1 <?xml-stylesheet type="text/xsl" href="http://docbook.sourceforge.net/release/xsl/current/xhtml/docbook.xsl"?>
2 <!DOCTYPE refentry PUBLIC "-//OASIS//DTD DocBook XML V4.2//EN"
3 "http://www.oasis-open.org/docbook/xml/4.2/docbookx.dtd">
6 This file is part of systemd.
8 Copyright 2010 Lennart Poettering
10 systemd is free software; you can redistribute it and/or modify it
11 under the terms of the GNU Lesser General Public License as published by
12 the Free Software Foundation; either version 2.1 of the License, or
13 (at your option) any later version.
15 systemd is distributed in the hope that it will be useful, but
16 WITHOUT ANY WARRANTY; without even the implied warranty of
17 MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
18 Lesser General Public License for more details.
20 You should have received a copy of the GNU Lesser General Public License
21 along with systemd; If not, see <http://www.gnu.org/licenses/>.
24 <refentry id="systemd.exec">
26 <title>systemd.exec</title>
27 <productname>systemd</productname>
31 <contrib>Developer</contrib>
32 <firstname>Lennart</firstname>
33 <surname>Poettering</surname>
34 <email>lennart@poettering.net</email>
40 <refentrytitle>systemd.exec</refentrytitle>
41 <manvolnum>5</manvolnum>
45 <refname>systemd.exec</refname>
46 <refpurpose>Execution environment configuration</refpurpose>
50 <para><filename><replaceable>service</replaceable>.service</filename>,
51 <filename><replaceable>socket</replaceable>.socket</filename>,
52 <filename><replaceable>mount</replaceable>.mount</filename>,
53 <filename><replaceable>swap</replaceable>.swap</filename></para>
57 <title>Description</title>
59 <para>Unit configuration files for services, sockets,
60 mount points, and swap devices share a subset of
61 configuration options which define the execution
62 environment of spawned processes.</para>
64 <para>This man page lists the configuration options
65 shared by these four unit types. See
66 <citerefentry><refentrytitle>systemd.unit</refentrytitle><manvolnum>5</manvolnum></citerefentry>
67 for the common options of all unit configuration
69 <citerefentry><refentrytitle>systemd.service</refentrytitle><manvolnum>5</manvolnum></citerefentry>,
70 <citerefentry><refentrytitle>systemd.socket</refentrytitle><manvolnum>5</manvolnum></citerefentry>,
71 <citerefentry><refentrytitle>systemd.swap</refentrytitle><manvolnum>5</manvolnum></citerefentry>,
73 <citerefentry><refentrytitle>systemd.mount</refentrytitle><manvolnum>5</manvolnum></citerefentry>
74 for more information on the specific unit
75 configuration files. The execution specific
76 configuration options are configured in the [Service],
77 [Socket], [Mount], or [Swap] sections, depending on the unit
82 <title>Options</title>
84 <variablelist class='unit-directives'>
87 <term><varname>WorkingDirectory=</varname></term>
89 <listitem><para>Takes an absolute
90 directory path. Sets the working
91 directory for executed processes. If
92 not set, defaults to the root directory
93 when systemd is running as a system
94 instance and the respective user's
95 home directory if run as
96 user.</para></listitem>
100 <term><varname>RootDirectory=</varname></term>
102 <listitem><para>Takes an absolute
103 directory path. Sets the root
104 directory for executed processes, with
106 <citerefentry project='man-pages'><refentrytitle>chroot</refentrytitle><manvolnum>2</manvolnum></citerefentry>
107 system call. If this is used, it must
108 be ensured that the process and all
109 its auxiliary files are available in
110 the <function>chroot()</function>
111 jail.</para></listitem>
115 <term><varname>User=</varname></term>
116 <term><varname>Group=</varname></term>
118 <listitem><para>Sets the Unix user
119 or group that the processes are executed
120 as, respectively. Takes a single user or group
121 name or ID as argument. If no group is
122 set, the default group of the user is
123 chosen.</para></listitem>
127 <term><varname>SupplementaryGroups=</varname></term>
129 <listitem><para>Sets the supplementary
130 Unix groups the processes are executed
131 as. This takes a space-separated list
132 of group names or IDs. This option may
133 be specified more than once in which
134 case all listed groups are set as
135 supplementary groups. When the empty
136 string is assigned the list of
137 supplementary groups is reset, and all
138 assignments prior to this one will
139 have no effect. In any way, this
140 option does not override, but extends
141 the list of supplementary groups
142 configured in the system group
144 user.</para></listitem>
148 <term><varname>Nice=</varname></term>
150 <listitem><para>Sets the default nice
151 level (scheduling priority) for
152 executed processes. Takes an integer
153 between -20 (highest priority) and 19
154 (lowest priority). See
155 <citerefentry><refentrytitle>setpriority</refentrytitle><manvolnum>2</manvolnum></citerefentry>
156 for details.</para></listitem>
160 <term><varname>OOMScoreAdjust=</varname></term>
162 <listitem><para>Sets the adjustment
163 level for the Out-Of-Memory killer for
164 executed processes. Takes an integer
165 between -1000 (to disable OOM killing
166 for this process) and 1000 (to make
167 killing of this process under memory
168 pressure very likely). See <ulink
169 url="https://www.kernel.org/doc/Documentation/filesystems/proc.txt">proc.txt</ulink>
170 for details.</para></listitem>
174 <term><varname>IOSchedulingClass=</varname></term>
176 <listitem><para>Sets the IO scheduling
177 class for executed processes. Takes an
178 integer between 0 and 3 or one of the
179 strings <option>none</option>,
180 <option>realtime</option>,
181 <option>best-effort</option> or
182 <option>idle</option>. See
183 <citerefentry><refentrytitle>ioprio_set</refentrytitle><manvolnum>2</manvolnum></citerefentry>
184 for details.</para></listitem>
188 <term><varname>IOSchedulingPriority=</varname></term>
190 <listitem><para>Sets the IO scheduling
191 priority for executed processes. Takes
192 an integer between 0 (highest
193 priority) and 7 (lowest priority). The
194 available priorities depend on the
195 selected IO scheduling class (see
197 <citerefentry><refentrytitle>ioprio_set</refentrytitle><manvolnum>2</manvolnum></citerefentry>
198 for details.</para></listitem>
202 <term><varname>CPUSchedulingPolicy=</varname></term>
204 <listitem><para>Sets the CPU
205 scheduling policy for executed
206 processes. Takes one of
207 <option>other</option>,
208 <option>batch</option>,
209 <option>idle</option>,
210 <option>fifo</option> or
211 <option>rr</option>. See
212 <citerefentry><refentrytitle>sched_setscheduler</refentrytitle><manvolnum>2</manvolnum></citerefentry>
213 for details.</para></listitem>
217 <term><varname>CPUSchedulingPriority=</varname></term>
219 <listitem><para>Sets the CPU
220 scheduling priority for executed
221 processes. The available priority
222 range depends on the selected CPU
223 scheduling policy (see above). For
224 real-time scheduling policies an
225 integer between 1 (lowest priority)
226 and 99 (highest priority) can be used.
227 See <citerefentry><refentrytitle>sched_setscheduler</refentrytitle><manvolnum>2</manvolnum></citerefentry>
233 <term><varname>CPUSchedulingResetOnFork=</varname></term>
235 <listitem><para>Takes a boolean
236 argument. If true, elevated CPU
237 scheduling priorities and policies
238 will be reset when the executed
239 processes fork, and can hence not leak
240 into child processes. See
241 <citerefentry><refentrytitle>sched_setscheduler</refentrytitle><manvolnum>2</manvolnum></citerefentry>
242 for details. Defaults to false.</para></listitem>
246 <term><varname>CPUAffinity=</varname></term>
248 <listitem><para>Controls the CPU
249 affinity of the executed
250 processes. Takes a space-separated
251 list of CPU indices. This option may
252 be specified more than once in which
253 case the specificed CPU affinity masks
254 are merged. If the empty string is
255 assigned, the mask is reset, all
256 assignments prior to this will have no
258 <citerefentry><refentrytitle>sched_setaffinity</refentrytitle><manvolnum>2</manvolnum></citerefentry>
259 for details.</para></listitem>
263 <term><varname>UMask=</varname></term>
265 <listitem><para>Controls the file mode
266 creation mask. Takes an access mode in
268 <citerefentry><refentrytitle>umask</refentrytitle><manvolnum>2</manvolnum></citerefentry>
269 for details. Defaults to
270 0022.</para></listitem>
274 <term><varname>Environment=</varname></term>
276 <listitem><para>Sets environment
277 variables for executed
278 processes. Takes a space-separated
279 list of variable assignments. This
280 option may be specified more than once
281 in which case all listed variables
282 will be set. If the same variable is
283 set twice, the later setting will
284 override the earlier setting. If the
285 empty string is assigned to this
286 option, the list of environment
287 variables is reset, all prior
288 assignments have no effect.
289 Variable expansion is not performed
290 inside the strings, however, specifier
291 expansion is possible. The $ character has
293 If you need to assign a value containing spaces
294 to a variable, use double quotes (")
295 for the assignment.</para>
298 <programlisting>Environment="VAR1=word1 word2" VAR2=word3 "VAR3=$word 5 6"</programlisting>
299 gives three variables <literal>VAR1</literal>,
300 <literal>VAR2</literal>, <literal>VAR3</literal>
301 with the values <literal>word1 word2</literal>,
302 <literal>word3</literal>, <literal>$word 5 6</literal>.
307 <citerefentry project='man-pages'><refentrytitle>environ</refentrytitle><manvolnum>7</manvolnum></citerefentry>
308 for details about environment variables.</para></listitem>
311 <term><varname>EnvironmentFile=</varname></term>
312 <listitem><para>Similar to
313 <varname>Environment=</varname> but
314 reads the environment variables from a
315 text file. The text file should
316 contain new-line-separated variable
317 assignments. Empty lines and lines
318 starting with ; or # will be ignored,
319 which may be used for commenting. A line
320 ending with a backslash will be concatenated
321 with the following one, allowing multiline variable
322 definitions. The parser strips leading
323 and trailing whitespace from the values
324 of assignments, unless you use
325 double quotes (").</para>
327 <para>The argument passed should be an
328 absolute filename or wildcard
329 expression, optionally prefixed with
330 <literal>-</literal>, which indicates
331 that if the file does not exist, it
332 will not be read and no error or warning
333 message is logged. This option may be
334 specified more than once in which case
335 all specified files are read. If the
336 empty string is assigned to this
337 option, the list of file to read is
338 reset, all prior assignments have no
341 <para>The files listed with this
342 directive will be read shortly before
343 the process is executed (more
344 specifically, after all
345 processes from a previous unit state
346 terminated. This means you can
347 generate these files in one unit
348 state, and read it with this option in
349 the next). Settings from these files
350 override settings made with
351 <varname>Environment=</varname>. If
352 the same variable is set twice from
353 these files, the files will be read in
354 the order they are specified and the
355 later setting will override the
356 earlier setting.</para></listitem>
360 <term><varname>StandardInput=</varname></term>
361 <listitem><para>Controls where file
362 descriptor 0 (STDIN) of the executed
363 processes is connected to. Takes one
364 of <option>null</option>,
365 <option>tty</option>,
366 <option>tty-force</option>,
367 <option>tty-fail</option> or
368 <option>socket</option>.</para>
370 <para>If <option>null</option> is
371 selected, standard input will be
373 <filename>/dev/null</filename>,
374 i.e. all read attempts by the process
375 will result in immediate EOF.</para>
377 <para>If <option>tty</option> is
378 selected, standard input is connected
379 to a TTY (as configured by
380 <varname>TTYPath=</varname>, see
381 below) and the executed process
382 becomes the controlling process of the
383 terminal. If the terminal is already
384 being controlled by another process,
385 the executed process waits until the
386 current controlling process releases
389 <para><option>tty-force</option> is similar
390 to <option>tty</option>, but the
391 executed process is forcefully and
392 immediately made the controlling
393 process of the terminal, potentially
394 removing previous controlling
398 <para><option>tty-fail</option> is
399 similar to <option>tty</option> but if
400 the terminal already has a controlling
401 process start-up of the executed
402 process fails.</para>
404 <para>The <option>socket</option>
405 option is only valid in
406 socket-activated services, and only
407 when the socket configuration file
409 <citerefentry><refentrytitle>systemd.socket</refentrytitle><manvolnum>5</manvolnum></citerefentry>
410 for details) specifies a single socket
411 only. If this option is set, standard
412 input will be connected to the socket
413 the service was activated from, which
414 is primarily useful for compatibility
415 with daemons designed for use with the
417 <citerefentry><refentrytitle>inetd</refentrytitle><manvolnum>8</manvolnum></citerefentry>
420 <para>This setting defaults to
421 <option>null</option>.</para></listitem>
424 <term><varname>StandardOutput=</varname></term>
425 <listitem><para>Controls where file
426 descriptor 1 (STDOUT) of the executed
427 processes is connected to. Takes one
428 of <option>inherit</option>,
429 <option>null</option>,
430 <option>tty</option>,
431 <option>journal</option>,
432 <option>syslog</option>,
433 <option>kmsg</option>,
434 <option>journal+console</option>,
435 <option>syslog+console</option>,
436 <option>kmsg+console</option> or
437 <option>socket</option>.</para>
439 <para><option>inherit</option>
440 duplicates the file descriptor of
441 standard input for standard
444 <para><option>null</option> connects
446 <filename>/dev/null</filename>,
447 i.e. everything written to it will be
450 <para><option>tty</option> connects
451 standard output to a tty (as
453 <varname>TTYPath=</varname>, see
454 below). If the TTY is used for output
455 only, the executed process will not
456 become the controlling process of the
457 terminal, and will not fail or wait
458 for other processes to release the
461 <para><option>journal</option>
462 connects standard output with the
463 journal which is accessible via
464 <citerefentry><refentrytitle>journalctl</refentrytitle><manvolnum>1</manvolnum></citerefentry>.
465 Note that everything that is written
466 to syslog or kmsg (see below) is
467 implicitly stored in the journal as
468 well, the specific two options listed
469 below are hence supersets of this
472 <para><option>syslog</option> connects
473 standard output to the <citerefentry
474 project='man-pages'><refentrytitle>syslog</refentrytitle><manvolnum>3</manvolnum></citerefentry>
475 system syslog service, in addition to
476 the journal. Note that the journal
477 daemon is usually configured to
478 forward everything it receives to
479 syslog anyway, in which case this
480 option is no different from
481 <option>journal</option>.</para>
483 <para><option>kmsg</option> connects
484 standard output with the kernel log
485 buffer which is accessible via
487 project='man-pages'><refentrytitle>dmesg</refentrytitle><manvolnum>1</manvolnum></citerefentry>,
488 in addition to the journal. The
489 journal daemon might be configured to
490 send all logs to kmsg anyway, in which
491 case this option is no different from
492 <option>journal</option>.</para>
494 <para><option>journal+console</option>,
495 <option>syslog+console</option> and
496 <option>kmsg+console</option> work in
497 a similar way as the three options
498 above but copy the output to the
499 system console as well.</para>
501 <para><option>socket</option> connects
502 standard output to a socket acquired
503 via socket activation. The semantics
504 are similar to the same option of
505 <varname>StandardInput=</varname>.</para>
507 <para>This setting defaults to the
509 <option>DefaultStandardOutput=</option>
511 <citerefentry><refentrytitle>systemd-system.conf</refentrytitle><manvolnum>5</manvolnum></citerefentry>,
513 <option>journal</option>.</para></listitem>
516 <term><varname>StandardError=</varname></term>
517 <listitem><para>Controls where file
518 descriptor 2 (STDERR) of the
519 executed processes is connected to.
520 The available options are identical to
522 <varname>StandardOutput=</varname>,
523 with one exception: if set to
524 <option>inherit</option> the file
525 descriptor used for standard output is
526 duplicated for standard error. This
527 setting defaults to the value set with
528 <option>DefaultStandardError=</option>
530 <citerefentry><refentrytitle>systemd-system.conf</refentrytitle><manvolnum>5</manvolnum></citerefentry>,
532 <option>inherit</option>.</para></listitem>
535 <term><varname>TTYPath=</varname></term>
536 <listitem><para>Sets the terminal
537 device node to use if standard input, output,
538 or error are connected to a
539 TTY (see above). Defaults to
540 <filename>/dev/console</filename>.</para></listitem>
543 <term><varname>TTYReset=</varname></term>
544 <listitem><para>Reset the terminal
545 device specified with
546 <varname>TTYPath=</varname> before and
547 after execution. Defaults to
548 <literal>no</literal>.</para></listitem>
551 <term><varname>TTYVHangup=</varname></term>
552 <listitem><para>Disconnect all clients
553 which have opened the terminal device
555 <varname>TTYPath=</varname>
556 before and after execution. Defaults
558 <literal>no</literal>.</para></listitem>
561 <term><varname>TTYVTDisallocate=</varname></term>
562 <listitem><para>If the terminal
563 device specified with
564 <varname>TTYPath=</varname> is a
565 virtual console terminal, try to
566 deallocate the TTY before and after
567 execution. This ensures that the
568 screen and scrollback buffer is
570 <literal>no</literal>.</para></listitem>
573 <term><varname>SyslogIdentifier=</varname></term>
574 <listitem><para>Sets the process name
575 to prefix log lines sent to syslog or
576 the kernel log buffer with. If not set,
577 defaults to the process name of the
578 executed process. This option is only
580 <varname>StandardOutput=</varname> or
581 <varname>StandardError=</varname> are
582 set to <option>syslog</option> or
583 <option>kmsg</option>.</para></listitem>
586 <term><varname>SyslogFacility=</varname></term>
587 <listitem><para>Sets the syslog
588 facility to use when logging to
589 syslog. One of <option>kern</option>,
590 <option>user</option>,
591 <option>mail</option>,
592 <option>daemon</option>,
593 <option>auth</option>,
594 <option>syslog</option>,
595 <option>lpr</option>,
596 <option>news</option>,
597 <option>uucp</option>,
598 <option>cron</option>,
599 <option>authpriv</option>,
600 <option>ftp</option>,
601 <option>local0</option>,
602 <option>local1</option>,
603 <option>local2</option>,
604 <option>local3</option>,
605 <option>local4</option>,
606 <option>local5</option>,
607 <option>local6</option> or
608 <option>local7</option>. See
609 <citerefentry project='man-pages'><refentrytitle>syslog</refentrytitle><manvolnum>3</manvolnum></citerefentry>
610 for details. This option is only
612 <varname>StandardOutput=</varname> or
613 <varname>StandardError=</varname> are
614 set to <option>syslog</option>.
616 <option>daemon</option>.</para></listitem>
619 <term><varname>SyslogLevel=</varname></term>
620 <listitem><para>Default syslog level
621 to use when logging to syslog or the
622 kernel log buffer. One of
623 <option>emerg</option>,
624 <option>alert</option>,
625 <option>crit</option>,
626 <option>err</option>,
627 <option>warning</option>,
628 <option>notice</option>,
629 <option>info</option>,
630 <option>debug</option>. See
631 <citerefentry project='man-pages'><refentrytitle>syslog</refentrytitle><manvolnum>3</manvolnum></citerefentry>
632 for details. This option is only
634 <varname>StandardOutput=</varname> or
635 <varname>StandardError=</varname> are
636 set to <option>syslog</option> or
637 <option>kmsg</option>. Note that
638 individual lines output by the daemon
639 might be prefixed with a different log
640 level which can be used to override
641 the default log level specified
642 here. The interpretation of these
643 prefixes may be disabled with
644 <varname>SyslogLevelPrefix=</varname>,
645 see below. For details see
646 <citerefentry><refentrytitle>sd-daemon</refentrytitle><manvolnum>3</manvolnum></citerefentry>.
649 <option>info</option>.</para></listitem>
653 <term><varname>SyslogLevelPrefix=</varname></term>
654 <listitem><para>Takes a boolean
655 argument. If true and
656 <varname>StandardOutput=</varname> or
657 <varname>StandardError=</varname> are
658 set to <option>syslog</option>,
659 <option>kmsg</option> or
660 <option>journal</option>, log lines
661 written by the executed process that
662 are prefixed with a log level will be
663 passed on to syslog with this log
664 level set but the prefix removed. If
665 set to false, the interpretation of
666 these prefixes is disabled and the
667 logged lines are passed on as-is. For
668 details about this prefixing see
669 <citerefentry><refentrytitle>sd-daemon</refentrytitle><manvolnum>3</manvolnum></citerefentry>.
670 Defaults to true.</para></listitem>
674 <term><varname>TimerSlackNSec=</varname></term>
675 <listitem><para>Sets the timer slack
676 in nanoseconds for the executed
677 processes. The timer slack controls
678 the accuracy of wake-ups triggered by
680 <citerefentry><refentrytitle>prctl</refentrytitle><manvolnum>2</manvolnum></citerefentry>
681 for more information. Note that in
682 contrast to most other time span
683 definitions this parameter takes an
684 integer value in nano-seconds if no
685 unit is specified. The usual time
687 too.</para></listitem>
691 <term><varname>LimitCPU=</varname></term>
692 <term><varname>LimitFSIZE=</varname></term>
693 <term><varname>LimitDATA=</varname></term>
694 <term><varname>LimitSTACK=</varname></term>
695 <term><varname>LimitCORE=</varname></term>
696 <term><varname>LimitRSS=</varname></term>
697 <term><varname>LimitNOFILE=</varname></term>
698 <term><varname>LimitAS=</varname></term>
699 <term><varname>LimitNPROC=</varname></term>
700 <term><varname>LimitMEMLOCK=</varname></term>
701 <term><varname>LimitLOCKS=</varname></term>
702 <term><varname>LimitSIGPENDING=</varname></term>
703 <term><varname>LimitMSGQUEUE=</varname></term>
704 <term><varname>LimitNICE=</varname></term>
705 <term><varname>LimitRTPRIO=</varname></term>
706 <term><varname>LimitRTTIME=</varname></term>
707 <listitem><para>These settings control
708 various resource limits for executed
710 <citerefentry><refentrytitle>setrlimit</refentrytitle><manvolnum>2</manvolnum></citerefentry>
711 for details. Use the string
712 <varname>infinity</varname> to
713 configure no limit on a specific
714 resource.</para></listitem>
718 <term><varname>PAMName=</varname></term>
719 <listitem><para>Sets the PAM service
720 name to set up a session as. If set,
721 the executed process will be
722 registered as a PAM session under the
723 specified service name. This is only
724 useful in conjunction with the
725 <varname>User=</varname> setting. If
726 not set, no PAM session will be opened
727 for the executed processes. See
728 <citerefentry project='man-pages'><refentrytitle>pam</refentrytitle><manvolnum>8</manvolnum></citerefentry>
729 for details.</para></listitem>
733 <term><varname>CapabilityBoundingSet=</varname></term>
735 <listitem><para>Controls which
736 capabilities to include in the
737 capability bounding set for the
738 executed process. See
739 <citerefentry project='man-pages'><refentrytitle>capabilities</refentrytitle><manvolnum>7</manvolnum></citerefentry>
740 for details. Takes a whitespace-separated
741 list of capability names as read by
742 <citerefentry><refentrytitle>cap_from_name</refentrytitle><manvolnum>3</manvolnum></citerefentry>,
743 e.g. <constant>CAP_SYS_ADMIN</constant>,
744 <constant>CAP_DAC_OVERRIDE</constant>,
745 <constant>CAP_SYS_PTRACE</constant>.
746 Capabilities listed will be included
747 in the bounding set, all others are
748 removed. If the list of capabilities
749 is prefixed with <literal>~</literal>,
750 all but the listed capabilities will
751 be included, the effect of the
752 assignment inverted. Note that this
753 option also affects the respective
754 capabilities in the effective,
755 permitted and inheritable capability
757 <varname>Capabilities=</varname>
758 does. If this option is not used, the
759 capability bounding set is not
760 modified on process execution, hence
761 no limits on the capabilities of the
762 process are enforced. This option may
763 appear more than once in which case
764 the bounding sets are merged. If the
765 empty string is assigned to this
766 option, the bounding set is reset to
767 the empty capability set, and all
768 prior settings have no effect. If set
769 to <literal>~</literal> (without any
770 further argument), the bounding set is
771 reset to the full set of available
772 capabilities, also undoing any
773 previous settings.</para></listitem>
777 <term><varname>SecureBits=</varname></term>
778 <listitem><para>Controls the secure
779 bits set for the executed process.
780 Takes a space-separated combination of
781 options from the following list:
782 <option>keep-caps</option>,
783 <option>keep-caps-locked</option>,
784 <option>no-setuid-fixup</option>,
785 <option>no-setuid-fixup-locked</option>,
786 <option>noroot</option>, and
787 <option>noroot-locked</option>. This
788 option may appear more than once in
789 which case the secure bits are ORed.
790 If the empty string is assigned to
791 this option, the bits are reset to 0.
793 project='man-pages'><refentrytitle>capabilities</refentrytitle><manvolnum>7</manvolnum></citerefentry>
794 for details.</para></listitem>
798 <term><varname>Capabilities=</varname></term>
799 <listitem><para>Controls the
800 <citerefentry project='man-pages'><refentrytitle>capabilities</refentrytitle><manvolnum>7</manvolnum></citerefentry>
801 set for the executed process. Take a
802 capability string describing the
803 effective, permitted and inherited
804 capability sets as documented in
805 <citerefentry><refentrytitle>cap_from_text</refentrytitle><manvolnum>3</manvolnum></citerefentry>.
806 Note that these capability sets are
807 usually influenced (and filtered) by the capabilities
808 attached to the executed file. Due to
810 <varname>CapabilityBoundingSet=</varname>
811 is probably a much more useful
812 setting.</para></listitem>
816 <term><varname>ReadWriteDirectories=</varname></term>
817 <term><varname>ReadOnlyDirectories=</varname></term>
818 <term><varname>InaccessibleDirectories=</varname></term>
820 <listitem><para>Sets up a new file
821 system namespace for executed
822 processes. These options may be used
823 to limit access a process might have
824 to the main file system
825 hierarchy. Each setting takes a
826 space-separated list of absolute
827 directory paths. Directories listed in
828 <varname>ReadWriteDirectories=</varname>
829 are accessible from within the
830 namespace with the same access rights
831 as from outside. Directories listed in
832 <varname>ReadOnlyDirectories=</varname>
833 are accessible for reading only,
834 writing will be refused even if the
835 usual file access controls would
836 permit this. Directories listed in
837 <varname>InaccessibleDirectories=</varname>
838 will be made inaccessible for
839 processes inside the namespace. Note
840 that restricting access with these
841 options does not extend to submounts
842 of a directory that are created later
843 on. These options may be specified
844 more than once in which case all
845 directories listed will have limited
846 access from within the namespace. If
847 the empty string is assigned to this
848 option, the specific list is reset,
849 and all prior assignments have no
852 <varname>ReadOnlyDirectories=</varname>
854 <varname>InaccessibleDirectories=</varname>
856 <literal>-</literal>, in which case
857 they will be ignored when they do not
858 exist. Note that using this
859 setting will disconnect propagation of
860 mounts from the service to the host
861 (propagation in the opposite direction
862 continues to work). This means that
863 this setting may not be used for
864 services which shall be able to
865 install mount points in the main mount
866 namespace.</para></listitem>
870 <term><varname>PrivateTmp=</varname></term>
872 <listitem><para>Takes a boolean
873 argument. If true, sets up a new file
874 system namespace for the executed
875 processes and mounts private
876 <filename>/tmp</filename> and
877 <filename>/var/tmp</filename>
878 directories inside it that is not
879 shared by processes outside of the
880 namespace. This is useful to secure
881 access to temporary files of the
882 process, but makes sharing between
884 <filename>/tmp</filename> or
885 <filename>/var/tmp</filename>
886 impossible. If this is enabled, all
887 temporary files created by a service
888 in these directories will be removed
889 after the service is stopped. Defaults
890 to false. It is possible to run two or
891 more units within the same private
892 <filename>/tmp</filename> and
893 <filename>/var/tmp</filename>
894 namespace by using the
895 <varname>JoinsNamespaceOf=</varname>
897 <citerefentry><refentrytitle>systemd.unit</refentrytitle><manvolnum>5</manvolnum></citerefentry>
898 for details. Note that using this
899 setting will disconnect propagation of
900 mounts from the service to the host
901 (propagation in the opposite direction
902 continues to work). This means that
903 this setting may not be used for
904 services which shall be able to install
905 mount points in the main mount
906 namespace.</para></listitem>
910 <term><varname>PrivateDevices=</varname></term>
912 <listitem><para>Takes a boolean
913 argument. If true, sets up a new /dev
914 namespace for the executed processes
915 and only adds API pseudo devices such
916 as <filename>/dev/null</filename>,
917 <filename>/dev/zero</filename> or
918 <filename>/dev/random</filename> (as
919 well as the pseudo TTY subsystem) to
920 it, but no physical devices such as
921 <filename>/dev/sda</filename>. This is
922 useful to securely turn off physical
923 device access by the executed
924 process. Defaults to false. Enabling
925 this option will also remove
926 <constant>CAP_MKNOD</constant> from
927 the capability bounding set for the
928 unit (see above), and set
929 <varname>DevicePolicy=closed</varname>
931 <citerefentry><refentrytitle>systemd.resource-control</refentrytitle><manvolnum>5</manvolnum></citerefentry>
932 for details). Note that using this
933 setting will disconnect propagation of
934 mounts from the service to the host
935 (propagation in the opposite direction
936 continues to work). This means that
937 this setting may not be used for
938 services which shall be able to
939 install mount points in the main mount
940 namespace.</para></listitem>
944 <term><varname>PrivateNetwork=</varname></term>
946 <listitem><para>Takes a boolean
947 argument. If true, sets up a new
948 network namespace for the executed
949 processes and configures only the
950 loopback network device
951 <literal>lo</literal> inside it. No
952 other network devices will be
953 available to the executed process.
954 This is useful to securely turn off
955 network access by the executed
956 process. Defaults to false. It is
957 possible to run two or more units
958 within the same private network
959 namespace by using the
960 <varname>JoinsNamespaceOf=</varname>
962 <citerefentry><refentrytitle>systemd.unit</refentrytitle><manvolnum>5</manvolnum></citerefentry>
963 for details. Note that this option
964 will disconnect all socket families
965 from the host, this includes
966 AF_NETLINK and AF_UNIX. The latter has
967 the effect that AF_UNIX sockets in the
968 abstract socket namespace will become
969 unavailable to the processes (however,
970 those located in the file system will
972 accessible).</para></listitem>
976 <term><varname>ProtectSystem=</varname></term>
978 <listitem><para>Takes a boolean
980 <literal>full</literal>. If true,
981 mounts the <filename>/usr</filename>
982 directory read-only for processes
983 invoked by this unit. If set to
984 <literal>full</literal>, the
985 <filename>/etc</filename> directory is mounted
986 read-only, too. This setting ensures
987 that any modification of the vendor
988 supplied operating system (and
989 optionally its configuration) is
990 prohibited for the service. It is
991 recommended to enable this setting for
992 all long-running services, unless they
993 are involved with system updates or
994 need to modify the operating system in
995 other ways. Note however that
996 processes retaining the CAP_SYS_ADMIN
997 capability can undo the effect of this
998 setting. This setting is hence
999 particularly useful for daemons which
1000 have this capability removed, for
1002 <varname>CapabilityBoundingSet=</varname>. Defaults
1003 to off.</para></listitem>
1007 <term><varname>ProtectHome=</varname></term>
1009 <listitem><para>Takes a boolean
1011 <literal>read-only</literal>. If true,
1013 <filename>/home</filename> and
1014 <filename>/run/user</filename> are
1015 made inaccessible and empty for
1016 processes invoked by this unit. If set
1017 to <literal>read-only</literal>, the
1018 two directories are made read-only
1019 instead. It is recommended to enable
1020 this setting for all long-running
1021 services (in particular network-facing
1022 ones), to ensure they cannot get access
1023 to private user data, unless the
1024 services actually require access to
1025 the user's private data. Note however
1026 that processes retaining the
1027 CAP_SYS_ADMIN capability can undo the
1028 effect of this setting. This setting
1029 is hence particularly useful for
1030 daemons which have this capability
1031 removed, for example with
1032 <varname>CapabilityBoundingSet=</varname>. Defaults
1033 to off.</para></listitem>
1037 <term><varname>MountFlags=</varname></term>
1039 <listitem><para>Takes a mount
1041 <option>shared</option>,
1042 <option>slave</option> or
1043 <option>private</option>, which
1044 control whether mounts in the file
1045 system namespace set up for this
1046 unit's processes will receive or
1047 propagate mounts or unmounts. See
1048 <citerefentry><refentrytitle>mount</refentrytitle><manvolnum>2</manvolnum></citerefentry>
1049 for details. Defaults to
1050 <option>shared</option>. Use
1051 <option>shared</option> to ensure that
1052 mounts and unmounts are propagated
1053 from the host to the container and
1054 vice versa. Use <option>slave</option>
1055 to run processes so that none of their
1056 mounts and unmounts will propagate to
1057 the host. Use <option>private</option>
1058 to also ensure that no mounts and
1059 unmounts from the host will propagate
1060 into the unit processes'
1061 namespace. Note that
1062 <option>slave</option> means that file
1063 systems mounted on the host might stay
1064 mounted continuously in the unit's
1065 namespace, and thus keep the device
1066 busy. Note that the file system
1067 namespace related options
1068 (<varname>PrivateTmp=</varname>,
1069 <varname>PrivateDevices=</varname>,
1070 <varname>ProtectSystem=</varname>,
1071 <varname>ProtectHome=</varname>,
1072 <varname>ReadOnlyDirectories=</varname>,
1073 <varname>InaccessibleDirectories=</varname>
1075 <varname>ReadWriteDirectories=</varname>)
1076 require that mount and unmount
1077 propagation from the unit's file
1078 system namespace is disabled, and
1080 <option>shared</option> to
1081 <option>slave</option>.
1086 <term><varname>UtmpIdentifier=</varname></term>
1088 <listitem><para>Takes a four
1089 character identifier string for an
1090 utmp/wtmp entry for this service. This
1091 should only be set for services such
1092 as <command>getty</command>
1093 implementations where utmp/wtmp
1094 entries must be created and cleared
1095 before and after execution. If the
1096 configured string is longer than four
1097 characters, it is truncated and the
1098 terminal four characters are
1099 used. This setting interprets %I style
1100 string replacements. This setting is
1101 unset by default, i.e. no utmp/wtmp
1102 entries are created or cleaned up for
1103 this service.</para></listitem>
1107 <term><varname>SELinuxContext=</varname></term>
1109 <listitem><para>Set the SELinux
1110 security context of the executed
1111 process. If set, this will override
1112 the automated domain
1113 transition. However, the policy still
1114 needs to autorize the transition. This
1115 directive is ignored if SELinux is
1116 disabled. If prefixed by
1117 <literal>-</literal>, all errors will
1119 <citerefentry><refentrytitle>setexeccon</refentrytitle><manvolnum>3</manvolnum></citerefentry>
1120 for details.</para></listitem>
1124 <term><varname>AppArmorProfile=</varname></term>
1126 <listitem><para>Takes a profile name as argument.
1127 The process executed by the unit will switch to
1128 this profile when started. Profiles must already
1129 be loaded in the kernel, or the unit will fail.
1130 This result in a non operation if AppArmor is not
1131 enabled. If prefixed by <literal>-</literal>, all errors
1137 <term><varname>IgnoreSIGPIPE=</varname></term>
1139 <listitem><para>Takes a boolean
1140 argument. If true, causes <constant>SIGPIPE</constant> to be
1141 ignored in the executed
1142 process. Defaults to true because
1143 <constant>SIGPIPE</constant> generally is useful only in
1144 shell pipelines.</para></listitem>
1148 <term><varname>NoNewPrivileges=</varname></term>
1150 <listitem><para>Takes a boolean
1151 argument. If true, ensures that the
1152 service process and all its children
1153 can never gain new privileges. This
1154 option is more powerful than the respective
1155 secure bits flags (see above), as it
1156 also prohibits UID changes of any
1157 kind. This is the simplest, most
1158 effective way to ensure that a process
1159 and its children can never elevate
1160 privileges again.</para></listitem>
1164 <term><varname>SystemCallFilter=</varname></term>
1166 <listitem><para>Takes a
1167 space-separated list of system call
1168 names. If this setting is used, all
1169 system calls executed by the unit
1170 processes except for the listed ones
1171 will result in immediate process
1172 termination with the
1173 <constant>SIGSYS</constant> signal
1174 (whitelisting). If the first character
1175 of the list is <literal>~</literal>,
1176 the effect is inverted: only the
1177 listed system calls will result in
1178 immediate process termination
1179 (blacklisting). If running in user
1180 mode and this option is used,
1181 <varname>NoNewPrivileges=yes</varname>
1182 is implied. This feature makes use of the
1183 Secure Computing Mode 2 interfaces of
1184 the kernel ('seccomp filtering') and
1185 is useful for enforcing a minimal
1186 sandboxing environment. Note that the
1187 <function>execve</function>,
1188 <function>rt_sigreturn</function>,
1189 <function>sigreturn</function>,
1190 <function>exit_group</function>,
1191 <function>exit</function> system calls
1192 are implicitly whitelisted and do not
1193 need to be listed explicitly. This
1194 option may be specified more than once
1195 in which case the filter masks are
1196 merged. If the empty string is
1197 assigned, the filter is reset, all
1198 prior assignments will have no
1201 <para>If you specify both types of
1202 this option (i.e. whitelisting and
1203 blacklisting), the first encountered
1204 will take precedence and will dictate
1205 the default action (termination or
1206 approval of a system call). Then the
1207 next occurrences of this option will
1208 add or delete the listed system calls
1209 from the set of the filtered system
1210 calls, depending of its type and the
1211 default action. (For example, if you have started
1212 with a whitelisting of
1213 <function>read</function> and
1214 <function>write</function>, and right
1215 after it add a blacklisting of
1216 <function>write</function>, then
1217 <function>write</function> will be
1218 removed from the set.)
1223 <term><varname>SystemCallErrorNumber=</varname></term>
1225 <listitem><para>Takes an
1226 <literal>errno</literal> error number
1227 name to return when the system call
1228 filter configured with
1229 <varname>SystemCallFilter=</varname>
1230 is triggered, instead of terminating
1231 the process immediately. Takes an
1233 <constant>EPERM</constant>,
1234 <constant>EACCES</constant> or
1235 <constant>EUCLEAN</constant>. When this
1236 setting is not used, or when the empty
1237 string is assigned, the process will be
1238 terminated immediately when the filter
1239 is triggered.</para></listitem>
1243 <term><varname>SystemCallArchitectures=</varname></term>
1245 <listitem><para>Takes a space
1246 separated list of architecture
1247 identifiers to include in the system
1248 call filter. The known architecture
1250 <constant>x86</constant>,
1251 <constant>x86-64</constant>,
1252 <constant>x32</constant>,
1253 <constant>arm</constant> as well as
1254 the special identifier
1255 <constant>native</constant>. Only
1256 system calls of the specified
1257 architectures will be permitted to
1258 processes of this unit. This is an
1259 effective way to disable compatibility
1260 with non-native architectures for
1261 processes, for example to prohibit
1262 execution of 32-bit x86 binaries on
1263 64-bit x86-64 systems. The special
1264 <constant>native</constant> identifier
1265 implicitly maps to the native
1266 architecture of the system (or more
1267 strictly: to the architecture the
1268 system manager is compiled for). If
1269 running in user mode and this option
1271 <varname>NoNewPrivileges=yes</varname>
1272 is implied. Note that setting this
1273 option to a non-empty list implies
1274 that <constant>native</constant> is
1275 included too. By default, this option
1276 is set to the empty list, i.e. no
1277 architecture system call filtering is
1278 applied.</para></listitem>
1282 <term><varname>RestrictAddressFamilies=</varname></term>
1284 <listitem><para>Restricts the set of
1285 socket address families accessible to
1286 the processes of this unit. Takes a
1287 space-separated list of address family
1288 names to whitelist, such as
1289 <constant>AF_UNIX</constant>,
1290 <constant>AF_INET</constant> or
1291 <constant>AF_INET6</constant>. When
1292 prefixed with <constant>~</constant>
1293 the listed address families will be
1294 applied as blacklist, otherwise as
1295 whitelist. Note that this restricts
1297 <citerefentry><refentrytitle>socket</refentrytitle><manvolnum>2</manvolnum></citerefentry>
1298 system call only. Sockets passed into
1299 the process by other means (for
1300 example, by using socket activation
1301 with socket units, see
1302 <citerefentry><refentrytitle>systemd.socket</refentrytitle><manvolnum>5</manvolnum></citerefentry>)
1303 are unaffected. Also, sockets created
1304 with <function>socketpair()</function>
1305 (which creates connected AF_UNIX
1306 sockets only) are unaffected. Note
1307 that this option has no effect on
1308 32-bit x86 and is ignored (but works
1309 correctly on x86-64). If running in user
1310 mode and this option is used,
1311 <varname>NoNewPrivileges=yes</varname>
1312 is implied. By default, no
1313 restriction applies, all address
1314 families are accessible to
1315 processes. If assigned the empty
1316 string, any previous list changes are
1319 <para>Use this option to limit
1320 exposure of processes to remote
1321 systems, in particular via exotic
1322 network protocols. Note that in most
1324 <constant>AF_UNIX</constant> address
1325 family should be included in the
1326 configured whitelist as it is
1327 frequently used for local
1328 communication, including for
1329 <citerefentry><refentrytitle>syslog</refentrytitle><manvolnum>2</manvolnum></citerefentry>
1330 logging.</para></listitem>
1334 <term><varname>Personality=</varname></term>
1336 <listitem><para>Controls which
1338 <citerefentry><refentrytitle>uname</refentrytitle><manvolnum>2</manvolnum></citerefentry>
1339 shall report, when invoked by unit
1340 processes. Takes one of
1341 <constant>x86</constant> and
1342 <constant>x86-64</constant>. This is
1343 useful when running 32-bit services on
1344 a 64-bit host system. If not specified,
1345 the personality is left unmodified and
1346 thus reflects the personality of the
1348 kernel.</para></listitem>
1352 <term><varname>RuntimeDirectory=</varname></term>
1353 <term><varname>RuntimeDirectoryMode=</varname></term>
1355 <listitem><para>Takes a list of
1356 directory names. If set, one or more
1357 directories by the specified names
1358 will be created below
1359 <filename>/run</filename> (for system
1361 <varname>$XDG_RUNTIME_DIR</varname>
1362 (for user services) when the unit is
1363 started, and removed when the unit is
1364 stopped. The directories will have the
1365 access mode specified in
1366 <varname>RuntimeDirectoryMode=</varname>,
1367 and will be owned by the user and
1369 <varname>User=</varname> and
1370 <varname>Group=</varname>. Use this to
1371 manage one or more runtime directories
1372 of the unit and bind their lifetime to
1373 the daemon runtime. The specified
1374 directory names must be relative, and
1376 <literal>/</literal>, i.e. must refer
1377 to simple directories to create or
1378 remove. This is particularly useful
1379 for unprivileged daemons that cannot
1380 create runtime directories in
1381 <filename>/run</filename> due to lack
1382 of privileges, and to make sure the
1383 runtime directory is cleaned up
1384 automatically after use. For runtime
1385 directories that require more complex
1386 or different configuration or lifetime
1387 guarantees, please consider using
1388 <citerefentry><refentrytitle>tmpfiles.d</refentrytitle><manvolnum>5</manvolnum></citerefentry>.</para></listitem>
1395 <title>Environment variables in spawned processes</title>
1397 <para>Processes started by the system are executed in
1398 a clean environment in which select variables
1399 listed below are set. System processes started by systemd
1400 do not inherit variables from PID 1, but processes
1401 started by user systemd instances inherit all
1402 environment variables from the user systemd instance.
1405 <variablelist class='environment-variables'>
1407 <term><varname>$PATH</varname></term>
1409 <listitem><para>Colon-separated list
1410 of directiories to use when launching
1411 executables. Systemd uses a fixed
1413 <filename>/usr/local/sbin</filename>:<filename>/usr/local/bin</filename>:<filename>/usr/sbin</filename>:<filename>/usr/bin</filename>:<filename>/sbin</filename>:<filename>/bin</filename>.
1418 <term><varname>$LANG</varname></term>
1420 <listitem><para>Locale. Can be set in
1421 <citerefentry><refentrytitle>locale.conf</refentrytitle><manvolnum>5</manvolnum></citerefentry>
1422 or on the kernel command line (see
1423 <citerefentry><refentrytitle>systemd</refentrytitle><manvolnum>1</manvolnum></citerefentry>
1425 <citerefentry><refentrytitle>kernel-command-line</refentrytitle><manvolnum>7</manvolnum></citerefentry>).
1430 <term><varname>$USER</varname></term>
1431 <term><varname>$LOGNAME</varname></term>
1432 <term><varname>$HOME</varname></term>
1433 <term><varname>$SHELL</varname></term>
1435 <listitem><para>User name (twice), home
1436 directory, and the login shell.
1437 The variables are set for the units that
1438 have <varname>User=</varname> set,
1440 <command>systemd</command> instances.
1442 <citerefentry><refentrytitle>passwd</refentrytitle><manvolnum>5</manvolnum></citerefentry>.
1447 <term><varname>$XDG_RUNTIME_DIR</varname></term>
1449 <listitem><para>The directory for volatile
1450 state. Set for the user <command>systemd</command>
1451 instance, and also in user sessions.
1453 <citerefentry><refentrytitle>pam_systemd</refentrytitle><manvolnum>8</manvolnum></citerefentry>.
1458 <term><varname>$XDG_SESSION_ID</varname></term>
1459 <term><varname>$XDG_SEAT</varname></term>
1460 <term><varname>$XDG_VTNR</varname></term>
1462 <listitem><para>The identifier of the
1463 session, the seat name, and
1464 virtual terminal of the session. Set
1466 <citerefentry><refentrytitle>pam_systemd</refentrytitle><manvolnum>8</manvolnum></citerefentry>
1468 <varname>$XDG_SEAT</varname> and
1469 <varname>$XDG_VTNR</varname> will
1470 only be set when attached to a seat and a
1471 tty.</para></listitem>
1475 <term><varname>$MAINPID</varname></term>
1477 <listitem><para>The PID of the units
1478 main process if it is known. This is
1479 only set for control processes as
1481 <varname>ExecReload=</varname> and
1482 similar. </para></listitem>
1486 <term><varname>$MANAGERPID</varname></term>
1488 <listitem><para>The PID of the user
1489 <command>systemd</command> instance,
1490 set for processes spawned by it.
1495 <term><varname>$LISTEN_FDS</varname></term>
1496 <term><varname>$LISTEN_PID</varname></term>
1498 <listitem><para>Information about file
1499 descriptors passed to a service for
1500 socket activation. See
1501 <citerefentry><refentrytitle>sd_listen_fds</refentrytitle><manvolnum>3</manvolnum></citerefentry>.
1506 <term><varname>$TERM</varname></term>
1508 <listitem><para>Terminal type, set
1509 only for units connected to a terminal
1510 (<varname>StandardInput=tty</varname>,
1511 <varname>StandardOutput=tty</varname>,
1513 <varname>StandardError=tty</varname>).
1515 <citerefentry project='man-pages'><refentrytitle>termcap</refentrytitle><manvolnum>5</manvolnum></citerefentry>.
1520 <para>Additional variables may be configured by the
1521 following means: for processes spawned in specific
1522 units, use the <varname>Environment=</varname> and
1523 <varname>EnvironmentFile=</varname> options above; to
1524 specify variables globally, use
1525 <varname>DefaultEnvironment=</varname> (see
1526 <citerefentry><refentrytitle>systemd-system.conf</refentrytitle><manvolnum>5</manvolnum></citerefentry>)
1527 or the kernel option
1528 <varname>systemd.setenv=</varname> (see
1529 <citerefentry><refentrytitle>systemd</refentrytitle><manvolnum>1</manvolnum></citerefentry>). Additional
1530 variables may also be set through PAM,
1531 cf. <citerefentry project='man-pages'><refentrytitle>pam_env</refentrytitle><manvolnum>8</manvolnum></citerefentry>.</para>
1535 <title>See Also</title>
1537 <citerefentry><refentrytitle>systemd</refentrytitle><manvolnum>1</manvolnum></citerefentry>,
1538 <citerefentry><refentrytitle>systemctl</refentrytitle><manvolnum>1</manvolnum></citerefentry>,
1539 <citerefentry><refentrytitle>journalctl</refentrytitle><manvolnum>8</manvolnum></citerefentry>,
1540 <citerefentry><refentrytitle>systemd.unit</refentrytitle><manvolnum>5</manvolnum></citerefentry>,
1541 <citerefentry><refentrytitle>systemd.service</refentrytitle><manvolnum>5</manvolnum></citerefentry>,
1542 <citerefentry><refentrytitle>systemd.socket</refentrytitle><manvolnum>5</manvolnum></citerefentry>,
1543 <citerefentry><refentrytitle>systemd.swap</refentrytitle><manvolnum>5</manvolnum></citerefentry>,
1544 <citerefentry><refentrytitle>systemd.mount</refentrytitle><manvolnum>5</manvolnum></citerefentry>,
1545 <citerefentry><refentrytitle>systemd.kill</refentrytitle><manvolnum>5</manvolnum></citerefentry>,
1546 <citerefentry><refentrytitle>systemd.resource-control</refentrytitle><manvolnum>5</manvolnum></citerefentry>,
1547 <citerefentry><refentrytitle>systemd.directives</refentrytitle><manvolnum>7</manvolnum></citerefentry>,
1548 <citerefentry><refentrytitle>tmpfiles.d</refentrytitle><manvolnum>5</manvolnum></citerefentry>,
1549 <citerefentry project='man-pages'><refentrytitle>exec</refentrytitle><manvolnum>3</manvolnum></citerefentry>