Created attachment 156331 [details] journalctl log I'm probably reporting this against the wrong component, but who knows, it looks like a tricky one. Logs indicate that it could be related to the new multiscreen support. Anyway - ever since upgrading from plasma 5.26 to 5.27.0 on a dual-monitor setup, I have started getting a really strange issue in about 1/3 of cases. It goes like this: * First, you need to leave the machine idle long enough for the powersaver to kick in and turn off the monitors * Then, mouse/keyboard events enable the monitors correctly, but "enter password" dialog doesn't appear, the lock screen behaves as if it were frozen. You can still enter the password blindly and press enter, however * When you do this, you are greeted by a long list of dr konqi reports about crashed applications. In every single case, the backtrace looks like this: [KCrash Handler] #4 0x00007f90efef506d in () at /lib64/libQt5XcbQpa.so.5 #5 0x00007f90efef90ff in () at /lib64/libQt5XcbQpa.so.5 #6 0x00007f90efef8400 in () at /lib64/libQt5XcbQpa.so.5 #7 0x00007f90efed4524 in QXcbConnection::handleXcbEvent(xcb_generic_event_t*) () at /lib64/libQt5XcbQpa.so.5 #8 0x00007f90efed54b6 in QXcbConnection::processXcbEvents(QFlags<QEventLoop::ProcessEventsFlag>) () at /lib64/libQt5XcbQpa.so.5 #9 0x00007f90efefb143 in () at /lib64/libQt5XcbQpa.so.5 #10 0x00007f90f7f27a90 in g_main_context_dispatch () at /lib64/libglib-2.0.so.0 #11 0x00007f90f7f27e48 in () at /lib64/libglib-2.0.so.0 #12 0x00007f90f7f27edc in g_main_context_iteration () at /lib64/libglib-2.0.so.0 #13 0x00007f90fbf33c16 in QEventDispatcherGlib::processEvents(QFlags<QEventLoop::ProcessEventsFlag>) () at /lib64/libQt5Core.so.5 #14 0x00007f90fbedac5b in QEventLoop::exec(QFlags<QEventLoop::ProcessEventsFlag>) () at /lib64/libQt5Core.so.5 #15 0x00007f90fbee2dc6 in QCoreApplication::exec() () at /lib64/libQt5Core.so.5 #16 0x000055d2152ef701 in () #17 0x00007f90fb62caf0 in __libc_start_call_main () at /lib64/libc.so.6 #18 0x00007f90fb62cbb9 in __libc_start_main_impl () at /lib64/libc.so.6 #19 0x000055d2152f0065 in () [Inferior 1 (process 2930) detached] Non-KDE apps like firefox survive. What further complicates matters is that this doesn't seem to happen every time, only about 1/3. There's a lot of seemingly relevant messages in the systemd journal, I'm attaching these in a separate file. OBSERVED RESULT Many crashes, as described above. EXPECTED RESULT No crashes. SOFTWARE/OS VERSIONS Linux/KDE Plasma: 5.27.0 (X11) KDE Plasma Version: KDE Frameworks Version: 5.103.0 Qt Version: 5.15.8
Created attachment 156332 [details] For comparison, log when the crashes do not happen
No idea how to even begin debugging this one. Looks like a head-scratcher indeed. Starting with kscreenlocker since that's the first system component that starts to go wrong.
Created attachment 156401 [details] log (saved from KSystemLog) Operating System: EndeavourOS KDE Plasma Version: 5.27.0 KDE Frameworks Version: 5.103.0 Qt Version: 5.15.8 Kernel Version: 6.0.11-zen3-rt14-xanmod1-1-rt (64-bit) Graphics Platform: X11 Processors: 16 × AMD Ryzen 7 5800X3D 8-Core Processor Memory: 31.3 GiB of RAM Graphics Processor: AMD Radeon RX 6900 XT Product Name: B550 Taichi This seems to reflect the behavior I am seeing: Description: When switching to a different input on primary display and returning to the KDE system, KDED dies. Dual Monitor with Daisy Chained DP Steps: 1. Switch input on my primary to a different system 2. Switch back to the input connected to the KDE 5.27 machine Results: KDE 5.27 system is at a black screen Switching to TTY and back will restart SDDM (workaround) (Session is killed in the background it seems) Sample lines: 2/17/23 2:31 PM kscreen_backend_launcher The X11 connection broke (error 1). Did the X11 server die? 2/17/23 2:31 PM ksmserver The X11 connection broke (error 1). Did the X11 server die? 2/17/23 2:31 PM xdg-desktop-portal-kde The X11 connection broke (error 1). Did the X11 server die? 2/17/23 2:31 PM kwalletd5 The X11 connection broke (error 1). Did the X11 server die?
>The X11 connection broke (error 1). Did the X11 server die? Did the X11 server die? Please look for your X11 log files and coredumps
(In reply to David Edmundson from comment #4) > >The X11 connection broke (error 1). Did the X11 server die? > > Did the X11 server die? > > Please look for your X11 log files and coredumps I have gone from 100% repro yesterday to no repros today - only change was I took 5.27.1 updates. I will update if I get another repro.
(In reply to Daniel Ellis from comment #5) > (In reply to David Edmundson from comment #4) > > >The X11 connection broke (error 1). Did the X11 server die? > > > > Did the X11 server die? > > > > Please look for your X11 log files and coredumps > > I have gone from 100% repro yesterday to no repros today - only change was I > took 5.27.1 updates. > I will update if I get another repro. My issue appears to have actually been: https://bugs.kde.org/show_bug.cgi?id=464590 Which was fixed by: http://commits.kde.org/libkscreen/d42dea00822c1cfb7478d1d55746e0750156918b
Please let us know if it stops happening after you upgrade the system to Plasma 5.27.1 (already released) or 5.27.2 (will be released in a few days).
Aye, will test once the distro updates. For what it's worth, if I (still on 5.27.0) try to use powersaving without the screenlocker, I get the same backtrace as in 464590, so chances are good that my issue will be fixed. Will report back.
After 2 days with 5.27.1, there have been no more crashes. A bit early to be entirely sure, but I'd say it's probably good now.
Yay! Thanks for following up.