Created attachment 157609 [details] backtrace SUMMARY Moving a window to top edge with pen, to maximize, crashes kwin occasionally STEPS TO REPRODUCE 1. click&hold a titlebar with pen 2. move it to top edge 3. shake up and down, touching and not touching, see the maximize animation 4. within 2 to 22 moves, kwin crashes (random number) - a little video attachment just in case OBSERVED RESULT kwin maximizes the window every time EXPECTED RESULT kwin sometimes crash while maximizing SOFTWARE/OS VERSIONS Linux: Fedora 37 kernel 6.2 KDE Plasma Version: 5.27.3 KDE Frameworks Version: 5.104.0 Qt Version: 5.15.8 ADDITIONAL INFORMATION Just before it crashes, I see that it sticks to the pen while I'm hovering (not holding) without releasing the click and it goes black The additional video crashed too early without even the maximizing effect, so I have 2 traces with one being a quick crash and one taking a bit of up and down effort. I doubt they're much different but just trying to help :P
Created attachment 157610 [details] backtrace 2
Created attachment 157611 [details] quick video demonstration
Created attachment 157612 [details] Quick demonstration of expected behaviour (done with mouse instead, since pen crashes, you can see by the regular cursor shape)
Can reproduce on Wayland, did not test X11. Same thing happens like in OP, except they swapped observed and expected result. It's not too bad to workaround, but it kinda sucks for people that don't know about it if they're working on something and they accidentally do this, since the "apps survive compositor crash" thing hasn't been implemented yet. ------- Operating System: EndeavourOS KDE Plasma Version: 5.27.3 KDE Frameworks Version: 5.104.0 Qt Version: 5.15.8 Kernel Version: 6.2.8-zen1-1-zen (64-bit) Graphics Platform: Wayland Processors: 16 × AMD Ryzen 7 1700 Eight-Core Processor Memory: 15.6 GiB of RAM Graphics Processor: AMD Radeon RX 6600 XT Manufacturer: Micro-Star International Co., Ltd. Product Name: MS-7B79 System Version: 3.0
(In reply to bird_d from comment #4) > Can reproduce on Wayland, did not test X11. Same thing happens like in OP, > except they swapped observed and expected result. I couldn't edit, too late when I noticed, it's obviously the other way 😅
Interesting. Cannot reproduce on Wayland. Pasting backtrace inline for searchability: Thread 1 (Thread 0x7fb6e209a980 (LWP 31638) "kwin_wayland"): #0 0x0000003000080e53 in () #1 0x00007fb6e8de94ac in KWin::Workspace::takeActivity(KWin::Window*, QFlags<KWin::Workspace::ActivityFlag>) () at /lib64/libkwin.so.5 #2 0x00007fb6e8f41cc6 in KWin::Workspace::delayFocus() () at /lib64/libkwin.so.5 #3 0x00007fb6e74d0e96 in void doActivate<false>(QObject*, int, void**) () at /lib64/libQt5Core.so.5 #4 0x00007fb6e74d421e in QTimer::timeout(QTimer::QPrivateSignal) () at /lib64/libQt5Core.so.5 #5 0x00007fb6e74c7fc5 in QObject::event(QEvent*) () at /lib64/libQt5Core.so.5 #6 0x00007fb6e6baed62 in QApplicationPrivate::notify_helper(QObject*, QEvent*) () at /lib64/libQt5Widgets.so.5 #7 0x00007fb6e749d4e8 in QCoreApplication::notifyInternal2(QObject*, QEvent*) () at /lib64/libQt5Core.so.5 #8 0x00007fb6e74ed981 in QTimerInfoList::activateTimers() () at /lib64/libQt5Core.so.5 #9 0x00007fb6e74ebf30 in QEventDispatcherUNIX::processEvents(QFlags<QEventLoop::ProcessEventsFlag>) () at /lib64/libQt5Core.so.5 #10 0x00005635f7eaf5f1 in QUnixEventDispatcherQPA::processEvents(QFlags<QEventLoop::ProcessEventsFlag>) () #11 0x00007fb6e749bf3a in QEventLoop::exec(QFlags<QEventLoop::ProcessEventsFlag>) () at /lib64/libQt5Core.so.5 #12 0x00007fb6e74a4002 in QCoreApplication::exec() () at /lib64/libQt5Core.so.5 #13 0x00005635f7dcfd21 in main(int, char**) (argc=<optimized out>, argv=<optimized out>) at /usr/src/debug/kwin-5.27.3-2.fc37.x86_64/src/main_wayland.cpp:628 A debugging session is active.
Backtrace looks very much like the one from bug #459860. Therefore I mark it as a dup. *** This bug has been marked as a duplicate of bug 459860 ***
(In reply to postix from comment #7) > Backtrace looks very much like the one from bug #459860. Therefore I mark it > as a dup. > > *** This bug has been marked as a duplicate of bug 459860 *** STEPS TO REPRODUCE does not match at all but if it's the same bug, well, thanks. Should I add anything to it that could be helpful?