10 months agomeson: avoid ternary op in .format() 14003/head
Zbigniew Jędrzejewski-Szmek [Wed, 13 Nov 2019 21:22:58 +0000 (22:22 +0100)] 
meson: avoid ternary op in .format()

meson 0.49 can't parse that for some reason. I'm keeping this separate so it
can be reverted easily when we bump required meson version.

10 months agomeson: make user $PATH configurable
Zbigniew Jędrzejewski-Szmek [Tue, 12 Nov 2019 14:38:19 +0000 (15:38 +0100)] 
meson: make user $PATH configurable

This partially reverts db11487d1062655f17db54c4d710653f16c87313 (the logic to
calculate the correct value is removed, we always use the same setting as for
the system manager). Distributions have an easy mechanism to override this if
they wish.

I think making this configurable is better, because different distros clearly
want different defaults here, and making this configurable is nice and clean.
If we don't make it configurable, distros which either have to carry patches,
or what would be worse, rely on some other configuration mechanism, like
/etc/profile. Those other solutions do not apply everywhere (they usually
require the shell to be used at some point), so it is better if we provide
a nice way to override the default.

Fixes  #13469.

10 months agoMerge pull request #13862 from zachsmith/systemd-tmpfiles-deprecate-for-force
Zbigniew Jędrzejewski-Szmek [Tue, 12 Nov 2019 09:28:59 +0000 (10:28 +0100)] 
Merge pull request #13862 from zachsmith/systemd-tmpfiles-deprecate-for-force

systemd-tmpfiles: deprecate F for f+

10 months agoudev: do not append newline when write attributes
Yu Watanabe [Sun, 10 Nov 2019 05:42:55 +0000 (14:42 +0900)] 
udev: do not append newline when write attributes

Before 25de7aa7b90c23d33ea50ada1e50c5834a414237, the content is written
by `fprintf()` without new line. So WRITE_STRING_FILE_AVOID_NEWLINE flag
is necessary.

Fixes #13985.

10 months agoMerge pull request #13997 from khfeng/hwdb-dell-vostro5581-ish
Anita Zhang [Tue, 12 Nov 2019 01:56:56 +0000 (17:56 -0800)] 
Merge pull request #13997 from khfeng/hwdb-dell-vostro5581-ish

hwdb: Mark Intel Sensor Hub's accel sensor on Vostro 5581 as being in…

10 months agoMerge pull request #13996 from poettering/utc-fix
Anita Zhang [Tue, 12 Nov 2019 01:47:43 +0000 (17:47 -0800)] 
Merge pull request #13996 from poettering/utc-fix

accept UTC timezone explicitly, even if timezone data is missing

10 months agohwdb: Mark Intel Sensor Hub's accel sensor on Vostro 5581 as being in the base 13997/head
Kai-Heng Feng [Mon, 11 Nov 2019 16:30:46 +0000 (00:30 +0800)] 
hwdb: Mark Intel Sensor Hub's accel sensor on Vostro 5581 as being in the base

This laptop uses the accelerometer as a freefall sensor, so mark it as
in base to prevent screen rotation.

10 months agotime-util: uniquify timezone list, in case UTC is listed in, too 13996/head
Lennart Poettering [Mon, 11 Nov 2019 15:56:31 +0000 (16:56 +0100)] 
time-util: uniquify timezone list, in case UTC is listed in, too

10 months agotime-util: always accept UTC as valid timezone
Lennart Poettering [Mon, 11 Nov 2019 15:53:52 +0000 (16:53 +0100)] 
time-util: always accept UTC as valid timezone

We already handle it specially in get_timezones(), hence we should OK it
here too, even if the timezone file doesn't actually exist.

Prompted by:

(Yes, Ubuntu should install the UTC timezone data unconditionally: it
should not be an option, even if all other timezone data is excluded,
but since it's our business to validate user input but not out business
to validate distros, let's just accept "UTC" unconditionally, it's magic
after all)

10 months agohwdb: Add accel orientation quirk for Wortmann Terra Pad 1061
Hans de Goede [Wed, 6 Nov 2019 09:22:21 +0000 (10:22 +0100)] 
hwdb: Add accel orientation quirk for Wortmann Terra Pad 1061

Add a quirk to fix the accelerometer orientation on the Wortmann
Terra Pad 1061 tablet.

10 months agojournal-file: delete some unnecessary braces
Vito Caputo [Sun, 10 Nov 2019 07:00:19 +0000 (23:00 -0800)] 
journal-file: delete some unnecessary braces

Trivial change, just something I noticed skimming the code.

10 months agoMerge pull request #13975 from keszybz/more-seccomp-syscalls
Yu Watanabe [Sat, 9 Nov 2019 14:27:34 +0000 (23:27 +0900)] 
Merge pull request #13975 from keszybz/more-seccomp-syscalls

Add more syscalls to the seccomp lists

10 months agoshared/sleep-config: fix potential SEGV
Zbigniew Jędrzejewski-Szmek [Fri, 8 Nov 2019 09:25:31 +0000 (10:25 +0100)] 
shared/sleep-config: fix potential SEGV

We were looking at the wrong variable, and would always crash if this
comparison was reached. Fixes #13965.

Also, fix crash (_cleanup_ called on uninitialized variable) if we failed in
error path.

While at it, let's shorten some messages.

10 months agosystemd-tmpfiles: cleanup man page program listing 13862/head
Zach Smith [Fri, 1 Nov 2019 05:10:10 +0000 (22:10 -0700)] 
systemd-tmpfiles: cleanup man page program listing

10 months agoAdd @pkey syscall group 13975/head
Zbigniew Jędrzejewski-Szmek [Fri, 8 Nov 2019 11:56:56 +0000 (12:56 +0100)] 
Add @pkey syscall group

Inspired by
This change doesn't solve the issue, but makes it easier to whitelist the
syscall group.

10 months agoseccomp: add all *time64 syscalls
Zbigniew Jędrzejewski-Szmek [Fri, 8 Nov 2019 13:00:10 +0000 (14:00 +0100)] 
seccomp: add all *time64 syscalls

> utime is an obsolete system call. The current kernel interface is
> utimensat_time64. New 32-bit architectures do not even provide the utime
> system call.

Also add all other *time64 syscalls listed in

10 months agoMerge pull request #13554 from keur/systemctl_status_timer
Zbigniew Jędrzejewski-Szmek [Fri, 8 Nov 2019 13:19:40 +0000 (14:19 +0100)] 
Merge pull request #13554 from keur/systemctl_status_timer

systemctl: Add timer activation to status

10 months agolibblkid: open device in nonblock mode.
Michal Suchanek [Mon, 4 Nov 2019 20:23:15 +0000 (21:23 +0100)] 
libblkid: open device in nonblock mode.

When autoclose is set (kernel default but many distributions reverse the
setting) opening a CD-rom device causes the tray to close.

The function of blkid is to report the current state of the device and
not to change it. Hence it should use O_NONBLOCK when opening the
device to avoid closing a CD-rom tray.

blkid is used liberally in scripts so it can potentially interfere with
the user operating the CD-rom hardware.

[ add O_NONBLOCK also to:
                  - wipefs
                  - blkid_new_probe_from_filename()
                  - blkid_evaluate_tag()]

Signed-off-by: Michal Suchanek <>
Signed-off-by: Karel Zak <>
(cherry picked from commit 39f5af25982d8b0244000e92a9d0e0e6557d0e17)

10 months agoMerge pull request #13962 from keszybz/man-ordering
Zbigniew Jędrzejewski-Szmek [Fri, 8 Nov 2019 09:18:46 +0000 (10:18 +0100)] 
Merge pull request #13962 from keszybz/man-ordering

Describe ordering in case of Conflicts=

10 months agoman: describe ordering in case of Conflicts= 13962/head
Zbigniew Jędrzejewski-Szmek [Wed, 6 Nov 2019 12:42:13 +0000 (13:42 +0100)] 
man: describe ordering in case of Conflicts=

Fixes #13421.

10 months agodoc: Fix missing parenthesis
Alcaro [Thu, 7 Nov 2019 13:19:18 +0000 (14:19 +0100)] 
doc: Fix missing parenthesis

10 months agoinclude missing_fcntl.h where needed
Anita Zhang [Thu, 7 Nov 2019 06:25:43 +0000 (22:25 -0800)] 
include missing_fcntl.h where needed

f5947a5e925117c55b390460d592f57504277bf9 dropped missing.h and
replaced with the more specific headers but did not add
missing_fcntl.h in places that use O_TMPFILE. This is needed for
some older versions of glibc.

10 months agoMerge pull request #13904 from keur/job_mode_triggering
Zbigniew Jędrzejewski-Szmek [Thu, 7 Nov 2019 07:36:26 +0000 (08:36 +0100)] 
Merge pull request #13904 from keur/job_mode_triggering

Job mode triggering

10 months agocore: change top-level drop-in from -.service.d to service.d
Anita Zhang [Tue, 5 Nov 2019 02:29:55 +0000 (18:29 -0800)] 
core: change top-level drop-in from -.service.d to service.d

Discussed in #13743, the -.service semantic conflicts with the
existing root mount and slice names, making this feature not
uniformly extensible to all types. Change the name to be
<type>.d instead.

Updating to this format also extends the top-level dropin to
unit types.

10 months agoman: put description of Wants= above Requires=
Zbigniew Jędrzejewski-Szmek [Wed, 6 Nov 2019 12:26:29 +0000 (13:26 +0100)] 
man: put description of Wants= above Requires=

We want users to use Wants, but we'd describe Requires first and ask users to
look for Wants instead. While at it, let's split the wall of text into sensible
paragraphs: syntax first, followed by semantics and longer description, and
finally hints and comparison to other configuration items last.

10 months agoMerge pull request #13960 from keszybz/meson-loop-fix
Anita Zhang [Wed, 6 Nov 2019 19:44:09 +0000 (11:44 -0800)] 
Merge pull request #13960 from keszybz/meson-loop-fix

meson: remove strange dep that causes meson to enter infinite loop

10 months agoman: fix option typo in pam_systemd man page
Dominique Martinet [Wed, 6 Nov 2019 06:32:17 +0000 (07:32 +0100)] 
man: fix option typo in pam_systemd man page

The session= option does not exist, XDG_SESSION_TYPE overrides the type=

10 months agomeson: remove strange dep that causes meson to enter infinite loop 13960/head
Zbigniew Jędrzejewski-Szmek [Wed, 6 Nov 2019 11:44:39 +0000 (12:44 +0100)] 
meson: remove strange dep that causes meson to enter infinite loop

The value is obviously bogus, but didn't seem to cause problems so far.
With meson-0.52.0, it causes a hang. The number of aliases is always rather
small (usually just one or two, possibly up to a dozen in a few cases), so
even if this causes some looping, it is strange that it has such a huge impact.
But let's just remove it.

Fixes #13742.

Tested with meson-0.52.0-1.module_f31+6771+f5d842eb.noarch,

10 months agoman: save pull-raw example file without underscores
Anita Zhang [Wed, 6 Nov 2019 00:04:27 +0000 (16:04 -0800)] 
man: save pull-raw example file without underscores

Destination file needs to be a valid hostname and underscores
are not valid hostname characters.

Closes #13542

10 months agoman: small grammatical/word choice fixes to crypttab man page
Anita Zhang [Wed, 6 Nov 2019 00:20:42 +0000 (16:20 -0800)] 
man: small grammatical/word choice fixes to crypttab man page

Closes #13608

10 months agoMerge pull request #13936 from keszybz/format-table-uninhibited
Anita Zhang [Tue, 5 Nov 2019 23:03:15 +0000 (15:03 -0800)] 
Merge pull request #13936 from keszybz/format-table-uninhibited

Output tables at full width if piped

10 months agoMerge pull request #13944 from yuwata/network-split-struct-prefix
Anita Zhang [Tue, 5 Nov 2019 23:02:01 +0000 (15:02 -0800)] 
Merge pull request #13944 from yuwata/network-split-struct-prefix

network: split struct Prefix into Prefix and RoutePrefix

10 months agosd-device-enumerator: do not return error when a device is removed
Yu Watanabe [Tue, 5 Nov 2019 14:02:31 +0000 (23:02 +0900)] 
sd-device-enumerator: do not return error when a device is removed

Fixes #13814.

10 months agosystemctl: simplify printing of Triggers/TriggeredBy 13554/head
Zbigniew Jędrzejewski-Szmek [Tue, 5 Nov 2019 21:01:46 +0000 (22:01 +0100)] 
systemctl: simplify printing of Triggers/TriggeredBy

10 months agosystemctl: Align all status outputs to TriggeredBy
Kevin Kuehler [Mon, 4 Nov 2019 22:52:13 +0000 (14:52 -0800)] 
systemctl: Align all status outputs to TriggeredBy

10 months agosystemctl: Add TriggeredBy and Triggers to status
Kevin Kuehler [Mon, 4 Nov 2019 22:48:06 +0000 (14:48 -0800)] 
systemctl: Add TriggeredBy and Triggers to status

For all units that aren't timers, if it is activated by another unit,
add the triggering unit under the  "TriggeredBy:" header. If a unit can
trigger other units, print the units it triggers other the "Triggers:"

10 months agoman: Document --job-mode=triggering switch 13904/head
Kevin Kuehler [Fri, 1 Nov 2019 08:47:37 +0000 (01:47 -0700)] 
man: Document --job-mode=triggering switch

10 months agocore: Add triggering job mode
Kevin Kuehler [Fri, 1 Nov 2019 07:54:03 +0000 (00:54 -0700)] 
core: Add triggering job mode

When used with systemctl stop, follows TRIGGERED_BY dependencies and
adds them to the same transaction.

Fixes: #3043

10 months agoshared/format-table: disable ellipsization when piped 13936/head
Zbigniew Jędrzejewski-Szmek [Mon, 4 Nov 2019 13:59:31 +0000 (14:59 +0100)] 
shared/format-table: disable ellipsization when piped

Fixes #13461.

$ systemd-inhibit --no-pager
WHO            UID  USER    PID   COMM           WHAT                                  WHY          MODE
ModemManager   0    root    1093  ModemManager   sleep                                 ModemManage… delay
NetworkManager 0    root    1400  NetworkManager sleep                                 NetworkMana… delay
UPower         0    root    5141  upowerd        sleep                                 Pause devic… delay
zbyszek        1000 zbyszek 10036 gsd-power      handle-lid-switch                     External mo… block
zbyszek        1000 zbyszek 10035 gsd-media-keys handle-power-key:handle-suspend-key:… GNOME handl… block
zbyszek        1000 zbyszek 10035 gsd-media-keys sleep                                 GNOME handl… delay
zbyszek        1000 zbyszek 10036 gsd-power      sleep                                 GNOME needs… delay

7 inhibitors listed.
$ systemd-inhibit --no-pager|grep suspend
$ systemd-inhibit --no-pager|cat
WHO            UID  USER    PID   COMM           WHAT                WHY   MODE
ModemManager   0    root    1093  ModemManager   sleep               Mode… delay
NetworkManager 0    root    1400  NetworkManager sleep               Netw… delay
UPower         0    root    5141  upowerd        sleep               Paus… delay
zbyszek        1000 zbyszek 10036 gsd-power      handle-lid-switch   Exte… block
zbyszek        1000 zbyszek 10035 gsd-media-keys handle-power-key:h… GNOM… block
zbyszek        1000 zbyszek 10035 gsd-media-keys sleep               GNOM… delay
zbyszek        1000 zbyszek 10036 gsd-power      sleep               GNOM… delay

$ build/systemd-inhibit --no-pager
(same as above)

$ build/systemd-inhibit --no-pager|grep suspend
zbyszek        1000 zbyszek 10035 gsd-media-keys handle-power-key:handle-suspend-key:handle-hibernate-key GNOME handling keypresses                                   block
$ build/systemd-inhibit --no-pager|cat
WHO            UID  USER    PID   COMM           WHAT                                                     WHY                                                         MODE
ModemManager   0    root    1093  ModemManager   sleep                                                    ModemManager needs to reset devices                         delay
NetworkManager 0    root    1400  NetworkManager sleep                                                    NetworkManager needs to turn off networks                   delay
UPower         0    root    5141  upowerd        sleep                                                    Pause device polling                                        delay
zbyszek        1000 zbyszek 10036 gsd-power      handle-lid-switch                                        External monitor attached or configuration changed recently block
zbyszek        1000 zbyszek 10035 gsd-media-keys handle-power-key:handle-suspend-key:handle-hibernate-key GNOME handling keypresses                                   block
zbyszek        1000 zbyszek 10035 gsd-media-keys sleep                                                    GNOME handling keypresses                                   delay
zbyszek        1000 zbyszek 10036 gsd-power      sleep                                                    GNOME needs to lock the screen                              delay

7 inhibitors listed.

Note that this affect all tools that use format-table.c: machinectl, busctl,
loginctl, systemd-analyze, networkctl, portablectl.

10 months agonetwork: split struct Prefix into Prefix and RoutePrefix 13944/head
Yu Watanabe [Tue, 5 Nov 2019 13:31:20 +0000 (22:31 +0900)] 
network: split struct Prefix into Prefix and RoutePrefix

10 months agoMerge pull request #13939 from yuwata/network-fix-memleak-and-13938
Yu Watanabe [Tue, 5 Nov 2019 11:16:06 +0000 (20:16 +0900)] 
Merge pull request #13939 from yuwata/network-fix-memleak-and-13938

network: fix memleak and invalid free function

10 months agojournal: refresh cached credentials of stdout streams
Lorenz Bauer [Mon, 4 Nov 2019 16:35:46 +0000 (16:35 +0000)] 
journal: refresh cached credentials of stdout streams

journald assumes that getsockopt(SO_PEERCRED) correctly identifies the
process on the remote end of the socket. However, this is incorrect
according to man 7 socket:

    The returned  credentials  are  those that were in effect at the
    time of the call to connect(2) or socketpair(2).

This becomes a problem when a new process inherits the stdout stream
from a parent. First, log messages from the child process will
be attributed to the parent. Second, the struct ucred used by journald
becomes invalid as soon as the parent exits. Further sendmsg calls then
fail with ENOENT. Logs for the child process then vanish from the journal.

Fix this by using recvmsg on the stdout stream, and refreshing the cached
struct ucred if SCM_CREDENTIALS indicate a new process.

Fixes #13708

10 months agohwdb: add XKB_FIXED_MODEL to the keyboard hwdb
Sebastian Wick [Thu, 31 Oct 2019 13:27:24 +0000 (14:27 +0100)] 
hwdb: add XKB_FIXED_MODEL to the keyboard hwdb

Chromebook keyboards have a top row which generates f1-f10 key codes but
the keys have media symbols printed on them. A simple scan code to key
code mapping to the correct media keys makes the f1-f10 inaccessible. To
properly use the keyboard a custom key code to symbol mapping in xbk is
required (a variant of the chromebook xkb model is already upstream).
Other devices have similar problems.
This commit makes it possible to specify which xkb model should be used
for a specific device by setting XKB_FIXED_MODEL.

10 months agoMerge pull request #13935 from poettering/bootctl-random-seed-mkdir
Anita Zhang [Tue, 5 Nov 2019 06:57:12 +0000 (22:57 -0800)] 
Merge pull request #13935 from poettering/bootctl-random-seed-mkdir

bootctl: create leading dirs when "bootctl random-seed" is called if …

10 months agosd-radv: fix memleak 13939/head
Yu Watanabe [Tue, 5 Nov 2019 04:30:00 +0000 (13:30 +0900)] 
sd-radv: fix memleak

10 months agotest: add testcase for issue #13938
Yu Watanabe [Tue, 5 Nov 2019 01:49:07 +0000 (10:49 +0900)] 
test: add testcase for issue #13938

10 months agonetwork: fix memleak in route_prefix_free()
Yu Watanabe [Tue, 5 Nov 2019 01:46:44 +0000 (10:46 +0900)] 
network: fix memleak in route_prefix_free()

10 months agonetwork: fix typo
Yu Watanabe [Tue, 5 Nov 2019 01:42:19 +0000 (10:42 +0900)] 
network: fix typo

10 months agonetwork: fix invalid cleanup function
Yu Watanabe [Tue, 5 Nov 2019 01:39:15 +0000 (10:39 +0900)] 
network: fix invalid cleanup function

10 months agonetwork: fix memleak
Yu Watanabe [Tue, 5 Nov 2019 00:39:44 +0000 (09:39 +0900)] 
network: fix memleak

10 months agonetwork: use fix invalid free function
Yu Watanabe [Tue, 5 Nov 2019 00:36:39 +0000 (09:36 +0900)] 
network: use fix invalid free function

Fixes #13938.

10 months agocore, job: fix breakage of ordering dependencies by systemctl reload command
HATAYAMA Daisuke [Mon, 28 Oct 2019 10:35:24 +0000 (19:35 +0900)] 
core, job: fix breakage of ordering dependencies by systemctl reload command

Currently, systemctl reload command breaks ordering dependencies if it's
executed when its target service unit is in activating state.

For example, prepare A.service, B.service and as follows:

    # systemctl cat A.service B.service
    # /etc/systemd/system/A.service

    ExecStart=/usr/bin/echo A1
    ExecStart=/usr/bin/sleep 60
    ExecStart=/usr/bin/echo A2
    ExecReload=/usr/bin/echo A reloaded

    # /etc/systemd/system/B.service

    ExecStart=/usr/bin/echo B

    # /etc/systemd/system/
    Wants=A.service B.service

Start them.

    # systemctl daemon-reload
    # systemctl start

Then, we have:

    # LANG=C journalctl --no-pager -u A.service -u B.service -u -b
    -- Logs begin at Mon 2019-09-09 00:25:06 EDT, end at Thu 2019-10-24 22:28:47 EDT. --
    Oct 24 22:27:47 localhost.localdomain systemd[1]: Starting A...
    Oct 24 22:27:47 localhost.localdomain systemd[1]: A.service: Child 967 belongs to A.service.
    Oct 24 22:27:47 localhost.localdomain systemd[1]: A.service: Main process exited, code=exited, status=0/SUCCESS
    Oct 24 22:27:47 localhost.localdomain systemd[1]: A.service: Running next main command for state start.
    Oct 24 22:27:47 localhost.localdomain systemd[1]: A.service: Passing 0 fds to service
    Oct 24 22:27:47 localhost.localdomain systemd[1]: A.service: About to execute: /usr/bin/sleep 60
    Oct 24 22:27:47 localhost.localdomain systemd[1]: A.service: Forked /usr/bin/sleep as 968
    Oct 24 22:27:47 localhost.localdomain systemd[968]: A.service: Executing: /usr/bin/sleep 60
    Oct 24 22:27:52 localhost.localdomain systemd[1]: A.service: Trying to enqueue job A.service/reload/replace
    Oct 24 22:27:52 localhost.localdomain systemd[1]: A.service: Merged into running job, re-running: A.service/reload as 1288
    Oct 24 22:27:52 localhost.localdomain systemd[1]: A.service: Enqueued job A.service/reload as 1288
    Oct 24 22:27:52 localhost.localdomain systemd[1]: A.service: Unit cannot be reloaded because it is inactive.
    Oct 24 22:27:52 localhost.localdomain systemd[1]: A.service: Job 1288 A.service/reload finished, result=invalid
    Oct 24 22:27:52 localhost.localdomain systemd[1]: B.service: Passing 0 fds to service
    Oct 24 22:27:52 localhost.localdomain systemd[1]: B.service: About to execute: /usr/bin/echo B
    Oct 24 22:27:52 localhost.localdomain systemd[1]: B.service: Forked /usr/bin/echo as 970
    Oct 24 22:27:52 localhost.localdomain systemd[970]: B.service: Executing: /usr/bin/echo B
    Oct 24 22:27:52 localhost.localdomain systemd[1]: B.service: Failed to send unit change signal for B.service: Connection reset by peer
    Oct 24 22:27:52 localhost.localdomain systemd[1]: B.service: Changed dead -> start
    Oct 24 22:27:52 localhost.localdomain systemd[1]: Starting B...
    Oct 24 22:27:52 localhost.localdomain echo[970]: B
    Oct 24 22:27:52 localhost.localdomain systemd[1]: B.service: Child 970 belongs to B.service.
    Oct 24 22:27:52 localhost.localdomain systemd[1]: B.service: Main process exited, code=exited, status=0/SUCCESS
    Oct 24 22:27:52 localhost.localdomain systemd[1]: B.service: Changed start -> exited
    Oct 24 22:27:52 localhost.localdomain systemd[1]: B.service: Job 1371 B.service/start finished, result=done
    Oct 24 22:27:52 localhost.localdomain systemd[1]: Started B.
    Oct 24 22:27:52 localhost.localdomain systemd[1]: Job 1287 finished, result=done
    Oct 24 22:27:52 localhost.localdomain systemd[1]: Reached target C.
    Oct 24 22:27:52 localhost.localdomain systemd[1]: Failed to send unit change signal for Connection reset by peer
    Oct 24 22:28:47 localhost.localdomain systemd[1]: A.service: Child 968 belongs to A.service.
    Oct 24 22:28:47 localhost.localdomain systemd[1]: A.service: Main process exited, code=exited, status=0/SUCCESS
    Oct 24 22:28:47 localhost.localdomain systemd[1]: A.service: Running next main command for state start.
    Oct 24 22:28:47 localhost.localdomain systemd[1]: A.service: Passing 0 fds to service
    Oct 24 22:28:47 localhost.localdomain systemd[1]: A.service: About to execute: /usr/bin/echo A2
    Oct 24 22:28:47 localhost.localdomain systemd[1]: A.service: Forked /usr/bin/echo as 972
    Oct 24 22:28:47 localhost.localdomain systemd[972]: A.service: Executing: /usr/bin/echo A2
    Oct 24 22:28:47 localhost.localdomain echo[972]: A2
    Oct 24 22:28:47 localhost.localdomain systemd[1]: A.service: Child 972 belongs to A.service.
    Oct 24 22:28:47 localhost.localdomain systemd[1]: A.service: Main process exited, code=exited, status=0/SUCCESS
    Oct 24 22:28:47 localhost.localdomain systemd[1]: A.service: Changed start -> exited

The issue occurs not only in reload command, i.e.:

  - reload
  - try-restart
  - reload-or-restart
  - reload-or-try-restart commands

The cause of this issue is that job_type_collapse() doesn't take care of the
activating state.

Fixes: #10464

10 months agojournalctl: allow running vacuum on remote journals, too
Jan Kundrát [Sat, 2 Nov 2019 15:42:01 +0000 (16:42 +0100)] 
journalctl: allow running vacuum on remote journals, too

Right now the `systemd-journal-remote` service does not constrain its
resource usage (I just run out of space on my 100GB partition, for
example). This patch does not change that, but it at least makes it
possible to run something like:

  journalctl --directory /var/log/journal/remote --rotate --vacuum-size=90G

fixes #2376

Co-authored-by: Mike Auty <>
10 months agobootctl: create leading dirs when "bootctl random-seed" is called if needed 13935/head
Lennart Poettering [Mon, 4 Nov 2019 13:25:57 +0000 (14:25 +0100)] 
bootctl: create leading dirs when "bootctl random-seed" is called if needed

Prompted by: #13603

10 months agoMerge pull request #13906 from yuwata/drop-missing-take2
Zbigniew Jędrzejewski-Szmek [Mon, 4 Nov 2019 12:41:37 +0000 (13:41 +0100)] 
Merge pull request #13906 from yuwata/drop-missing-take2

tree-wide: drop xxx.h when xxx-util.h is included

10 months agosd-boot: Skip adding boot entries when the loader does not exist
Jan Janssen [Fri, 1 Nov 2019 11:31:07 +0000 (12:31 +0100)] 
sd-boot: Skip adding boot entries when the loader does not exist

10 months agoallow an empty DefaultInstance= in configuration files
Jérémy Rosen [Fri, 1 Nov 2019 23:03:54 +0000 (00:03 +0100)] 
allow an empty DefaultInstance= in configuration files

It is currently possible to override the DefaultInstance via drop-ins but
not remove it completely. Allow to do that by specifying an empty

10 months agoman/ fix typo
Jan Synacek [Mon, 4 Nov 2019 09:43:05 +0000 (10:43 +0100)] 
man/ fix typo

10 months agoutil: drop unnecessary headers from util.c 13906/head
Yu Watanabe [Sun, 3 Nov 2019 15:40:37 +0000 (00:40 +0900)] 
util: drop unnecessary headers from util.c

10 months agotree-wide: drop double newline
Yu Watanabe [Fri, 1 Nov 2019 09:33:24 +0000 (18:33 +0900)] 
tree-wide: drop double newline

10 months agotree-wide: drop input.h when missing_input.h is included
Yu Watanabe [Fri, 1 Nov 2019 09:28:32 +0000 (18:28 +0900)] 
tree-wide: drop input.h when missing_input.h is included

10 months agotree-wide: drop stdio.h when stdio-util.h is included
Yu Watanabe [Fri, 1 Nov 2019 09:27:33 +0000 (18:27 +0900)] 
tree-wide: drop stdio.h when stdio-util.h is included

10 months agotree-wide: drop signal.h when signal-util.h is included
Yu Watanabe [Fri, 1 Nov 2019 09:27:01 +0000 (18:27 +0900)] 
tree-wide: drop signal.h when signal-util.h is included

10 months agotree-wide: drop mman.h when missing_mman.h is included
Yu Watanabe [Fri, 1 Nov 2019 09:26:26 +0000 (18:26 +0900)] 
tree-wide: drop mman.h when missing_mman.h is included

10 months agotree-wide: drop magic.h when missing_magic.h is included
Yu Watanabe [Fri, 1 Nov 2019 09:24:45 +0000 (18:24 +0900)] 
tree-wide: drop magic.h when missing_magic.h is included

10 months agotree-wide: drop stat.h or statfs.h when stat-util.h is included
Yu Watanabe [Fri, 1 Nov 2019 09:23:26 +0000 (18:23 +0900)] 
tree-wide: drop stat.h or statfs.h when stat-util.h is included

10 months agotree-wide: drop socket.h when socket-util.h is included
Yu Watanabe [Fri, 1 Nov 2019 09:22:48 +0000 (18:22 +0900)] 
tree-wide: drop socket.h when socket-util.h is included

10 months agotree-wide: drop mntent.h when fstab-util.h is included
Yu Watanabe [Fri, 1 Nov 2019 09:21:33 +0000 (18:21 +0900)] 
tree-wide: drop mntent.h when fstab-util.h is included

10 months agotree-wide: drop libkmod.h when module-util.h is included
Yu Watanabe [Fri, 1 Nov 2019 09:20:59 +0000 (18:20 +0900)] 
tree-wide: drop libkmod.h when module-util.h is included

10 months agotree-wide: drop blkid.h when blkid-util.h is included
Yu Watanabe [Fri, 1 Nov 2019 09:20:31 +0000 (18:20 +0900)] 
tree-wide: drop blkid.h when blkid-util.h is included

10 months agotree-wide: drop acl.h when acl-util.h is included
Yu Watanabe [Fri, 1 Nov 2019 09:19:54 +0000 (18:19 +0900)] 
tree-wide: drop acl.h when acl-util.h is included

10 months agotree-wide: drop pwd.h and grp.h when user-util.h is included
Yu Watanabe [Fri, 1 Nov 2019 09:18:31 +0000 (18:18 +0900)] 
tree-wide: drop pwd.h and grp.h when user-util.h is included

10 months agotree-wide: drop time.h when time-util.h is included
Yu Watanabe [Fri, 1 Nov 2019 09:17:53 +0000 (18:17 +0900)] 
tree-wide: drop time.h when time-util.h is included

10 months agotree-wide: drop capability.h when capability-util.h is included
Yu Watanabe [Fri, 1 Nov 2019 09:17:16 +0000 (18:17 +0900)] 
tree-wide: drop capability.h when capability-util.h is included

10 months agotree-wide: drop sched.h when missing_sched.h is included
Yu Watanabe [Fri, 1 Nov 2019 09:16:18 +0000 (18:16 +0900)] 
tree-wide: drop sched.h when missing_sched.h is included

10 months agotree-wide: drop gcrypt.h when gcrypt-util.h is included
Yu Watanabe [Fri, 1 Nov 2019 08:57:13 +0000 (17:57 +0900)] 
tree-wide: drop gcrypt.h when gcrypt-util.h is included

10 months agotree-wide: drop locale.h when locale-util.h is included
Yu Watanabe [Fri, 1 Nov 2019 08:56:33 +0000 (17:56 +0900)] 
tree-wide: drop locale.h when locale-util.h is included

10 months agotree-wide: drop glob.h when glob-util.h is included
Yu Watanabe [Fri, 1 Nov 2019 08:56:05 +0000 (17:56 +0900)] 
tree-wide: drop glob.h when glob-util.h is included

10 months agotree-wide: drop dirent.h when dirent-util.h is included
Yu Watanabe [Fri, 1 Nov 2019 08:55:16 +0000 (17:55 +0900)] 
tree-wide: drop dirent.h when dirent-util.h is included

10 months agotree-wide: drop alloca.h when alloc-util.h is included
Yu Watanabe [Fri, 1 Nov 2019 08:46:59 +0000 (17:46 +0900)] 
tree-wide: drop alloca.h when alloc-util.h is included

10 months agotree-wide: drop string.h when string-util.h or friends are included
Yu Watanabe [Fri, 1 Nov 2019 08:44:54 +0000 (17:44 +0900)] 
tree-wide: drop string.h when string-util.h or friends are included

10 months agoMerge pull request #13899 from poettering/in-gid-tweak
Yu Watanabe [Sun, 3 Nov 2019 13:54:32 +0000 (22:54 +0900)] 
Merge pull request #13899 from poettering/in-gid-tweak

user-util: tweak to in_gid()

10 months agoMerge pull request #13909 from poettering/env-copy-pid
Yu Watanabe [Sun, 3 Nov 2019 13:46:42 +0000 (22:46 +0900)] 
Merge pull request #13909 from poettering/env-copy-pid

Fixes for the "saved_env" copy logic

10 months agonspawn: respect quiet on capabilities warning
Justin Trudell [Fri, 1 Nov 2019 19:00:16 +0000 (12:00 -0700)] 
nspawn: respect quiet on capabilities warning

10 months agofs-util: let's avoid unnecessary strerror()
Lennart Poettering [Fri, 1 Nov 2019 10:43:34 +0000 (11:43 +0100)] 
fs-util: let's avoid unnecessary strerror()

strerror() is not thread safe. Let's avoid it where it is easy hence.

(Ideally we'd not use it at all anymore, but that's sometimes a bit
nasty, not in this case though, where it is very easy to avoid)

Follow-up for: 27c3112dcbd1b5f171c36c32550d9c6331375b0b

10 months agoMerge pull request #13916 from ddstreet/test-network
Yu Watanabe [Sun, 3 Nov 2019 12:42:15 +0000 (21:42 +0900)] 
Merge pull request #13916 from ddstreet/test-network

test-network: minor changes to work better on Debian/Ubuntu autopkgtest

10 months agonspawn: mangle slice name
Lennart Poettering [Fri, 1 Nov 2019 10:21:05 +0000 (11:21 +0100)] 
nspawn: mangle slice name

It's user-facing, parsed from the command line and we typically mangle
in these cases, let's do so here too. (In particular as the identical
switch for systemd-run already does it.)

10 months agobasic/signal-util: drop unnecessary parens
Zbigniew Jędrzejewski-Szmek [Wed, 30 Oct 2019 13:36:05 +0000 (14:36 +0100)] 
basic/signal-util: drop unnecessary parens

10 months agoMerge pull request #13905 from poettering/cpuset-fixes
Lennart Poettering [Fri, 1 Nov 2019 22:44:36 +0000 (23:44 +0100)] 
Merge pull request #13905 from poettering/cpuset-fixes

fixes to the cpuset cgroup logic

10 months agotest-network: Remove/replace non-capturing group regex 13916/head
Dan Streetman [Thu, 31 Oct 2019 16:19:23 +0000 (12:19 -0400)] 
test-network: Remove/replace non-capturing group regex

The has some regex that uses non-capturing
groups, but there is no need to use that with assertRegex; the
groups aren't referenced so it doesn't matter if it's capturing or
non-capturing.  However, there are a few places where optional groups
should have been used instead, so this changes that.

Specifically, groups like this:
(?:whatever |)
should actually be:
(whatever )?

Additionally, this is specifically needed for these tests to run on
Debian systems, because this assertRegex:
'Link File: (?:/usr)/lib/systemd/network/'
needs to be:
'Link File: (/usr)?/lib/systemd/network/'

10 months agotest: check /usr/lib and /lib for systemd binaries
Dan Streetman [Thu, 31 Oct 2019 10:44:20 +0000 (06:44 -0400)] 
test: check /usr/lib and /lib for systemd binaries

Fedora uses /usr/lib while Debian uses /lib; find the right location

10 months agotest/test-network/ suppress stderr for functionality checks
Dan Streetman [Thu, 31 Oct 2019 10:07:33 +0000 (06:07 -0400)] 
test/test-network/ suppress stderr for functionality checks

10 months agomkosi: Find hostname command on Arch Linux
Kevin Kuehler [Thu, 31 Oct 2019 22:41:32 +0000 (15:41 -0700)] 
mkosi: Find hostname command on Arch Linux

exec-specifier.service: Executing: /usr/bin/sh -c 'test mkosi-7d5e81c7b81c42338d060a6b98edd44a = $(hostname)'
/usr/bin/sh: hostname: command not found
/usr/bin/sh: line 0: test: mkosi-7d5e81c7b81c42338d060a6b98edd44a: unary operator expected
Received SIGCHLD from PID 7389 (sh).
Child 7389 (sh) died (code=exited, status=2/INVALIDARGUMENT)

gettext provides the hostname binary, but puts it in
/usr/lib/gettext/hostname, which is not part of the default $PATH. Using
inetutils instead puts the binary in /usr/bin/hostname.

10 months agoupdate TODO
Lennart Poettering [Fri, 1 Nov 2019 12:07:51 +0000 (13:07 +0100)] 
update TODO

10 months agostatic-destruct: add missing closing '(' in comment 13909/head
Lennart Poettering [Fri, 1 Nov 2019 10:31:20 +0000 (11:31 +0100)] 
static-destruct: add missing closing '(' in comment

10 months agopid1: rework environment block copy logic
Lennart Poettering [Fri, 1 Nov 2019 10:26:05 +0000 (11:26 +0100)] 
pid1: rework environment block copy logic

This reworks the logic introduced in
a5cede8c24fddda9b73f142e09b18b49adde1b9c (#13693).

First of all, let's move this out of util.c, since only PID 1 really
needs this, and there's no real need to have it in util.c.

Then, fix freeing of the variable. It previously relied on
STATIC_DESTRUCTOR_REGISTER() which however relies on static_destruct()
to be called explicitly. Currently only the main-func.h macros do that,
and PID 1 does not. (It might be worth investigating whether to do that,
but it's not trivial.) Hence the freeing wasn't applied.

Finally, an OOM check was missing, add it in.

10 months agocgroup: add missing OOM check, and shorten code a bit 13905/head
Lennart Poettering [Fri, 1 Nov 2019 09:22:03 +0000 (10:22 +0100)] 
cgroup: add missing OOM check, and shorten code a bit

cpu_set_to_range_string() can fail due to OOM. Handle that.

unit_write_settingf() exists, use it instead of formatting a string

cpu_set_add_all() can fail due to OOM. Let's avoid it if we don't have
to use it, just copy over the cpuset directly.

10 months agocpuset: fix indentation and log about OOM we otherwise ignore
Lennart Poettering [Fri, 1 Nov 2019 09:21:53 +0000 (10:21 +0100)] 
cpuset: fix indentation and log about OOM we otherwise ignore

10 months agocgroup: add some basic OOM safety where it was missing
Lennart Poettering [Fri, 1 Nov 2019 09:21:35 +0000 (10:21 +0100)] 
cgroup: add some basic OOM safety where it was missing

10 months agosystemd-tmpfiles: deprecate F for f+
Zach Smith [Mon, 28 Oct 2019 16:32:16 +0000 (09:32 -0700)] 
systemd-tmpfiles: deprecate F for f+