Let's reduce our attack surface by insisting that XBOOTLDR is vfat when
auto-probing, just like we do for the ESP. Given neither can
realistically be integrity protected (because firmware needs to access
them) let's insist on a vfat which has a much smaller attack surface,
and one we have to accept (for now) anyway, given that the ESP must be
VFAT.
This only applies to auto-probing of course. If people mount things
explicitly via fstab none of this matters. But we really shouldn't
automount a btrfs/xfs/ext4 partition as XBOOTLDR just because it looks
like one, as that would really defeat our otherwise possibly very strict
image policies.
This also introduces a new env var $SYSTEMD_DISSECT_FSTYPE_<DESIGNATOR>
environment variable that may override this hardcoding. This is in
particular useful in our testcases, since various actually do use ext4
as XBOOTLDR case. The tests are updated to make use of the new env var,
both as a mechanism to test this and to keep the tests working.
Previously, we'd take the image policy only into consideration when
dissecting the mage, but for the unlock/verity step we'd go via best
effort. Change that. This means we can now enforce policies such as
activating by root hash only even if a signature exists and similar.
Also, introduce a separate error code if we try to unlock a Verity
volume but have no root hash. Previously we'd return ENOKEY for that,
exactly like we do for encrypted volumes where we have no passparse. The
interctive unlock loop dissected_image_decrypt_interactively() is
otherwise very confused and will ask for a root hash, which makes no
sense. Hence use two distinct errors for this.
Sharing verity volumes is problematic for a veriety of reasons, for
example because it might pin the wrong backing device at the wrong time.
Let's hence turn this around: unless verity sharing is enabled, leave it
off, and turn $SYSTEMD_VERITY_SHARING into a true boolean that can be
set both ways.
The primary usecase for verity sharing is RootImage=, where it probably
makes sense to leave on, hence set the flag there.
This is crucial when putting together installers which install an OS on
a second disk: if verity sharing is always on we might mount the wrong
of the two disks at the wrong time.
Before:
Next elapse timestamp should be Sun Oct 12 00:10:00 UTC 2025 <= Sun 2025-10-12 05:43:04 UTC <= Sun Oct 12 22:10:00 UTC
After:
Next elapse timestamp should be Tue 2025-10-14 00:10:00 CEST <= Tue 2025-10-14 19:39:11 CEST <= Tue 2025-10-14 22:10:00 CEST
The EnterNamespace= feature currently doesn't work if the debuginfo is
separated from the crashing binary. Until that's resolved, let's run the
test only if the test binary has embedded debuginfo (.debug_info
section; e.g. when systemd is built without WITH_DEBUG=1) or it contains
MiniDebugInfo (.gnu_debugdata section; default on Fedora and CentOS).
See: https://github.com/systemd/systemd/pull/39268#issuecomment-3390745718
Cockpit's podman plugin needs to know the lingering status so the UI can
advertise enabling `podman-restart` (which depends on lingering to
work). Currently it relies on watching `/var/lib/systemd/linger/${user}`
but that isn't a public API.
Related: #22244
The ability to encrypt/authenticate encryption with a null key was
originally just a fallback concept for cases where during early boot we
have no host key, but the local system has no TPM2. Nowadays it is used
for other stuff as well, such as pcrlock data propagation (i.e. data
that needs no protection itself and required to properly to TPM key
derivation).
Let's give better, explicit control over null key usage, i.e. let's make
it a tristate both on the systemd-creds command line and in the Varlink
IPC to control three cases:
- the default that we allow it only if SecureBoot is off
- explicitly allowed
- explicitly refused (this is new)
Ideally systemd-creds --allow-null switch would take a boolean argument
to control this as a tristate. Alas, that would be a compat break, hence
I added --refuse-null instead (which also maps to the low-level flag for
this).
This also normalizes that the null key is always called "null key" in
messages, and not sometimes "empty key" or "fallback key".
This makes systemd-sysctl support following syntax, like
systemd-sysusers and systemd-tmpfiles:
```sh
$ /usr/lib/systemd/systemd-sysctl --inline 'foo.bar=42'
$ /usr/lib/systemd/systemd-sysctl - <<EOF
foo.bar=42
EOF
```
* From the TEST-74-AUX-UTILS.ssh.sh writing directly into
/etc/ssh/sshd_config can hide config from distro using /usr/etc.
* Using directory configuration /etc/ssh/sshd_config.d/* suits for all
cases.
Signed-off-by: val4oss <val4oss@pm.me>
Those were added before 0.61 which is our minimum version.
Dots were dropped from the end of some messages. Most messages did not have the
trailing dot.
0.63 added support for multi-line f-strings. We can't use those yet.
This adds what #39101 and #39070 did for the regular firstboot wizard
but for the homectl firstboot part: i.e. port to the generic prompt
loop, show the "chrome" bars, and mute the console.
And then it also makes querying for aux groups and shells optional,
because quite frankly, i am not sure what to answer there.
When -f/--force is used, wipefs does not call ioctl(BLKRRPART), hence
the partitions will remain in sysfs. Let's not use the option.
This also makes it use 'udevadm lock' for safety.
The keymap only really matters if there's local access to a system, i.e.
if there's actually a physical kbd directly connected to it, to apply it
to. If during firstboot we are not talked to via a VT (but via SSH,
container, or hypervisor console or so instead), then it's very unlikely
we ever are. Hence, don't ask for a keymap, and let#s shortcut the
questions asked at boot.
We only show this blurb on the first question, but the first question is
interactive anyway, hence no need to wait for input first explicitly, we
won't progress anyway without user input.
(similar for homectl's firstboot)
Before 23ac08115a, even if we failed to
create the cgroup for a unit, a cgroup runtime object for the cgroup is
created with the cgroup path. Hence, the creation of cgroup is failed,
execution of the unit will fail in posix_spawn_wrapper() and logged
something like the following:
```
systemd[1]: testservice.service: Failed to create cgroup /testslice.slice/testservice.service: Cannot allocate memory
systemd[1]: testservice.service: Failed to spawn executor: No such file or directory
systemd[1]: testservice.service: Failed to spawn 'start' task: No such file or directory
systemd[1]: testservice.service: Failed with result 'resources'.
systemd[1]: Failed to start testservice.service.
```
However, after the commit, when we failed to create the cgroup, a cgroup
runtime object is not created, hence NULL will be assigned to
ExecParameters.cgroup_path in unit_set_exec_params().
Hence, the unit process will be invoked in the init.scope.
```
systemd[1]: testservice.service: Failed to create cgroup /testslice.slice/testservice.service: Cannot allocate memory
systemd[1]: Starting testservice.service...
cat[1094]: 0::/init.scope
systemd[1]: testservice.service: Deactivated successfully.
systemd[1]: Finished testservice.service.
```
where the test service calls 'cat /proc/self/cgroup'.
To fix the issue, let's fail earlier when we failed to create cgroup.
Follow-up for 23ac08115a (v258).
We passes log level through kernel command line. It is not necessary to
set to debug level at the beginning, and set to info at the end.
This is important when a test has several subtests. If a subtest sets
log level to info at the end, then subsequent tests may not generate any
useful logs.
We intend to make self-registering machines an unprivileged operation,
but currently that would allow an unprivileged user to register a
process they own in the root namespace, and then login as any
user they like, including root, which is not ideal.
Forbid non-root from shelling into a machine that is running in
the root user namespace.
When deserializing a serialized timer unit with RandomizedDelaySec= set,
systemd should use the last inactive exit timestamp instead of current
realtime to calculate the new next elapse, so the timer unit actually
runs in the given calendar window.
Provides coverage for:
- https://github.com/systemd/systemd/issues/18678
- https://github.com/systemd/systemd/pull/27752
This also makes shebang always use env command, and drops unnecessary
'bash -c' or 'sh -c' when a signle command is invoked in the shell,
like sleep or echo.
It's not well-formed to begin with. And util-linux's mount(8)
is pretty much ubiquitously employed, hence it will be rejected
elsewhere too. Just stop pretending it is valid just because
glibc parser is sloppy.
I noticed in our NixOS packaging that we were working around the fact
that core/swap.c looks for swapon and swapoff in /sbin
Lets make it configurable just like all the other util-linux binaries
through meson and make it default to /usr/sbin/{swapon,swapoff}
This way mounts work on a systemd without the /sbin -> /usr/sbin
compatibility symlink. (And as a side-effect has NixOS be able to have
it in /nix/store too like the other util-linux tools).
Given that `unmerged-usr` support was dropped in 255 I think this is a
safe change?
Registering a process as a machine means a caller can get machined
to send sigterm to it, and more. If an unpriv user is registering,
ensure the registered process is actually owned by the user.
Follow-up for adaff8eb35
If the test VM reboots and the test re-runs, creating the images
fails as they already exist:
[ 218.227766] TEST-50-DISSECT.sh[889]: + mksquashfs testkit/ testkit.raw
[ 218.238754] TEST-50-DISSECT.sh[2964]: FATAL ERROR: Could not read $HOME, use -recovery-path or -no-recovery options
[ 218.239284] TEST-50-DISSECT.sh[2964]: Found a valid exportable SQUASHFS superblock on testkit.raw.
[ 218.239554] TEST-50-DISSECT.sh[2964]: Compression used gzip
[ 218.240176] TEST-50-DISSECT.sh[2964]: Inodes are compressed
[ 218.240459] TEST-50-DISSECT.sh[2964]: Data is compressed
[ 218.241072] TEST-50-DISSECT.sh[2964]: Fragments are compressed
[ 218.241526] TEST-50-DISSECT.sh[2964]: Xattrs are compressed
[ 218.241953] TEST-50-DISSECT.sh[2964]: Fragments are present in the filesystem
[ 218.242411] TEST-50-DISSECT.sh[2964]: Always-use-fragments option is not specified
[ 218.242843] TEST-50-DISSECT.sh[2964]: Duplicates are removed
[ 218.243560] TEST-50-DISSECT.sh[2964]: Xattrs are stored
[ 218.243889] TEST-50-DISSECT.sh[2964]: Filesystem size 0.38 Kbytes (0.00 Mbytes)
[ 218.244563] TEST-50-DISSECT.sh[2964]: Block size 131072
[ 218.245051] TEST-50-DISSECT.sh[2964]: Number of fragments 1
[ 218.245512] TEST-50-DISSECT.sh[2964]: Number of inodes 6
[ 218.245851] TEST-50-DISSECT.sh[2964]: Number of ids 1
[ 218.246393] TEST-50-DISSECT.sh[2964]: Parallel mksquashfs: Using 2 processors
[ 218.246820] TEST-50-DISSECT.sh[2964]: Scanning existing filesystem...
[ 218.247286] TEST-50-DISSECT.sh[2964]: Read existing filesystem, 5 inodes scanned
[ 218.252974] TEST-50-DISSECT.sh[2964]: Appending to existing 4.0 filesystem on testkit.raw, block size 131072
[ 218.253593] TEST-50-DISSECT.sh[2964]: All -b, -noI, -noD, -noF, -noX, -noId, -no-duplicates, -no-fragments,
[ 218.253848] TEST-50-DISSECT.sh[2964]: -always-use-fragments, -exportable and -comp options ignored
[ 218.257196] TEST-50-DISSECT.sh[2964]: If appending is not wanted, please re-run with -noappend specified!
https://github.com/systemd/systemd/actions/runs/17674609143/job/50233691148?pr=38867
The default is already to propagate the env vars, so this
was unnecessary and actually creates problem as it removes
custom PATHs
This reverts commit 994af53395.
When the test VM is accidentally rebooted, there exists the previously
created volume, and the command fails with the following:
```
TEST-64-UDEV-STORAGE.sh[282]: + lvm pvcreate -y /dev/md/mdlvm
TEST-64-UDEV-STORAGE.sh[442]: Can't initialize physical volume "/dev/md127" of volume group "mdlvm_vg" without -ff
TEST-64-UDEV-STORAGE.sh[442]: /dev/md127: physical volume not initialized.
[FAILED] Failed to start TEST-64-UDEV-STORAGE-mdadm_lvm.service.
```
Let's ignore the existence of previous volume and forcibly create new one.
Workaround for issue #38240.
Lennart Poettering
Antonio Alvarez Feijoo
Frantisek Sumsal
Jelle van der Waa
Yu Watanabe
val4oss
Zbigniew Jędrzejewski-Szmek
Luca Boccassi
Daan De Meyer
Mike Yuan