systemd/.git
12 months agoman: tone down claims on processes having exited already in ExecStop=
Lennart Poettering [Tue, 5 Nov 2024 12:51:00 +0000 (13:51 +0100)]
man: tone down claims on processes having exited already in ExecStop=

Processes can easily survive the first kill operation we execute, hence
we shouldn't make strong claims about them having exited already. Let's
just say "likely" hence.

Fixes: #15032
(cherry picked from commit ac804bc2f8d814d2afcdccd88f7469ac320da1c8)

12 months agoman: document that .path units don't care for hidden files
Lennart Poettering [Tue, 5 Nov 2024 12:36:19 +0000 (13:36 +0100)]
man: document that .path units don't care for hidden files

Fixes: #32751
(cherry picked from commit 5adc4337996f561df43c617e37ef4b157073b2e6)

12 months agoman: document that PrivateTmp= is unaffected by ProtectSystem=strict
Lennart Poettering [Tue, 5 Nov 2024 12:33:53 +0000 (13:33 +0100)]
man: document that PrivateTmp= is unaffected by ProtectSystem=strict

Fixes: #33130
(cherry picked from commit b71173709651102081c9d8c6d6e3d2a6ef5cf17e)

12 months agoman: don't claim SELinuxContext= only worked in the system service manager
Lennart Poettering [Tue, 5 Nov 2024 12:15:36 +0000 (13:15 +0100)]
man: don't claim SELinuxContext= only worked in the system service manager

Fixes: #34840
(cherry picked from commit ecbe9ae5a0952d968c57e59dda4a42b72e5b5e6d)

12 months agoman: document the timeout applied to /usr/lib/systemd/system-shutdown/ drop-in binaries
Lennart Poettering [Tue, 5 Nov 2024 12:12:20 +0000 (13:12 +0100)]
man: document the timeout applied to /usr/lib/systemd/system-shutdown/ drop-in binaries

Fixes: #34949
(cherry picked from commit af080967bad2524d2fa2fef6ba5e9fcf18d3017e)

12 months agotest: delete /swapfile after swapoff
Luca Boccassi [Tue, 5 Nov 2024 18:15:36 +0000 (18:15 +0000)]
test: delete /swapfile after swapoff

[   23.608342] TEST-55-OOMD.sh[689]: + btrfs filesystem mkswapfile -s 64M /swapfile
[   23.651930] TEST-55-OOMD.sh[704]: ERROR: cannot create new swapfile: File exists

(cherry picked from commit 78b032d727e8f9e925c10c6617a1e409307ffc24)

12 months agorun: handle gracefully if we can't find binary client-side due to perms
Lennart Poettering [Tue, 5 Nov 2024 10:54:14 +0000 (11:54 +0100)]
run: handle gracefully if we can't find binary client-side due to perms

Fixes: #35022
(cherry picked from commit 9810899ef2f28fbb42cf659e6892b1a5074cfc83)

12 months agoresolved: log error messages for openssl/gnutls context creation
Zbigniew Jędrzejewski-Szmek [Sun, 3 Nov 2024 11:58:12 +0000 (12:58 +0100)]
resolved: log error messages for openssl/gnutls context creation

In https://bugzilla.redhat.com/show_bug.cgi?id=2322937 we're getting
an error message:
Okt 29 22:21:03 fedora systemd-resolved[29311]: Could not create manager: Cannot allocate memory
I expect that this actually comes from dnstls_manager_init(), the
openssl version. But without real logs it's hard to know for sure.

Use EIO instead of ENOMEM, because the problem is unlikely to be actually
related to memory.

(cherry picked from commit ee95e86ae163e436384f1b782a77a7e18deba890)

12 months agoci: add coverage for builds without sd-boot
Luca Boccassi [Mon, 4 Nov 2024 19:43:02 +0000 (19:43 +0000)]
ci: add coverage for builds without sd-boot

This should catch compilation issues such as:
https://github.com/systemd/systemd/pull/35014

(cherry picked from commit 8a3ac7afa6be8447061dc6cbd98830bb332fed28)

12 months agotest: set nullglob to avoid failure when building without sd-boot
Luca Boccassi [Mon, 4 Nov 2024 20:26:34 +0000 (20:26 +0000)]
test: set nullglob to avoid failure when building without sd-boot

2024-11-04T20:13:17.3258095Z + for loader in build/src/boot/efi/*{.efi,.efi.stub}
2024-11-04T20:13:17.3258275Z ++ sbverify --list 'build/src/boot/efi/*.efi'
2024-11-04T20:13:17.3258525Z + [[ Error reading file build/src/boot/efi/*.efi: No such file or directory
2024-11-04T20:13:17.3258952Z Can't open image build/src/boot/efi/*.efi != \N\o\ \s\i\g\n\a\t\u\r\e\ \t\a\b\l\e\ \p\r\e\s\e\n\t ]]
(cherry picked from commit 441922336baf4e302e3b88849c5a08c73b602044)

12 months agotest: fix tool name in comment
Luca Boccassi [Mon, 4 Nov 2024 20:22:01 +0000 (20:22 +0000)]
test: fix tool name in comment

(cherry picked from commit c53df275d512a219806a181de3bd57f02dc38dab)

12 months agopcrlock: Pad pe hash to a multiple of 8 bytes
Daan De Meyer [Sun, 3 Nov 2024 20:45:29 +0000 (21:45 +0100)]
pcrlock: Pad pe hash to a multiple of 8 bytes

All other tools (sbsigntools, osslsigncode, sbctl, goblin) do this
as well so let's follow suite.

(cherry picked from commit e37701a8cd2db1e67d28bcf337467d8efc6de41e)

12 months agotest-dhcp-client: utilize log_info instead of printf
Colin Foster [Fri, 1 Nov 2024 20:39:11 +0000 (15:39 -0500)]
test-dhcp-client: utilize log_info instead of printf

log_info appears to be the preferred method to convey information from
tests. Convert all the printfs to log_info to follow this standard.

(cherry picked from commit 38557d9ffbc6351b8980faf90d54619790436d43)

12 months agoudev-builtin-path_id: SAS wide ports must have num_phys > 1
Martin Wilck [Wed, 30 Oct 2024 15:57:39 +0000 (16:57 +0100)]
udev-builtin-path_id: SAS wide ports must have num_phys > 1

Some kernel SAS drivers (e.g. smartpqi) expose ports with num_phys = 0. udev
shouldn't treat these ports as wide ports.  SAS wide ports always have
num_phys > 1. See comments for sas_port_add_phy() in the kernel sources.

Sample data from a smartpqi system to illustrate the issue below.
Here the phy device is attached to port 0:0, which has no end devices attached
and the SAS end device (where sda is attached) is associated with SAS
port 0:1, which has no associated phy device. Thus num_phys for port-0:1 is 0.
This is arguably wrong, but it's how smartpqi has always set up its devices in
sysfs.

/sys/class/sas_phy/phy-0:0 -> ../../devices/pci0000:46/0000:46:02.0/0000:47:00.0/host0/scsi_host/host0/phy-0:0/sas_phy/phy-0:0
/sys/devices/pci0000:46/0000:46:02.0/0000:47:00.0/host0/scsi_host/host0/port-0:0/phy-0:0 -> ../phy-0:0
/sys/devices/pci0000:46/0000:46:02.0/0000:47:00.0/host0/scsi_host/host0/phy-0:0/port -> ../port-0:0

/sys/class/sas_device/end_device-0:1 -> ../../devices/pci0000:46/0000:46:02.0/0000:47:00.0/host0/scsi_host/host0/port-0:1/end_device-0:1/sas_device/end_device-0:1
/sys/class/block/sda -> ../../devices/pci0000:46/0000:46:02.0/0000:47:00.0/host0/scsi_host/host0/port-0:1/end_device-0:1/target0:0:0/0:0:0:0/block/sda

Signed-off-by: Martin Wilck <mwilck@suse.com>
(cherry picked from commit 7f6674624eedc95db2a9877edc6ff424f4453531)

12 months agoTEST-64-UDEV-STORAGE: Don't hardcode device name in long-sysfs-path test
Daan De Meyer [Fri, 1 Nov 2024 20:27:52 +0000 (21:27 +0100)]
TEST-64-UDEV-STORAGE: Don't hardcode device name in long-sysfs-path test

There's no guarantee our device will be named /dev/vda, so give it
a serial so we can query for its devname inside the test.

(cherry picked from commit 2ec809dd3baf39b83b8f581e7ea837b9732f9964)

12 months agoTEST-17-UDEV: Don't hardcode root device name
Daan De Meyer [Fri, 1 Nov 2024 20:27:08 +0000 (21:27 +0100)]
TEST-17-UDEV: Don't hardcode root device name

There's no guarantee the root device will be /dev/sda, so let's use
bootctl to get the actual path instead of harcoding it.

(cherry picked from commit 29a8e71d9c0858aef502f091a0ef58d5569b1c70)

12 months agopam: quiet a spurious debug message
Ronan Pigott [Sat, 2 Nov 2024 02:44:18 +0000 (19:44 -0700)]
pam: quiet a spurious debug message

This singular debug message gets printed even if debug is not enabled.
Quiet this message when debug is not enabled for consistency.

(cherry picked from commit f4092cb9745cc2fc1f889eeaffa5cb5133969d85)

12 months agoposix_spawn_wrapper: do not set POSIX_SPAWN_SETSIGDEF flag
Mike Gilbert [Thu, 24 Oct 2024 16:24:35 +0000 (12:24 -0400)]
posix_spawn_wrapper: do not set POSIX_SPAWN_SETSIGDEF flag

Setting this flag is a noop without a corresponding call to
posix_spawnattr_setsigdefault.

If we call posix_spawnattr_setsigdefault with a full signal set,
it causes glibc's posix_spawn implementation to call sigaction 63 times,
once for each signal. That seems wasteful.

This feature is really only useful for signals which have their
disposition set to SIG_IGN. Otherwise the dispostion gets set to
SIG_DFL automatically, either by clone(CLONE_CLEAR_SIGHAND) or the
subsequent execve.

As far as I can tell, systemd does not have any signals set to SIG_IGN
under normal operating conditions.

(cherry picked from commit ff94426f8a2d6cd4ea2e370835db152917a1684e)

12 months agocore: make mount(8) and swapon(8) inherit SMACK label from systemd
Łukasz Stelmach [Tue, 29 Oct 2024 14:53:45 +0000 (15:53 +0100)]
core: make mount(8) and swapon(8) inherit SMACK label from systemd

By default mount(8), umount(8), swapon(8) and swapoff(8) should run with
with the SMACK label inherited from systemd rather than the default one
meant for services.

Fixes: aa5ae9711ef3cd0c69b7fcfbd65bca05fb704a8a
Follow-up-for: 20bbf5ee4c6c80599a91e7a4b7474e931a27db4a
(cherry picked from commit 8144537a81c7a815af3d4c63cd8545ee17b2715d)

12 months agotest-network: add test for DHCPv4 address removal on stop
Yu Watanabe [Wed, 23 Oct 2024 20:25:41 +0000 (05:25 +0900)]
test-network: add test for DHCPv4 address removal on stop

For issue #34837.

(cherry picked from commit 58a011ba4865d083bffdebe33126637bc5bba511)

12 months agonetwork: process queued remove requests before networkd is stopped
Yu Watanabe [Wed, 23 Oct 2024 19:40:45 +0000 (04:40 +0900)]
network: process queued remove requests before networkd is stopped

This makes networkd process all queued remove requests when a
terminating or restarting signal is received. Otherwise, e.g. DHCPv4
address will not be removed on stop, especially when
KeepConfiguration=no.

Fixes a bug introduced by 85a6f300c14d75d161cbfdb3eaf5af9594400ecd and
its subsequent commits.

Fixes #34837.

Co-authored-by: Will Fancher <elvishjerricco@gmail.com>
(cherry picked from commit db68e990467ffea2b6655e4726856ca7eb32b38d)

12 months agotest-sbat: separate the two sbat sections
Zbigniew Jędrzejewski-Szmek [Wed, 30 Oct 2024 14:19:24 +0000 (15:19 +0100)]
test-sbat: separate the two sbat sections

(cherry picked from commit 07000101eb9529c2d6b5f5402c9fe643c5f98420)

12 months agocryptenroll,homectl,journalctl: adjust messages before qrcodes
Zbigniew Jędrzejewski-Szmek [Mon, 28 Oct 2024 12:59:05 +0000 (13:59 +0100)]
cryptenroll,homectl,journalctl: adjust messages before qrcodes

Users will generally know what a qrcode is, so let's not treat them as dumb and
explain that it can be scanned. OTOH, we should say what the qrcode contains
and it is useful to give a hint why the users would want to scan it. Reword
messages accordingly.

(Also, don't say "to your phone", when somebody might be using a stolen phone,
or something else then a phone.)

(cherry picked from commit 10faa40ba781cf499258a3b37de02dd643822dc6)

12 months agobsod: make message for qrcode more useful
Zbigniew Jędrzejewski-Szmek [Mon, 28 Oct 2024 12:51:25 +0000 (13:51 +0100)]
bsod: make message for qrcode more useful

People know what a qrcode is. We don't need to tell them to scan it.
Instead, we should say what the code contains.

While at it, rename "stream" to "f" in line with the usual style.

(cherry picked from commit abf1cae0a75ca07f09afbb4eaa9f11fc429b1d02)

12 months agodocs: Align some comments in HACKING.md
Daan De Meyer [Wed, 30 Oct 2024 10:16:13 +0000 (11:16 +0100)]
docs: Align some comments in HACKING.md

(cherry picked from commit a33f453702e40f4b52226c6be1a01200aee69470)

12 months agoanalyze: Add times in seconds for Activating and Activated in tooltip
hugo303 [Fri, 25 Oct 2024 10:15:02 +0000 (12:15 +0200)]
analyze: Add times in seconds for Activating and Activated in tooltip

Print the times in seconds in the tooltip to remove the need to count
and trying to follow the lines in the svg diagram in order to see at
what times these events happen.

(cherry picked from commit f172dfddde3379319ee3a02666a7ecf11a5711f4)

12 months agoTEST-80-NOTIFYACCESS: don't specify --pid= if MAINPID= is provided explicitly
Mike Yuan [Tue, 29 Oct 2024 17:35:50 +0000 (18:35 +0100)]
TEST-80-NOTIFYACCESS: don't specify --pid= if MAINPID= is provided explicitly

Otherwise, with recent additions, the MAINPIDFDID= generated by
systemd-notify would mismatch with overridden MAINPID=.

(cherry picked from commit c3ecb747f1e35f609f15fc94ad4d5e5ca0bda4a2)

12 months agoqrcode-util: avoid memleak in error path
Zbigniew Jędrzejewski-Szmek [Mon, 28 Oct 2024 12:45:40 +0000 (13:45 +0100)]
qrcode-util: avoid memleak in error path

(cherry picked from commit 439306da8b3962f683f5359c461d1669c070f377)

12 months agotest-terminal-util: print value of colors_enabled()
Zbigniew Jędrzejewski-Szmek [Mon, 28 Oct 2024 12:39:36 +0000 (13:39 +0100)]
test-terminal-util: print value of colors_enabled()

This makes it easier to diagnose why colors are disabled.

(cherry picked from commit b137b2979868e2de5fb5c26e90bacee33597b8e7)

12 months agobsod: do not check for color support
Zbigniew Jędrzejewski-Szmek [Mon, 28 Oct 2024 12:38:58 +0000 (13:38 +0100)]
bsod: do not check for color support

When invoked on a running system, bsod would not print the qrcode.
The check for "color support" on stdout is pointless, since we're not
printing to stdout but to a terminal fd that is opened separately.

(cherry picked from commit 5a64c86936477ecea5cc1fb8dbc79faf522cf370)

12 months agoqrcode-util: add debug message to show why a qrcode wasn't printed
Zbigniew Jędrzejewski-Szmek [Mon, 28 Oct 2024 12:15:32 +0000 (13:15 +0100)]
qrcode-util: add debug message to show why a qrcode wasn't printed

(cherry picked from commit f0764b98e5c136cb948a8034949064f610acca24)

12 months agosysv-generator: break long message into lines
Zbigniew Jędrzejewski-Szmek [Fri, 18 Oct 2024 17:41:25 +0000 (19:41 +0200)]
sysv-generator: break long message into lines

The journal handles multi-line messages nicely, and they are easier
to read. Drop the recycling symbol, there is no circular process here,
we go from a to b and never back to a again.

(cherry picked from commit bb56c27fc81da2777cd7064a0b88ca011eced509)

12 months agosd-id128: mark functions as const, not pure
Zbigniew Jędrzejewski-Szmek [Mon, 28 Oct 2024 08:20:32 +0000 (09:20 +0100)]
sd-id128: mark functions as const, not pure

We would need to use pure if the funtion was getting pointers and
dereferencing them. But sd128_t is a structure and those functions
only access the parameters of the call.

(cherry picked from commit dc32b09b70c9bb20821df92ac82ace83d8a968e2)

12 months agosd-common: add __const__
Zbigniew Jędrzejewski-Szmek [Mon, 28 Oct 2024 08:13:10 +0000 (09:13 +0100)]
sd-common: add __const__

const is stronger than pure, see
https://gcc.gnu.org/onlinedocs/gcc/Common-Function-Attributes.html#index-pure-function-attribute
and
https://gcc.gnu.org/onlinedocs/gcc/Common-Function-Attributes.html#index-const-function-attribute.

(cherry picked from commit 955c51c087f1fb6d0b7a0091db943ad05ba3095e)

12 months agomeson: add loongarch64's definition to cpu_arch_defines
Zhou Qiankang [Mon, 28 Oct 2024 04:47:20 +0000 (12:47 +0800)]
meson: add loongarch64's definition to cpu_arch_defines

The default definition to add is `-D__loongarch64__`, which is not searched in [bpf_tracing.h](https://github.com/libbpf/libbpf/blob/09b9e83102eb8ab9e540d36b4559c55f3bcdb95d/src/bpf_tracing.h#L68)

This may avoid `error: Must specify a BPF target arch via __TARGET_ARCH_xxx` in loongarch64

Signed-off-by: Zhou Qiankang <wszqkzqk@qq.com>
(cherry picked from commit 85d0aff84c83182875bc564e295978efd76ab905)

12 months agocore/service: use log_unit_* where appropriate
Mike Yuan [Sat, 26 Oct 2024 15:38:06 +0000 (17:38 +0200)]
core/service: use log_unit_* where appropriate

(cherry picked from commit 1e8f0beee4272ddc8b25dfa9af8e54bafc4c061a)

12 months agosd-event: fix memleak when built without assertion
Yu Watanabe [Fri, 25 Oct 2024 18:25:26 +0000 (03:25 +0900)]
sd-event: fix memleak when built without assertion

Fixes a bug introduced by baf3fdec27f0b3a1f3d39c7def2a778824cbee51.

This also adds several assertions at the beginning of the function.

Fixes #34899.

(cherry picked from commit 5dc0668802cd07cdca2dc5bda52cc1e63b57f145)

12 months agotest: add quick test to verify the PAM stack really ran in all run0 modes of operation
Lennart Poettering [Thu, 24 Oct 2024 20:51:49 +0000 (22:51 +0200)]
test: add quick test to verify the PAM stack really ran in all run0 modes of operation

(cherry picked from commit f515ea1cd4981ceb1a2c85d676f279db2972283f)

12 months agocore: make sure that if PAMName= is set we always do the full user changing even...
Lennart Poettering [Thu, 24 Oct 2024 08:52:56 +0000 (10:52 +0200)]
core: make sure that if PAMName= is set we always do the full user changing even if no user is specified explicitly

When PAMName= is set this should be enough to go through our entire user
changing story, so that PAM is definitely run, and environment variables
definitely pulled in and so on.

Previously, it would happen that under some circumstances we might no do
this when transitioning from root to root itself even though PAM was
enabled.

Fixes: #34682
(cherry picked from commit e4b4d9cc7adf245950e8676be0e0f4a813069500)

12 months agoman/veritysetup-generator: document veritytab kernel command line option
Antonio Alvarez Feijoo [Thu, 24 Oct 2024 08:07:28 +0000 (10:07 +0200)]
man/veritysetup-generator: document veritytab kernel command line option

(cherry picked from commit e98e3f856d56bfb8dd8883f9e570bb56c4b1e3b2)

12 months agoman: fix links to veritysetup(8)
Antonio Alvarez Feijoo [Thu, 24 Oct 2024 07:54:48 +0000 (09:54 +0200)]
man: fix links to veritysetup(8)

(cherry picked from commit dcbfc7872e96c20b4d29b527d3467f547f6971d0)

12 months agocore: don't forget about fallback_smack_process_label
Łukasz Stelmach [Wed, 23 Oct 2024 10:58:36 +0000 (12:58 +0200)]
core: don't forget about fallback_smack_process_label

Call setup_smack() also when only fallback_smack_process_label is set.

Fixes: 75689fb2d41f
(cherry picked from commit 20bbf5ee4c6c80599a91e7a4b7474e931a27db4a)

12 months agologind: allow read/write to char-hvc devices
Luca Boccassi [Tue, 22 Oct 2024 16:04:41 +0000 (17:04 +0100)]
logind: allow read/write to char-hvc devices

virtio console uses /dev/hvc* so we need access to write wall
messages

(cherry picked from commit 5ff6841c2378ed83e645681cbd4ee145f68d72b7)

12 months agoTEST-02-UNITTESTS: reuse $TEST_MATCH_SUBTEST to specify unit tests to be run
Yu Watanabe [Tue, 22 Oct 2024 06:32:12 +0000 (15:32 +0900)]
TEST-02-UNITTESTS: reuse $TEST_MATCH_SUBTEST to specify unit tests to be run

Then, we can easily test specific unit tests in qemu or container.

(cherry picked from commit aa7507ea4a57508da0f2a7d1ae6e139793c5694a)

12 months agotest-execute: update permission of credstore
Yu Watanabe [Mon, 5 Jun 2023 05:18:47 +0000 (14:18 +0900)]
test-execute: update permission of credstore

Follow-up for 40fb9eebbc075ce1e63100386d2c5f177ad7d738.

(cherry picked from commit c443f6924fa3b02113da2536dd816a15ee708510)

12 months agotest: CET/EET are deprecated, use Europe/Berlin and Kyiv
Luca Boccassi [Mon, 21 Oct 2024 16:31:09 +0000 (17:31 +0100)]
test: CET/EET are deprecated, use Europe/Berlin and Kyiv

The links moved to the legacy dataset so they won't be available by
default, so stop using them and just use the city ones instead

(cherry picked from commit aa077884c13769ae3bd6aa98978b4ac9e64b5365)

12 months agotest-dhcp6: terminate fqdn option
Ronan Pigott [Thu, 25 Jan 2024 00:53:35 +0000 (17:53 -0700)]
test-dhcp6: terminate fqdn option

The encoded fqdn in this option must be properly terminated. We will
soon validate that this field is correctly encoded, so correct it in the
test.

(cherry picked from commit 2d9822b634680f1be1d20920aceddac76de110eb)

12 months agopcrlock: Take VirtualSize > SizeOfRawData into account
Daan De Meyer [Mon, 7 Oct 2024 15:39:27 +0000 (17:39 +0200)]
pcrlock: Take VirtualSize > SizeOfRawData into account

If VirtualSize > SizeOfRawData, measure extra zeros to take into
account the extra zeros also measured by the stub.

(cherry picked from commit b53f2d5ed8ad0e537e9086daf84f9c2bf69fb72b)

12 months agotest: exercise bypass mode on the sd-resolved stub
Ronan Pigott [Fri, 11 Oct 2024 05:40:03 +0000 (22:40 -0700)]
test: exercise bypass mode on the sd-resolved stub

A basic test will verify that we provide the right flags.

(cherry picked from commit b7b1c50c6afd2ab8c2c779df9b73c6dbbc773f70)

12 months agoresolved: when adding names to packet fails, remove them from label compression hash...
Lennart Poettering [Thu, 12 Sep 2024 15:05:24 +0000 (17:05 +0200)]
resolved: when adding names to packet fails, remove them from label compression hash table again

let's make sure we undo any pollution of the label compression hash
table.

Fixes: #33671
(cherry picked from commit 360105f1e748148ba17bdb3f47525f01aba4127f)

12 months agodns-domain: tweak hash table comparison function for DNS names
Lennart Poettering [Thu, 12 Sep 2024 15:00:04 +0000 (17:00 +0200)]
dns-domain: tweak hash table comparison function for DNS names

Currently, when comparing two DNS names when storing them in a
hashtable, and the DNS names are not actually valid we'll compare the
error codes.

This is not very smart however, since this means two invalid DNS names
that happen to be equally "invalid" will be considered identical, even
if their strings are entirely different.

Let's find a better solution for this niche case: let's simple compare
the domains as strings.

This matters in case of DNS label compression: if we already added added
an invalid DNS name into the label compression hash table, and lookup
any other invalid DNS name, this lookup will likely return what the
earlier one already returned, and that's confusing.

(cherry picked from commit 8ed2c62d46f93c2117d65a908c316a381073af16)

12 months agoresolved: update condition for caching full packets
Ronan Pigott [Sat, 12 Oct 2024 17:00:56 +0000 (10:00 -0700)]
resolved: update condition for caching full packets

Previously a full packet was cached only if the CD bit was set, but this
no longer corresponds to the cases where bypass is enabled.

Update the cache to retain a full packet in the cases where it might
actually be useful.

(cherry picked from commit fa02d04ee9a4787c2c39032185b6fe3e7146b26b)

12 months agoresolved: enable CD bit without DO set
Ronan Pigott [Sat, 19 Oct 2024 04:10:57 +0000 (21:10 -0700)]
resolved: enable CD bit without DO set

This is useful for a validating resolver to indicate to a non-validating
resolver when checking was disabled for the query. This matches the
behavior of the major public resovlers in response to queries with CD bu
tnot DO set.

(cherry picked from commit 36074e01499ea4fc89a0c642cef83bd650014e5a)

12 months agoresolved: authenticate bypass queries
Ronan Pigott [Mon, 7 Oct 2024 18:05:18 +0000 (11:05 -0700)]
resolved: authenticate bypass queries

Following 13e15dae9f0b, resolved does not forward the AD bit for bypass
queries, but resolved also didn't do it's own validation, making these
replies appear to never be authentic. We should enable validation for
bypass queries.

Let's disable our own validation when processing a +cd query, and also
ensure that it skips the cache so that we don't accidentally fail to
return inauthentic replies from upstream.

Previously, when we had a bypass transaction without cd, a cached,
authenticated, reply with cd could be served, leaving the cd bit
erroneously set in the reply. Only reply with a CD bit if the client
requested it.

Fixes: 13e15dae9f0b (resolved: clear the AD bit for bypass packets)
(cherry picked from commit 008f23b7c5f255e48c71c3b4da74a85448667817)

12 months agoman/systemd-nspawn: emphasise that user namespaces are strongly recommended
Zbigniew Jędrzejewski-Szmek [Tue, 15 Oct 2024 16:53:00 +0000 (18:53 +0200)]
man/systemd-nspawn: emphasise that user namespaces are strongly recommended

(cherry picked from commit 9b1a5bc365e379b4b13849adacfde3427f55ca38)

12 months agotest: customize /etc/os-release instead of /usr/lib/os-release
Luca Boccassi [Fri, 18 Oct 2024 14:02:03 +0000 (15:02 +0100)]
test: customize /etc/os-release instead of /usr/lib/os-release

As per spec image builders can create a local /etc/os-release
with per-image IDs, so modify that one instead of the original
one in /usr/lib. For example we do this when we build debian
unstable images in mkosi.

(cherry picked from commit 2f6fe4e1131d39fcafa9e00a7902919efb5361e1)

12 months agodissect-image: generate better log message for EUCLEAN dissect error
Lennart Poettering [Fri, 18 Oct 2024 12:16:53 +0000 (14:16 +0200)]
dissect-image: generate better log message for EUCLEAN dissect error

Fixes: #31799
(cherry picked from commit 2186334e00acba6e6c1a4564bce60474eecfbf16)

12 months agodissect-image: uppercase first char of dissect error message systematically
Lennart Poettering [Fri, 18 Oct 2024 12:16:13 +0000 (14:16 +0200)]
dissect-image: uppercase first char of dissect error message systematically

Some of the log message stricts used proper uppercasing, others didn't.
Fix that to make it uniform.

(cherry picked from commit 620a03f669a9075f2d78c2fcf7db45f7046481bc)

12 months agoresolved: refresh resolv.conf files when link goes away
Lennart Poettering [Fri, 18 Oct 2024 09:30:12 +0000 (11:30 +0200)]
resolved: refresh resolv.conf files when link goes away

This might have the effect that some DNS server or search domain
disappears, hence rewrite the relevant files.

See: #27543
(cherry picked from commit 562f7bde8872b4fd03db11bf25c9dd294fd2c186)

12 months agonetworkd: raise limits on number of address 8x
Lennart Poettering [Fri, 18 Oct 2024 07:19:23 +0000 (09:19 +0200)]
networkd: raise limits on number of address 8x

Limits should be enforced, but not in a way real setups collide with
them.

There have been multiple reports that current limits are too low, hence
raise them 8x.

Fixes: #24852
(cherry picked from commit af7674f4ad30e83efc84f04c45f01e6eff137702)

12 months agoGREEDY_REALLOC_APPEND: Make more type safe
Adrian Vovk [Wed, 4 Sep 2024 17:44:26 +0000 (13:44 -0400)]
GREEDY_REALLOC_APPEND: Make more type safe

Previously, GREEDY_REALLOC_APPEND would compile perfectly fine and cause
subtle memory corruption if the caller messes up the type they're passing
in (i.e. by forgetting to pass-by-reference when appending a Type* to an
array of Type*). Now this will lead to compilation failure

(cherry picked from commit fafc3c2d5c7fae6bad0f6dc51611ae9390589ade)

12 months agoudev: consider serial ports as unconfigured only if both port and iomem_base sysattr...
Lennart Poettering [Thu, 17 Oct 2024 07:57:41 +0000 (09:57 +0200)]
udev: consider serial ports as unconfigured only if both port and iomem_base sysattr is zero

Various archs do not know the "io port" concept, hence check for the
"iomem" configuration, too.

Fixes: #34800
(cherry picked from commit 9da582984b632a75f562a304ff96dd1043281430)

12 months agoTEST-60-MOUNT-RATELIMIT: disable journal ratelimiting
Yu Watanabe [Wed, 16 Oct 2024 13:34:48 +0000 (22:34 +0900)]
TEST-60-MOUNT-RATELIMIT: disable journal ratelimiting

Otherwise, journal check in testcase_mount_ratelimit() may fail and we
need to wait 2 minutes.

(cherry picked from commit 73970cdcd1cebe49a2858ce016c0cd4ce6037b2e)

12 months agoTEST-60-MOUNT-RATEMINIT: split into small test cases
Yu Watanabe [Wed, 16 Oct 2024 13:32:09 +0000 (22:32 +0900)]
TEST-60-MOUNT-RATEMINIT: split into small test cases

Then, use run_testcases().

(cherry picked from commit 805d0eed0b51e2ef761fc522f092b71dc028f700)

12 months agoTEST-60-MOUNT-RATELIMIT: wait for mount unit being started or stopped
Yu Watanabe [Wed, 16 Oct 2024 12:17:15 +0000 (21:17 +0900)]
TEST-60-MOUNT-RATELIMIT: wait for mount unit being started or stopped

(cherry picked from commit c5928a768417b298eb2741107fa7492e93d637fc)

12 months agoFix maybe-uninitialized warnings with gcc 14.2
Luca Boccassi [Wed, 16 Oct 2024 10:42:06 +0000 (11:42 +0100)]
Fix maybe-uninitialized warnings with gcc 14.2

../src/resolve/resolved-bus.c: In function ‘call_link_method’:
../src/resolve/resolved-bus.c:1769:16: warning: ‘l’ may be used uninitialized [-Wmaybe-uninitialized]
 1769 |         return handler(message, l, error);
      |                ^~~~~~~~~~~~~~~~~~~~~~~~~~
../src/resolve/resolved-bus.c:1755:15: note: ‘l’ was declared here
 1755 |         Link *l;
      |               ^
../src/resolve/resolved-bus.c: In function ‘bus_method_get_link’:
../src/resolve/resolved-bus.c:1828:13: warning: ‘l’ may be used uninitialized [-Wmaybe-uninitialized]
 1828 |         p = link_bus_path(l);
      |             ^~~~~~~~~~~~~~~~
../src/resolve/resolved-bus.c:1816:15: note: ‘l’ was declared here
 1816 |         Link *l;
      |               ^

(cherry picked from commit 5f911aca8434b4163514019fcb4c1c967a50617c)

12 months agojournalctl: erase verify key before free
Yu Watanabe [Wed, 16 Oct 2024 10:27:36 +0000 (19:27 +0900)]
journalctl: erase verify key before free

Even optarg is erased, copied string was not erased.
Let's erase the copied key for safety.

(cherry picked from commit d0ad4e88d4e6b5e312c359a6505125f7e088f3e3)

12 months agoTEST-55-OOMD: check slice property before stressing slice
Yu Watanabe [Wed, 16 Oct 2024 06:20:08 +0000 (15:20 +0900)]
TEST-55-OOMD: check slice property before stressing slice

(cherry picked from commit a6092b532e1de0e1f9702f432c0c4c7de495bf4f)

12 months agoTEST-55-OOMD: stop test units when unnecessary
Yu Watanabe [Wed, 16 Oct 2024 06:18:41 +0000 (15:18 +0900)]
TEST-55-OOMD: stop test units when unnecessary

Then, sleep becomes not necessary anymore. This greatly improve performance.

(cherry picked from commit 98a0bb8a6487a1bd0767223f3738525c2b3eae52)

12 months agoTEST-55-OOMD: split into small testcases
Yu Watanabe [Wed, 16 Oct 2024 06:04:07 +0000 (15:04 +0900)]
TEST-55-OOMD: split into small testcases

Then, we can run each small test cases separately.

(cherry picked from commit 6bba058076cf4ce91fd470841d5ced47c8d32d6a)

12 months agoTEST-55-OOMD: check global config earlier
Yu Watanabe [Wed, 16 Oct 2024 05:52:49 +0000 (14:52 +0900)]
TEST-55-OOMD: check global config earlier

'Default Memory Pressure Duration' field in oomctl, which can be configured
with DefaultMemoryPressureDurationSec= in oomd.conf, is a global config.
Let's check it earlier.

This also drops unnecessary cleanup at the beginning.

(cherry picked from commit 23fb9b63cd12f2040419dbc846d7a87e5193935c)

12 months agoTEST-55-OOMD: set ManagedOOMMemoryPressure= and friends in a drop-in config
Yu Watanabe [Fri, 11 Oct 2024 07:09:11 +0000 (16:09 +0900)]
TEST-55-OOMD: set ManagedOOMMemoryPressure= and friends in a drop-in config

Fedora and friends has a drop-in config for the settings in
/usr/lib/systemd/user/slice.d/ . Hence, settings in the main .slice may be
overridden. Let's set below in a drop-in with higher decimal prefix.

Also, rename override.conf -> 99-managed-oom-preference.conf for the same reason.

(cherry picked from commit 1473836a1e3c528221439e3b0ce90e729b31adb3)

12 months agoTEST-19-CGROUP: add test cases for IPAddressAllow=/IPAddressDeny=
Yu Watanabe [Wed, 16 Oct 2024 05:31:16 +0000 (14:31 +0900)]
TEST-19-CGROUP: add test cases for IPAddressAllow=/IPAddressDeny=

(cherry picked from commit 5f3cfb9d5ee334c53cc407308ba677401a6ba1cd)

12 months agocore/cgroup: fix IPAddressAllow=/IPAddressDeny= set through DBus
Yu Watanabe [Wed, 16 Oct 2024 03:45:34 +0000 (12:45 +0900)]
core/cgroup: fix IPAddressAllow=/IPAddressDeny= set through DBus

Fixes a regression caused by 84ebe6f01381c21b88e37e856956c9c9ee6781d6 (v250).
Fixes #34773.

(cherry picked from commit 77bbd9f1bd2b01bcb2a49ed42c6dc06613532bcf)

12 months agogpt-auto: remove directory check for ESP mount
Michael Ferrari [Tue, 15 Oct 2024 16:42:20 +0000 (18:42 +0200)]
gpt-auto: remove directory check for ESP mount

Ensure that we always attempt to mount the `ESP` partition to `/boot`
when there is no `XBOOTLDR` partition.

Fixes an issue when booting without a `XBOOTLDR` partition and an empty
root partition, since it would mount the `ESP` partition to `/efi/`
unconditionally causing boot entries to not be under `/boot/` as
recommended by the Boot Loader Specification.

(cherry picked from commit 28dbee46d5bfdcfa601e80e248dbde22b8d65664)

12 months agocore: do not fail if ignorable img.v/ vpick dir is missing
Luca Boccassi [Tue, 15 Oct 2024 13:05:50 +0000 (14:05 +0100)]
core: do not fail if ignorable img.v/ vpick dir is missing

Do not fail if the directory is missing entirely, other than just empty

Follow-up for 00f546e25e8
Follow-up for 5e79dd96a88
Follow-up for 622efc544dc

(cherry picked from commit e1efa9d8044a7564844b8d599eda4b59eb1152e9)

12 months agocore: do not fail if ignorable img.v/ vpick dir is empty
Luca Boccassi [Thu, 12 Sep 2024 16:46:04 +0000 (18:46 +0200)]
core: do not fail if ignorable img.v/ vpick dir is empty

If the vpick directory is configured to be ignored if missing, do not
fail and just skip ahead.

Follow-up for 5e79dd96a88
Follow-up for 622efc544dc

(cherry picked from commit 00f546e25e8e46e5e289e1f2fc2ed3dcc6999223)

12 months agoTEST-74-AUX-UTILS: add tests for 'run0 -D'
John A. Leuenhagen [Tue, 15 Oct 2024 06:00:18 +0000 (02:00 -0400)]
TEST-74-AUX-UTILS: add tests for 'run0 -D'

(cherry picked from commit 4ca75488d935431e772c25eb6cf060ea85553158)

12 months agorun: fix bug causing run0 to ignore `-D /`.
John A. Leuenhagen [Tue, 15 Oct 2024 04:57:52 +0000 (00:57 -0400)]
run: fix bug causing run0 to ignore `-D /`.

Since the root directory was being suppressed to NULL, the subsequent
check would erroneously think that no working directory was specified.
This caused the default working directory to be applied instead.

(cherry picked from commit 0054a2acc3894e45171806bd64887211820b8eaf)

12 months agonetwork/dhcp6: set hostname even if UseAddress=no
Yu Watanabe [Sun, 15 Sep 2024 19:45:13 +0000 (04:45 +0900)]
network/dhcp6: set hostname even if UseAddress=no

Follow-up for f963f8953daeab03b892616ce0c65f7572932187 and
1536b7b2d00819615bf8eba194de7ccd20c3689f.

(cherry picked from commit 8fead9c9e46e5f71ae6f6b038ff7f72c5a13b663)

12 months agoudev: do not try to lock whole block device on remove event
Yu Watanabe [Mon, 14 Oct 2024 21:22:24 +0000 (06:22 +0900)]
udev: do not try to lock whole block device on remove event

As another device may be created with the same device node while udevd
is processing the remove event of the previous owner of the device node.

This also adds comment why we skip watching device node on remove.

(cherry picked from commit e8df18c9e171c87aebb2df8ac3bdd8f116236892)

12 months agodocs: add a missing : character
Gaël PORTAY [Mon, 14 Oct 2024 17:16:07 +0000 (19:16 +0200)]
docs: add a missing : character

This adds the missing colon character to the section systemd-sysusers.

(cherry picked from commit 8ef5ea2bf6d8c9ef12b00ab3838b9af81e30bf5d)

12 months agoman: document preference for secure_getenv() in coding style
Lennart Poettering [Mon, 14 Oct 2024 09:55:59 +0000 (11:55 +0200)]
man: document preference for secure_getenv() in coding style

(cherry picked from commit c9b477415a6293b74df67c8118bafb0ef8662819)

12 months agotest: add test for local outbounds with preferred source address
Yu Watanabe [Sat, 12 Oct 2024 22:24:08 +0000 (07:24 +0900)]
test: add test for local outbounds with preferred source address

(cherry picked from commit 0343ef97644c38e215e97ef9d1b6d942b08bb8ef)

12 months agolocal-addresses: honor RTA_PREFSRC field of gateway
Yu Watanabe [Sat, 12 Oct 2024 21:56:38 +0000 (06:56 +0900)]
local-addresses: honor RTA_PREFSRC field of gateway

Fixes #34739.

(cherry picked from commit 4adf2653e2a68d4d593b46734fd9e24721a8d449)

12 months agolocal-addresses: use FOREACH_ARRAY() macro
Yu Watanabe [Sat, 12 Oct 2024 21:30:58 +0000 (06:30 +0900)]
local-addresses: use FOREACH_ARRAY() macro

(cherry picked from commit 418641f73f78e651c6b0ecd05e055f790cb1eae7)

12 months agoTEST-64-UDEV-STORAGE: insert udevadm settle more
Yu Watanabe [Fri, 11 Oct 2024 18:09:10 +0000 (03:09 +0900)]
TEST-64-UDEV-STORAGE: insert udevadm settle more

Hopefully fixes #34073.

(cherry picked from commit acadc9a84925fa725de1a4dbcfe2355432c0769a)

12 months agoTEST-58-REPART: drop duplicated inclusion of util.sh
Yu Watanabe [Fri, 11 Oct 2024 05:08:49 +0000 (14:08 +0900)]
TEST-58-REPART: drop duplicated inclusion of util.sh

(cherry picked from commit 4ca7b553295315c759a622a47240af56e2d06b46)

12 months agomkosi: replace PackageManagerTrees= with SandboxTrees=
Yu Watanabe [Fri, 11 Oct 2024 06:52:48 +0000 (15:52 +0900)]
mkosi: replace PackageManagerTrees= with SandboxTrees=

(cherry picked from commit 4e3f16b5ef122996f9545d883af74ff1235833eb)

12 months agoukify: Require both key and cert be set in generate_keys
Jörg Behrmann [Fri, 11 Oct 2024 12:34:15 +0000 (14:34 +0200)]
ukify: Require both key and cert be set in generate_keys

If either is None the write in this branch of the code will fail.

(cherry picked from commit 9ba53499b97d59cc57ce763b8ba7f272a876dd25)

12 months agosystemd-update-helper: Show executed commands if debug logging is enabled
Daan De Meyer [Fri, 11 Oct 2024 14:51:04 +0000 (16:51 +0200)]
systemd-update-helper: Show executed commands if debug logging is enabled

(cherry picked from commit 8b8668b9e71837cb541cd432bc37e4c9405e49cd)

12 months agoseccomp: allowlist uretprobe() syscall
Lennart Poettering [Fri, 11 Oct 2024 07:46:14 +0000 (09:46 +0200)]
seccomp: allowlist uretprobe() syscall

This is a new syscall provided by the kernel used to implement faster
uprobes. It's not supposed to be called by userspace, but only by kernel
generated uprobe code.

It should be fine to allow this, as the kernel authenticates the
invocation itself, and we shouldn't break compat with things.

Note that this allowlisting is not sufficient to make ureprobe() work.
libseccomp must be tought the syscall too, but this can happen
independently.

Fixes: #34615
(cherry picked from commit d693c483a2bb3eae490fd78d68fc16d0a731fee2)

12 months agomkosi: Fix up ownership of testuser home directory on first boot
Daan De Meyer [Wed, 9 Oct 2024 12:49:07 +0000 (14:49 +0200)]
mkosi: Fix up ownership of testuser home directory on first boot

When building unprivileged, the testuser home directory ends up
owned by root:root because mkosi can't chown directories to other
owners when running unprivileged. So let's fix up the testuser
ownership on first boot with tmpfiles instead.

(cherry picked from commit 0d0ecaab000cf2768a3edf1e73119bf2fce952b0)

12 months agodocs/DESKTOP_ENVIRONMENTS: clarify name aliases
David Rheinsberg [Fri, 11 Oct 2024 08:46:05 +0000 (10:46 +0200)]
docs/DESKTOP_ENVIRONMENTS: clarify name aliases

Add a note to the service-file naming scheme that reminds developers
that those names might be aliases. Hence, when parsing such unit names,
the entire name-array of a unit must be parsed, rather than just the
unit ID.

The service-name of existing applications might be already part of their
API. Hence, not all applications can switch the service ID to this new
naming scheme, but can provide suitable aliases. Document this behavior.

(cherry picked from commit b3b7cf8b7c35df14c6eb4f79da1a241dc0aa8c7e)

12 months agodocs/DESKTOP_ENVIRONMENT: clarify <RANDOM> usage
David Rheinsberg [Fri, 11 Oct 2024 08:07:40 +0000 (10:07 +0200)]
docs/DESKTOP_ENVIRONMENT: clarify <RANDOM> usage

The <RANDOM> part is optional in the naming scheme of application units.
However, this is only true for service files. Scope units must include
the <RANDOM> part, otherwise it would be impossible to parse:

The schema would be:

    `app[-<launcher>]-<ApplicationID>[-<RANDOM>].scope`

in which case a two-part name would be impossible to parse, since it is
unclear whether the launcher of the random bit where omitted.

(cherry picked from commit cbaebf811e6edb79279f72950b2d63189be314d3)

12 months agomeson: sort includes
Lennart Poettering [Thu, 10 Oct 2024 09:49:47 +0000 (11:49 +0200)]
meson: sort includes

(cherry picked from commit 882032faaf9e2d2524936e82ccc770903d4c38d6)

12 months agocore/manager: still send out STATUS=Ready for user manager
Mike Yuan [Thu, 10 Oct 2024 19:16:05 +0000 (21:16 +0200)]
core/manager: still send out STATUS=Ready for user manager

This effectively reverts 37d15cd132f3a8a0bf42fb252c1604e804171ff2.

The offending commit wrongly assumed that the second READY=1
notification is for system scope only, but it also serves the purpose
of flushing out previous STATUS= containing user unit job status.

(cherry picked from commit da81a108653e2ef19102698dbc0184bd18b084d9)

12 months agodocs/DESKTOP_ENVIRONMENTS: fix formatting
David Rheinsberg [Fri, 11 Oct 2024 07:53:25 +0000 (09:53 +0200)]
docs/DESKTOP_ENVIRONMENTS: fix formatting

The annotation about omittance is meant to be about the `RANDOM` string.
However, the current formatting makes it look like the entire naming
scheme is optional. Fix this.

(cherry picked from commit 5697bce82b1b37602255ac12e4a8f4a12305dc03)

12 months agodocs: don't mention split-usr path anymore
Lennart Poettering [Fri, 11 Oct 2024 07:55:07 +0000 (09:55 +0200)]
docs: don't mention split-usr path anymore

We don't support split /usr/ anymore. Hence fix the paths. This
apparently matters because of PK validating the binary path.

Fixes: #34712
(cherry picked from commit c28a13b5afcf021191494bfa565c55b784dc08fa)

12 months agomkosi: fix sections for settings
Yu Watanabe [Thu, 10 Oct 2024 19:19:03 +0000 (04:19 +0900)]
mkosi: fix sections for settings

Follow-up for 963157ca78429c51feb3103828d8cc94440956f3.

(cherry picked from commit 727dc1f23a2e16da7f1e24810157d5b7c9136525)