udevadm trigger should not fail

(Closes: #917633)

Thanks a lot.

Can you move the "Closes: #xxxxx" to a separate line so it works with "gbp dch"

added 1 commit

• acdb48c4 - udevadm trigger should not fail

Compare with previous version

changed title from udevadm trigger should not fail (Closes: #917633) to udevadm trigger should not fail

changed the description

@agaida-guest Michael was talking about your commit message, not the bug description.

both are changed - so what - right now i don't use gbp

Hm, so I tried to reproduce this failure, but failed so far. Inside a pbuilder chroot I ran

root@pluto:/# udevadm trigger --subsystem-match=block --action=change
root@pluto:/# echo $?
0
root@pluto:/# mount
/proc on /proc type proc (rw,relatime)
sysfs on /sys type sysfs (rw,nosuid,nodev,noexec,relatime)
tmpfs on /dev/shm type tmpfs (rw,relatime)
devpts on /dev/pts type devpts (rw,nosuid,noexec,relatime,gid=5,mode=620,ptmxmode=666)
devpts on /dev/ptmx type devpts (rw,nosuid,noexec,relatime,gid=5,mode=620,ptmxmode=666)
devpts on /dev/console type devpts (rw,nosuid,noexec,relatime,gid=5,mode=620,ptmxmode=000)

@agaida-guest I assume your chroot did not have /sys mounted?

In older udev releases, a missing sysfs did not trigger an error. The relevant change was introduced in https://github.com/systemd/systemd/commit/13aca847695#diff-e3e322f67caf1b94a0008af64eadfaf5R303

I wonder if this change in behaviour was intended

@biebl - you are right

@agaida-guest You haven't yet done git push -f with an updated commit message with Closes on a separate line.

That's fine that you don't use gbp, but the person you want to accept this merge proposal does and the requested change helps him.

I've asked for clarification regarding the changed udevadm trigger behaviour. https://github.com/systemd/systemd/issues/11333

If this change in was intentional, we'll probably have to update quite a few packages more.

@jbicha - erm, please have a look again - if i was not clear enough, again: I've changed both request and commit thing. If the braces should be removed this is a different thing.

So - if one want anything else changed - please write exactly what you want. This is boring for a some characters change.

I probably should have included a link to the gbp dch doc: http://honk.sigxcpu.org/projects/git-buildpackage/manual-html/man.gbp.dch.html

added 1 commit

• 71f81536 - udevadm trigger should not fail

Compare with previous version

@biebl - nice, i should switch to gbp too with LXQt

There is some movement on the systemd/udev upstream bug report. See https://github.com/systemd/systemd/pull/11346

I'm currently leaning towards reassigning #917633 to udev and handling this in udevadm directly.

Upstream has merged this now.

Yup, will prepare an upload of systemd/udev with this PR rsn.

Reassigned 917633 to udev, closing this MR.

@agaida-guest Even if I didn't merge this MR, I hope you don't consider your time was wasted as we found a better solution and maybe you even learned something new :-)

closed

No, the time wasn't wasted in no way - as i wrote - it was a dirty hack, having this solved upstream is far better. Thanks a lot for your efforts.