Yu Watanabe
86aa7cb4b8
network/macsec: use hash_ops with destructor
2025-04-16 23:29:45 +09:00
Yu Watanabe
c64c71fbe5
network/macsec: replace security_association_init() with structured initializer
2025-04-16 23:29:45 +09:00
Yu Watanabe
2af1f13c6f
network/wireguard: use hash_ops with destructor for managing WireguardPeer
2025-04-16 23:29:45 +09:00
Yu Watanabe
643f1c1306
network/l2tp: use hash_ops with destructor for managing sessions
2025-04-16 23:29:45 +09:00
Daan De Meyer
1a4c2e8807
mkosi: Two improvements ( #37155 )
2025-04-16 15:12:15 +02:00
Daan De Meyer
52a82bcfc9
mkosi: Reuse main image prepare scripts in subimages
...
In the subimages we also want to make sure all dependencies of the
systemd packages are cached so reuse the same prepare scripts from
the main image to do that.
We only want required dependencies in the subimages, not recommended
or suggested dependendencies, so add an environment variable
$SYSTEMD_REQUIRED_DEPS_ONLY which the prepare scripts can check for
and enable it for the subimages.
2025-04-16 14:37:58 +02:00
Daan De Meyer
0ac6fc2917
mkosi: Make sure coreutils is installed in initrd/exitrd
...
This is already installed but Fedora/CentOS systems are nudged towards
installing coreutils-single which then later causes issues when we try
to install coreutils as a dependency of systemd so let's make sure we
pick coreutils from the beginning.
2025-04-16 14:21:45 +02:00
Lennart Poettering
bbeeea4362
bootctl: replace --no-variables by --variables=BOOL
...
I think the current behaviour of not doing EFI variables when we are run
in a container makes a ton of sense, but in some cases it's useful to
do EFI var setup even when a set of namespaces is set up for us, for
example to recover a hosed installation from a rescue disk.
While we are at it, let's remove some duplicate checks, and
systematically output information why we skip various operations.
Fixes : #36174 #35005
2025-04-16 13:39:12 +02:00
Daan De Meyer
53628612b7
elf2efi: Add back python 3.7 compat
...
The walrus operator was introduced in python 3.9 so using it
accidentally bumped the minimum python version to 3.9 from 3.7 and
bluca still cares about keeping this working on 3.8 so let's remove
the usage of the walrus operator.
Follow up for d6a29eee2b
2025-04-16 13:06:27 +02:00
Yu Watanabe
ef79411c31
use hashmap_dump_sorted() and friends at more several places ( #37150 )
2025-04-16 19:20:19 +09:00
Lennart Poettering
3d3d8f1dd2
update TODO
2025-04-16 10:48:31 +02:00
Yu Watanabe
f477751b10
docs: fix mkosi section for Environment= setting
2025-04-16 08:30:25 +09:00
Yu Watanabe
667fc1d993
conf-files: use hashmap_dump_sorted() at one more place
2025-04-16 08:17:52 +09:00
Yu Watanabe
dcf1c14b95
resolve: use set_dump_sorted() at one more place
2025-04-16 08:16:21 +09:00
Yu Watanabe
a0fa3f1616
resolve: fix memleak
...
Follow-up for 5c274555c3 .
2025-04-15 23:34:49 +01:00
Luca Boccassi
86cfbfae97
xattr: fix assert that breaks importctl ( #37146 )
...
getxattr_at_malloc() calls getxattr_pinned_internal() with size 0 to get
the size of the xattr, and then asserts because n > 0
```
Assertion '(size_t) n <= size' failed at src/basic/xattr-util.c:107, function getxattr_pinned_internal().
```
Follow-up for b8df25dcfe
2025-04-16 06:44:24 +09:00
Yu Watanabe
8a05dbb31a
core: remove cgroup v1 properties ( #37026 )
2025-04-16 06:22:46 +09:00
Yu Watanabe
b2b270e1d1
cgroup-setup: use path_join() to get path to cgroup attribute
...
We already have the path to the cgroup. Hence, it is not necessary to
call cg_get_path() again.
2025-04-16 03:56:04 +09:00
Yu Watanabe
188286eec6
cgroup-setup: drop unnecessary controller argument
...
This also removes unnecessary cg_{create,attach,trim}_everywhere(),
and rename cg_enable_everywhere() -> cg_enable().
2025-04-16 03:56:04 +09:00
Yu Watanabe
c35a585763
cgroup-setup: drop cgroup v1 support
2025-04-16 03:56:04 +09:00
Yu Watanabe
3988e2489a
core: drop several more cgroup version check
2025-04-16 03:56:04 +09:00
Yu Watanabe
3fcb98cbff
core/bpf-foreign: drop unnecessary check for cgroup v1
2025-04-16 03:56:04 +09:00
Yu Watanabe
ff7f99db30
core/bpf-devices: drop cgroup v1 support
2025-04-16 03:55:00 +09:00
Yu Watanabe
4ee64e4365
core/cgroup: drop cgroup v1 specific code in unit_update_cgroup()
2025-04-16 03:42:11 +09:00
Yu Watanabe
35f88201f2
core/cgroup: drop unnecessary cgroup version check in several getters
...
After 2b61489e5a , these checks are not
necessary anymore.
2025-04-16 03:25:46 +09:00
Yu Watanabe
0818300288
core/cgroup: drop unnecessary _unified_ modifier from function names
2025-04-16 03:25:45 +09:00
Yu Watanabe
f1c5534eb6
core/cgroup: remove logic of CGroup v1 attribute assignment
2025-04-16 03:19:16 +09:00
Yu Watanabe
a7b06f6c30
core/cgroup: remove unused legacy parameters in CGroupContext
2025-04-15 22:34:22 +09:00
Yu Watanabe
3e316a6699
core/cgroup: remove CGroup v1 settings in dump message
2025-04-15 22:34:22 +09:00
Yu Watanabe
5da476ac77
bus-unit-util: refuse setting CGroup v1 properties through DBus
2025-04-15 22:34:22 +09:00
Yu Watanabe
db6986e02c
core: deprecate CGroup v1 DBus properties
2025-04-15 22:34:22 +09:00
Yu Watanabe
98d64ff500
core: remove deprecated CGroup v1 settings
2025-04-15 22:34:22 +09:00
Yu Watanabe
051675ea80
test: replace deprecated CGroup v1 settings with v2 ones
2025-04-15 22:34:22 +09:00
Yu Watanabe
5d902cc21f
journal: hash_ops related cleanups ( #37115 )
2025-04-15 22:32:15 +09:00
Yu Watanabe
9fdb01ed66
resolve: hash_ops related cleanups ( #37116 )
2025-04-15 22:31:57 +09:00
Yu Watanabe
7a7a9bbed3
network-tools: hash_ops related cleanups ( #37118 )
2025-04-15 22:31:40 +09:00
Yu Watanabe
3cb9ac88a0
journal: drop old kernels support ( #37137 )
...
This drops several codes for supporting old kernels.
Also, this cleanups code a bit.
2025-04-15 22:30:51 +09:00
Luca Boccassi
bd9c55ebe2
import/export: add support for zstd
2025-04-15 12:21:30 +01:00
Yu Watanabe
7566bbc5cb
journal/kmsg: drop old kernels (<3.6) support
...
The flags field was added by kernel v3.6:
https://github.com/torvalds/linux/commit/d39f3d77c9b1fe7cc33a14beb4a4849af0a4ac22
Now our baseline on the kernel is v5.4, so we can assume there exists
the flags field after the monotonic timestamp.
This also adds several short comments.
2025-04-15 06:19:07 +09:00
Yu Watanabe
9ec1e87bf3
journal/kmsg: modernize server_open_dev_kmsg()
...
Let's assign fd and event source only on success.
2025-04-15 06:18:52 +09:00
Yu Watanabe
117aa1cab4
journal/kmsg: drop Server.dev_kmsg_readable flag and use Server.read_kmsg
...
When /dev/kmsg is opened, the two flags have the same value.
This also adds an assertion in server_read_dev_kmsg().
2025-04-15 06:17:22 +09:00
Yu Watanabe
da741d9fd9
journal/kmsg: drop support old kernels (<3.5) without reading /dev/kmsg support
...
/dev/kmsg is readable since kernel v3.5:
https://github.com/torvalds/linux/commit/e11fea92e13fb91c50bacca799a6131c81929986
Now our baseline on the kernel is v5.4. Let's assume /dev/kmsg is always
readable.
2025-04-15 05:22:28 +09:00
Yu Watanabe
d127423ce6
journal: remove old kernels (<3.2) support in monitoring hostname
...
Similar to cc983fc9dc .
2025-04-15 04:24:12 +09:00
Yu Watanabe
b01f00e9c5
test: drop error conditions for old kernels (<3.2)
...
Now our baseline on the kernel is 5.4.
2025-04-15 03:41:40 +09:00
Yu Watanabe
d50cfa98aa
tree-wide: IFA_FLAGS is supported since kernel v3.14
...
Now our baseline on the kernel is v5.4.
Let's unconditionally use the IFA_FLAGS attribute, and drop the getter
and setter for the deprecated header element.
Continuation of 29c2d4e158 .
2025-04-15 03:41:16 +09:00
Yu Watanabe
6dbf2c1beb
sd-netlink,sd-device: drop old kernels (<4.2) support
...
NETLINK_LIST_MEMBERSHIPS is supported since kernel v4.2, specifically
https://github.com/torvalds/linux/commit/b42be38b2778eda2237fc759e55e3b698b05b315
Our baseline on the kernel is v5.4. Let's drop unnecessary conditions.
2025-04-15 03:40:40 +09:00
Yu Watanabe
d3d375a686
udev-builtin-net_id: drop old kernels (<4.20) support
...
Since kernel v4.20, specifically since
https://github.com/torvalds/linux/commit/9b8b2a323008aedd39a8debb861b825707f01420
IP-over-Infiniband driver also uses dev_port.
As our baseline on the kernel is v5.4. Let's remove the fallback code.
For more details about the fallback, see
https://github.com/systemd/systemd/pull/10082#issuecomment-421416461
2025-04-15 03:40:14 +09:00
Yu Watanabe
6d203bd830
test-condition: drop cgroup version check
...
After f74ca66e68 , we do not check the
cgroup version of the runtime environment, and unconditionally returns 0
for "v1", and 1 for "v2".
2025-04-14 17:22:46 +01:00
Yu Watanabe
cc983fc9dc
resolve: drop code for supporting old kernels (<3.2)
...
Now our baseline on the kernel is 5.4.
Note, the removed code triggered assert_return() in
sd_event_source_set_description().
2025-04-14 22:26:40 +09:00
Yu Watanabe
5e3887e6c1
loop-util: drop workaround for kernel older than v5.0
...
Now our baseline on the kernel is v5.4.
2025-04-14 21:57:53 +09:00