FS#66458 - [systemd] unit GC regression

Attached to Project: Arch Linux
Opened by Kevin Kuehler (keur) - Tuesday, 28 April 2020, 22:07 GMT
Last edited by Christian Hesse (eworm) - Tuesday, 23 June 2020, 11:14 GMT
Task Type Bug Report
Category Packages: Core
Status Closed
Assigned To Dave Reisner (falconindy)
Christian Hesse (eworm)
Architecture All
Severity Medium
Priority Normal
Reported Version
Due in Version Undecided
Due Date Undecided
Percent Complete 100%
Votes 0
Private No

Details

Description:

Revert https://github.com/systemd/systemd/pull/14086 downstream as it causing regressions with existing systemd units. This has been reverted in both Fedora and Debian.

Additional info:
* package version(s) systemd >= 245
* Arch Linux user report: https://github.com/systemd/systemd/pull/14086#issuecomment-620336661

* debian bug report: https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=953670

* fedora bug report: https://bugzilla.redhat.com/show_bug.cgi?id=1803293

Steps to reproduce: N/A
This task depends upon

Closed by  Christian Hesse (eworm)
Tuesday, 23 June 2020, 11:14 GMT
Reason for closing:  Fixed
Additional comments about closing:  systemd 245.6-8
Comment by Andreas (bjoe2k4) - Tuesday, 02 June 2020, 23:01 GMT
Please really consider to revert this commit in the package. Problems have been reported upstream[1][2], but there is nothing happening but tumbleweed rolling. This issue has been around almost 3 months and it looks like it will be around for many more months to go. Arch Linux should really not fall behind Debian, which got it's package fixed almost 2 moths ago!

[1] https://github.com/systemd/systemd/issues/15242
[2] https://github.com/systemd/systemd/issues/15188
Comment by Dave Reisner (falconindy) - Tuesday, 02 June 2020, 23:12 GMT
If you'd like to see real progress here, please submit a pull request upstream that reverts pull 14086. That fixes it for everyone, rather than asking Arch to carry these reverts with no clear upstream solution.
Comment by Rado Sław (rdslw) - Sunday, 07 June 2020, 11:04 GMT
Dave, your comment is not useful.
It's clear, also for you, that it is not a lack of PR on systemd which is halting it.

Let me summarise, facts:
1. This bug affects a lot of arch users.
2. This bug is 3 months old
3. systemd maintainers are aware of it
4. some other distros (incl. SLOW ones and BIG ones) introduced reverts
5. even systemd maintainer employer distro reverted it.

I think that 1-5 warrants action on arch.
Comment by loqs (loqs) - Sunday, 07 June 2020, 11:54 GMT
If upstream agreed the cause of the issue was 14086 it would have been reverted.
An upstream response to [1] may help clarify upstream's view.

[1] https://github.com/systemd/systemd/issues/15188#issuecomment-639809505
Comment by Andreas (bjoe2k4) - Thursday, 11 June 2020, 13:23 GMT
It doesn't look like I can get any reaction or even getting this issue labeled with a proper milestone tag. Given that most distros reverted this and don't have to suffer the consequences I don't think many more people are going to complain until the really slow distros are catching up. I'm out if ideas. ¯\_(ツ)_/¯

Maybe someone else can try reach out?
Comment by Dave Reisner (falconindy) - Thursday, 11 June 2020, 14:37 GMT
I've created https://github.com/systemd/systemd/pull/16149 and added the v246 tag to it.
Comment by Dave Reisner (falconindy) - Tuesday, 23 June 2020, 11:03 GMT
My non-useful behavior has resulted in a merged PR. Arch can happily revert this now.

Loading...