]> git.ipfire.org Git - thirdparty/systemd.git/log
thirdparty/systemd.git
11 months agobootspec: Look at /loader/addons in XBOOTLDR 34959/head
Adrian Vovk [Tue, 1 Oct 2024 20:54:22 +0000 (22:54 +0200)] 
bootspec: Look at /loader/addons in XBOOTLDR

The bootspec util-lib's handling of global addons didn't previously
match the behavior of sd-stub, and this commit corrects that.

First, bootspec didn't load global addons from the XBOOTLDR dir, but the
stub does. So, bootspec now enumerates addons in XBOOTLDR, not just ESP

Second, the stub only loads resources (including addons) from the
partition that it was found on. Thus, we must keep track of which
partition the global addons come from, and which partition each boot
entry comes from. In other words: global addons found on the ESP will
NOT apply to UKIs found in XBOOTLDR, and bootspec now reflects that.

11 months agobootspec: Fixup loading of local addons for UKIs
Adrian Vovk [Wed, 30 Oct 2024 19:38:49 +0000 (15:38 -0400)] 
bootspec: Fixup loading of local addons for UKIs

Follow-up for 59b3df9

11 months agobootspec: Fixup memory leak
Adrian Vovk [Tue, 1 Oct 2024 20:53:02 +0000 (22:53 +0200)] 
bootspec: Fixup memory leak

This would previously leak memory: the array was deleted but contents
inside of the array were not

11 months agosd-stub: Fixup typo & measurement order
Adrian Vovk [Tue, 1 Oct 2024 20:29:47 +0000 (22:29 +0200)] 
sd-stub: Fixup typo & measurement order

A previous commit accidentally reversed the measurement order of the
confext initrds and sysext initrds via a minor typo. This commit fixes
the typo and restores the original measurement order

Follow-up: ac32323

11 months agosysext: set SELinux context for hierarchies and workdir
gerblesh [Thu, 14 Nov 2024 19:12:40 +0000 (19:12 +0000)] 
sysext: set SELinux context for hierarchies and workdir

11 months agoman: several more assorted fixes
Yu Watanabe [Tue, 26 Nov 2024 13:29:36 +0000 (22:29 +0900)] 
man: several more assorted fixes

Continuation of 4ebbb5bfe88ac3d793c395472648660c33251546.
Closes #35307.

11 months agomachine: increase timeouts in attempt to fix #35115 (#35117)
Luca Boccassi [Tue, 26 Nov 2024 16:12:56 +0000 (16:12 +0000)] 
machine: increase timeouts in attempt to fix #35115 (#35117)

An attempt to fix https://github.com/systemd/systemd/issues/35115

11 months agosd-varlink: fix bug when enqueuing messages with fds asynchronously
Lennart Poettering [Wed, 20 Nov 2024 12:13:41 +0000 (13:13 +0100)] 
sd-varlink: fix bug when enqueuing messages with fds asynchronously

When determining the poll events to wait for we need to take the queue
of pending messages that carry fds into account. Otherwise we might end
up not waking up if such an fd-carrying message is enqueued
asynchronously (i.e. not from a dispatch callback).

11 months agoman/systemd-system.conf: Correct "struct" to "strict" (#35364)
Winterhuman [Tue, 26 Nov 2024 13:41:49 +0000 (13:41 +0000)] 
man/systemd-system.conf: Correct "struct" to "strict" (#35364)

11 months agocryptsetup: convert pkcs11/fido2 to iovec for key handling
Luca Boccassi [Mon, 25 Nov 2024 22:31:01 +0000 (22:31 +0000)] 
cryptsetup: convert pkcs11/fido2 to iovec for key handling

key-data might be NULL. Fixes crash:

0  0x0000559c62120530 in attach_luks_or_plain_or_bitlk (cd=0x559c6b192830, name=0x7ffd57981dc4 "root", token_type=TOKEN_FIDO2, key_file=0x0, key_data=0x0, passwords=0x0, flags=524296, until=0)
    at ../src/cryptsetup/cryptsetup.c:2234
        pass_volume_key = false
        r = 1469577760
        __func__ = '\000' <repeats 29 times>
1  0x0000559c6212279c in run (argc=6, argv=0x7ffd5797fe98) at ../src/cryptsetup/cryptsetup.c:2597
        discovered_key_data = {iov_base = 0x0, iov_len = 0}
        key_data = 0x0
        token_type = TOKEN_FIDO2
        destroy_key_file = 0x0
        flags = 524296
        until = 0
        passphrase_type = PASSPHRASE_NONE
        volume = 0x7ffd57981dc4 "root"
        source = 0x7ffd57981dc9 "/dev/disk/by-uuid/8372fb39-9ba4-461a-a618-07dcaae66280"
        status = CRYPT_INACTIVE
        tries = 0
        key_file = 0x0
        config = 0x7ffd57981e05 "luks,discard,fido2-device=auto,x-initrd.attach"
        use_cached_passphrase = true
        try_discover_key = true
        discovered_key_fn = 0x7ffd5797fa70 "root.key"
        passwords = 0x0
        cd = 0x559c6b192830
        verb = 0x7ffd57981dbd "attach"
        r = 0
        __func__ = "\000\000\000"
2  0x0000559c621231e6 in main (argc=6, argv=0x7ffd5797fe98) at ../src/cryptsetup/cryptsetup.c:2674
        r = 32553
        __func__ = "\000\000\000\000"

Follow-up for 53b6c99018f918a5d2c9000ac5fe3a2440115ea7

11 months agoupdatectl: fix DBus method signature for SetFeatureEnabled
Abderrahim Kitouni [Tue, 26 Nov 2024 10:55:33 +0000 (11:55 +0100)] 
updatectl: fix DBus method signature for SetFeatureEnabled

The signature was changed to 'sit' in sysupdated during review, but updatectl
kept using 'sbt'

11 months agoRevert "Revert "man: use MIT-0 license for example codes in daemon(7)""
Yu Watanabe [Tue, 26 Nov 2024 10:47:41 +0000 (19:47 +0900)] 
Revert "Revert "man: use MIT-0 license for example codes in daemon(7)""

This reverts commit 7a9d0abe4df81a020e2e457405f8509b2afb7c04.

11 months agoRevert "man: use MIT-0 license for example codes in daemon(7)"
Luca Boccassi [Tue, 26 Nov 2024 10:43:19 +0000 (10:43 +0000)] 
Revert "man: use MIT-0 license for example codes in daemon(7)"

This reverts commit 6046cc3660810efcc6fe50b1c850ea642218245b.

11 months agoman: use MIT-0 license for example codes in daemon(7)
Yu Watanabe [Tue, 26 Nov 2024 09:21:29 +0000 (18:21 +0900)] 
man: use MIT-0 license for example codes in daemon(7)

This page contains many short example codes. I do not think we should
add SPDX-License-Identifier for all codes.

Closes #35356.

11 months agoman: assorted fixes (#35326)
Luca Boccassi [Mon, 25 Nov 2024 15:02:08 +0000 (15:02 +0000)] 
man: assorted fixes (#35326)

Closes #35307.

11 months agotest: Use env in testsuite readme
Daan De Meyer [Mon, 25 Nov 2024 14:47:51 +0000 (15:47 +0100)] 
test: Use env in testsuite readme

Let's make sure we use env when we're setting environment variables
to rely less on shell specifics.

11 months agocore/device: ignore ID_PROCESSING udev property on enumerate (#35332)
Zbigniew Jędrzejewski-Szmek [Mon, 25 Nov 2024 13:21:36 +0000 (14:21 +0100)] 
core/device: ignore ID_PROCESSING udev property on enumerate (#35332)

Fixes #35329.

11 months agoCheck inode number to see if we are in init namespace (#35306)
Zbigniew Jędrzejewski-Szmek [Mon, 25 Nov 2024 13:13:36 +0000 (14:13 +0100)] 
Check inode number to see if we are in init namespace (#35306)

This is a more comprehensive fix compared to #35273. Also adds a minimal
test only.

Based on Luca's #35273 but generalizes the code a bit.

In v258 we really should get rid of the old heuristics around userns and
cgroupns detection, but given we are late in the v257 cycle this keeps
them in.

11 months agotest: Dump coredumps from journal in the integration test wrapper
Daan De Meyer [Fri, 22 Nov 2024 21:51:45 +0000 (22:51 +0100)] 
test: Dump coredumps from journal in the integration test wrapper

Fixes #35277

11 months agonetworkd-test.py: disable IPv6AcceptRA= if not necessary
Yu Watanabe [Mon, 25 Nov 2024 07:15:44 +0000 (16:15 +0900)] 
networkd-test.py: disable IPv6AcceptRA= if not necessary

To speed up the test. Otherwise, it takes about few seconds interfaces
to enter the configured state. And may networkd-wait-online timeouts.

11 months agoTEST-17: add reproducer for issue #35329 35332/head
Yu Watanabe [Sun, 24 Nov 2024 04:57:37 +0000 (13:57 +0900)] 
TEST-17: add reproducer for issue #35329

Without the previous commit, the test case will fail.

11 months agocore/device: ignore ID_PROCESSING udev property on enumerate
Yu Watanabe [Sun, 24 Nov 2024 04:19:27 +0000 (13:19 +0900)] 
core/device: ignore ID_PROCESSING udev property on enumerate

This partially reverts the commit 405be62f05d76f1845f347737b5972158c79dd3e
"tree-wide: refuse enumerated device with ID_PROCESSING=1".

Otherwise, when systemd-udev-trigger.service is (re)started just before
daemon-reexec, which can be easily happen on systemd package update, then
udev database files for many devices may have ID_PROCESSING=1 property,
thus devices may not be enumerated on daemon-reexec. That causes many
units especially mount units being deactivated after daemon-reexec.

Fixes #35329.

11 months agotest: mask tmpfiles.d file shipped by selinux policy package in containers
Luca Boccassi [Sat, 23 Nov 2024 13:28:03 +0000 (13:28 +0000)] 
test: mask tmpfiles.d file shipped by selinux policy package in containers

This tmpfiles.d wants to write to sysfs, which is read-only in containers,
so systemd-tmpfiles --create fails in TEST-22-TMPFILES when ran in nspawn
if the selinux policy package is instealled. Mask it, as it's not our
config file, we don't need it in the test.

11 months agomkosi: Use shared extra tree between initrd and main image
Daan De Meyer [Sun, 24 Nov 2024 18:19:52 +0000 (19:19 +0100)] 
mkosi: Use shared extra tree between initrd and main image

Let's share more between initrd and main system and use a shared
extra tree to achieve that.

11 months agomkosi: Make sure mkosi.clangd always runs on the host
Daan De Meyer [Sun, 24 Nov 2024 22:19:01 +0000 (23:19 +0100)] 
mkosi: Make sure mkosi.clangd always runs on the host

If the editor that invokes mkosi.clangd is a flatpak, let's make sure
that mkosi is run on the host and not in the flatpak sandbox since it
won't be installed there.

11 months agoVarious mkosi and integration test fixes (#35336)
Luca Boccassi [Sun, 24 Nov 2024 18:10:03 +0000 (18:10 +0000)] 
Various mkosi and integration test fixes (#35336)

11 months agotest: Lint integration-test-wrapper.py 35336/head
Daan De Meyer [Sat, 23 Nov 2024 11:54:02 +0000 (12:54 +0100)] 
test: Lint integration-test-wrapper.py

11 months agotest: Fix typing errors in integration-test-wrapper.py
Daan De Meyer [Sat, 23 Nov 2024 12:01:00 +0000 (13:01 +0100)] 
test: Fix typing errors in integration-test-wrapper.py

11 months agotest: Format integration-test-wrapper.py
Daan De Meyer [Sat, 23 Nov 2024 11:59:32 +0000 (12:59 +0100)] 
test: Format integration-test-wrapper.py

11 months agoukify: Fix typing error
Daan De Meyer [Sat, 23 Nov 2024 13:04:27 +0000 (14:04 +0100)] 
ukify: Fix typing error

11 months agoMove mypy.ini and ruff.toml to top level
Daan De Meyer [Sat, 23 Nov 2024 11:58:47 +0000 (12:58 +0100)] 
Move mypy.ini and ruff.toml to top level

This allows reusing them for integration-test-wrapper.py as well.

11 months agointegration-test-wrapper: Remove unneeded format strings
Daan De Meyer [Fri, 22 Nov 2024 21:37:34 +0000 (22:37 +0100)] 
integration-test-wrapper: Remove unneeded format strings

11 months agomkosi: Add github CLI to tools
Daan De Meyer [Sun, 24 Nov 2024 09:50:02 +0000 (10:50 +0100)] 
mkosi: Add github CLI to tools

11 months agomkosi: Add dnf and dnf5 to sanitizer workaround list
Daan De Meyer [Sat, 23 Nov 2024 23:03:50 +0000 (00:03 +0100)] 
mkosi: Add dnf and dnf5 to sanitizer workaround list

11 months agomkosi: Install clangd everywhere
Daan De Meyer [Sat, 23 Nov 2024 22:33:54 +0000 (23:33 +0100)] 
mkosi: Install clangd everywhere

11 months agomkosi: Use bash to execute command -v
Daan De Meyer [Sat, 23 Nov 2024 21:29:56 +0000 (22:29 +0100)] 
mkosi: Use bash to execute command -v

command is only an executable on Fedora due to a downstream patch,
on Arch for example it's only a builtin so we have to use bash to
execute command -v to get proper results on Arch.

11 months agomkosi: Add shellcheck to tools
Daan De Meyer [Sat, 23 Nov 2024 21:29:31 +0000 (22:29 +0100)] 
mkosi: Add shellcheck to tools

11 months agomkosi: update to latest
Daan De Meyer [Sat, 23 Nov 2024 15:11:36 +0000 (16:11 +0100)] 
mkosi: update to latest

11 months agoNEWS: add blurb thanking Nick Owens
Vito Caputo [Sun, 24 Nov 2024 05:54:56 +0000 (21:54 -0800)] 
NEWS: add blurb thanking Nick Owens

Nick's largely responsible for nerd-sniping me into fixing #34516
and did most of the testing.

11 months agohwdb: add entry for Chuwi Hi10 X1 (#35331)
白一百 [Sun, 24 Nov 2024 07:30:33 +0000 (15:30 +0800)] 
hwdb: add entry for Chuwi Hi10 X1 (#35331)

https://www.chuwi.com/product/items/chuwi-hi10-x1.html
Rotated -90 degrees in the Z axis.

11 months agocurl-util: do not configure new io event source when the event loop is already dead
Yu Watanabe [Sat, 23 Nov 2024 16:05:47 +0000 (01:05 +0900)] 
curl-util: do not configure new io event source when the event loop is already dead

Similar to c5ecf0949460dd0bf3211db128a385ce6375252e, but for io event source.

Fixes #35322.

11 months agoman: update documentation about basic .netdev file handling 35326/head
Yu Watanabe [Sat, 23 Nov 2024 15:46:47 +0000 (00:46 +0900)] 
man: update documentation about basic .netdev file handling

Follow-up for #34909 and later PRs.

11 months agoman: asorted fixes
Yu Watanabe [Sat, 23 Nov 2024 13:07:56 +0000 (22:07 +0900)] 
man: asorted fixes

Closes #35307.

11 months agomeasure: add 'dtbauto' option in help message
Ani Sinha [Sat, 23 Nov 2024 06:15:14 +0000 (11:45 +0530)] 
measure: add 'dtbauto' option in help message

'dtbauto' command line was missing from the help string. Add it.

11 months agopo: Translated using Weblate (French)
Léane GRASSER [Sat, 23 Nov 2024 10:38:35 +0000 (11:38 +0100)] 
po: Translated using Weblate (French)

Currently translated at 100.0% (257 of 257 strings)

Co-authored-by: Léane GRASSER <leane.grasser@proton.me>
Translate-URL: https://translate.fedoraproject.org/projects/systemd/main/fr/
Translation: systemd/main

11 months agonetworkd-test.py: fix interface state checker
Yu Watanabe [Fri, 22 Nov 2024 18:32:34 +0000 (03:32 +0900)] 
networkd-test.py: fix interface state checker

After 259125d53d98541623b69e83000b5543f2352f5e, network interfaces
declared by .netdev files are created after systemd-networkd sends READY
notification. So, even when networkd is started, the netdevs may not
be created yet, and 'ip' command may fail. Let's also check the return
code of the command.

This also
- drops never worked stdout checks,
- makes the test fail if the interface is not created within the timeout.

11 months agonspawn: improve log message on bad incoming sd_notify() message
Lennart Poettering [Fri, 22 Nov 2024 23:04:43 +0000 (00:04 +0100)] 
nspawn: improve log message on bad incoming sd_notify() message

It's the PID that is wrong, not the UID/GID, be precise.

11 months agonspawn: fix userns_mkdir() invocation
Lennart Poettering [Thu, 21 Nov 2024 11:16:57 +0000 (12:16 +0100)] 
nspawn: fix userns_mkdir() invocation

The wrong error code was logged.

But actually given that userns_mkdir() is fine with existing dirs, let's
drop the redundant conditionalization.

Follow-up for: a1fcaa1549d86098d0ba75254b6afc96c786b3b6

11 months agoshutdown: propagate one more error from sync_making_progress()
Yu Watanabe [Fri, 22 Nov 2024 21:40:22 +0000 (06:40 +0900)] 
shutdown: propagate one more error from sync_making_progress()

No functional change, just refactoring, as anyway all errors will be
ignored by the caller.

11 months agonamespace-util: handle -ENOSPC by userns_acquire() gracefully in is_idmapping_support...
Yu Watanabe [Sat, 23 Nov 2024 08:32:23 +0000 (17:32 +0900)] 
namespace-util: handle -ENOSPC by userns_acquire() gracefully in is_idmapping_supported() (#35313)

Follow-up for edae62120f13b24d51812d1d7c0ab24acb420305.
Fixes #35311.

11 months agoshutdown: close DM block device before issuing DM_DEV_REMOVE ioctl
Yu Watanabe [Fri, 22 Nov 2024 20:47:40 +0000 (05:47 +0900)] 
shutdown: close DM block device before issuing DM_DEV_REMOVE ioctl

Otherwise, the ioctl() may fail with EBUSY.

Follow-up for b4b66b26620bfaf5818c95d5cffafd85207694e7.
Hopefully fixes #35243.

11 months agobasic/linux: update kernel headers from v6.12
Yu Watanabe [Fri, 22 Nov 2024 19:33:57 +0000 (04:33 +0900)] 
basic/linux: update kernel headers from v6.12

11 months agoUndeprecate commandline params forcequotacheck, fastboot, and forcefsck
Zbigniew Jędrzejewski-Szmek [Thu, 21 Nov 2024 19:36:51 +0000 (20:36 +0100)] 
Undeprecate commandline params forcequotacheck, fastboot, and forcefsck

Those are historical names, but there is nothing wrong with them. The files on
/ (/fastboot, /forcefsck, and /forcequotacheck) are problematic because they
require a modification of the root file system. But the commandline params work
fine. They have the obvious advantage compared to our "modern" option that they
are much easier to type without looking up the spelling in the docs. Undeprecate
them to avoid unnecessary churn.

11 months agotest-namespace: tweak log message a bit 35306/head
Lennart Poettering [Fri, 22 Nov 2024 16:43:17 +0000 (17:43 +0100)] 
test-namespace: tweak log message a bit

11 months agovirt: make use of ns inode check in running_in_userns() and running_in_cgroupns(...
Lennart Poettering [Fri, 22 Nov 2024 16:37:35 +0000 (17:37 +0100)] 
virt: make use of ns inode check in running_in_userns() and running_in_cgroupns() too

11 months agodetect-virt: check the inode number of the pid namespace
Luca Boccassi [Thu, 21 Nov 2024 00:47:39 +0000 (00:47 +0000)] 
detect-virt: check the inode number of the pid namespace

The indoe number of root pid namespace is hardcoded in the kernel to
0xEFFFFFFC since 3.8, so check the inode number of our pid namespace
if all else fails. If it's not 0xEFFFFFFC then we are in a pid
namespace, hence a container environment.

Fixes https://github.com/systemd/systemd/issues/35249

[Reworked by Lennart, to make use of namespace_is_init()]

11 months agonamespace-util: add generic namespace_is_init() call
Lennart Poettering [Fri, 22 Nov 2024 16:11:29 +0000 (17:11 +0100)] 
namespace-util: add generic namespace_is_init() call

11 months agonamespace-util: update log messages 35313/head
Yu Watanabe [Fri, 22 Nov 2024 21:10:32 +0000 (06:10 +0900)] 
namespace-util: update log messages

11 months agonamespace-util: handle -ENOSPC by userns_acquire() gracefully in is_idmapping_supported()
Yu Watanabe [Fri, 22 Nov 2024 21:08:48 +0000 (06:08 +0900)] 
namespace-util: handle -ENOSPC by userns_acquire() gracefully in is_idmapping_supported()

Follow-up for edae62120f13b24d51812d1d7c0ab24acb420305.
Fixes #35311.

11 months agolink README.logs from tmpfiles.d/legacy.conf only if available
Christian Hesse [Fri, 22 Nov 2024 10:59:26 +0000 (11:59 +0100)] 
link README.logs from tmpfiles.d/legacy.conf only if available

The file README.logs is installed only if SysVInit support is enabled.
Thus the link should depend on it as well.

11 months agovarlink: apparently on old kernels SO_PEERPIDFD returns EINVAL
Lennart Poettering [Fri, 22 Nov 2024 11:14:59 +0000 (12:14 +0100)] 
varlink: apparently on old kernels SO_PEERPIDFD returns EINVAL

11 months agohwdb: add support for PineTab2 to 60-sensor.hwdb (#35304)
Marco Tomaschett [Fri, 22 Nov 2024 18:08:06 +0000 (13:08 -0500)] 
hwdb: add support for PineTab2 to 60-sensor.hwdb (#35304)

Add accelerometer support for PineTab2

11 months agouserdbctl: two trivial fixlets (#35296)
Lennart Poettering [Fri, 22 Nov 2024 15:06:01 +0000 (16:06 +0100)] 
userdbctl: two trivial fixlets (#35296)

Fixes: #35294
11 months agotpm2-util: fix parameter name
Antonio Alvarez Feijoo [Fri, 22 Nov 2024 13:21:58 +0000 (14:21 +0100)] 
tpm2-util: fix parameter name

11 months agoman: split cryptenroll man page into sections (#35297)
Luca Boccassi [Fri, 22 Nov 2024 12:01:07 +0000 (12:01 +0000)] 
man: split cryptenroll man page into sections (#35297)

11 months agouserdbctl: respect selected disposition also when showing gid boundaries 35296/head
Lennart Poettering [Fri, 22 Nov 2024 10:26:34 +0000 (11:26 +0100)] 
userdbctl: respect selected disposition also when showing gid boundaries

Follow-up for: ad5de3222f7

11 months agouserdbctl: fix counting
Lennart Poettering [Fri, 22 Nov 2024 09:11:32 +0000 (10:11 +0100)] 
userdbctl: fix counting

Fixes: #35294
11 months agouserbdctl: show 'mapped' user range only inside of userns
Lennart Poettering [Fri, 22 Nov 2024 08:34:56 +0000 (09:34 +0100)] 
userbdctl: show 'mapped' user range only inside of userns

Outside of userns the concept makes no sense, there cannot be users
mapped from further outside.

11 months agocryptenroll: it's called PKCS#11, not PKCS11 35297/head
Lennart Poettering [Fri, 22 Nov 2024 09:11:04 +0000 (10:11 +0100)] 
cryptenroll: it's called PKCS#11, not PKCS11

In the --help text we really should use the official spelling, just like
in the man page.

11 months agoman: add enrollment type sections to cryptenroll man page
Lennart Poettering [Fri, 22 Nov 2024 09:10:11 +0000 (10:10 +0100)] 
man: add enrollment type sections to cryptenroll man page

We have the same sections in the --help text, hence we even more so
should have them in the man page.

11 months agobash-completion: add --list-devices to systemd-cryptenroll
Antonio Alvarez Feijoo [Thu, 21 Nov 2024 11:50:26 +0000 (12:50 +0100)] 
bash-completion: add --list-devices to systemd-cryptenroll

And also use it to list suitable block devices.

11 months agocore/service: service_add_fd_store() consumes passed fd
Yu Watanabe [Thu, 21 Nov 2024 20:15:49 +0000 (05:15 +0900)] 
core/service: service_add_fd_store() consumes passed fd

Without this change, the fd is closed twice on failure.

Fixes a bug introduced by dff9808a628c31b7ecb1f1aba8fdc3be06ce8372.

Fixes #35288.

11 months agohwdb: Add quirk for Logitech MX Keys for Mac
Tobias Zimmermann [Wed, 20 Nov 2024 22:16:21 +0000 (23:16 +0100)] 
hwdb: Add quirk for Logitech MX Keys for Mac

The KEY_102ND and KEY_GRAVE keys are switched on the
Logitech MX Keys for Mac, so switch them back

11 months agoKeep tmpfiles/legacy.conf even if SysVInit support is dropped (#35278)
Zbigniew Jędrzejewski-Szmek [Thu, 21 Nov 2024 20:13:50 +0000 (21:13 +0100)] 
Keep tmpfiles/legacy.conf even if SysVInit support is dropped (#35278)

11 months agopo: Translated using Weblate (Ukrainian)
Yuri Chornoivan [Thu, 21 Nov 2024 19:38:33 +0000 (20:38 +0100)] 
po: Translated using Weblate (Ukrainian)

Currently translated at 100.0% (257 of 257 strings)

Co-authored-by: Yuri Chornoivan <yurchor@ukr.net>
Translate-URL: https://translate.fedoraproject.org/projects/systemd/main/uk/
Translation: systemd/main

11 months agotmpfiles: add period at end of the sentence 35278/head
Zbigniew Jędrzejewski-Szmek [Thu, 21 Nov 2024 09:25:38 +0000 (10:25 +0100)] 
tmpfiles: add period at end of the sentence

The license that is immediately above is properly punctuated and it looks
sloppy when our line below isn't.

11 months agotmpfiles: narrow scope of HAVE_SYSV_COMPAT condition for legacy.conf
Zbigniew Jędrzejewski-Szmek [Thu, 21 Nov 2024 09:12:56 +0000 (10:12 +0100)] 
tmpfiles: narrow scope of HAVE_SYSV_COMPAT condition for legacy.conf

That file contains a bunch of entries of which only some are related to SysV.
The rest are just "traditional APIs" that need to stay. In particular,
/var/lock a.k.a. /run/lock is used by many programs (LVM, iscsi, alsactl).
Similarly, the README about /var/log is something that should stay as long as
we have people migrating from older systems or using the copiuos documentation
that mentions /var/log/messages.txt on the Internet.

/var/lock/subsys is only used by sysvinit, and our code to support /forcefsck,
/fastboot, and /forcequotacheck is conditionalized on HAVE_SYSV_COMPAT, so
conditionalize those here on HAVE_SYSV_COMPAT too.

11 months agocgroup-util: fix memory leak on error
Luca Boccassi [Thu, 21 Nov 2024 00:53:20 +0000 (00:53 +0000)] 
cgroup-util: fix memory leak on error

CID#1565824

Follow-up for f6793bbcf0e3f0a6daa77add96183b88d5ec2117

11 months agonetwork: update state files before replying bus method (#35255)
Luca Boccassi [Wed, 20 Nov 2024 23:36:35 +0000 (23:36 +0000)] 
network: update state files before replying bus method (#35255)

Follow-up for 2b07a3211ba8b1b81d6cebb9650d5cb24554b08a.

Fixes the failure found in
https://autopkgtest.ubuntu.com/results/autopkgtest-noble-upstream-systemd-ci-systemd-ci/noble/amd64/s/systemd-upstream/20241115_182040_92382@/log.gz
. Relevant logs:
```
Nov 16 02:48:36 systemd-networkd[2706]: veth99: Reconfiguring with /run/systemd/network/25-dhcp-client-ipv6-only.network.
Nov 16 02:48:36 systemd-networkd[2706]: veth99: NDISC: Started IPv6 Router Solicitation client
Nov 16 02:48:36 systemd-networkd[2706]: veth99: IPv6 Router Discovery is configured and started.
Nov 16 02:48:36 systemd-networkd[2706]: veth99: NDISC: Sent Router Solicitation, next solicitation in 3s
Nov 16 02:48:36 systemd-networkd[2706]: veth99: NDISC: Received Router Advertisement from fe80::1034:56ff:fe78:9abd: flags=0xc0(managed, other), preference=medium, lifetime=30min
Nov 16 02:48:36 systemd-networkd[2706]: veth99: NDISC: Invoking callback for 'router' event.
Nov 16 02:48:36 systemd-networkd[2706]: veth99: link_check_ready(): dynamic addressing protocols are enabled but none of them finished yet.
Nov 16 02:48:36 systemd-networkd[2706]: veth99: DHCPv6 client: Starting in Solicit mode
Nov 16 02:48:36 systemd-networkd[2706]: veth99: DHCPv6 client: State changed: stopped -> solicitation
Nov 16 02:48:36 systemd-networkd[2706]: veth99: Acquiring DHCPv6 lease on NDisc request
Nov 16 02:48:36 systemd-networkd[2706]: veth99: DHCPv6 client: Sent Solicit
Nov 16 02:48:36 systemd-networkd[2706]: veth99: DHCPv6 client: Next retransmission in 1s
Nov 16 02:48:37 systemd-networkd[2706]: veth99: DHCPv6 client: Sent Solicit
Nov 16 02:48:37 systemd-networkd[2706]: veth99: DHCPv6 client: Next retransmission in 1s
Nov 16 02:48:39 systemd-networkd[2706]: veth99: NDISC: Received Neighbor Advertisement from fe80::1034:56ff:fe78:9abd: Router=yes, Solicited=yes, Override=no
Nov 16 02:48:39 systemd-networkd[2706]: veth99: NDISC: Invoking callback for 'neighbor' event.
Nov 16 02:48:39 systemd-networkd[2706]: veth99: DHCPv6 client: Processed Reply message
Nov 16 02:48:39 systemd-networkd[2706]: veth99: DHCPv6 client: T1 expires in 50s
Nov 16 02:48:39 systemd-networkd[2706]: veth99: DHCPv6 client: T2 expires in 55s
Nov 16 02:48:39 systemd-networkd[2706]: veth99: DHCPv6 client: Valid lifetime expires in 2min
Nov 16 02:48:39 systemd-networkd[2706]: veth99: DHCPv6 client: State changed: solicitation -> bound
Nov 16 02:48:39 systemd-networkd[2706]: veth99: DHCPv6 address 2600::15/128 (valid for 1min 59s, preferred for 1min 59s)
Nov 16 02:48:41 systemd-networkd[2706]: veth99: Received updated DHCPv6 address (configured): 2600::15/128 (valid for 1min 58s, preferred for 1min 58s), flags: no-prefixroute, scope: global
Nov 16 02:48:41 systemd-networkd[2706]: veth99: DHCPv6 addresses and routes set.
Nov 16 02:48:41 systemd-networkd[2706]: veth99: link_check_ready(): IPv4LL:no DHCPv4:no DHCPv6:yes DHCP-PD:no NDisc:no
Nov 16 02:48:41 systemd-networkd[2706]: veth99: State changed: configuring -> configured
```
The interface veth99 entered the configured state after 5 seconds, but
at the same time, the `wait_online()` in the test script considered the
test failed.
The function `wait_online()` first invokes
`systemd-networkd-wait-online` with `--timeout=20`, then check setup
states of interfaces with 5 seconds timeout. So, the failure suggests
that `systemd-networkd-wait-online` finishes immediately, as the state
file was not updated when it is invoked, and thus it handles the
interface veth99 already in the configured state.

11 months agopo: Translated using Weblate (Slovenian)
Martin Srebotnjak [Wed, 20 Nov 2024 19:13:08 +0000 (20:13 +0100)] 
po: Translated using Weblate (Slovenian)

Currently translated at 100.0% (257 of 257 strings)

Co-authored-by: Martin Srebotnjak <miles@filmsi.net>
Translate-URL: https://translate.fedoraproject.org/projects/systemd/main/sl/
Translation: systemd/main

11 months agopo: Translated using Weblate (Finnish)
Jiri Grönroos [Wed, 20 Nov 2024 19:13:07 +0000 (20:13 +0100)] 
po: Translated using Weblate (Finnish)

Currently translated at 100.0% (257 of 257 strings)

Co-authored-by: Jiri Grönroos <jiri.gronroos@iki.fi>
Translate-URL: https://translate.fedoraproject.org/projects/systemd/main/fi/
Translation: systemd/main

11 months agopo: Translated using Weblate (Ukrainian)
Dmytro Markevych [Wed, 20 Nov 2024 19:13:07 +0000 (20:13 +0100)] 
po: Translated using Weblate (Ukrainian)

Currently translated at 100.0% (257 of 257 strings)

Co-authored-by: Dmytro Markevych <hotr1pak@gmail.com>
Translate-URL: https://translate.fedoraproject.org/projects/systemd/main/uk/
Translation: systemd/main

11 months agopo: Translated using Weblate (French)
Léane GRASSER [Wed, 20 Nov 2024 19:13:07 +0000 (20:13 +0100)] 
po: Translated using Weblate (French)

Currently translated at 100.0% (257 of 257 strings)

Co-authored-by: Léane GRASSER <leane.grasser@proton.me>
Translate-URL: https://translate.fedoraproject.org/projects/systemd/main/fr/
Translation: systemd/main

11 months agotest-network: actually check metric and preference 35255/head
Yu Watanabe [Wed, 20 Nov 2024 18:43:32 +0000 (03:43 +0900)] 
test-network: actually check metric and preference

Otherwise, nexthop ID may contain e.g. 300, then
===
AssertionError: '300' unexpectedly found in
'default nhid 3860882700 via fe80::1034:56ff:fe78:9a99 proto ra metric 512 expires 1798sec pref high\n
 default nhid 2639230080 via fe80::1034:56ff:fe78:9a98 proto ra metric 2048 expires 1798sec pref low'
===

11 months agonetwork: update state files before replying bus method
Yu Watanabe [Wed, 20 Nov 2024 13:30:55 +0000 (22:30 +0900)] 
network: update state files before replying bus method

Follow-up for 2b07a3211ba8b1b81d6cebb9650d5cb24554b08a.

11 months agokillall: gracefully handle processes inserted into containers via nsenter -a
Lennart Poettering [Wed, 20 Nov 2024 11:02:46 +0000 (12:02 +0100)] 
killall: gracefully handle processes inserted into containers via nsenter -a

"nsenter -a" doesn't migrate the specified process into the target
cgroup (it really should). Thus the cgroup will remain in a cgroup
that is (due to cgroup ns) outside our visibility. The kernel will
report the cgroup path of such cgroups as starting with "/../". Detect
that and print a reasonably error message instead of trying to resolve
that.

11 months agoprocess-util: refuse to operate on remote PidRef
Mike Yuan [Wed, 20 Nov 2024 14:11:03 +0000 (15:11 +0100)] 
process-util: refuse to operate on remote PidRef

Follow-up for 7e3e540b88db5546d0c63103619d96b033871b7b

11 months agocryptenroll: fix typo
Antonio Alvarez Feijoo [Wed, 20 Nov 2024 14:39:46 +0000 (15:39 +0100)] 
cryptenroll: fix typo

11 months agotest-audit-util: Migrate to new assertion macros
Daan De Meyer [Wed, 20 Nov 2024 13:44:38 +0000 (14:44 +0100)] 
test-audit-util: Migrate to new assertion macros

11 months agocore/exec-invoke: suppress placeholder home only in build_environment() (#35219)
Yu Watanabe [Wed, 20 Nov 2024 08:34:25 +0000 (17:34 +0900)] 
core/exec-invoke: suppress placeholder home only in build_environment() (#35219)

Alternative to https://github.com/systemd/systemd/pull/34789
Closes #34789

11 months agounits: add initrd directory to list of conditions for systemd-confext
Luca Boccassi [Tue, 19 Nov 2024 23:34:00 +0000 (23:34 +0000)] 
units: add initrd directory to list of conditions for systemd-confext

systemd-sysext has the same check, but it was forgotten for confexts.
Needed to activate confexts from the ESP in the initrd.

11 months agohwdb: add new Cube Mix Plus (i18D) rotation info
Xuanjun Wen [Tue, 19 Nov 2024 17:20:01 +0000 (01:20 +0800)] 
hwdb: add new Cube Mix Plus (i18D) rotation info

Added rotation information for the new version of Cube Mix Plus (i18D).

11 months agosd-varlink: mark sd_varlink_server_{ref,unref} as _public_ (#35241)
Mike Yuan [Tue, 19 Nov 2024 20:21:15 +0000 (21:21 +0100)] 
sd-varlink: mark sd_varlink_server_{ref,unref} as _public_ (#35241)

Co-authored-by: Thorsten Kukuk <kukuk@suse.com>
11 months agomachine: add debug for systemd-nspawn@.service 35117/head
Ivan Kruglov [Tue, 19 Nov 2024 18:12:32 +0000 (19:12 +0100)] 
machine: add debug for systemd-nspawn@.service

11 months agomachine: increase timeouts in attempt to fix #35115
Ivan Kruglov [Mon, 11 Nov 2024 14:13:55 +0000 (15:13 +0100)] 
machine: increase timeouts in attempt to fix #35115

11 months agouser-record: don't synthesize default list of self-modfiable fields for non-regular...
Lennart Poettering [Tue, 19 Nov 2024 13:32:21 +0000 (14:32 +0100)] 
user-record: don't synthesize default list of self-modfiable fields for non-regular users. (#35133)

A follow-up for a192250eda1e5cc1f8fc799cf9b85d37e7fa0519

/cc @AdrianVovk

11 months agoman/kernel-command-line: fix typo
Antonio Alvarez Feijoo [Tue, 19 Nov 2024 12:36:38 +0000 (13:36 +0100)] 
man/kernel-command-line: fix typo

11 months agonetwork/ndisc: process zero lifetime options at first (#35212)
Luca Boccassi [Tue, 19 Nov 2024 12:42:03 +0000 (12:42 +0000)] 
network/ndisc: process zero lifetime options at first (#35212)

Fixes two issues reported at #33468.

11 months agocryptenroll: show better log message if slot to wipe does not exist
Antonio Alvarez Feijoo [Tue, 19 Nov 2024 10:26:49 +0000 (11:26 +0100)] 
cryptenroll: show better log message if slot to wipe does not exist

```
$ systemd-cryptenroll /dev/vda3
SLOT TYPE
   0 password
$ systemd-cryptenroll --wipe-slot 1 /dev/vda3
Failed to wipe slot 1, continuing: No such file or directory
```

11 months agoupdate TODO
Lennart Poettering [Tue, 19 Nov 2024 10:03:02 +0000 (11:03 +0100)] 
update TODO