| Summary: | crash in KWin::Window::setMoveResizeOutput() while system was idle on wayland | ||
|---|---|---|---|
| Product: | [Plasma] kwin | Reporter: | stakanov |
| Component: | generic-crash | Assignee: | KWin default assignee <kwin-bugs-null> |
| Status: | RESOLVED FIXED | ||
| Severity: | crash | CC: | 0kusekushi0, alex+kde, andrzej, dave, dino, eldobhato.adrss, farchord, ibarona, jona, kdedev, lastminutejanedoe, marco.magliona, nate, r.maurizzi, shastao, thiago, tres.passer, valmenu59a, vlad.zahorodnii, wadja, wolf |
| Priority: | VHI | Keywords: | drkonqi |
| Version First Reported In: | 6.2.2 | ||
| Target Milestone: | --- | ||
| Platform: | openSUSE | ||
| OS: | Linux | ||
| See Also: |
https://bugs.kde.org/show_bug.cgi?id=489632 https://bugs.kde.org/show_bug.cgi?id=510066 |
||
| Latest Commit: | https://invent.kde.org/plasma/kwin/-/commit/a0c63b75dae838d305586c0deec92afa65e1200e | Version Fixed/Implemented In: | 6.4.6 |
| Sentry Crash Report: | |||
|
Description
stakanov
2024-11-01 20:06:26 UTC
Backtrace is similar to the one in Bug 489632, but not identical. *** Bug 504573 has been marked as a duplicate of this bug. *** Judging by other crash reports, it seems like the Window is fine now, but the Output is deleted without notifying the Workspace. *** Bug 510066 has been marked as a duplicate of this bug. *** *** Bug 510639 has been marked as a duplicate of this bug. *** A possibly relevant merge request was started @ https://invent.kde.org/plasma/kwin/-/merge_requests/8277 Git commit 5aa070d5f4a4c0ce1802aa94eec6ebf9bc9edca9 by Xaver Hugl, on behalf of Vlad Zahorodnii. Committed on 20/10/2025 at 12:14. Pushed by zamundaaa into branch 'master'. Fix Workspace::activateNextWindow() corrupting active output on hotplug After 4ee32d7f71de7f7f9a24c44b9dc4dd31fd3be98f, layer shell windows are removed when the Workspace::outputRemoved() signal is emitted. The side-effect of that change is that the activateNextWindow() may be called when the Workspace::outputRemoved() signal is emitted. The activateNextWindow() function is interesting because it can set the active output to the one of the Window. If the next candidate window is on a just removed output, the Workspace can end up with a dangling active output. To prevent that case, this change moves active output cleanup code so it is after the outputRemoved() signal is emitted and after windows are re-arranged. M +116 -0 autotests/integration/workspace_test.cpp M +5 -1 src/workspace.cpp https://invent.kde.org/plasma/kwin/-/commit/5aa070d5f4a4c0ce1802aa94eec6ebf9bc9edca9 Git commit 63edb977e99a860a5a167e210e6222a24115c593 by Xaver Hugl, on behalf of Vlad Zahorodnii. Committed on 20/10/2025 at 13:11. Pushed by zamundaaa into branch 'Plasma/6.4'. Fix Workspace::activateNextWindow() corrupting active output on hotplug After 4ee32d7f71de7f7f9a24c44b9dc4dd31fd3be98f, layer shell windows are removed when the Workspace::outputRemoved() signal is emitted. The side-effect of that change is that the activateNextWindow() may be called when the Workspace::outputRemoved() signal is emitted. The activateNextWindow() function is interesting because it can set the active output to the one of the Window. If the next candidate window is on a just removed output, the Workspace can end up with a dangling active output. To prevent that case, this change moves active output cleanup code so it is after the outputRemoved() signal is emitted and after windows are re-arranged. (cherry picked from commit 5aa070d5f4a4c0ce1802aa94eec6ebf9bc9edca9) M +116 -0 autotests/integration/workspace_test.cpp M +5 -1 src/workspace.cpp https://invent.kde.org/plasma/kwin/-/commit/63edb977e99a860a5a167e210e6222a24115c593 Git commit a0c63b75dae838d305586c0deec92afa65e1200e by Xaver Hugl, on behalf of Vlad Zahorodnii. Committed on 20/10/2025 at 21:12. Pushed by zamundaaa into branch 'Plasma/6.5'. Fix Workspace::activateNextWindow() corrupting active output on hotplug After 4ee32d7f71de7f7f9a24c44b9dc4dd31fd3be98f, layer shell windows are removed when the Workspace::outputRemoved() signal is emitted. The side-effect of that change is that the activateNextWindow() may be called when the Workspace::outputRemoved() signal is emitted. The activateNextWindow() function is interesting because it can set the active output to the one of the Window. If the next candidate window is on a just removed output, the Workspace can end up with a dangling active output. To prevent that case, this change moves active output cleanup code so it is after the outputRemoved() signal is emitted and after windows are re-arranged. (cherry picked from commit 5aa070d5f4a4c0ce1802aa94eec6ebf9bc9edca9) M +116 -0 autotests/integration/workspace_test.cpp M +5 -1 src/workspace.cpp https://invent.kde.org/plasma/kwin/-/commit/a0c63b75dae838d305586c0deec92afa65e1200e *** Bug 510917 has been marked as a duplicate of this bug. *** *** Bug 511059 has been marked as a duplicate of this bug. *** Is there a way I can test this commit on a live CD to check whether it resolves the screen lock / display waking from sleep bug reported elsewhere? https://bugs.kde.org/show_bug.cgi?id=511059 *** Bug 511168 has been marked as a duplicate of this bug. *** *** Bug 511166 has been marked as a duplicate of this bug. *** *** Bug 511125 has been marked as a duplicate of this bug. *** *** Bug 511215 has been marked as a duplicate of this bug. *** *** Bug 511230 has been marked as a duplicate of this bug. *** *** Bug 511289 has been marked as a duplicate of this bug. *** *** Bug 511374 has been marked as a duplicate of this bug. *** *** Bug 511403 has been marked as a duplicate of this bug. *** *** Bug 511438 has been marked as a duplicate of this bug. *** With Debian's packaging of kwin 6.5.1 in testing, I was able to lock my screen and disconnect my dock without a kwin crash now. *** Bug 511539 has been marked as a duplicate of this bug. *** *** Bug 511446 has been marked as a duplicate of this bug. *** *** Bug 512675 has been marked as a duplicate of this bug. *** |