_SYSCALL_FILTER_SET_MAX] = { + [SYSCALL_FILTER_SET_BASIC_IO] = { + /* Basic IO */ + .name = "@basic-io", + .value = + "close\0" + "dup2\0" + "dup3\0" + "dup\0" + "lseek\0" + "pread64\0" + "preadv\0" + "pwrite64\0" + "pwritev\0" + "read\0" + "readv\0" + "write\0" + "writev\0" + }, [SYSCALL_FILTER_SET_CLOCK] = { /* Clock */ .name = "@clock", @@ -196,15 +271,22 @@ const SyscallFilterSet syscall_filter_sets[_SYSCALL_FILTER_SET_MAX] = { "sys_debug_setcontext\0" }, [SYSCALL_FILTER_SET_DEFAULT] = { - /* Default list */ + /* Default list: the most basic of operations */ .name = "@default", .value = + "clock_getres\0" + "clock_gettime\0" + "clock_nanosleep\0" "execve\0" "exit\0" "exit_group\0" "getrlimit\0" /* make sure processes can query stack size and such */ + "gettimeofday\0" + "nanosleep\0" + "pause\0" "rt_sigreturn\0" "sigreturn\0" + "time\0" }, [SYSCALL_FILTER_SET_IO_EVENT] = { /* Event loop use */ @@ -226,9 +308,10 @@ const SyscallFilterSet syscall_filter_sets[_SYSCALL_FILTER_SET_MAX] = { "select\0" }, [SYSCALL_FILTER_SET_IPC] = { - /* Message queues, SYSV IPC or other IPC: unusual */ + /* Message queues, SYSV IPC or other IPC */ .name = "@ipc", .value = "ipc\0" + "memfd_create\0" "mq_getsetattr\0" "mq_notify\0" "mq_open\0" @@ -239,6 +322,8 @@ const SyscallFilterSet syscall_filter_sets[_SYSCALL_FILTER_SET_MAX] = { "msgget\0" "msgrcv\0" "msgsnd\0" + "pipe2\0" + "pipe\0" "process_vm_readv\0" "process_vm_writev\0" "semctl\0" @@ -379,7 +464,6 @@ const SyscallFilterSet syscall_filter_sets[_SYSCALL_FILTER_SET_MAX] = { .value = "arch_prctl\0" "clone\0" - "execve\0" "execveat\0" "fork\0" "kill\0" @@ -406,6 +490,22 @@ const SyscallFilterSet syscall_filter_sets[_SYSCALL_FILTER_SET_MAX] = { "s390_pci_mmio_write\0" #endif }, + [SYSCALL_FILTER_SET_RESOURCES] = { + /* Alter resource settings */ + .name = "@resources", + .value = + "sched_setparam\0" + "sched_setscheduler\0" + "sched_setaffinity\0" + "setpriority\0" + "setrlimit\0" + "set_mempolicy\0" + "migrate_pages\0" + "move_pages\0" + "mbind\0" + "sched_setattr\0" + "prlimit64\0" + }, }; const SyscallFilterSet *syscall_filter_set_find(const char *name) { diff --git a/src/shared/seccomp-util.h b/src/shared/seccomp-util.h index 8050fc6fbf..8e209efef2 100644 --- a/src/shared/seccomp-util.h +++ b/src/shared/seccomp-util.h @@ -38,6 +38,7 @@ typedef struct SyscallFilterSet { } SyscallFilterSet; enum { + SYSCALL_FILTER_SET_BASIC_IO, SYSCALL_FILTER_SET_CLOCK, SYSCALL_FILTER_SET_CPU_EMULATION, SYSCALL_FILTER_SET_DEBUG, @@ -52,6 +53,7 @@ enum { SYSCALL_FILTER_SET_PRIVILEGED, SYSCALL_FILTER_SET_PROCESS, SYSCALL_FILTER_SET_RAW_IO, + SYSCALL_FILTER_SET_RESOURCES, _SYSCALL_FILTER_SET_MAX }; diff --git a/src/shared/switch-root.c b/src/shared/switch-root.c index 47d3a5a1fa..4eff4f692e 100644 --- a/src/shared/switch-root.c +++ b/src/shared/switch-root.c @@ -75,17 +75,29 @@ int switch_root(const char *new_root, const char *oldroot, bool detach_oldroot, NULSTR_FOREACH(i, move_mounts) { char new_mount[PATH_MAX]; struct stat sb; + size_t n; - xsprintf(new_mount, "%s%s", new_root, i); + n = snprintf(new_mount, sizeof new_mount, "%s%s", new_root, i); + if (n >= sizeof new_mount) { + bool move = mountflags & MS_MOVE; + + log_warning("New path is too long, %s: %s%s", + move ? "forcing unmount instead" : "ignoring", + new_root, i); + + if (move) + if (umount2(i, MNT_FORCE) < 0) + log_warning_errno(errno, "Failed to unmount %s: %m", i); + continue; + } mkdir_p_label(new_mount, 0755); - if ((stat(new_mount, &sb) < 0) || + if (stat(new_mount, &sb) < 0 || sb.st_dev != new_root_stat.st_dev) { /* Mount point seems to be mounted already or - * stat failed. Unmount the old mount - * point. */ + * stat failed. Unmount the old mount point. */ if (umount2(i, MNT_DETACH) < 0) log_warning_errno(errno, "Failed to unmount %s: %m", i); continue; @@ -97,10 +109,9 @@ int switch_root(const char *new_root, const char *oldroot, bool detach_oldroot, if (umount2(i, MNT_FORCE) < 0) log_warning_errno(errno, "Failed to unmount %s: %m", i); - } - if (mountflags & MS_BIND) - log_error_errno(errno, "Failed to bind mount %s to %s: %m", i, new_mount); + } else if (mountflags & MS_BIND) + log_error_errno(errno, "Failed to bind mount %s to %s: %m", i, new_mount); } } diff --git a/src/sysctl/sysctl.c b/src/sysctl/sysctl.c index fbc1e0eb1a..b3587e249d 100644 --- a/src/sysctl/sysctl.c +++ b/src/sysctl/sysctl.c @@ -51,19 +51,46 @@ static int apply_all(OrderedHashmap *sysctl_options) { k = sysctl_write(property, value); if (k < 0) { - log_full_errno(k == -ENOENT ? LOG_INFO : LOG_WARNING, k, - "Couldn't write '%s' to '%s', ignoring: %m", value, property); - - if (r == 0 && k != -ENOENT) - r = k; + /* If the sysctl is not available in the kernel or we are running with reduced privileges and + * cannot write it, then log about the issue at LOG_NOTICE level, and proceed without + * failing. (EROFS is treated as a permission problem here, since that's how container managers + * usually protected their sysctls.) In all other cases log an error and make the tool fail. */ + + if (IN_SET(k, -EPERM, -EACCES, -EROFS, -ENOENT)) + log_notice_errno(k, "Couldn't write '%s' to '%s', ignoring: %m", value, property); + else { + log_error_errno(k, "Couldn't write '%s' to '%s': %m", value, property); + if (r == 0) + r = k; + } } } return r; } +static bool test_prefix(const char *p) { + char **i; + + if (strv_isempty(arg_prefixes)) + return true; + + STRV_FOREACH(i, arg_prefixes) { + const char *t; + + t = path_startswith(*i, "/proc/sys/"); + if (!t) + t = *i; + if (path_startswith(p, t)) + return true; + } + + return false; +} + static int parse_file(OrderedHashmap *sysctl_options, const char *path, bool ignore_enoent) { _cleanup_fclose_ FILE *f = NULL; + unsigned c = 0; int r; assert(path); @@ -77,7 +104,7 @@ static int parse_file(OrderedHashmap *sysctl_options, const char *path, bool ign } log_debug("Parsing %s", path); - while (!feof(f)) { + for (;;) { char l[LINE_MAX], *p, *value, *new_value, *property, *existing; void *v; int k; @@ -89,6 +116,8 @@ static int parse_file(OrderedHashmap *sysctl_options, const char *path, bool ign return log_error_errno(errno, "Failed to read file '%s', ignoring: %m", path); } + c++; + p = strstrip(l); if (!*p) continue; @@ -98,7 +127,7 @@ static int parse_file(OrderedHashmap *sysctl_options, const char *path, bool ign value = strchr(p, '='); if (!value) { - log_error("Line is not an assignment in file '%s': %s", path, value); + log_error("Line is not an assignment at '%s:%u': %s", path, c, value); if (r == 0) r = -EINVAL; @@ -111,26 +140,15 @@ static int parse_file(OrderedHashmap *sysctl_options, const char *path, bool ign p = sysctl_normalize(strstrip(p)); value = strstrip(value); - if (!strv_isempty(arg_prefixes)) { - char **i, *t; - STRV_FOREACH(i, arg_prefixes) { - t = path_startswith(*i, "/proc/sys/"); - if (t == NULL) - t = *i; - if (path_startswith(p, t)) - goto found; - } - /* not found */ + if (!test_prefix(p)) continue; - } -found: existing = ordered_hashmap_get2(sysctl_options, p, &v); if (existing) { if (streq(value, existing)) continue; - log_debug("Overwriting earlier assignment of %s in file '%s'.", p, path); + log_debug("Overwriting earlier assignment of %s at '%s:%u'.", p, path, c); free(ordered_hashmap_remove(sysctl_options, p)); free(v); } @@ -229,12 +247,12 @@ static int parse_argv(int argc, char *argv[]) { } int main(int argc, char *argv[]) { + OrderedHashmap *sysctl_options = NULL; int r = 0, k; - OrderedHashmap *sysctl_options; r = parse_argv(argc, argv); if (r <= 0) - return r < 0 ? EXIT_FAILURE : EXIT_SUCCESS; + goto finish; log_set_target(LOG_TARGET_AUTO); log_parse_environment(); diff --git a/src/systemctl/systemctl.c b/src/systemctl/systemctl.c index d311bbec1a..dd3b931cd6 100644 --- a/src/systemctl/systemctl.c +++ b/src/systemctl/systemctl.c @@ -410,23 +410,24 @@ static bool output_show_unit(const UnitInfo *u, char **patterns) { } 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; + unsigned circle_len = 0, id_len, max_id_len, load_len, active_len, sub_len, job_len, desc_len, max_desc_len; const UnitInfo *u; unsigned n_shown = 0; - int job_count = 0, desc_len; + 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; + max_desc_len = strlen("DESCRIPTION"); 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)); + max_desc_len = MAX(max_desc_len, strlen(u->description)); if (u->job_id != 0) { job_len = MAX(job_len, strlen(u->job_type)); @@ -442,7 +443,7 @@ static int output_units_list(const UnitInfo *unit_infos, unsigned c) { if (!arg_full && original_stdout_is_tty) { unsigned basic_len; - id_len = MIN(max_id_len, 25u); + id_len = MIN(max_id_len, 25u); /* as much as it needs, but at most 25 for now */ basic_len = circle_len + 5 + id_len + 5 + active_len + sub_len; if (job_count) @@ -455,19 +456,21 @@ static int output_units_list(const UnitInfo *unit_infos, unsigned c) { /* Either UNIT already got 25, or is fully satisfied. * Grant up to 25 to DESC now. */ incr = MIN(extra_len, 25u); - desc_len += incr; + desc_len = incr; extra_len -= incr; - /* split the remaining space between UNIT and DESC, - * but do not give UNIT more than it needs. */ + /* Of the remainder give as much as the ID needs to the ID, and give the rest to the + * description but not more than it needs. */ if (extra_len > 0) { - incr = MIN(extra_len / 2, max_id_len - id_len); + incr = MIN(max_id_len - id_len, extra_len); id_len += incr; - desc_len += extra_len - incr; + desc_len += MIN(extra_len - incr, max_desc_len - desc_len); } } - } else + } else { id_len = max_id_len; + desc_len = max_desc_len; + } for (u = unit_infos; u < unit_infos + c; u++) { _cleanup_free_ char *e = NULL, *j = NULL; @@ -493,8 +496,9 @@ static int output_units_list(const UnitInfo *unit_infos, unsigned c) { if (job_count) printf("%-*s ", job_len, "JOB"); - printf("%.*s%s\n", - !arg_full && arg_no_pager ? desc_len : -1, + printf("%-*.*s%s\n", + desc_len, + !arg_full && arg_no_pager ? (int) desc_len : -1, "DESCRIPTION", ansi_normal()); } @@ -513,13 +517,13 @@ static int output_units_list(const UnitInfo *unit_infos, unsigned c) { off_circle = ansi_normal(); circle = true; on_loaded = underline ? ansi_highlight_red_underline() : ansi_highlight_red(); - off_loaded = on_underline; + off_loaded = underline ? on_underline : ansi_normal(); } else if (streq(u->active_state, "failed") && !arg_plain) { on_circle = ansi_highlight_red(); off_circle = ansi_normal(); circle = true; on_active = underline ? ansi_highlight_red_underline() : ansi_highlight_red(); - off_active = on_underline; + off_active = underline ? on_underline : ansi_normal(); } if (u->machine) { @@ -550,8 +554,9 @@ static int output_units_list(const UnitInfo *unit_infos, unsigned c) { sub_len, u->sub_state, off_active, job_count ? job_len + 1 : 0, u->job_id ? u->job_type : ""); - printf("%.*s%s\n", - desc_len > 0 ? desc_len : -1, + printf("%-*.*s%s\n", + desc_len, + !arg_full && arg_no_pager ? (int) desc_len : -1, u->description, off_underline); } @@ -5272,7 +5277,7 @@ static int cat(int argc, char *argv[], void *userdata) { else puts(""); - if (need_daemon_reload(bus, *name)) + if (need_daemon_reload(bus, *name) > 0) /* ignore errors (<0), this is informational output */ fprintf(stderr, "%s# Warning: %s changed on disk, the version systemd has loaded is outdated.\n" "%s# This output shows the current version of the unit's original fragment and drop-in files.\n" diff --git a/src/test/test-seccomp.c b/src/test/test-seccomp.c index 0060ecdf02..43d1567288 100644 --- a/src/test/test-seccomp.c +++ b/src/test/test-seccomp.c @@ -25,6 +25,8 @@ #include "macro.h" #include "process-util.h" #include "seccomp-util.h" +#include "string-util.h" +#include "util.h" static void test_seccomp_arch_to_string(void) { uint32_t a, b; @@ -38,6 +40,36 @@ static void test_seccomp_arch_to_string(void) { assert_se(a == b); } +static void test_architecture_table(void) { + const char *n, *n2; + + NULSTR_FOREACH(n, + "native\0" + "x86\0" + "x86-64\0" + "x32\0" + "arm\0" + "arm64\0" + "mips\0" + "mips64\0" + "mips64-n32\0" + "mips-le\0" + "mips64-le\0" + "mips64-le-n32\0" + "ppc\0" + "ppc64\0" + "ppc64-le\0" + "s390\0" + "s390x\0") { + uint32_t c; + + assert_se(seccomp_arch_from_string(n, &c) >= 0); + n2 = seccomp_arch_to_string(c); + log_info("seccomp-arch: %s → 0x%"PRIx32" → %s", n, c, n2); + assert_se(streq_ptr(n, n2)); + } +} + static void test_syscall_filter_set_find(void) { assert_se(!syscall_filter_set_find(NULL)); assert_se(!syscall_filter_set_find("")); @@ -96,6 +128,7 @@ static void test_filter_sets(void) { int main(int argc, char *argv[]) { test_seccomp_arch_to_string(); + test_architecture_table(); test_syscall_filter_set_find(); test_filter_sets(); diff --git a/src/test/test-unit-file.c b/src/test/test-unit-file.c index 7ef087a2e3..12f48bf435 100644 --- a/src/test/test-unit-file.c +++ b/src/test/test-unit-file.c @@ -589,7 +589,7 @@ static void test_install_printf(void) { assert_se(specifier_machine_id('m', NULL, NULL, &mid) >= 0 && mid); assert_se(specifier_boot_id('b', NULL, NULL, &bid) >= 0 && bid); assert_se((host = gethostname_malloc())); - assert_se((user = getusername_malloc())); + assert_se((user = uid_to_name(getuid()))); assert_se(asprintf(&uid, UID_FMT, getuid()) >= 0); #define expect(src, pattern, result) \ diff --git a/src/udev/collect/collect.c b/src/udev/collect/collect.c index 349585b634..0e973cd521 100644 --- a/src/udev/collect/collect.c +++ b/src/udev/collect/collect.c @@ -85,16 +85,16 @@ static void usage(void) */ static int prepare(char *dir, char *filename) { - char buf[512]; + char buf[PATH_MAX]; int r, fd; r = mkdir(dir, 0700); if (r < 0 && errno != EEXIST) return -errno; - xsprintf(buf, "%s/%s", dir, filename); + snprintf(buf, sizeof buf, "%s/%s", dir, filename); - fd = open(buf,O_RDWR|O_CREAT|O_CLOEXEC, S_IRUSR|S_IWUSR); + fd = open(buf, O_RDWR|O_CREAT|O_CLOEXEC, S_IRUSR|S_IWUSR); if (fd < 0) fprintf(stderr, "Cannot open %s: %m\n", buf); diff --git a/src/udev/udev-builtin-net_id.c b/src/udev/udev-builtin-net_id.c index a7be2a4eed..fe9d6f4482 100644 --- a/src/udev/udev-builtin-net_id.c +++ b/src/udev/udev-builtin-net_id.c @@ -35,10 +35,12 @@ * Type of names: * b<number> — BCMA bus core number * c<bus_id> — CCW bus group name, without leading zeros [s390] - * o<index>[d<dev_port>] — on-board device index number - * s<slot>[f<function>][d<dev_port>] — hotplug slot index number + * o<index>[n<phys_port_name>|d<dev_port>] + * — on-board device index number + * s<slot>[f<function>][n<phys_port_name>|d<dev_port>] + * — hotplug slot index number * x<MAC> — MAC address - * [P<domain>]p<bus>s<slot>[f<function>][d<dev_port>] + * [P<domain>]p<bus>s<slot>[f<function>][n<phys_port_name>|d<dev_port>] * — PCI geographical location * [P<domain>]p<bus>s<slot>[f<function>][u<port>][..][c<config>][i<interface>] * — USB port number chain @@ -137,7 +139,7 @@ static int dev_pci_onboard(struct udev_device *dev, struct netnames *names) { unsigned dev_port = 0; size_t l; char *s; - const char *attr; + const char *attr, *port_name; int idx; /* ACPI _DSM — device specific method for naming a PCI or PCI Express device */ @@ -164,10 +166,15 @@ static int dev_pci_onboard(struct udev_device *dev, struct netnames *names) { if (attr) dev_port = strtol(attr, NULL, 10); + /* kernel provided front panel port name for multiple port PCI device */ + port_name = udev_device_get_sysattr_value(dev, "phys_port_name"); + s = names->pci_onboard; l = sizeof(names->pci_onboard); l = strpcpyf(&s, l, "o%d", idx); - if (dev_port > 0) + if (port_name) + l = strpcpyf(&s, l, "n%s", port_name); + else if (dev_port > 0) l = strpcpyf(&s, l, "d%d", dev_port); if (l == 0) names->pci_onboard[0] = '\0'; @@ -202,9 +209,9 @@ static int dev_pci_slot(struct udev_device *dev, struct netnames *names) { unsigned domain, bus, slot, func, dev_port = 0; size_t l; char *s; - const char *attr; + const char *attr, *port_name; struct udev_device *pci = NULL; - char slots[256], str[256]; + char slots[PATH_MAX]; _cleanup_closedir_ DIR *dir = NULL; struct dirent *dent; int hotplug_slot = 0, err = 0; @@ -217,6 +224,9 @@ static int dev_pci_slot(struct udev_device *dev, struct netnames *names) { if (attr) dev_port = strtol(attr, NULL, 10); + /* kernel provided front panel port name for multiple port PCI device */ + port_name = udev_device_get_sysattr_value(dev, "phys_port_name"); + /* compose a name based on the raw kernel's PCI bus, slot numbers */ s = names->pci_path; l = sizeof(names->pci_path); @@ -225,7 +235,9 @@ static int dev_pci_slot(struct udev_device *dev, struct netnames *names) { l = strpcpyf(&s, l, "p%us%u", bus, slot); if (func > 0 || is_pci_multifunction(names->pcidev)) l = strpcpyf(&s, l, "f%u", func); - if (dev_port > 0) + if (port_name) + l = strpcpyf(&s, l, "n%s", port_name); + else if (dev_port > 0) l = strpcpyf(&s, l, "d%u", dev_port); if (l == 0) names->pci_path[0] = '\0'; @@ -236,7 +248,8 @@ static int dev_pci_slot(struct udev_device *dev, struct netnames *names) { err = -ENOENT; goto out; } - xsprintf(slots, "%s/slots", udev_device_get_syspath(pci)); + + snprintf(slots, sizeof slots, "%s/slots", udev_device_get_syspath(pci)); dir = opendir(slots); if (!dir) { err = -errno; @@ -245,8 +258,7 @@ static int dev_pci_slot(struct udev_device *dev, struct netnames *names) { for (dent = readdir(dir); dent != NULL; dent = readdir(dir)) { int i; - char *rest; - char *address; + char *rest, *address, str[PATH_MAX]; if (dent->d_name[0] == '.') continue; @@ -255,7 +267,8 @@ static int dev_pci_slot(struct udev_device *dev, struct netnames *names) { continue; if (i < 1) continue; - xsprintf(str, "%s/%s/address", slots, dent->d_name); + + snprintf(str, sizeof str, "%s/%s/address", slots, dent->d_name); if (read_one_line_file(str, &address) >= 0) { /* match slot address with device by stripping the function */ if (strneq(address, udev_device_get_sysname(names->pcidev), strlen(address))) @@ -275,7 +288,9 @@ static int dev_pci_slot(struct udev_device *dev, struct netnames *names) { l = strpcpyf(&s, l, "s%d", hotplug_slot); if (func > 0 || is_pci_multifunction(names->pcidev)) l = strpcpyf(&s, l, "f%d", func); - if (dev_port > 0) + if (port_name) + l = strpcpyf(&s, l, "n%s", port_name); + else if (dev_port > 0) l = strpcpyf(&s, l, "d%d", dev_port); if (l == 0) names->pci_slot[0] = '\0'; diff --git a/src/udev/udev-node.c b/src/udev/udev-node.c index 5d2997fd8f..43004bc0bc 100644 --- a/src/udev/udev-node.c +++ b/src/udev/udev-node.c @@ -337,7 +337,7 @@ out: void udev_node_add(struct udev_device *dev, bool apply, mode_t mode, uid_t uid, gid_t gid, struct udev_list *seclabel_list) { - char filename[UTIL_PATH_SIZE]; + char filename[sizeof("/dev/block/:") + 2*DECIMAL_STR_MAX(unsigned)]; struct udev_list_entry *list_entry; log_debug("handling device node '%s', devnum=%s, mode=%#o, uid="UID_FMT", gid="GID_FMT, @@ -360,7 +360,7 @@ void udev_node_add(struct udev_device *dev, bool apply, void udev_node_remove(struct udev_device *dev) { struct udev_list_entry *list_entry; - char filename[UTIL_PATH_SIZE]; + char filename[sizeof("/dev/block/:") + 2*DECIMAL_STR_MAX(unsigned)]; /* remove/update symlinks, remove symlinks from name index */ udev_list_entry_foreach(list_entry, udev_device_get_devlinks_list_entry(dev)) diff --git a/src/udev/udev-watch.c b/src/udev/udev-watch.c index 9ce5e975de..bc9096ed0c 100644 --- a/src/udev/udev-watch.c +++ b/src/udev/udev-watch.c @@ -89,7 +89,7 @@ unlink: } void udev_watch_begin(struct udev *udev, struct udev_device *dev) { - char filename[UTIL_PATH_SIZE]; + char filename[sizeof("/run/udev/watch/") + DECIMAL_STR_MAX(int)]; int wd; int r; @@ -116,7 +116,7 @@ void udev_watch_begin(struct udev *udev, struct udev_device *dev) { void udev_watch_end(struct udev *udev, struct udev_device *dev) { int wd; - char filename[UTIL_PATH_SIZE]; + char filename[sizeof("/run/udev/watch/") + DECIMAL_STR_MAX(int)]; if (inotify_fd < 0) return; @@ -135,7 +135,7 @@ void udev_watch_end(struct udev *udev, struct udev_device *dev) { } struct udev_device *udev_watch_lookup(struct udev *udev, int wd) { - char filename[UTIL_PATH_SIZE]; + char filename[sizeof("/run/udev/watch/") + DECIMAL_STR_MAX(int)]; char device[UTIL_NAME_SIZE]; ssize_t len; diff --git a/test/TEST-04-JOURNAL/test-journal.sh b/test/TEST-04-JOURNAL/test-journal.sh index 6646eccfa7..493ff00ce0 100755 --- a/test/TEST-04-JOURNAL/test-journal.sh +++ b/test/TEST-04-JOURNAL/test-journal.sh @@ -59,4 +59,12 @@ sleep 3 systemctl stop forever-print-hola [[ ! -f "/i-lose-my-logs" ]] +# https://github.com/systemd/systemd/issues/4408 +rm -f /i-lose-my-logs +systemctl start forever-print-hola +sleep 3 +systemctl kill --signal=SIGKILL systemd-journald +sleep 3 +[[ ! -f "/i-lose-my-logs" ]] + touch /testok |