| Summary: | Logind occupies 100% CPU after resume | ||
|---|---|---|---|
| Product: | systemd | Reporter: | Fabian Kosmale <inkane> |
| Component: | general | Assignee: | systemd-bugs |
| Status: | RESOLVED FIXED | QA Contact: | systemd-bugs |
| Severity: | normal | ||
| Priority: | medium | ||
| Version: | unspecified | ||
| Hardware: | x86-64 (AMD64) | ||
| OS: | Linux (All) | ||
| Whiteboard: | |||
| i915 platform: | i915 features: | ||
This if fixed in git. |
Use of freedesktop.org services, including Bugzilla, is subject to our Code of Conduct. How we collect and use information is described in our Privacy Policy.
With systemd 198, (probably 199), 200 and 202 , systemd occupies 100% CPU when suspending (to RAM or to disk) from KDE – and only when NetworkManager is used. Restarting systemd-logind solves this, though one needs to manually restart NetworkManager too. stracing logind shows epoll_wait(4, {}, 1, 0) = 0 being called multiple times per second. When I attached gdb to systemd-logind to locate the issue, it showied that logind is cycling around in login/logind.c (manager_run). The line numbers are from logind as of 202. The output is available at http://paste.chakra-project.org/4560/ So far I've seen this only affecting us (Chakra Linux), though I don't know how many distributions besides Arch and us and us have packaged systemd > 197 so far. As stated above this could only be reproduced on systems suspending from KDE and using NetworkManager. Users running wicd or using systemctl suspend were not affected. Any help with finding the cause of this would be highly appreciated.