I assume you're looking at "lightdm.service: Consumed 3.007s CPU time." I think that's just how long the CPU was working on issuing commands to kill it. I don't think that includes the 90sec waiting periods for the sigterm and sigkill timeouts.
I'm looking at the timestamps on the logs. This was 11 o'clock, so 23:02:44 is hh:mm:ss. I interpreted the timeline as...
Lightdm is also responsible for killing plymouth, and I'm not sure how that's getting handled after this. Regardless, I can no longer recreate this.
I'm looking at the timestamps on the logs. This was 11 o'clock, so 23:02:44 is hh:mm:ss. I interpreted the timeline as...
- Sometime around 23:01:14 or earlier I gave a "sudo reboot" command.
- Around 23:01:14 (11:01pm and 14 seconds) it tries to sigterm lightdm. I inferred this from the log of the sigterm 90sec timeout.
- At 23:02:44 (90sec later) sigterm times out, kernel tries to sigkill lightdm.
- At 23:04:14 (another 90sec later) sigkill times out.
Lightdm is also responsible for killing plymouth, and I'm not sure how that's getting handled after this. Regardless, I can no longer recreate this.
Statistics: Posted by mgrocketman — Tue Apr 02, 2024 2:06 am