]> git.ipfire.org Git - thirdparty/systemd.git/log
thirdparty/systemd.git
5 years agotest: print test of name in output 14338/head
Zbigniew Jędrzejewski-Szmek [Wed, 25 Mar 2020 12:15:37 +0000 (13:15 +0100)] 
test: print test of name in output

5 years agomeson: use find_program() for add-git-hook.sh
Zbigniew Jędrzejewski-Szmek [Sat, 28 Mar 2020 08:30:51 +0000 (09:30 +0100)] 
meson: use find_program() for add-git-hook.sh

On azure systemd.systemd ci, the build would fail with:
meson.build:53:0: ERROR: Program or command '/home/appuser/fuzzer/tools/add-git-hook.sh' not found or not executable

We use find_program() for all helpers, so let's do it for this one too.
This should solve the issue, whatever it exactly is.

5 years agotest: run unittests with newlines in /proc/cmdline
Zbigniew Jędrzejewski-Szmek [Fri, 13 Dec 2019 13:20:35 +0000 (14:20 +0100)] 
test: run unittests with newlines in /proc/cmdline

This should help catch this class of errors.

5 years agotest: stop caching loopdev
Zbigniew Jędrzejewski-Szmek [Fri, 20 Mar 2020 19:17:11 +0000 (20:17 +0100)] 
test: stop caching loopdev

It is more trouble than it is worth. The setup is of a loopback device
is very quick, so it's better to always create it when needed and
immediately drop afterwards.

5 years agotest: perform partial cleanup after each test is run
Zbigniew Jędrzejewski-Szmek [Fri, 20 Mar 2020 18:09:35 +0000 (19:09 +0100)] 
test: perform partial cleanup after each test is run

This causes the unprivileged-nspawn-root directory to be removed
after running one test. The advantage is that we reduce the maximum
disk-space use quite a bit (47*400 MB → about 18GB).

5 years agotest: simplify the Makefiles
Zbigniew Jędrzejewski-Szmek [Fri, 20 Mar 2020 17:31:10 +0000 (18:31 +0100)] 
test: simplify the Makefiles

has-overflow was a temporary hack that was removed in
844da987ef8b8c98f837d3328eeb3ed481f43835 (Oct. 2016). All the makefiles
can be the same, and all the targets can be handled identically.

5 years agotest: run tests directly from the loopback device
Zbigniew Jędrzejewski-Szmek [Fri, 13 Dec 2019 13:21:31 +0000 (14:21 +0100)] 
test: run tests directly from the loopback device

Before, we'd copy the test tree into nspawn-root, and run the tests from there.
This is OK, and doesn't actually take much extra time. But it uses quite a lot
of extra disk space. So let's make things a bit more efficient by running
directly from the image file.

We still run the unprivileged nspawn tests from a copy. Once the kernel
implements fs shift, we can do away with that too.

5 years agotest: wipe journal files after test
Zbigniew Jędrzejewski-Szmek [Fri, 20 Mar 2020 16:46:04 +0000 (17:46 +0100)] 
test: wipe journal files after test

Otherwise they accumulate and create noise.

5 years agotest: echo a message when the tests fails
Zbigniew Jędrzejewski-Szmek [Fri, 20 Mar 2020 16:16:30 +0000 (17:16 +0100)] 
test: echo a message when the tests fails

Sometimes it is not clear from the preceding output that there was a
failure.

5 years agotest/TEST-01: allow running w/o -Dinstall-tests=true
Zbigniew Jędrzejewski-Szmek [Tue, 17 Dec 2019 11:52:35 +0000 (12:52 +0100)] 
test/TEST-01: allow running w/o -Dinstall-tests=true

This is useful for sanitizer builds under CI, see
https://github.com/systemd/systemd-centos-ci/pull/202#issuecomment-566476592.

5 years agotest: rework how images are created
Zbigniew Jędrzejewski-Szmek [Thu, 12 Dec 2019 08:37:19 +0000 (09:37 +0100)] 
test: rework how images are created

Before, we'd create a separate image for each test, in
/var/tmp/systemd-test.XXXXX/rootdisk.img. Most of the images
where very similar, except that each one had some unit files installed
specifically for the test. The installation of those custom unit files
was removed in previous commits (all the unit files are always installed).

The new approach is to only create as few distinct images as possible.
We have:
default.img: the "normal" image suitable for almost all the tests
basic.img: the same as default image but doesn't mask any services
cryptsetup.img: p2 is used for encrypted /var
badid.img: /etc/machine-id is overwritten with stuff
selinux.img: with selinux added for fun and fun
and a few others:

ls -l build/test/*img
lrwxrwxrwx 1 root root 38 Mar 21 21:23 build/test/badid.img -> /var/tmp/systemd-test.PJFFeo/badid.img
lrwxrwxrwx 1 root root 38 Mar 21 21:17 build/test/basic.img -> /var/tmp/systemd-test.na0xOI/basic.img
lrwxrwxrwx 1 root root 43 Mar 21 21:18 build/test/cryptsetup.img -> /var/tmp/systemd-test.Tzjv06/cryptsetup.img
lrwxrwxrwx 1 root root 40 Mar 21 21:19 build/test/default.img -> /var/tmp/systemd-test.EscAsS/default.img
lrwxrwxrwx 1 root root 39 Mar 21 21:22 build/test/nspawn.img -> /var/tmp/systemd-test.HSebKo/nspawn.img
lrwxrwxrwx 1 root root 40 Mar 21 21:20 build/test/selinux.img -> /var/tmp/systemd-test.daBjbx/selinux.img
lrwxrwxrwx 1 root root 39 Mar 21 21:21 build/test/test08.img -> /var/tmp/systemd-test.OgnN8Z/test08.img

I considered trying to use the same image everywhere. It would probably be
possible, but it would be very brittle. By using separate images where it is
necessary we keep various orthogonal modifications independent.

The way that images are cached is complicated by the fact that we still
want to keep them in /var/tmp. Thus, an image is created on first use and
linked to from build/test/ so it can be found by other tests.

Tests cannot be run in parallel. I think that is an acceptable limitation.
Creation of the images was probably taking more resources then the actual
tests, so we should be better off anyway.

5 years agosemaphoreci: trivial cleanup
Zbigniew Jędrzejewski-Szmek [Tue, 17 Dec 2019 09:34:31 +0000 (10:34 +0100)] 
semaphoreci: trivial cleanup

5 years agotest/TEST-28: avoid race in handling of /testok
Zbigniew Jędrzejewski-Szmek [Tue, 17 Dec 2019 09:35:37 +0000 (10:35 +0100)] 
test/TEST-28: avoid race in handling of /testok

5 years agotest: make sure to reset state before starting integration tests
Zbigniew Jędrzejewski-Szmek [Fri, 13 Dec 2019 11:34:41 +0000 (12:34 +0100)] 
test: make sure to reset state before starting integration tests

5 years agotest: mount root rw and drop /etc/fstab
Zbigniew Jędrzejewski-Szmek [Thu, 12 Dec 2019 14:17:30 +0000 (15:17 +0100)] 
test: mount root rw and drop /etc/fstab

We had an fstab for the sole purpose of remounting "/" rw. Mounting root ro
is a pointless excercise in obsolete approaches. More importantly, the nspawn
image is now the same as the qemu one.

5 years agotest: drop cargo-cult line
Zbigniew Jędrzejewski-Szmek [Thu, 12 Dec 2019 07:35:20 +0000 (08:35 +0100)] 
test: drop cargo-cult line

p2 is only used in the cryptsetup test...

5 years agotest: improve reporting a bit
Zbigniew Jędrzejewski-Szmek [Tue, 10 Dec 2019 16:33:58 +0000 (17:33 +0100)] 
test: improve reporting a bit

5 years agotest: de-duplicate test_setup() functions
Zbigniew Jędrzejewski-Szmek [Tue, 10 Dec 2019 15:40:31 +0000 (16:40 +0100)] 
test: de-duplicate test_setup() functions

Now we will run setup_nspawn_root() in some case where do don't need to.

5 years agotest: move TEST-47-* to static files
Zbigniew Jędrzejewski-Szmek [Fri, 20 Mar 2020 18:57:54 +0000 (19:57 +0100)] 
test: move TEST-47-* to static files

5 years agotest: move TEST-46-HOMED to static files
Zbigniew Jędrzejewski-Szmek [Fri, 20 Mar 2020 18:52:41 +0000 (19:52 +0100)] 
test: move TEST-46-HOMED to static files

5 years agotest: convert TEST-45 to a normal meson test
Zbigniew Jędrzejewski-Szmek [Fri, 20 Mar 2020 19:22:42 +0000 (20:22 +0100)] 
test: convert TEST-45 to a normal meson test

On my laptop, it now takes 0.23 s. We don't need the overhead of creating
an image and spawning a machine to mangle a file in /tmp.

5 years agotest: move TEST-44-* to static files
Zbigniew Jędrzejewski-Szmek [Fri, 20 Mar 2020 18:15:09 +0000 (19:15 +0100)] 
test: move TEST-44-* to static files

5 years agotest: move TEST-43-* setup to static files
Zbigniew Jędrzejewski-Szmek [Fri, 20 Mar 2020 18:05:36 +0000 (19:05 +0100)] 
test: move TEST-43-* setup to static files

Setup of lingering is dropped. I don't think it's necessary for anything,
because the unit is still pulled in from the testsuite service.

5 years agotest: move TEST-42-EXECSTOPPOST setup to static files
Zbigniew Jędrzejewski-Szmek [Tue, 10 Dec 2019 10:49:39 +0000 (11:49 +0100)] 
test: move TEST-42-EXECSTOPPOST setup to static files

5 years agotest/TEST-41: use a file in /tmp and reset the contents
Zbigniew Jędrzejewski-Szmek [Fri, 20 Mar 2020 20:59:54 +0000 (21:59 +0100)] 
test/TEST-41: use a file in /tmp and reset the contents

If the test was run a second time from the same root, it would fail.

5 years agotest: move TEST-41-ONESHOT-RESTART setup to static files
Zbigniew Jędrzejewski-Szmek [Tue, 10 Dec 2019 09:09:53 +0000 (10:09 +0100)] 
test: move TEST-41-ONESHOT-RESTART setup to static files

5 years agotest: move TEST-40-EXEC-COMMAND-EX setup to static files
Zbigniew Jędrzejewski-Szmek [Tue, 10 Dec 2019 09:08:09 +0000 (10:08 +0100)] 
test: move TEST-40-EXEC-COMMAND-EX setup to static files

5 years agotest: move TEST-39-EXECRELOAD setup to static files
Zbigniew Jędrzejewski-Szmek [Tue, 10 Dec 2019 09:02:10 +0000 (10:02 +0100)] 
test: move TEST-39-EXECRELOAD setup to static files

5 years agotest: move TEST-37-RUNTIMEDIRECTORYPRESERVE setup to static files
Zbigniew Jędrzejewski-Szmek [Tue, 10 Dec 2019 08:58:50 +0000 (09:58 +0100)] 
test: move TEST-37-RUNTIMEDIRECTORYPRESERVE setup to static files

5 years agotest/TEST-36: move the config files to /run
Zbigniew Jędrzejewski-Szmek [Sat, 21 Mar 2020 10:17:30 +0000 (11:17 +0100)] 
test/TEST-36: move the config files to /run

This way we always start with a clean slate when running the test
repeatedly.

5 years agotest: move TEST-36-NUMAPOLICY setup to static files
Zbigniew Jędrzejewski-Szmek [Tue, 10 Dec 2019 08:54:36 +0000 (09:54 +0100)] 
test: move TEST-36-NUMAPOLICY setup to static files

No need to call setup_nspawn_root() if we plan to only run under qemu.

5 years agotest: convert TEST-35-NETWORK-GENERATOR into a unit test
Zbigniew Jędrzejewski-Szmek [Tue, 10 Dec 2019 12:04:39 +0000 (13:04 +0100)] 
test: convert TEST-35-NETWORK-GENERATOR into a unit test

5 years agotest: move TEST-34-DYNAMICUSERMIGRATE setup to static files
Zbigniew Jędrzejewski-Szmek [Tue, 10 Dec 2019 08:49:08 +0000 (09:49 +0100)] 
test: move TEST-34-DYNAMICUSERMIGRATE setup to static files

5 years agotest: move TEST-33-CLEAN-UNIT setup to static files
Zbigniew Jędrzejewski-Szmek [Tue, 10 Dec 2019 08:47:14 +0000 (09:47 +0100)] 
test: move TEST-33-CLEAN-UNIT setup to static files

5 years agotest: move TEST-32-OOMPOLICY setup to static files
Zbigniew Jędrzejewski-Szmek [Tue, 10 Dec 2019 08:43:14 +0000 (09:43 +0100)] 
test: move TEST-32-OOMPOLICY setup to static files

5 years agotest: move TEST-31-DEVICE-ENUMERATION setup to static files
Zbigniew Jędrzejewski-Szmek [Tue, 10 Dec 2019 08:41:07 +0000 (09:41 +0100)] 
test: move TEST-31-DEVICE-ENUMERATION setup to static files

5 years agotest: move TEST-30-ONCLOCKCHANGE setup to static files
Zbigniew Jędrzejewski-Szmek [Tue, 10 Dec 2019 12:45:48 +0000 (13:45 +0100)] 
test: move TEST-30-ONCLOCKCHANGE setup to static files

The two timezone files are now installed in the global setup. I am not too
happy about this, but it still seems better than to create a completely
separate image just for this.

5 years agotest: move TEST-29-UDEV-ID_RENAMING setup to static files
Zbigniew Jędrzejewski-Szmek [Tue, 10 Dec 2019 08:35:49 +0000 (09:35 +0100)] 
test: move TEST-29-UDEV-ID_RENAMING setup to static files

5 years agotest: move TEST-28-PERCENTJ-WANTEDBY setup to static files
Zbigniew Jędrzejewski-Szmek [Tue, 10 Dec 2019 08:32:44 +0000 (09:32 +0100)] 
test: move TEST-28-PERCENTJ-WANTEDBY setup to static files

5 years agotest: move TEST-27-STDOUTFILE setup to static files
Zbigniew Jędrzejewski-Szmek [Tue, 10 Dec 2019 08:19:13 +0000 (09:19 +0100)] 
test: move TEST-27-STDOUTFILE setup to static files

5 years agotest: move TEST-26-SETENV setup to static files
Zbigniew Jędrzejewski-Szmek [Tue, 10 Dec 2019 08:13:20 +0000 (09:13 +0100)] 
test: move TEST-26-SETENV setup to static files

Also run mask_supporting_services() here.

5 years agotest/TEST-25: remove all images at the end of the test
Zbigniew Jędrzejewski-Szmek [Sat, 21 Mar 2020 14:29:43 +0000 (15:29 +0100)] 
test/TEST-25: remove all images at the end of the test

scratch5 would be left behind, and trying to run the test again on the same
image would fail.

5 years agotest: move TEST-25-IMPORT setup to static files
Zbigniew Jędrzejewski-Szmek [Tue, 10 Dec 2019 08:10:28 +0000 (09:10 +0100)] 
test: move TEST-25-IMPORT setup to static files

Also run mask_supporting_services() here.

5 years agotest: move TEST-24-UNIT-TESTS setup to static files
Zbigniew Jędrzejewski-Szmek [Tue, 10 Dec 2019 14:31:09 +0000 (15:31 +0100)] 
test: move TEST-24-UNIT-TESTS setup to static files

I picked the list of zone files to install by grepping through the code. This
is is a bit brittle, but installing all of them takes a while, and more
importantly, writes a lot of lines to the log.

5 years agotest-fileio: fix bogus error when /proc/cmdline contains newlines
Zbigniew Jędrzejewski-Szmek [Fri, 13 Dec 2019 11:21:11 +0000 (12:21 +0100)] 
test-fileio: fix bogus error when /proc/cmdline contains newlines

The kernel does not sanitize /proc/cmdline. E.g. when running under qemu, it is
easy to pass a string with newline by mistake. We use read_one_line_file(), so
we would read only the first list of the file, and
write_string_file(WRITE_STRING_FILE_VERIFY_ON_FAILURE) would fail because the
target file is obviously different. Change to a kernel-generated file to avoid
the issue.

v2:
- use /proc/version instead of /proc/uptime for attempted writes, so the test
  test passes even if test_write_string_file_verify() takes more than 10 ms ;]

5 years agotest: move TEST-23-TYPE-EXEC setup to static files
Zbigniew Jędrzejewski-Szmek [Tue, 10 Dec 2019 07:59:09 +0000 (08:59 +0100)] 
test: move TEST-23-TYPE-EXEC setup to static files

5 years agotest: move TEST-22-TMPFILES setup to static files
Zbigniew Jędrzejewski-Szmek [Tue, 10 Dec 2019 12:32:20 +0000 (13:32 +0100)] 
test: move TEST-22-TMPFILES setup to static files

5 years agotest: move TEST-20-MAINPIDGAMES setup to static files
Zbigniew Jędrzejewski-Szmek [Tue, 10 Dec 2019 07:32:28 +0000 (08:32 +0100)] 
test: move TEST-20-MAINPIDGAMES setup to static files

5 years agotest: move TEST-19-DELEGATE setup to static files
Zbigniew Jędrzejewski-Szmek [Tue, 10 Dec 2019 07:29:43 +0000 (08:29 +0100)] 
test: move TEST-19-DELEGATE setup to static files

5 years agotest: move TEST-18-FAILUREACTION setup to static files
Zbigniew Jędrzejewski-Szmek [Tue, 10 Dec 2019 07:27:33 +0000 (08:27 +0100)] 
test: move TEST-18-FAILUREACTION setup to static files

5 years agotest: move TEST-17-WANTS setup to static files
Zbigniew Jędrzejewski-Szmek [Tue, 10 Dec 2019 07:21:02 +0000 (08:21 +0100)] 
test: move TEST-17-WANTS setup to static files

5 years agotest/TEST-16: simplify how variables are set
Zbigniew Jędrzejewski-Szmek [Thu, 12 Dec 2019 20:06:58 +0000 (21:06 +0100)] 
test/TEST-16: simplify how variables are set

The amount of lines doesn't change much, but I think it's much more
straightforward this way.

5 years agotest: move TEST-16-EXTEND-TIMEOUT setup to static files
Zbigniew Jędrzejewski-Szmek [Tue, 10 Dec 2019 07:16:40 +0000 (08:16 +0100)] 
test: move TEST-16-EXTEND-TIMEOUT setup to static files

5 years agotest/TEST-15: call daemon-reload in two more places
Zbigniew Jędrzejewski-Szmek [Sat, 21 Mar 2020 16:01:09 +0000 (17:01 +0100)] 
test/TEST-15: call daemon-reload in two more places

Sometimes the test would fail there, nondeterministically. I'm not sure why,
but relying on PID1 not caching the file is clearly very brittle. Let's instead
call daemon-reload.

5 years agotest/TEST-15: remove all created unit files
Zbigniew Jędrzejewski-Szmek [Sat, 21 Mar 2020 15:30:27 +0000 (16:30 +0100)] 
test/TEST-15: remove all created unit files

We would miss anything created under a template instance.

5 years agotest: move TEST-15-DROPIN setup to static files
Zbigniew Jędrzejewski-Szmek [Mon, 9 Dec 2019 22:29:18 +0000 (23:29 +0100)] 
test: move TEST-15-DROPIN setup to static files

Unfortunately the services needed to be renamed because of conflicts with other
services that exist in the shared dir.

5 years agotest: move most of TEST-14-MACHINE-ID setup to static files
Zbigniew Jędrzejewski-Szmek [Mon, 9 Dec 2019 21:02:06 +0000 (22:02 +0100)] 
test: move most of TEST-14-MACHINE-ID setup to static files

5 years agotest: move most of TEST-13-NSPAWN-SMOKE setup to static files
Zbigniew Jędrzejewski-Szmek [Mon, 9 Dec 2019 20:45:46 +0000 (21:45 +0100)] 
test: move most of TEST-13-NSPAWN-SMOKE setup to static files

5 years agotest: move TEST-12-ISSUE-3171 setup to static files
Zbigniew Jędrzejewski-Szmek [Mon, 9 Dec 2019 20:26:59 +0000 (21:26 +0100)] 
test: move TEST-12-ISSUE-3171 setup to static files

5 years agotest: move TEST-11-ISSUE-3166 setup to static files
Zbigniew Jędrzejewski-Szmek [Mon, 9 Dec 2019 19:53:38 +0000 (20:53 +0100)] 
test: move TEST-11-ISSUE-3166 setup to static files

5 years agotest: move TEST-10-ISSUE-2467 setup to static files
Zbigniew Jędrzejewski-Szmek [Mon, 9 Dec 2019 19:42:48 +0000 (20:42 +0100)] 
test: move TEST-10-ISSUE-2467 setup to static files

5 years agotest: use shell arrays and remove duplicates
Zbigniew Jędrzejewski-Szmek [Mon, 9 Dec 2019 19:35:51 +0000 (20:35 +0100)] 
test: use shell arrays and remove duplicates

The tool lists are obviously not empty, so let's remove the test if they are.

5 years agotest: move TEST-09-ISSUE-2691 setup to static files
Zbigniew Jędrzejewski-Szmek [Mon, 9 Dec 2019 19:27:40 +0000 (20:27 +0100)] 
test: move TEST-09-ISSUE-2691 setup to static files

5 years agotest: move TEST-08-ISSUE-2730 setup to static files and meson scripts
Zbigniew Jędrzejewski-Szmek [Tue, 10 Dec 2019 15:05:48 +0000 (16:05 +0100)] 
test: move TEST-08-ISSUE-2730 setup to static files and meson scripts

Unfortunately meson does not install symlinks, but copies the symlink
destination instead. So symlinks need to be created by a script.
This commit adds both symlinks in test/testsuite-08.units/ and meson
scriptlet calls. Strictly speaking, the first is not necessary, since nothing
reads stuff directly from the source tree.

5 years agotest: move TEST-07-ISSUE-1981 setup to static files
Zbigniew Jędrzejewski-Szmek [Mon, 9 Dec 2019 18:51:17 +0000 (19:51 +0100)] 
test: move TEST-07-ISSUE-1981 setup to static files

5 years agotest/TEST-06: make autorelabel.service a static file too
Zbigniew Jędrzejewski-Szmek [Tue, 10 Dec 2019 19:46:48 +0000 (20:46 +0100)] 
test/TEST-06: make autorelabel.service a static file too

5 years agotest/TEST-06: skip test if no selinux
Zbigniew Jędrzejewski-Szmek [Thu, 19 Mar 2020 15:57:50 +0000 (16:57 +0100)] 
test/TEST-06: skip test if no selinux

I put SELINUX=disabled on my laptop, and the test fails with ENOENT when trying
to write to /sys/fs/selinux/enforce. It's a bit of a special case, but let's
avoid the failure.

5 years agotest: move most of TEST-06-* setup to static files
Zbigniew Jędrzejewski-Szmek [Mon, 9 Dec 2019 18:42:57 +0000 (19:42 +0100)] 
test: move most of TEST-06-* setup to static files

5 years agotest: move TEST-05-RLIMITS setup to static files
Zbigniew Jędrzejewski-Szmek [Tue, 10 Dec 2019 13:09:59 +0000 (14:09 +0100)] 
test: move TEST-05-RLIMITS setup to static files

The test currently fails in the check for LimitNOFILESoft/LimitNOFILE. I see
default values there. This doesn't seem to be related to the changes in the
test suite, but rather to the recent changes to pid1.

5 years agotest: convert TEST-04-JOURNAL to generic image
Zbigniew Jędrzejewski-Szmek [Mon, 9 Dec 2019 18:19:05 +0000 (19:19 +0100)] 
test: convert TEST-04-JOURNAL to generic image

5 years agotest: move TEST-03-JOBS setup to static files
Zbigniew Jędrzejewski-Szmek [Mon, 9 Dec 2019 18:09:32 +0000 (19:09 +0100)] 
test: move TEST-03-JOBS setup to static files

5 years agotest: move part of TEST-02-CRYPTSETUP setup to static files
Zbigniew Jędrzejewski-Szmek [Thu, 12 Dec 2019 08:59:53 +0000 (09:59 +0100)] 
test: move part of TEST-02-CRYPTSETUP setup to static files

Since we create an encrypted partition for this test, let's create a separate image here.

5 years agotest: move TEST-01-BASIC setup to static files
Zbigniew Jędrzejewski-Szmek [Mon, 9 Dec 2019 17:56:13 +0000 (18:56 +0100)] 
test: move TEST-01-BASIC setup to static files

After this commit, only tests that were converted will work, because the
interface to test_run() changed.

5 years agotest: allow overriding EFI_MOUNT like other variables
Zbigniew Jędrzejewski-Szmek [Mon, 9 Dec 2019 17:52:35 +0000 (18:52 +0100)] 
test: allow overriding EFI_MOUNT like other variables

5 years agomeson: report -Dinstall-tests in summary
Zbigniew Jędrzejewski-Szmek [Mon, 16 Dec 2019 16:01:17 +0000 (17:01 +0100)] 
meson: report -Dinstall-tests in summary

5 years agomeson: use install_subdir() to install files
Zbigniew Jędrzejewski-Szmek [Sun, 8 Dec 2019 10:24:39 +0000 (11:24 +0100)] 
meson: use install_subdir() to install files

This gives us slightly less control, but we don't need to update the
file list...

5 years agotest: replace symlinks with actual unit files
Zbigniew Jędrzejewski-Szmek [Fri, 20 Mar 2020 15:12:12 +0000 (16:12 +0100)] 
test: replace symlinks with actual unit files

During installation, meson complains:
> Installing /home/zbyszek/src/systemd-work/test/units/sysinit.target to /var/tmp/systemd-test.Q1FSuj/root/usr/lib/systemd/tests/testdata/units
> Warning: trying to copy a symlink that points to a file. This will copy the file,
> but this will be changed in a future version of Meson to copy the symlink as is. Please update your
> build definitions so that it will not break when the change happens.

It *is* convenient to have those files as symlinks, but it is also confusing,
because symlinks create aliases, and it seems that in those cases we actually
don't want aliases (at least in the case of loopy*.service that'd make the test
pointless).

5 years agotests: move unit files to units/ subdirectory
Zbigniew Jędrzejewski-Szmek [Fri, 6 Dec 2019 20:51:59 +0000 (21:51 +0100)] 
tests: move unit files to units/ subdirectory

We have a bazillion of those unit files, and keeping them all directly in tests/
has become rather unwieldy.

5 years agosd-bus: sd_bus_set/get_method_call_timeout docs + cleanups.
Daan De Meyer [Wed, 18 Mar 2020 18:29:57 +0000 (19:29 +0100)] 
sd-bus: sd_bus_set/get_method_call_timeout docs + cleanups.

5 years agonetworkctl: Add support to display bond
Susant Sahani [Wed, 18 Mar 2020 07:50:15 +0000 (08:50 +0100)] 
networkctl: Add support to display bond

5 years agoMerge pull request #14749 from msekletar/cpu-aff-numa-v3
Zbigniew Jędrzejewski-Szmek [Wed, 18 Mar 2020 12:35:58 +0000 (13:35 +0100)] 
Merge pull request #14749 from msekletar/cpu-aff-numa-v3

core: add support for setting CPUAffinity= to special "numa" value

5 years agosd-bus: sd_bus_call/sd_bus_call_async_docs + cleanups.
Daan De Meyer [Tue, 17 Mar 2020 20:52:30 +0000 (21:52 +0100)] 
sd-bus: sd_bus_call/sd_bus_call_async_docs + cleanups.

5 years agoMerge pull request #15143 from ssahani/networkctl-tunnel-geneve
Yu Watanabe [Wed, 18 Mar 2020 09:33:38 +0000 (18:33 +0900)] 
Merge pull request #15143 from ssahani/networkctl-tunnel-geneve

networkctl: Add support to display geneve tunnel info

5 years agonetworkctl: Add support to display vti and vti6 tunnels 15143/head
Susant Sahani [Wed, 18 Mar 2020 04:21:41 +0000 (05:21 +0100)] 
networkctl: Add support to display vti and vti6 tunnels

5 years agonetworkctl: add support to display ip6gre, ip6gretap, ip6erspan
Susant Sahani [Wed, 18 Mar 2020 03:27:09 +0000 (04:27 +0100)] 
networkctl: add support to display ip6gre, ip6gretap, ip6erspan

5 years agodocs: Add syntax for templated units to systemd.preset man page
Joerg Behrmann [Tue, 10 Mar 2020 15:34:13 +0000 (16:34 +0100)] 
docs: Add syntax for templated units to systemd.preset man page

This documents the syntax

     enable template@.service foo bar baz

that was introduced in #9901 to preset templated units.

5 years agoMerge pull request #15136 from yuwata/network-dhcp4-use-gateway
Zbigniew Jędrzejewski-Szmek [Tue, 17 Mar 2020 15:51:18 +0000 (16:51 +0100)] 
Merge pull request #15136 from yuwata/network-dhcp4-use-gateway

network: add a flag to ignore gateway provided by DHCP server

5 years agoMerge pull request #15108 from keszybz/systemctl-set-default-notices
Zbigniew Jędrzejewski-Szmek [Tue, 17 Mar 2020 15:39:45 +0000 (16:39 +0100)] 
Merge pull request #15108 from keszybz/systemctl-set-default-notices

Make systemctl set-default hint about some common issues

5 years agonetworkctl: Add support to display gre gretap and erspan
Susant Sahani [Tue, 17 Mar 2020 13:31:54 +0000 (14:31 +0100)] 
networkctl: Add support to display gre gretap and erspan

5 years agonetworkctl: Add support to display geneve tunnel info
Susant Sahani [Tue, 17 Mar 2020 09:56:57 +0000 (10:56 +0100)] 
networkctl: Add support to display geneve tunnel info

5 years agohwdb: add support for Lenovo Legion Y9000X2020 (#15127)
duguxy [Tue, 17 Mar 2020 11:24:00 +0000 (19:24 +0800)] 
hwdb: add support for Lenovo Legion Y9000X2020 (#15127)

5 years agopo: update Polish translation
Piotr Drąg [Sun, 15 Mar 2020 12:14:20 +0000 (13:14 +0100)] 
po: update Polish translation

5 years agoFix volume control keys on Lenovo Yoga S940
Alexander Malafeev [Sun, 15 Mar 2020 19:24:56 +0000 (20:24 +0100)] 
Fix volume control keys on Lenovo Yoga S940

With the default config repeated press of fn+VolumeUp or fn+VolumeDown lead to
repeated presses even after release, which leads to volume going to 100% or 0%.

5 years agosystemctl: print a notice when set-default is not effective 15108/head
Zbigniew Jędrzejewski-Szmek [Fri, 13 Mar 2020 16:57:06 +0000 (17:57 +0100)] 
systemctl: print a notice when set-default is not effective

$ sudo ln -svf multi-user.target /run/systemd/generator.early/default.target
'/run/systemd/generator.early/default.target' -> 'multi-user.target'
$ sudo build/systemctl set-default --root=/ sysinit.target
Removed /etc/systemd/system/default.target.
Created symlink /etc/systemd/system/default.target → /usr/lib/systemd/system/sysinit.target.
Note: "multi-user.target" is the default unit (possibly a runtime override).

The output is not super informative, but it should be enough to point the user in
the right direction.

Fixes #3645.

5 years agonetworkctl: Add support to show tunnel (ipip/sit) local and remote address
Susant Sahani [Tue, 17 Mar 2020 05:08:05 +0000 (06:08 +0100)] 
networkctl: Add support to show tunnel (ipip/sit) local and remote address

5 years agosd-bus: Add sd_bus_method_call/sd_bus_method_call_async docs + cleanups.
Daan De Meyer [Mon, 16 Mar 2020 19:41:57 +0000 (20:41 +0100)] 
sd-bus: Add sd_bus_method_call/sd_bus_method_call_async docs + cleanups.

5 years agoMerge pull request #15138 from ssahani/networkctl-vlan
Anita Zhang [Mon, 16 Mar 2020 19:56:07 +0000 (12:56 -0700)] 
Merge pull request #15138 from ssahani/networkctl-vlan

networkctl: Add support to display vlan ID

5 years agoMerge pull request #15139 from DaanDeMeyer/sd-bus-remove-unused-cite
Anita Zhang [Mon, 16 Mar 2020 19:50:26 +0000 (12:50 -0700)] 
Merge pull request #15139 from DaanDeMeyer/sd-bus-remove-unused-cite

sd-bus: Remove unused man cite of `sd_bus_slot_new_signal`.

5 years agosd-bus: Remove unused man cite of `sd_bus_slot_new_signal`. 15139/head
Daan De Meyer [Mon, 16 Mar 2020 18:52:40 +0000 (19:52 +0100)] 
sd-bus: Remove unused man cite of `sd_bus_slot_new_signal`.

5 years agonetwork: do not re-attach sd-event object
Yu Watanabe [Mon, 16 Mar 2020 07:17:04 +0000 (16:17 +0900)] 
network: do not re-attach sd-event object

Fixes #15126.