https://bugs.kde.org/show_bug.cgi?id=527720
Franck Maurice <[email protected]> changed: What |Removed |Added ---------------------------------------------------------------------------- CC| |[email protected] --- Comment #1 from Franck Maurice <[email protected]> --- I can reproduce this 100% of the time on EndeavourOS with Plasma 6.8.1 (Wayland), and I think I found the cause: the system tray Printers applet (print-manager). VERSIONS - plasma-workspace / print-manager 6.8.1, kwin 6.8.1-2 - cups 2.4.20, qt6-base 6.12.0 - Kernel 7.2.11-arch1-1, SDDM 0.21.0 WHAT HAPPENS When you reboot, systemd stops cups.service and the user session at the same time. CUPS goes down first, and plasmashell immediately logs, in a loop: plasmashell: IPP_INTERNAL_ERROR: clearing cookies and reconnecting plasmashell: Failed to reconnect Invalid argument (repeated) plasmashell: QObject::disconnect: Unexpected nullptr parameter Then plasmashell stops responding to SIGTERM: plasma-plasmashell.service: State 'stop-sigterm' timed out. Killing. plasma-plasmashell.service: Failed with result 'timeout'. The console shows "A stop job is running for User Manager for UID 1000". ISOLATION - Plain image wallpaper: still hangs (I ruled out the wallpaper plugins). - Printers entry disabled in System Tray settings: no hang on any reboot. - Printers applet re-enabled, plus a drop-in that makes the user session stop before CUPS: /etc/systemd/system/[email protected]/after-cups.conf [Unit] After=cups.service cups.socket Two reboots in a row: clean, plasmashell exits in under a second, no IPP_INTERNAL_ERROR. So the applet's (or libkcups') reconnect logic seems to block plasmashell's shutdown when the CUPS server goes away. Maybe this report should be moved to print-manager / applet. ~ -- You are receiving this mail because: You are watching all bug changes.
