https://bugs.kde.org/show_bug.cgi?id=525573
Oleg <[email protected]> changed: What |Removed |Added ---------------------------------------------------------------------------- CC| |[email protected] --- Comment #2 from Oleg <[email protected]> --- Title (if filing a new issue): Repeated 40-second plasmashell shutdown timeout after upgrading to Plasma 6.8.0 on Arch Linux Suggested destination: comment on existing KDE Bug 525573 https://bugs.kde.org/show_bug.cgi?id=525573 --- COPY THE TEXT BELOW INTO A COMMENT --- I can reproduce the same 40-second shutdown delay on Arch Linux with KDE Plasma 6.8.0 (Wayland). The problem became noticeable after upgrading from Plasma 6.7.5 to 6.8.0 on October 15, 2026, at 21:18 local time (+03:00). Environment: - Distribution: Arch Linux, x86_64 - KDE Plasma: 6.8.0 - KWin: 6.8.0 - Session: Wayland - Graphics hardware: Intel UHD (12th Gen Intel Core i5-12450H) - plasma-plasmashell.service: stock unit at /usr/lib/systemd/user/plasma-plasmashell.service; no drop-ins - TimeoutStopUSec=40s; KillMode=control-group - One locally installed third-party plasmoid: com.tokwell.wd-weather (not yet tested as a cause) Observed result: After selecting Shut Down, the graphical desktop becomes unresponsive, including mouse interaction. The power-off sequence is delayed while systemd waits for plasmashell to stop. After 40 seconds, systemd kills plasmashell with SIGKILL; during the October 16 event it also killed four kioworker processes in the same service cgroup. The shutdown then proceeded to filesystem unmounting and system power-off. I have not identified why plasmashell does not stop in time. Expected result: The Plasma session should exit promptly and shutdown should complete without a 40-second timeout or forced SIGKILL. Reproduction steps: 1. Log into a Plasma 6.8.0 Wayland session on Arch Linux. 2. Use the normal graphical Shut Down action. 3. Intermittently the session becomes unresponsive and the shutdown waits for the plasma-plasmashell.service stop timeout. 4. On the next boot, inspect the previous boot's journal with journalctl -b -1. Repeated occurrences shown by the journal (local time +03:00): - 2026-10-15 22:21:43 - 2026-10-15 23:12:56 - 2026-10-16 00:10:26 - 2026-10-16 01:15:43 - 2026-10-16 14:35:35 - 2026-10-16 15:35:40 Selected journal output from October 16, 15:35: 15:35:00 plasmashell[1048]: Failed to fetch all devices "org.freedesktop.DBus.Error.NameHasNoOwner" "Could not activate remote peer 'org.freedesktop.UDisks2': activation request failed: unit is invalid" 15:35:40 systemd[973]: plasma-plasmashell.service: State 'stop-sigterm' timed out. Killing. 15:35:40 systemd[973]: plasma-plasmashell.service: Killing process 1048 (plasmashell) with signal SIGKILL. 15:35:40 systemd[973]: plasma-plasmashell.service: Killing process 16613 (kioworker) with signal SIGKILL. 15:35:40 systemd[973]: plasma-plasmashell.service: Killing process 16618 (kioworker) with signal SIGKILL. 15:35:40 systemd[973]: plasma-plasmashell.service: Killing process 16626 (kioworker) with signal SIGKILL. 15:35:40 systemd[973]: plasma-plasmashell.service: Killing process 16628 (kioworker) with signal SIGKILL. 15:35:40 systemd[973]: plasma-plasmashell.service: Main process exited, code=killed, status=9/KILL 15:35:40 systemd[973]: plasma-plasmashell.service: Failed with result 'timeout'. 15:35:41 systemd[1]: Finished System Power Off. Package upgrade excerpts from /var/log/pacman.log: [2026-10-15T21:18:48+0300] [ALPM] upgraded kwin (6.7.5-3 -> 6.8.0-1) [2026-10-15T21:18:48+0300] [ALPM] upgraded plasma-workspace (6.7.5-3 -> 6.8.0-1) [2026-10-15T21:18:49+0300] [ALPM] upgraded plasma-desktop (6.7.5-1 -> 6.8.0-1) Additional details: - No coredumps were found for the affected October 16 session. - The original systemd user unit has TimeoutSec=40sec and no local override. - The timing strongly correlates with the upgrade, but I cannot prove that Plasma 6.8.0 itself, rather than a third-party plasmoid or another interaction, is the root cause. - I can provide further targeted journal logs or test steps if needed. --- END OF COMMENT --- -- You are receiving this mail because: You are watching all bug changes.
