| Summary: | kwin hangs systemd until timeout | ||
|---|---|---|---|
| Product: | [Plasma] kwin | Reporter: | tudor.iacobescu.2 |
| Component: | general | Assignee: | KWin default assignee <kwin-bugs-null> |
| Status: | RESOLVED UNMAINTAINED | ||
| Severity: | normal | CC: | katyaberezyaka, nate, rounakpropubg |
| Priority: | NOR | ||
| Version: | 5.19.5 | ||
| Target Milestone: | --- | ||
| Platform: | Arch Linux | ||
| OS: | Linux | ||
| Latest Commit: | Version Fixed In: | ||
| Sentry Crash Report: | |||
| Attachments: | 'journalctl -rb -2' log, covering a boot where this issue happened | ||
Git commit 08263b534d4a73781fb06b2f875e7557965c729c by Vlad Zahorodnii. Committed on 09/10/2020 at 16:33. Pushed by vladz into branch 'master'. Safely remove all internal clients and override-redirect windows When we destroy all internal clients and override-redirect windows, we assume that corresponding lists will be implicitly detached. However, in some cases, that might be not the case. For example, if the list is not shared, neither begin() nor end() will detach. Therefore, it is possible to hit invalidated iterators if the list is modified inside the loop. This change prevents hitting invalidated iterators by making explicit list copies. M +6 -2 workspace.cpp https://invent.kde.org/plasma/kwin/commit/08263b534d4a73781fb06b2f875e7557965c729c Git commit 5df72c6613e588b2cc920863dbefba16ba0b89c7 by Vlad Zahorodnii. Committed on 09/10/2020 at 16:35. Pushed by vladz into branch 'Plasma/5.20'. Safely remove all internal clients and override-redirect windows When we destroy all internal clients and override-redirect windows, we assume that corresponding lists will be implicitly detached. However, in some cases, that might be not the case. For example, if the list is not shared, neither begin() nor end() will detach. Therefore, it is possible to hit invalidated iterators if the list is modified inside the loop. This change prevents hitting invalidated iterators by making explicit list copies. (cherry picked from commit 08263b534d4a73781fb06b2f875e7557965c729c) M +6 -2 workspace.cpp https://invent.kde.org/plasma/kwin/commit/5df72c6613e588b2cc920863dbefba16ba0b89c7 We are on the same page. I am having the same issue- https://bugs.kde.org/show_bug.cgi?id=453928 look at that This bug was reported against an outdated version of KWin. We have made many changes since the. If the issue persists in newer versions can you reopen the bug report updating the version number. |
Created attachment 132142 [details] 'journalctl -rb -2' log, covering a boot where this issue happened Attempting to shut down or restart will often (but not always) make it so that systemd has to wait 90s for kwin_x11 and QDBusController to close. Looking through the logs, it seems they're spammed with kwin attempting to start over and over prior to being killed. STEPS TO REPRODUCE 1. Shut down or restart OBSERVED RESULT Systemd waits for 90s for kwin_x11 and QDBusController to close, then kills them. EXPECTED RESULT It should close itself. SOFTWARE/OS VERSIONS Linux/KDE Plasma: KDE Plasma Version: 5.19.5 KDE Frameworks Version: 5.74.0 Qt Version: 5.15.1 Kernel version: 5.8.12-arch1-1