Fixes: #11422
Oneshots going to inactive directly without ever entering UNIT_ACTIVE is
considered success. This however means that if something both Requires=
and Requisites= a unit of such nature, the verify-active job getting
merged into the start job makes it lose this property of failing the
depending jobs, as there, the start job has the result JOB_DONE on
success, so we never walk over RequisiteOf units.
This change makes sure that such units always go down. It is also only
meaningful with After=, but so is Requisite= itself. Also, we also catch
cases like a oneshot having RemainAfterExit= true making us start up
properly in such a setting, but then removing it, reloading the unit,
and restarting it. In such a case, we go down due to restart propagation
before them, and our start job waits on theirs, properly failing with
the JOB_DEPENDENCY result.
This covers cases where ConditionXYZ= creates a similar situation as
well.
Fixes: #11499
Let's return -EAGAIN so that on state change, unit_process_job tries to
add our job to run_queue again so that all the reloads that coalesced
into the installed reload (which itself merged into a running one)
inititally atleast runs *once*. This should ensure service picks up all
config changes reliably.
See the issue being fixed for a detailed explanation.
Fixes: #11305Fixes: #3260
Related: #11456
So, here's what happens in the described scenario in #11305. A unit goes
down, and that triggeres stop jobs for the other two units as they were
bound to it. Now, the timer for manager triggered restarts kicks in and
schedules a restart job with the JOB_FAIL job mode. This means there is
a stop job installed on those units, and now due to them being bound to
us they also get a restart job enqueued. This however is a conflicts, as
neither stop can merge into restart, nor restart into stop. However,
restart should be able to replace stop in any case. If the stop
procedure is ongoing, it can cancel the stop job, install itself, and
then after reaching dead finish and convert itself to a start job.
However, if we increase the timer, then it can always take those units
from inactive -> auto-restart.
We change the job mode to JOB_REPLACE so the restart job cancels the
stop job and installs itself.
Also, the original bug could be worked around by bumping RestartSec= to
avoid the conflicting.
This doesn't seem to be something that is going to break uses. That is
because for those who already had it working, there must have never been
conflicting jobs, as that would result in a desctructive transaction by
virtue of the job mode used.
After this change, the test case is able to work nicely without issues.
Makes it easier to understand what was merged (and easier to realize why).
Example is a start job running, and another unit triggering a verify-active job. It is not clear what job was it that from baz.service that merged into the installed job for bar.service in the debug logs. This makes it useful when debugging issues.
Jan 15 11:45:58 jupiter systemd[1218]: baz.service: Trying to enqueue job baz.service/start/replace
Jan 15 11:45:58 jupiter systemd[1218]: baz.service: Installed new job baz.service/start as 498
Jan 15 11:45:58 jupiter systemd[1218]: bar.service: Merged into installed job bar.service/start as 497
Jan 15 11:45:58 jupiter systemd[1218]: baz.service: Enqueued job baz.service/start as 498
It becomes:
Jan 15 11:45:58 jupiter systemd[1218]: bar.service: Merged bar.service/verify-active into installed job bar.service/start as 497
Jonathon Kowalski