David Tardon [Thu, 25 Jul 2024 08:06:34 +0000 (10:06 +0200)]
sd-event: do not assert on invalid signal
The signalfd_siginfo struct is received from outside via a FD, hence
assert() is not appropriate way to check it. Just do a normal runtime
check.
(cherry picked from commit
7a64c5f23efbb51fe4f1229c1a8aed6dd858a0a9)
David Tardon [Thu, 25 Jul 2024 07:47:56 +0000 (09:47 +0200)]
logind-dbus: check auth. for all inhibitor operations
Fixes #33834
(cherry picked from commit
639719e01065c3a2f557d70e4d8088c2ec71c7c6)
Jeffrey Bosboom [Thu, 25 Jul 2024 01:14:24 +0000 (18:14 -0700)]
docs/CONTROL_GROUP_INTERFACE.md: document accounting information available via D-Bus
(cherry picked from commit
f7fa6326820bbb75a2a7112072a7771944ad8ce3)
Mike Yuan [Wed, 24 Jul 2024 14:28:48 +0000 (16:28 +0200)]
basic/log: do not treat all negative errnos as synthetic
Currently, IS_SYNTHETIC_ERRNO() evaluates to true for all negative errnos,
because of the two's-complement negative value representation.
Subsequently, ERRNO= is not logged for most of our own code.
Let's fix this, by formatting all synthetic errnos as positive.
Then, treat all negative values as non-synthetic.
While at it, mark the evaluation order explicitly, and remove
unneeded comment.
Fixes #33800
(cherry picked from commit
268f58076f7e0258dce75f521d08199092279853)
Luca Boccassi [Thu, 25 Jul 2024 09:43:48 +0000 (10:43 +0100)]
test: fix D-Bus policy override for TEST-73-LOCALE
We don't need to allow non-root, and the policy needs to specify destination
and interface too, to narrow it down
Follow-up for
7b5c38a91def6cf236605010a0a93a1cd4c137e9
(cherry picked from commit
a4c436c9d8ae20baf9d1d08d43e64e6355bcce83)
Daan De Meyer [Wed, 24 Jul 2024 08:50:38 +0000 (10:50 +0200)]
test-fs-util: Modernize openat_report_new() test
(cherry picked from commit
b91ad562280ec4f37ba970f1f167d2d626da2692)
Daan De Meyer [Wed, 24 Jul 2024 07:56:25 +0000 (09:56 +0200)]
test-id128: Use new assertion macros
(cherry picked from commit
372e96a5a10630eb2e02111672221428e0be3fb8)
Daan De Meyer [Wed, 24 Jul 2024 07:56:01 +0000 (09:56 +0200)]
test: Add ASSERT_EQ_ID128() and ASSERT_NE_ID128()
(cherry picked from commit
60a4505da4cfb57653fac79a79745e32ab0bc311)
Daan De Meyer [Wed, 24 Jul 2024 07:55:21 +0000 (09:55 +0200)]
test: Add ASSERT_FAIL()
(cherry picked from commit
8eb06fd162c8828739f8619437351d98d7105ec9)
Daan De Meyer [Wed, 24 Jul 2024 11:27:46 +0000 (13:27 +0200)]
mkosi: Update to latest
Includes the required fix to make mkosi copy NOCOW disk images properly.
(cherry picked from commit
3d6c6f9b9e957735a801b8d1587508dea9599795)
Daniel P. Berrangé [Fri, 2 Aug 2024 12:07:13 +0000 (13:07 +0100)]
man/systemd-detect-virt: fix row spanning for VM header
This fixes
commit
9b0688f491674b53ef7a52bdf561a430c53673d6
Author: Yu Watanabe <watanabe.yu+github@gmail.com>
Date: Tue Jan 9 10:52:49 2024 +0900
virt: add Google Compute Engine support
Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>
(cherry picked from commit
9ffdfc67c6aedcb66c2b18c2c61bc32e585e6d6e)
Luca Boccassi [Wed, 24 Jul 2024 12:49:39 +0000 (13:49 +0100)]
meson: Bump version to 256.4
Luca Boccassi [Wed, 24 Jul 2024 09:13:44 +0000 (10:13 +0100)]
mkosi: update debian commit reference
*
c004a150e7 Configure default DNS servers for upstream CI builds
*
a27b1516c9 autopkgtest: set Release= in mkosi.local.conf to distinguish testing vs unstable
*
dddce818b2 autopkgtest: add allow-stderr to timedated test
*
2a7b111506 Install valrinkctl zsh completion file
*
862497a139 d/t/control: add Depends: lib{systemd,udev}-dev for upstream
*
cb78791b85 d/t/upstream: ensure correct ubuntu codename is used
*
fe7445b94d d/t/boot-and-services: fix a couple python sytax warnings
*
2a875917cc d/t/boot-and-services: skip test_tmp_cleanup if tmp.mount is overridden
*
6fe81acbc9 Install zsh completion for run0
Luca Boccassi [Wed, 24 Jul 2024 09:06:42 +0000 (10:06 +0100)]
hwdb: update to main@{2024-07-24}
git restore -s origin/main hwdb.d/ test/hwdb.d
vdovhanych [Mon, 22 Jul 2024 20:57:25 +0000 (22:57 +0200)]
add udev rules for trezor hw wallet devices
(cherry picked from commit
2cf425ec573b8f67025c5e74cd267015129e7349)
Mike Yuan [Tue, 23 Jul 2024 15:55:12 +0000 (17:55 +0200)]
core: reliably check if varlink socket has been deserialized
Follow-up for
6906c028e83b77b35eaaf87b27d0fe5c6e1984b7
The mentioned commit uses access() to check if varlink socket
already exists in the filesystem, but that isn't sufficient.
> Varlink sockets are not serialized until v252, so upgrading from
> v251 or older means we will not listen anymore on the varlink sockets.
>
> See https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=
1074789
> for more details as this was found when updating from Debian Bullseye to a new version.
After this commit, the set up of varlink_server is effectively
split into two steps. manager_varlink_init_system(), which is
called after deserialization, would no longer skip listening
even if Manager.varlink_server is in place, but actually
check if we're listening on desired sockets.
Then, manager_deserialize() can be switched back to using
manager_setup_varlink_server().
Alternative to #33817
Co-authored-by: Luca Boccassi <bluca@debian.org>
(cherry picked from commit
d4e5c66ed469c822ca5346c7a445ec1446b1d17f)
Daan De Meyer [Tue, 23 Jul 2024 12:16:23 +0000 (14:16 +0200)]
mkosi: Bump device timeout even more
I still manage to hit it in some cases so let's bump again.
(cherry picked from commit
fffbfb4ed1b949a853773e0445197c2aba4e5f91)
Zbigniew Jędrzejewski-Szmek [Sun, 14 Jul 2024 12:25:59 +0000 (14:25 +0200)]
man/systemd-repart: extend description and reword some sentences
The page was written when systemd-repart was primarily intended to be used on a
running system. But nowadays it's more often used to create images, so extend
that part of the description.
While at it, fix some whitespace issues and trim some overly complicated sentences.
(cherry picked from commit
d202ea57549248c4246c8f453a2ff88a4c2a7e1e)
Steve Traylen [Wed, 3 Jul 2024 12:19:08 +0000 (14:19 +0200)]
Document that MemorySwapMax supports % configuration
Certainly on systemd 252 at least a configuration of
```
MemorySwapMax=40%
```
is supported but this was missing from the man page.
Only MemoryMax was documented as supporting a %.
(cherry picked from commit
8af38e5b0475f514141d314088dcf9fffd7edc37)
Daan De Meyer [Fri, 19 Jul 2024 11:05:11 +0000 (13:05 +0200)]
Make vcs-tag do something useful for non-developer mode as well
When building packages of arbitrary commits of systemd-stable,
distributors might want to include a git sha of the exact commit
they're on. Let's extend vcs-tag a little to make this possible.
If we're on a commit matching a tag, don't generate a git sha at all.
If we're not on a commit matching a tag, generate a vcs tag as usually.
However, if we're not in developer mode, don't append a '^' if the tree
is dirty to accomodate package builds applying various patches to the
tree which shouldn't be considered as "dirty" edits.
(cherry picked from commit
944faf65986f36d7ed3a4ba5cb4cc763011f0f77)
Zbigniew Jędrzejewski-Szmek [Wed, 3 Jul 2024 14:11:43 +0000 (16:11 +0200)]
tools/fetch-distro: switch to the target branch
We switch opensuse from "factory" to "devel". I had an old checkout that was
using the stale branch.
(cherry picked from commit
1c85d56349da7943967377e45da595e32887e0a2)
Zbigniew Jędrzejewski-Szmek [Wed, 3 Jul 2024 14:10:49 +0000 (16:10 +0200)]
tools/fetch-distro: only fetch the configured branch
We don't need the other branches. This mostly cuts down on the
noise in output. But add '-v' to show what we're fetching.
(cherry picked from commit
8e84e154457b0d55e823f670c377233b0211d2b7)
Zbigniew Jędrzejewski-Szmek [Wed, 3 Jul 2024 14:01:03 +0000 (16:01 +0200)]
tools/update-distro-hash: rename, fetch the repository if appropriate
Let's rename the tool to tools/fetch-distro. It's useful to be able to fetch
the distro directly. But when that functionality is added, the old name is
confusing.
Now --update/-u must be specified to update the commits.
--reference-if-able is used to speed up the clone of debian.
It saves about 75% of the download.
(cherry picked from commit
82c459f910c3036b22c9f41472b622ef4bfdab1d)
Daan De Meyer [Mon, 22 Jul 2024 15:10:11 +0000 (17:10 +0200)]
Merge pull request #33803 from bluca/v256-stable
v256 batch
Daan De Meyer [Mon, 22 Jul 2024 08:28:57 +0000 (10:28 +0200)]
mkosi: update arch commit reference
*
1d577a6268 refresh the keys
*
12383ba712 make libarchive a regular optional dependency...
*
f51d5e04bf ignore rc releases in nvchecker
*
0b096a2baa add nvchecker
*
5f0ced863c upgpkg: 256.2-1: new upstream release
(cherry picked from commit
4ecd6c3552aff72a40924560d7dfe2557b6e7c8e)
Daan De Meyer [Mon, 22 Jul 2024 08:28:50 +0000 (10:28 +0200)]
mkosi: update fedora commit reference
*
00babccdea Simplify BFQ scheduler enablement
*
ef8ddb130b Rebuilt for https://fedoraproject.org/wiki/Fedora_41_Mass_Rebuild
*
5b4a5461d6 Fix changelog
*
a8c5c736f6 Only apply shorter shutdown timer changes on Fedora
*
f4e284cd7a Merge #150 `Deal with systemd-timesyncd backport in EPEL`
|\
| *
9378a0733a Deal with systemd-timesyncd backport in EPEL
* |
12d1f05029 Don't claim /sbin/installkernel if building for CentOS Stream 9
|/
*
79828f2753 spec: use "positive" conditions in conditionals
*
c5d3af1638 Add build dependency on rsync on CentOS Stream 9
*
8d080fb5cb Backport udma buffer access patch
*
6084453807 Add support for building from a specific branch
*
cb9d631ca0 Update PR patch metadata
*
3889da947e In standalone subpackages, suggest coreutils-single
*
b7800e3e66 Drop versions from Conflicts for standalone packages
(cherry picked from commit
1d8f16df903a5567e2359bac3b8ca0caf937d333)
Daan De Meyer [Mon, 22 Jul 2024 08:25:46 +0000 (10:25 +0200)]
mkosi: Use the Fedora Rawhide spec for CentOS
These are now practically identical, with the only differences between
the two having no effect on the rpm builds we do with mkosi, so let's
cut out the middle man and just use the Fedora Rawhide spec for CentOS
as well.
(cherry picked from commit
ebbae3d6328725767e23afef8692596e03f9d3d1)
Daan De Meyer [Mon, 22 Jul 2024 08:22:48 +0000 (10:22 +0200)]
mkosi: Switch back to PKG_SUBDIR instead of symlinks
Previously I thought it would make sense to allow running the build
scripts from within the VM/container to rebuild the packages. Instead
we ended up making it possible to rerun mkosi outside of the container/VM
to rebuild the packages, so let's switch back to $PKG_SUBDIR to tell the
build scripts where to look for the packaging sources.
(cherry picked from commit
762b23e86a8bd45879335af96006573a99884a71)
Collin L [Mon, 22 Jul 2024 09:36:47 +0000 (17:36 +0800)]
zsh/_networkctl: remove duplicated argument for completion (#31926)
It is unnecessary, which will mess the completion.
(cherry picked from commit
733518b41350ce781c7e41a4c866eafb9e549e1f)
Lennart Poettering [Thu, 11 Jul 2024 12:32:09 +0000 (14:32 +0200)]
import-creds: when we hit ENOENT on SMBIOS 11 do not even debug log
We'll *always* hit ENEOENT when iterating through SMBIOS type #11
fields, on the last one. it's very confusing to debug log about that,
let's just not do it.
(cherry picked from commit
5202ee42d5da0ae3a6655d2bc959a19d8c347e9d)
Mike Yuan [Mon, 8 Jul 2024 15:12:20 +0000 (17:12 +0200)]
core/unit: ignore dropins for masked units completely when checking need_reload
Follow-up for
19a44dfe4525ab01caf593a9c2beada4b412910d
If a drop-in is set from upper level, e.g. global unit_type.d/,
even if a unit is masked, its dropin_paths would still be partially
populated. However, unit_need_daemon_reload() would always
compare u->dropin_paths with empty strv in case of masked units,
resulting in it always returning true. Instead, let's ignore
dropins entirely here.
Fixes #33672
(cherry picked from commit
11b3775f514f521f353741ff6ac4d66cf0e928e8)
David Tardon [Fri, 12 Jul 2024 12:38:08 +0000 (14:38 +0200)]
test: add a reproducer for #33672
(cherry picked from commit
8b6de9e6381b39f59c936d2b0c6ce47f1b70a19e)
Mauri de Souza Meneguzzo [Tue, 25 Jun 2024 02:47:15 +0000 (23:47 -0300)]
shared: log error when execve fail
If there is an error with the execv call in fork_agent the
program exits without any meaningful log message. Log the
command and errno so the user gets more information about
the failure.
Fixes: #33418
Signed-off-by: Mauri de Souza Meneguzzo <mauri870@gmail.com>
(cherry picked from commit
a408d4453145621902b9a3ef78a552f83b09bd8d)
Mike Yuan [Tue, 30 Apr 2024 13:55:05 +0000 (21:55 +0800)]
test-execute: ExecStop= and friends should not get credentials
See #32583
(cherry picked from commit
60b218e36e96236e6ba386686f130fe34d3e1c2a)
Luca Boccassi [Sat, 20 Jul 2024 10:34:51 +0000 (11:34 +0100)]
test: override blocking localed policy in TEST-73-LOCALE
On Debian and derivatives writing calls to localed are blocked as other
tools are used to change settings, override that policy for the tests
(cherry picked from commit
7b5c38a91def6cf236605010a0a93a1cd4c137e9)
Dominique Martinet [Sat, 20 Jul 2024 11:38:37 +0000 (20:38 +0900)]
meson: fix missing failure if bpf-framework was enabled
If building with clang and clang does not support bpf, then enabling
-Dbpf-framework=enabled would silently drop the feature (even printing
bpf-framework: enabled in the meson build recap, and no message anywhere
that'd hint at the failure!)
This is unexpected, so add check to fail hard in this case.
All other code paths (gcc, missing bpftool) properly check for the
option, but it is not as easy for a custom command so check explicitly
(cherry picked from commit
8da20e3fe2a544979922cea457de3031aa74d64c)
Ronan Pigott [Fri, 19 Jul 2024 08:20:15 +0000 (01:20 -0700)]
zsh: add varlinkctl completions
(cherry picked from commit
af63b4b769bfb86ff7848af980268901e9c0b47d)
Daan De Meyer [Fri, 19 Jul 2024 12:00:04 +0000 (14:00 +0200)]
docs: Document how to do stable releases
(cherry picked from commit
518ea76ecdf4855a2db2c0d31ee45f4db48e3242)
Lucas Werkmeister [Tue, 16 Jul 2024 16:15:32 +0000 (18:15 +0200)]
man: Mention Type=oneshot timeout directive
Make the warning for oneshot services (where RuntimeMaxSec= has no
effect) more actionable by pointing to the directive people can use
instead to effectively limit their runtime.
(cherry picked from commit
8c4aa0f1c6a78b35712fa6a7acf6d755d0c0bd86)
Léane GRASSER [Thu, 18 Jul 2024 13:15:16 +0000 (15:15 +0200)]
l10n: fix credits for the French translation
(cherry picked from commit
0d8b38415c4b192c1a4e608da8dc16c7340e31ba)
Lennart Poettering [Wed, 17 Jul 2024 13:56:59 +0000 (15:56 +0200)]
sd-device: remove debug log message when dirs are missing
This is a common case, and nothing noteworthy at all. For example, if we
establish an enumerator for listing all devices tagged by some tag, then
the per-tag dir is not going to exist if there are currently no devices
tagged that way, but that's a really common case, and doesn't really
deserve any mention, not even at debug level.
(cherry picked from commit
a68c97a54527cacaeeac0c117493639fc455ef5e)
Mike Yuan [Tue, 16 Jul 2024 19:10:46 +0000 (21:10 +0200)]
id128: refuse --app-specific= if we're listing GPT types
Prompted by #33737
The intention of
b37e8184a5a376749fbf68674ed6d7a4fc9901aa
is to expose sd_id128_get_app_specific() on command line.
But combining that with GPT type list makes little sense.
(cherry picked from commit
fa96c55b7b0d19a7f72908ee7d3f8a1ef630be96)
Luca Boccassi [Tue, 16 Jul 2024 16:46:54 +0000 (17:46 +0100)]
gpt: add more architecture aliases
Same as the other aliases. Allows chaining commands like:
$ systemd-id128 show -P root-$(dpkg-architecture --query DEB_HOST_ARCH)
4f68bce3e8cd4db196e7fbcaf984b709
(cherry picked from commit
f0b151ce864371da06a4d4a63a2a8b5282817b7e)
Jörg Behrmann [Fri, 12 Jul 2024 08:43:54 +0000 (10:43 +0200)]
kernel-install: Remove existing loader entries and UKIs
When boot counting is enabled, adding a new loader entry or UKI can conflict
with an existing one that has booted successfully and therefore has its boot
counter removed. systemd-bless-boot will fail to bless the new successful boot,
since a file without a boot counter already exists. Since kernel-install will
clobber existing files without boot counting, we should therefore remove files
without a boot count as well, when we add a file with one.
Fixes: #33504
(cherry picked from commit
99d4575e541fa1fb00dc80f7aad572f3a66db461)
Lennart Poettering [Fri, 5 Jul 2024 08:13:35 +0000 (10:13 +0200)]
bootspec: correct log level for fatal errors
(cherry picked from commit
7b7f4c26089ab7c34c854a2713f35ccd262cc15a)
Lennart Poettering [Fri, 5 Jul 2024 07:52:58 +0000 (09:52 +0200)]
bootspec: implement sorting by tries left/done, to match what sd-boot does
(cherry picked from commit
35451a32043504013eed5725c8be46b36ccdf71a)
Lennart Poettering [Thu, 4 Jul 2024 15:07:59 +0000 (17:07 +0200)]
boot: compare filename suffixes without case
This is VFAT world after all.
(cherry picked from commit
764faf60400bafb1764b728aafe0dcf4cbf07364)
Lennart Poettering [Fri, 5 Jul 2024 16:52:07 +0000 (18:52 +0200)]
man: some fixes
Addons are called addons, say so. And some other fixes.
(cherry picked from commit
40d9c16d1ee98c16af5804bab256b62c37feac3b)
Kuntal Majumder [Tue, 9 Jul 2024 17:11:52 +0000 (19:11 +0200)]
man: fix typo in unit options section
Fixes #32918
(cherry picked from commit
f9572d2b89341dfb224aa2c7222a316e59627bc9)
Luca Boccassi [Mon, 8 Jul 2024 15:23:06 +0000 (16:23 +0100)]
fsck: do not pull down mount units on soft-reboot
Otherwise they will pull down the disk too, which we don't want on soft-reboot
(cherry picked from commit
bbb0b72849ebbeeb8e252d9aeed94521df4f0ae8)
ZHANG Yuntian [Fri, 5 Jul 2024 09:04:44 +0000 (17:04 +0800)]
man: fix typo in the alias symlink name
Symlink created by Alias will use the value as the file name.
(cherry picked from commit
3f0e7fd4fd1d20e3f4be18f485c76d25ce10f41b)
Henry Chen [Fri, 5 Jul 2024 03:19:35 +0000 (11:19 +0800)]
vmspawn: define default machines for mips targets
All mips variants of qemu-system default to malta.
Signed-off-by: Henry Chen <henry.chen@oss.cipunited.com>
Signed-off-by: Henry Chen <chenx97@aosc.io>
(cherry picked from commit
3bff57b7651caab41021dbcdf658c1a92c8fdf52)
Ronan Pigott [Mon, 1 Jul 2024 21:07:28 +0000 (14:07 -0700)]
path: drop IN_ATTRIB from parent directory watches
When watching a given pathspec, systemd unconditionally installs
IN_ATTRIB watches to track the link count of the resolved file. This
way, we are notified if the watched path disappears, even if the
resolved file inode is not removed.
Similarly, systemd installs inotify watches on each parent directory, to
be notified when the specified path appears. However, for these watches
IN_ATTRIB is an unnecessary addition to the mask. In inotify, IN_ATTRIB
on a directory is emitted whenever the attributes of any child changes,
which, for many paths, has the potential to cause a high number of
spurious wakeups in systemd. Let's remove IN_ATTRIB from the mask when
installing watches on the parent directories of the specified path.
(cherry picked from commit
8bf8c7d83dcffffa55b5f534fb98db6b01315dc1)
fwfy [Sat, 6 Jul 2024 09:17:20 +0000 (05:17 -0400)]
Remove extra period at the end of systemd-bsod's unit description. (#33632)
* Remove extra period at end of unit description.
Having an extra period at the end of this unit description makes log entries pertaining to it appear weirdly, as it seems the default expectation is that there is not to be a period at the end of a unit description.
e.g.: `systemd[1]: Started Displays emergency message in full screen..`
(cherry picked from commit
496b4fa0e974d7a1b10b8af966da445a28f512c5)
Daan De Meyer [Fri, 5 Jul 2024 20:31:55 +0000 (22:31 +0200)]
test: Set priority for TEST-73-LOCALE
I don't know why yet, but TEST-73-LOCALE can take more than 10
minutes. Until we figure out why, let's give it a higher priority
so it doesn't bottleneck the test run.
(cherry picked from commit
14963ea90d72300d3e27b5776d1aa45594b10a94)
Daan De Meyer [Fri, 5 Jul 2024 17:20:42 +0000 (19:20 +0200)]
TEST-06-SELINUX: Disable RuntimeBuildSources=
Otherwise fixfiles will try to relabel it which could potentially
lead to disaster. We also change the recommendation in HACKING.md
to set the default so that TEST-06-SELINUX can override it.
(cherry picked from commit
575f954b5d0e03d41fe08ce2a7d85b868e89b904)
Daan De Meyer [Fri, 5 Jul 2024 14:06:34 +0000 (16:06 +0200)]
test: Add missing --no-rebuild to doc
Otherwise meson will try to rebuild all targets.
(cherry picked from commit
b5eaa52d4958f24aba21a4a34e585156079d4abb)
Daan De Meyer [Fri, 19 Jul 2024 13:58:04 +0000 (15:58 +0200)]
meson: Bump version to 256.3
Mike Yuan [Thu, 13 Jun 2024 09:00:26 +0000 (11:00 +0200)]
core/cgroup: check root cgroup earlier for unit_get_memory_accounting
(cherry picked from commit
d3d035395eeced715c9609aeb8bfac7157dd0aa9)
Mike Yuan [Sat, 15 Jun 2024 11:15:08 +0000 (13:15 +0200)]
core/cgroup: make unit_has_host_root_cgroup take const Unit*
(cherry picked from commit
4442aef08e0fe8ba381b580455f7eb281c5a28a1)
Mike Yuan [Sat, 1 Jun 2024 23:50:09 +0000 (07:50 +0800)]
core: unify reset_accounting handling
Since the introduction of CGroupRuntime, there's no need
to call *_reset_accounting in unit_new(), hence make those
static. While at it, refrain from hardcoding default values
in cgroup_runtime_new(), but call the corresponding funcs.
This also corrects the default value of io_accounting_base.
Fixes #33482
(cherry picked from commit
bc347edfe075f36bada0e1e7aefd62b10fcad4a7)
Mike Yuan [Wed, 5 Jun 2024 18:06:46 +0000 (20:06 +0200)]
core: do not drop CGroupRuntime when unit stops, but only on GC
Fixes #33149
Replaces #33145
(cherry picked from commit
4918f14adabba71913cf3abb8b5d50a8ca0fe844)
Mike Yuan [Sat, 1 Jun 2024 22:37:08 +0000 (06:37 +0800)]
core/cgroup: use > 0 comparison rather than == 1
(cherry picked from commit
8153be97c46dec725fdfaa3358d518ce350bb2ec)
Mike Yuan [Thu, 13 Jun 2024 09:01:03 +0000 (11:01 +0200)]
core/cgroup: actually make use of the cached accounting values
If cgroup is already gone, i.e. CGRuntime.cgroup_path is NULL,
do not return -ENODATA prematurely, but check for cached values
first.
For #33149
(cherry picked from commit
3849d1f56b7ff9db94656ca84156eb9fe1574ecb)
Mike Yuan [Wed, 19 Jun 2024 19:28:05 +0000 (21:28 +0200)]
test-install-root: introduce test case for #33411
(cherry picked from commit
9fb5a8ca24e677e10f8c2b8973b5e2a11676bda0)
Mike Yuan [Wed, 19 Jun 2024 16:45:14 +0000 (18:45 +0200)]
shared/install: correctly report changes in install_info_symlink_alias()
Follow-up for
b2751cf0394d36c24590b5f7b33e9f864b57ba0d
Also make the conditions consistent for install_info_symlink_wants().
Fixes #33411
(cherry picked from commit
4441cf330b3847d6c553fb230e8e4c86aa75ebb9)
Mike Yuan [Wed, 19 Jun 2024 16:59:15 +0000 (18:59 +0200)]
shared/install: propagate all errors in install_info_apply()
Currently, install_info_apply() only updates r if it's 0,
meaning that if one of the earlier install_info_symlink_alias/wants()
calls returns > 0, errors generated by later calls will be discarded.
Fix that.
(cherry picked from commit
a159aa07e1548367d2fde80cb0d45b869c591864)
Mike Yuan [Wed, 19 Jun 2024 16:44:26 +0000 (18:44 +0200)]
shared/install: drop unneeded initialization
(cherry picked from commit
dd6b325a05c4caccd1a17dd4147f48a916eee386)
Mike Yuan [Sun, 30 Jun 2024 11:12:45 +0000 (13:12 +0200)]
systemctl: do not try to acquire triggering units for template units
(cherry picked from commit
09d6038d833468ba7c24c658597387ef699ca4fd)
Mike Yuan [Thu, 29 Feb 2024 15:45:54 +0000 (23:45 +0800)]
systemctl: skip triggering unit warning if unit vanished
(cherry picked from commit
701bd9d08ac1d16f74e2b453ca0826e85b1c8491)
Mike Yuan [Fri, 28 Jun 2024 13:56:44 +0000 (15:56 +0200)]
man/systemctl: --no-reload is honored by mask/unmask/preset too
(cherry picked from commit
399646faac8ca91a46287e40280f56055c9e0fe8)
Tom Yan [Mon, 8 Jul 2024 14:10:43 +0000 (22:10 +0800)]
logind-dbus: set gc_mode to USER_GC_BY_PIN when disable linger
The mode switch from any to pin is currently done in create_session().
However, if no (pinning) session is created before (or after) linger
is disabled, the user will not be gc'd after that. Therefore, also
perform the mode switch when linger is being disabled.
(cherry picked from commit
1df981a74ae19513b40167c6b320c30bd166ac34)
Mike Yuan [Sat, 6 Jul 2024 17:05:57 +0000 (19:05 +0200)]
logind-user: take gc_mode into account when reporting user state
Prompted by #33650
Previously, if a user manually starts user@.service (which is
something we support), we'd track it as 'manager' session.
However, since user_get_state() ignores all non-pinning sessions,
if lingering is not enabled, the user state would always be
reported as 'closing', which is spurious.
Let's instead take gc_mode into consideration, and ignore
non-pinning sessions only if USER_GC_BY_PIN.
(cherry picked from commit
56f4ab290643547853276fb72b4e6960586c3e7c)
Zbigniew Jędrzejewski-Szmek [Thu, 18 Jul 2024 12:43:32 +0000 (14:43 +0200)]
test_ukify: do not use files from /boot
They might not be readable to the unprivileged user running the tests
and it shouldn't really matter what is used. OTOH, we need a real kernel
because we look at the header.
(cherry picked from commit
987f4bce938e790622a4b4b89d37daa7adfdc141)
Zbigniew Jędrzejewski-Szmek [Thu, 18 Jul 2024 12:48:37 +0000 (14:48 +0200)]
test_ukify: use sha384 in the signing tests
On Fedora, with crypto policy TEST-FEDORA41, sha1 is not allowed:
$ SYSTEMD_LOG_LEVEL=debug build/systemd-measure sign
--linux=/lib/modules/6.9.7-200.fc40.x86_64/vmlinuz
--osrel=/tmp/tmp.osrelbl2sr77f
--cmdline=/tmp/tmp.cmdlineouc7hqtj
--uname=/tmp/tmp.unamecbjgesty
--pcrpkey=/tmp/tmpufiadu8l
--initrd=/boot/
3a9d668b4db749398a4a5e78a03bffa5/6.9.7-200.fc40.x86_64/initrd
--sbat=/tmp/tmp.sbataz9arpy0
--private-key=/tmp/tmppyf0gx6w
--public-key=/tmp/tmpufiadu8l
--bank=sha1
Measuring boot phases: enter-initrd, enter-initrd:leave-initrd, enter-initrd:leave-initrd:sysinit, enter-initrd:leave-initrd:sysinit:ready
Loaded 'libtss2-esys.so.0' via dlopen()
Loaded 'libtss2-rc.so.0' via dlopen()
Loaded 'libtss2-mu.so.0' via dlopen()
PolicyPCR calculated digest:
cec1a2ccb188ddd171a2be7bfa6b31cb9148776647354eb1069e0f891ed2dbe7
Failed to initialize signature context: error:
03000098:digital envelope routines::invalid digest
Failed to sign PCR policy: Input/output error
(cherry picked from commit
87204601df305c270ffa05430081bc5b76dede04)
Zbigniew Jędrzejewski-Szmek [Thu, 18 Jul 2024 12:50:43 +0000 (14:50 +0200)]
test_ukify: add instructions
Copied directly from
a1d6dbb1c94685d7972f63ed2762fe4ba0251287.
(cherry picked from commit
a669091a436c4f0c8537c83bba05a1e33d263dad)
Lennart Poettering [Thu, 11 Jul 2024 07:34:11 +0000 (09:34 +0200)]
terminal-util: don't issue "ESC c" sequence on reset, but only when erasing the screen
ESC c is a (vaguely defined) "reset to initial state" ANSI sequence.
Many terminals clear the screen in this case, but that's a bit drastic I
think for most resets.
ESC c was added to the reset logic in
00bc83a275fa3ca8d90579fe9597d8b651d47332 (i.e. very recently), and I
don't think the effect was clear at that time.
Let's keep the ESC c in place however when we actually want to clear the
screen. Hence move it from reset_terminal_fd() into vt_disallocate().
Fixes: #33689
(cherry picked from commit
841eb9c186816acfef8f3f724647a7f0649f214d)
Daan De Meyer [Fri, 19 Jul 2024 11:24:35 +0000 (13:24 +0200)]
mkosi: Bump default device timeout a little
We've been getting some integration test failures due to timeouts
on finding the root partition device. Let's bump the default device
timeout a little to see if it mitigates these failures.
(cherry picked from commit
70510bf2823b6fcad504d67c3e3eb2b11c5ef6c9)
Daan De Meyer [Fri, 19 Jul 2024 11:49:01 +0000 (13:49 +0200)]
meson: Drop version from 256.3 to v256.2
Let's keep the stable version at the version of the previous release
until we do a new stable release. Unlike the main branch, there's no
need to immediately bump the version to the next version as the only
reason we do that in the main branch is to make sure the version from
the main branch sorts higher than any stable releases for the previous
version. This does not apply for stable releases and so we can safely
keep the old version in meson.version until we do a new stable release.
Daan De Meyer [Fri, 19 Jul 2024 09:41:28 +0000 (11:41 +0200)]
meson: Drop ~devel suffix from systemd-stable version
Every commit in systemd-stable can be considered stable since CI has
to pass for every merge to stable and we don't do any RC releases or
extra testing when doing a new stable release. So let's reflect that
in the versioning by dropping the "~devel" suffix from meson.version.
This allows distributions to do releases from arbitrary commits in
systemd-stable without ending up with ~devel in their versions.
Daan De Meyer [Thu, 18 Jul 2024 20:35:11 +0000 (22:35 +0200)]
Merge pull request #33774 from DaanDeMeyer/mkosi-backport
v256-stable: Backport various mkosi changes
Daan De Meyer [Thu, 18 Jul 2024 11:01:27 +0000 (13:01 +0200)]
mkosi: Fix indentation
(cherry picked from commit
2315c6996f198b44034fab7a02dc4b8a21c13248)
Daan De Meyer [Thu, 18 Jul 2024 10:53:53 +0000 (12:53 +0200)]
mkosi: Fix typo
Our config parsing is flexible enough that this kind of worked surprisingly
enough.
(cherry picked from commit
ac31804db7b679ae4a4ea028642b87e528ed3a6b)
Daan De Meyer [Thu, 18 Jul 2024 10:51:44 +0000 (12:51 +0200)]
mkosi: Drop util-linux from centos/fedora packages
It's already included in the packages list in mkosi.conf.
(cherry picked from commit
91cc65c5ad6135f428cd6500d19268581ece70b2)
Daan De Meyer [Thu, 18 Jul 2024 10:48:48 +0000 (12:48 +0200)]
mkosi: Drop udev from Packages= list
It's pulled in via VolatilePackages=, no need to put it in Packages=.
(cherry picked from commit
1d914b268c116f5491f02bb554e45483404ed82a)
Daan De Meyer [Thu, 18 Jul 2024 10:47:51 +0000 (12:47 +0200)]
mkosi: Fix formatting
All of our lists start on the next line, so let's make KernelCommandLine=
fit that as well.
(cherry picked from commit
1250dc6a1d4fe1dbb8bcd21fc642c4ffdc60e0c7)
Daan De Meyer [Thu, 18 Jul 2024 06:46:14 +0000 (08:46 +0200)]
mkosi: Build CentOS Stream 10 images by default
CentOS Stream 10 has a newer util-linux which means the terminal
gets correctly resized to the size specified by mkosi. This is a
much nicer experience than CentOS Stream 9 where you're stuck on
80x24 so let's make CentOS Stream 10 the default release to build.
(cherry picked from commit
773f2e9633677298985c1f2991556aeed60c2843)
Daan De Meyer [Thu, 11 Jul 2024 14:09:47 +0000 (16:09 +0200)]
mkosi: Streamline running the integration tests without building systemd
Let's document in detail how to build the integration test image and run
the integration tests without building systemd. To streamline the process,
we stop automatically using binaries from build/ when invoking mkosi directly
and don't automatically use a tools tree anymore if systemd on the host is too
old. Instead, we document these options in HACKING.md and change the mkosi meson
target to automatically use the current build directory as an extra binary search
path for mkosi.
(cherry picked from commit
d279ec4a50d83ab44200ea410fa09d7ba2e5c824)
Daan De Meyer [Fri, 28 Jun 2024 07:58:56 +0000 (09:58 +0200)]
meson: Drop genkey target
In mkosi.images/system/mkosi.conf, we configure the certificate as
an extra tree so it's available inside the image. However, we pick up
the certificate from the top level repository directory and not from the
build directory where it is generated by the genkey meson target.
We currently have no way to access the build directory that mkosi was
invoked from when parsing the configuration file. Thus we have no way to
specify the correct location to the certificate when it's located in the
build directory.
For now, let's look for the key and certificate in the top level repository
root directory and drop the genkey target.
We don't have to change the Github Actions CI because it already runs genkey
manually before the image build (which is something we forgot to remove when
introducing the genkey target and is the reason this didn't cause issues before).
(cherry picked from commit
b00f44c3cff083972b0458359bbb5b80f0d10931)
Daan De Meyer [Thu, 11 Jul 2024 14:09:15 +0000 (16:09 +0200)]
mkosi: Skip sync script if NO_BUILD is enabled
If we're not doing a build, there's no point in syncing either so
lets skip it.
(cherry picked from commit
bb8f6ac445b80f9621503bec7a22caa57bb10ee4)
Daan De Meyer [Wed, 17 Jul 2024 16:56:02 +0000 (18:56 +0200)]
mkosi: Remove enforcing=0 from default kernel command line
We already have selinux=0 in the default kernel command line so
enforcing=0 is redundant. Instead, pass in enforcing=0 when we
enable selinux in TEST-06-SELINUX.
(cherry picked from commit
f5c44df92922d74cfff40989d6617e3a9e4c5904)
Daan De Meyer [Wed, 17 Jul 2024 16:55:24 +0000 (18:55 +0200)]
mkosi: Stop setting apparmor=0
It doesn't get pulled in as a dependency anyway and kernel command
line space is precious so let's remove apparmor=0 as it's a noop
anyway.
(cherry picked from commit
5319be5f0c68d7edb09438419ca4171e4692c859)
migleeson [Wed, 17 Jul 2024 07:45:26 +0000 (03:45 -0400)]
docs: update mkosi version mentioned in HACKING.md (#33723)
* fix: update docs since the default config uses a setting only available in v23
* fix: update docs to only refer to installing from the mkosi repo
(cherry picked from commit
cb2d6214caa25b8297bde01bb8bb925957b04ba7)
Daan De Meyer [Tue, 16 Jul 2024 15:51:55 +0000 (17:51 +0200)]
Merge pull request #33735 from DaanDeMeyer/backport-mkosi
v256-stable: Backport mkosi and test related changes
Daan De Meyer [Tue, 2 Jul 2024 16:34:39 +0000 (18:34 +0200)]
repart: Allow overriding fstype per partition designator
$SYSTEMD_REPART_OVERRIDE_FSTYPE is too invasive. Often you want to
override the fstype only for a specific designator, so let's support
that as well.
(cherry picked from commit
90a255779d1f8e6697e08e91918df88bb52274ad)
Daan De Meyer [Tue, 16 Jul 2024 12:15:38 +0000 (14:15 +0200)]
mkosi: List library packages explicitly in VolatilePackages=
These should already get pulled in as dependencies, but it doesn't
hurt to be explicit.
(cherry picked from commit
b209fff3a51b8e57ea5970575bf8d4db4790b620)
Daan De Meyer [Tue, 16 Jul 2024 10:07:38 +0000 (12:07 +0200)]
mkosi: Build initrd as a subimage
Let's make things a little more consistent and build the initrd
explicitly as a subimage as well instead of relying on mkosi building
it as part of the main image build.
We drop the opensuse initrd postinst script as we don't use erofs by
default anymore. We can always reintroduce it again later if needed.
(cherry picked from commit
68ee9771145464949d36d62a0ec687ecfb39290a)
Daan De Meyer [Tue, 16 Jul 2024 09:45:43 +0000 (11:45 +0200)]
mkosi: Drop CacheOnly=always from two subimages
This is a universal setting and can only be configured in the main
image so let's drop it from the subimages.
(cherry picked from commit
244952b1ead7bb14058a873fded2cb9709c82a95)
Daan De Meyer [Mon, 15 Jul 2024 21:07:52 +0000 (23:07 +0200)]
TEST-13-NSPAWN: make sure we don't load libnss_systemd
Let's make sure we don't load libnss_systemd.so from bash as the
necessary environment variables aren't set to make that work when
we're running with sanitizers enabled.
We can't add a sanitizer wrapper for bash as the wrapper runs using
bash so you end up in a loop.
(cherry picked from commit
5ac0dc70ff330a531539a92552fbeb0f542edfb5)
Daan De Meyer [Mon, 15 Jul 2024 15:15:43 +0000 (17:15 +0200)]
mkosi: Disable unique debug source names
We use -fdebug-prefix-map= because debugedit doesn't work for us (for
a currently unknown reason since it's the most obtuse code I've ever
had the pleasure of reading). With all the unique macros enabled, the
destination directory we pass to -fdebug-prefix-map= includes the package
release. The release is either the timestamp of the current commit or
the current time if the working tree is dirty. This means it generally
changes every time we rerun the build script. However, meson only reads
compiler arguments the first time it is invoked or if --wipe is specified.
This means that on a rerun -fdebug-prefix-map= will be configured wrong
and the build will fail.
Let's prevent this from happening by disabling the unique debug source
names by overriding the --unique-debug-src-base option that is passed to
find-debuginfo.sh by rpm via the _find_debuginfo_opts macro.
(cherry picked from commit
88221219a3290ee20de27aa030c2b23f03d8238a)