Created attachment 180391 [details] relevant logfiles after the upgrade to fedora 42 (kde plasma edition) the kde plasma-powerdevil won't start anymore. Reproducible: Always Steps to Reproduce: boot system OR try to restart plasma-powerdevil.service manually (via "systemctl --user restart plasma-powerdevil.service") Actual Results: Job for plasma-powerdevil.service failed because a timeout was exceeded. Expected Results: plasma-powerdevil should start. Additional Information: Operating System: Fedora Linux 42 KDE Plasma Version: 6.3.4 KDE Frameworks Version: 6.13.0 Qt Version: 6.8.2 Kernel Version: 6.14.2-300.fc42.x86_64 (64-bit) Graphics Platform: Wayland Processors: 16 × AMD Ryzen 7 2700X Eight-Core Processor Memory: 31,3 GiB of RAM Graphics Processor: AMD Radeon R9 380 Series
*** Bug 502979 has been marked as a duplicate of this bug. ***
What does `journalctl --user-unit --reverse plasma-powerdevil.service` say?
(In reply to Nate Graham from comment #2) > What does `journalctl --user-unit --reverse plasma-powerdevil.service` say? Failed to add match 'plasma-powerdevil.service': Invalid argument
maybe instead of ´´´journalctl --user-unit --reverse plasma-powerdevil.service´´´ ´´´journalctl --no-hostname -b 0 --reverse | grep 'plasma-powerdevil.service´´´ does also help: $ journalctl --no-hostname -b 0 --reverse | grep 'plasma-powerdevil.service Apr 24 11:44:56 systemd[1648]: Starting plasma-powerdevil.service - Powerdevil... Apr 24 11:44:56 systemd[1648]: plasma-powerdevil.service: Scheduled restart job, restart counter is at 3. Apr 24 11:44:56 systemd[1648]: plasma-powerdevil.service: Consumed 43.270s CPU time, 11.3G memory peak. Apr 24 11:44:56 systemd[1648]: Failed to start plasma-powerdevil.service - Powerdevil. Apr 24 11:44:56 systemd[1648]: plasma-powerdevil.service: Failed with result 'timeout'. Apr 24 11:44:54 systemd[1648]: plasma-powerdevil.service: start operation timed out. Terminating. Apr 24 11:44:32 kernel: cgroup: fork rejected by pids controller in /user.slice/user-1000.slice/user@1000.service/background.slice/plasma-powerdevil.service Apr 24 11:44:09 systemd[1648]: Starting plasma-powerdevil.service - Powerdevil... Apr 24 11:44:09 systemd[1648]: plasma-powerdevil.service: Scheduled restart job, restart counter is at 2. Apr 24 11:44:07 systemd[1648]: plasma-powerdevil.service: Consumed 43.319s CPU time, 11.3G memory peak. Apr 24 11:44:07 systemd[1648]: Failed to start plasma-powerdevil.service - Powerdevil. Apr 24 11:44:07 systemd[1648]: plasma-powerdevil.service: Failed with result 'timeout'. Apr 24 11:44:06 systemd[1648]: plasma-powerdevil.service: start operation timed out. Terminating. Apr 24 11:43:45 kernel: cgroup: fork rejected by pids controller in /user.slice/user-1000.slice/user@1000.service/background.slice/plasma-powerdevil.service Apr 24 11:43:21 systemd[1648]: Starting plasma-powerdevil.service - Powerdevil... Apr 24 11:43:21 systemd[1648]: plasma-powerdevil.service: Scheduled restart job, restart counter is at 1. Apr 24 11:43:20 systemd[1648]: plasma-powerdevil.service: Consumed 43.125s CPU time, 11.3G memory peak. Apr 24 11:43:20 systemd[1648]: Failed to start plasma-powerdevil.service - Powerdevil. Apr 24 11:43:20 systemd[1648]: plasma-powerdevil.service: Failed with result 'timeout'. Apr 24 11:43:19 systemd[1648]: plasma-powerdevil.service: start operation timed out. Terminating. Apr 24 11:42:58 kernel: cgroup: fork rejected by pids controller in /user.slice/user-1000.slice/user@1000.service/background.slice/plasma-powerdevil.service Apr 24 11:42:34 systemd[1648]: Starting plasma-powerdevil.service - Powerdevil...
> Apr 24 11:44:32 kernel: cgroup: fork rejected by pids controller in /user.slice/user-1000.slice/user@1000.service/background.slice/plasma-powerdevil.service Well this looks relevant.
PowerDevil to the best of my knowledge does not generally spawn other processes for its basic function. Two ideas: * Perhaps libddcutil uses a subprocess to query monitors through the I2C bus. Try disabling powerdevil's use of libddcutil via `POWERDEVIL_NO_DDCUTIL=1` environment variable to verify whether this is the case, more detail in the PowerDevil README: https://invent.kde.org/plasma/powerdevil/-/blob/master/README.md#troubleshooting--reporting-bugs * Perhaps your PowerDevil configuration sets a custom script command (for the current power state profile) which tries indefinitely to spawn a subprocess, fails and tries again indefinitely. Look in your ~/.config/powerdevilrc file for any "[RunScript]" sections; if any exist, comment them out for the time being to verify if this is the culprit. If neither of these provide any good information, please increase output verbosity as also noted in the PowerDevil README. Thanks!
Updating the status here, pending some more info based on Jakob's ideas - thanks!
ok, i did the following: $ systemctl --user edit plasma-powerdevil.service then i added the following entry at the right place in the file: [Service] Environment="POWERDEVIL_NO_DDCUTIL=1" $ systemctl --user daemon-reload $ systemctl --user start plasma-powerdevil.service => it works. i am not THAT suprised that it is related to my monitor (philips 325E). it makes always problems, especially when connected via display port. it has totally messed up EDID entries, that make my gpus turn up clock speeds at idle and it takes ages to get recognized when booting...
There's not much else we can do on our side other than to report it to ddcutil
This change in Plasma should make it a lot easier to work around these issues in the future, at least! https://invent.kde.org/plasma/kscreen/-/merge_requests/360 And the ddcutil issue tracker is located at https://github.com/rockowitz/ddcutil/issues - I didn't see an existing issue for that specific monitor, so for the reporter, it might be helpful to provide what you know about its EDID "uniqueness" :-)