Follow-up for 14631951ce
Before this commit, if WorkingDirectory= is empty or literally "-",
'simplified' is not populated, resulting in the ASSERT_PTR
in unit_write_settingf() below getting triggered.
Also, do not accept "-", so that the parser is consistent
with load-fragment.c
Fixes#33015
Currently, when service_set_main_pidref() is called
without specifying start_timestamp, exec_status_start()
always uses dual_timestamp_now(). This is not ideal,
though, as when the main pid changes halfway due to
e.g. sd_notify + MAINPID=, it's definitely spurious.
For manager test runs, the generator output paths are located in
/tmp, which means that if we mount a private /tmp for generators,
we lose all the generated units (actually the generators will just
fail because the directories don't exist, but if they did exist,
we'd still lose all the units).
Let's avoid the problem by skipping the private /tmp for manager
test runs. This also avoids any possible privilege issues with
mounting a private /tmp that might happen in this scenario.
Follow-up for 3c0a1b1e70
Before this commit, DefaultDependencies=no is set in
mount_add_extras(). However, when generating mount units
from /proc/self/mountinfo, we don't have a unit in memory
yet, and mount_setup_new_unit() doesn't call into
mount_add_extras().
Fixes#32838
Locking of the tty device and then /dev/console was added to synchronize
vconsole-setup with other writers to the console. But it turns out that often
the locking doesn't work and we carved out various cases where we ignore
failure:
- lack of permissions (in the user manager)
- missing device node
It turns out that there's at least one more failure mode: we get -EIO when the
console is (mis-)configured to point to an invalid device. E.g. in
rhbug#2273069 the reporter has a VM in Proxmox without a virtual console
configured and has 'console=tty console=ttyS0' on the kernel cmdline. I
couldn't reproduce this under libvirt, but failure with EIO has been reported
by at least four users in #30501.
Note that in systemd-vconsole-setup we report this is a hard failure, while
in the manager, we only do a debug line. So it's possible that the failure
also occured there, causing the rest of the setup of the tty to be skipped
without further notice.
Ignore the locking failure, since there's just too many ways it can fail. If we
proceed without a lock, we're back to the situation before we started locking,
which wasn't too bad. OTOH, skipping setup of the console is problematic for
users, and it seems better to try to do the setup without locking.
Fixes https://github.com/systemd/systemd/issues/30501,
https://bugzilla.redhat.com/show_bug.cgi?id=2273069.
Follow-up for 6b34871f5d.
The general idea is that the list of credentials to load can and will specify
credentials which actually aren't provided, so a warning is too much. Let's
downgrade this to "info". If it turns out to be too noisy, we can downgrade
further in the future.
Currently, on soft-reboot, /run/credentials/@system is unmounted
because it has DefaultDependencies=yes and as such will have
Conflicts=umount.target and Before=umount.target. Let's make sure
credential mounts survive soft-reboot by implying DefaultDependencies=no
for credential mounts.
Otherwise there's a chance the handover timestamp is less than the
start timestamp depending on the whims of the scheduler.
For non-forking services, we also propagate the start timestamp into
Service->main_exec_status to make sure it starts earlier than the
handoff timestamp.
Otherwise, if a service unit that requests LogNamespace= stopped before
systemd-journald@.service is started, logs generated by the service will be
lost, as systemd-journald@.socket is stopped and
systemd-journald@.service will never started.
To prevent the issue, let's introduce another implicit dependency to
a oneshot service that explicitly synchronizes a namespaced journal file
when the log namespace is not needed anymore.
Fixes#32604.
some changes succeeded
Follow-up for a82b8b3dc8
We don't need to invalidate the unit file state
if all operations failed.
Also, emit UnitFilesChanged signal as long as
some operations succeeded.
- drop unnecessary SYNTHETIC_ERRNO() when the logger does not propagate
error code,
- drop unnecessary '%m' in error message when the error code is
specified with SYNTHETIC_ERRNO(),
- add missing full stop at the end of log message,
- use RET_GATHER(),
- add missing ", ignoring.",
- upeercase the first letter, etc., etc...
Required for integration tests to power off on PID 1 crashes. We
deprecate systemd.crash_reboot and related options by removing them
from the documentation but still parsing them.
If mac_selinux_access_check() or bus_verify_bypass_dump_ratelimit_async()
fail, we goto "ratelimited" where we set a custom D-BUS error. In
"ratelimited", we call sd_bus_error_setf() which eventually hits an
assert_return(!bus_error_is_dirty()). Avoid hitting this assertion by
passing NULL as the error to mac_selinux_access_check() and
bus_verify_bypass_dump_ratelimit_async() since we will override the error
immediately anyway if either fails.
We modify both functions as well to allow passing a NULL error and fix
the argument name as well while we're at it.
Mike Yuan
Zbigniew Jędrzejewski-Szmek
Yu Watanabe
Daan De Meyer
Luca Boccassi
Timo Rothenpieler
David Tardon
Dmitry V. Levin