Summary: | Kwin Crash at screen shutdown | ||
---|---|---|---|
Product: | [Plasma] kwin | Reporter: | Daniël Huisman <d.d.huisman> |
Component: | general | Assignee: | KWin default assignee <kwin-bugs-null> |
Status: | RESOLVED UPSTREAM | ||
Severity: | crash | CC: | axb48, olaf.the.lost.viking |
Priority: | NOR | Keywords: | drkonqi |
Version First Reported In: | 5.6.3 | ||
Target Milestone: | --- | ||
Platform: | openSUSE | ||
OS: | Linux | ||
See Also: |
https://bugs.kde.org/show_bug.cgi?id=361439 https://bugreports.qt.io/browse/QTBUG-42803 |
||
Latest Commit: | Version Fixed In: | ||
Sentry Crash Report: |
Description
Daniël Huisman
2016-04-23 13:04:07 UTC
Well, now that the deliberate global QScreen crasher is gone, it can no longer stash the defects in QtQuick... => Upstream, compare bug #361439 or rather bug #361774 which is even more similar. Unfortunately we don't have debug values, but I guess that QOpenGLContext is dangeling/null in (rather than the dptr or it's screen member) since that's why QScreen was broken itfp... and to a certain degree we're likely just back on QTBUG-39996. The problem (here) would be that a signal causes ::setScreen on a dying GL context which largely seems to be the one of the QtQuick scene. Possibe QtQuick users in KWin (core) are - tabboxes except coverflow/slipswitch - the virtual desktop indicator - the buttons in presentwindows (close) and desktopgrid (+/-) effects - all aurorae driven decorations - things i might have forgotten. *** Bug 362754 has been marked as a duplicate of this bug. *** *** Bug 363108 has been marked as a duplicate of this bug. *** |