Bug 460834

Summary: When plasmashell crashes DrKonqi only has 40 seconds to do its thing before getting killed by systemd
Product: [Applications] drkonqi Reporter: Prajna Sariputra <putr4.s>
Component: generalAssignee: Plasma Bugs List <plasma-bugs-null>
Status: RESOLVED FIXED    
Severity: crash CC: kde, nate, qydwhotmail, sitter
Priority: NOR    
Version First Reported In: 5.26.1   
Target Milestone: ---   
Platform: Arch Linux   
OS: Linux   
Latest Commit: Version Fixed/Implemented In:
Sentry Crash Report:

Description Prajna Sariputra 2022-10-22 02:05:57 UTC
SUMMARY
With the systemd startup feature enabled if plasmashell crashes DrKonqi has a limited amount of time (appears to be 40s according to the contents of `/usr/lib/systemd/user/plasma-plasmashell.service`) before systemd ends up killing it since systemd thinks it's holding up the automatic service restart process.

That is especially a problem if debug symbols need to be downloaded, in my case DrKonqi did not even finish downloading plasmashell's debug symbols, let alone for all the libraries that it uses. Furthermore, the crash dump also did not appear in the "Crashed Processes Viewer" app, so it seems like that also got lost as a result.

STEPS TO REPRODUCE
1. Get `plasmashell` (or I guess any other Plasma component that runs as a service?) to crash
2. Wait until DrKonqi opens

OBSERVED RESULT
DrKonqi disappears after about 40s, and `plasmashell` comes back.

EXPECTED RESULT
DrKonqi does not disappear, allowing time to download the debug symbols and submit the bug report.

SOFTWARE/OS VERSIONS
Operating System: Arch Linux
KDE Plasma Version: 5.26.1
KDE Frameworks Version: 5.99.0
Qt Version: 5.15.6
Kernel Version: 6.0.2-arch1-1 (64-bit)
Graphics Platform: Wayland

ADDITIONAL INFORMATION
When DrKonqi disappears I see these two lines in `journalctl`:

> systemd[903]: plasma-plasmashell.service: State 'stop-sigterm' timed out. Killing.
> systemd[903]: plasma-plasmashell.service: Killing process 3088 (drkonqi) with signal SIGKILL.
Comment 1 Nate Graham 2022-10-24 20:24:41 UTC
Interesting.
Comment 2 Harald Sitter 2024-08-19 01:42:44 UTC
This defacto can't happen anymore with postmortem debugging via coredumpctl.