Bug#890833: at-spi2-core: Intermittent 90s shutdown delay: at-spi-dbus-bus.service: State 'stop-sigterm' timed out
Control: found -1 2.38.0-2 Hey, I have a sid system with at-spi2-core 2.38.0-2 running KDE Plasma Desktop. I still can see the problem. One thing I see is that this is only a problem if I start chromium. If I do not use chromium, the log is quite uninteresting. hefee Start chromium: 13:03:58 xxx systemd[5372]: Starting Accessibility services bus... 13:03:58 xxx systemd[5372]: Started Accessibility services bus. 15:16:09 xxx at-spi-bus-launcher[36537]: dbus-daemon[36537]: Activating service name='org.a11y.atspi.Registry' requested by ':1.0' (uid=1000 pid=36459 comm="/usr/lib/chromium/chromium --sho> 15:16:09 xxx at-spi-bus-launcher[36537]: dbus-daemon[36537]: Successfully activated service 'org.a11y.atspi.Registry' 15:16:09 xxx at-spi-bus-launcher[36552]: SpiRegistry daemon is running with well-known name - org.a11y.atspi.Registry 22:13:48 xxx at-spi-bus-launcher[36552]: X connection to :0 broken (explicit kill or server shutdown). 22:13:51 xxx systemd[5372]: Stopping Accessibility services bus... 22:15:21 xxx systemd[5372]: at-spi-dbus-bus.service: State 'stop-sigterm' timed out. Killing. 22:15:21 xxx systemd[5372]: at-spi-dbus-bus.service: Killing process 5689 (at-spi-bus-laun) with signal SIGKILL. 22:15:21 xxx systemd[5372]: at-spi-dbus-bus.service: Killing process 5696 (gdbus) with signal SIGKILL. 22:15:21 xxx systemd[5372]: at-spi-dbus-bus.service: Main process exited, code=killed, status=9/KILL 22:15:21 xxx systemd[5372]: at-spi-dbus-bus.service: Failed with result 'timeout'. 22:15:21 xxx systemd[5372]: Stopped Accessibility services bus. 22:15:21 xxx systemd[5372]: at-spi-dbus-bus.service: Consumed 4.294s CPU time. Without chromium: 22:25:26 xxx systemd[5284]: Starting Accessibility services bus... 22:25:27 xxx systemd[5284]: Started Accessibility services bus. 22:27:36 xxx systemd[5284]: Stopping Accessibility services bus... 22:27:36 xxx systemd[5284]: at-spi-dbus-bus.service: Succeeded. 22:27:36 xxx systemd[5284]: Stopped Accessibility services bus. signature.asc Description: This is a digitally signed message part.
Bug#890833: at-spi2-core: Intermittent 90s shutdown delay: at-spi-dbus-bus.service: State 'stop-sigterm' timed out
Hello, We have some new on this 90s-delay computer shutdown: upstream has committed something which might be fixing it. I have uploaded at-spi2-core version 2.34.0-4 which includes it. Could people who are regularly having the issue test this version? If it does fix it, I'll probably propose uploading the fix to buster. Samuel Maciej, le lun. 19 févr. 2018 17:18:07 +0100, a ecrit: > I shut the system down by clicking the "off switch" button in Gnome > DE and then clicking the "Shut down" or "Restart" button. In about 50% of > cases > the shutdown process takes 90 seconds longer than in the other 50%. Sergey Antonov, le lun. 26 févr. 2018 04:51:43 +0300, a ecrit: > Hi. I have same problem with AT SPI... > > lut 06 14:30:47 B2-46-AA-2018 systemd[612]: at-spi-dbus-bus.service: State > 'stop-sigterm' timed out. Killing. > lut 06 14:30:47 B2-46-AA-2018 systemd[613]: pam_unix(systemd-user:session): > session closed for user Debian-gdm > lut 06 14:30:47 B2-46-AA-2018 systemd[612]: at-spi-dbus-bus.service: Killing > process 648 (at-spi-bus-laun) with signal SIGKILL. Simon McVittie, le lun. 26 févr. 2018 08:19:55 +, a ecrit: > On Mon, 26 Feb 2018 at 04:51:43 +0300, Sergey Antonov wrote: > > Hi. I have same problem with AT SPI... > > > > lut 06 14:30:47 B2-46-AA-2018 systemd[612]: at-spi-dbus-bus.service: State > > 'stop-sigterm' timed out. Killing. > > lut 06 14:30:47 B2-46-AA-2018 systemd[613]: pam_unix(systemd-user:session): > > session closed for user Debian-gdm > > lut 06 14:30:47 B2-46-AA-2018 systemd[612]: at-spi-dbus-bus.service: > > Killing process 648 (at-spi-bus-laun) with signal SIGKILL. > > The per-user instance of systemd sends SIGTERM to at-spi-dbus-bus.service > (at-spi-bus-launcher). If the processes in that service still haven't > exited 90 seconds later, systemd logs this message and follows up > with SIGKILL. That means at-spi-bus-launcher isn't exiting promptly, > which seems like a bug in at-spi-bus-launcher.
