Created attachment 183865 [details] gdb backtrace of plasmashell at the time of the bug happening SUMMARY Every few days into a Plasma session, my Meta key suddenly stops opening the application launcher in my panel. When this happens, a bunch of other things break as well. Killing plasmashell is enough to fix things for a few days without needing to log out completely. STEPS TO REPRODUCE 1. Leave my computer on doing nothing overnight 2. Wake up in the morning 3. Try to use my Meta key. If it doesn't work, *the bug* is happening. OBSERVED RESULT Along with the pesky meta key issue, here's a list of weird things that also happen. - Sometimes my panel is completely hung, I can't click on anything or hover over things, nothing reacts. - Receiving a Discord message is enough to hang Discord. It keeps rendering frames, but I can't click on anything or type in the message field. - Trying to launch most apps from my panel (if it's responsive at all) causes it to hang, until the app eventually opens. - Trying to do the same, but from the application launcher, causes the application launcher itself to hang. That is, of course, until an app eventually launches. - Krunner still works fine, but sometimes it doesn't respond to the shortcut. - Although krunner works fine, trying to search inside the *app launcher* does not. It always returns no results, no matter what I search. EXPECTED RESULT My Meta key should always open the application launcher and none of the other weird issues should be happening. SOFTWARE/OS VERSIONS Linux/KDE Plasma: Plasma 6.4.3 on Arch Linux. Also observed on Fedora 42 Plasma Edition. KDE Plasma Version: 6.4.3 KDE Frameworks Version: 6.16.0 Qt Version: 6.9.1 ADDITIONAL INFORMATION I was told by another Plasma developer to take the backtrace of plasmashell the next time I see the bug happening. It's a lot of text, so I've attached the backtrace as a file. His reaction when reading it: "That's, umm, not good."
Sounds like bug #506892
> Thread 1 (Thread 0x7f8ea1b59a40 (LWP 2043570) "plasmashell"): > #0 0x00007f8ea72ade22 in ??? () at /usr/lib/libc.so.6 > #1 0x00007f8ea72a1fda in ??? () at /usr/lib/libc.so.6 > #2 0x00007f8ea72a264c in ??? () at /usr/lib/libc.so.6 > #3 0x00007f8ea72a4d1e in pthread_cond_wait () at /usr/lib/libc.so.6 > #4 0x00007f8ea7b2ddb0 in QWaitCondition::wait(QMutex*, QDeadlineTimer) () at /usr/lib/libQt6Core.so.6 > #5 0x00007f8ea97eabdc in ??? () at /usr/lib/libQt6DBus.so.6 > #6 0x00007f8ea97b253c in ??? () at /usr/lib/libQt6DBus.so.6 > #7 0x00007f8ea978591f in ??? () at /usr/lib/libQt6DBus.so.6 > #8 0x0000557cf02c401a in ??? () > #9 0x00007f8ea820d875 in QWindow::event(QEvent*) () at /usr/lib/libQt6Gui.so.6 The backtrace is missing debug symbols at the important spot. Can you fix that?
(In reply to Harald Sitter from comment #1) > Sounds like bug #506892 Yep, it absolutely does.
*** This bug has been marked as a duplicate of bug 506892 ***