Bug 493027

Created attachment 173590 [details]
Initial backtrace output after running the command reccommended at: https://community.kde.org/KWin/Debugging

SUMMARY
KWin segfaults upon moving my drawing tablet pen close to the screen, enough to hover the cursor. This has happened under various circumstances, but the exact cause is still unknown to me, as trying to reproduce the crash has been inconsistent. I initially thought this may be related to Nvidia sleep or system idling, however this has happened now multiple times while actively using the computer (for example, browsing the internet, and then immediately after picking up my pen and moving to draw on the tablet)

A backtrace has been attached, however it is incomplete and probably not as useful. I tried to retrieve one with debugging symbols to the best of my ability, however could not get `gdb` to attach to kwin_wayland with symbols active. See Additional Information below.

STEPS TO REPRODUCE
1. Move a drawing tablet pen close to the display,

OBSERVED RESULT
Upon moving the drawing tablet pen close to the screen, Kwin crashes and subsequently takes down other apps with it (Waterfox, Krita, Discord, Telegram, Konsole). System Settings and Dolphin file explorer does not crash however.

EXPECTED RESULT
Kwin should not crash.

SOFTWARE/OS VERSIONS
Windows: 
macOS: 
(available in the Info Center app, or by running `kinfo` in a terminal window)
Operating System: Garuda Linux 
KDE Plasma Version: 6.1.4
KDE Frameworks Version: 6.5.0
Qt Version: 6.7.2
Kernel Version: 6.10.9-zen1-2-zen (64-bit)
Graphics Platform: Wayland
Processors: 12 × AMD Ryzen 5 2600X Six-Core Processor
Memory: 15.5 GiB of RAM
Graphics Processor: NVIDIA GeForce GTX 1070 Ti/PCIe/SSE2


ADDITIONAL INFORMATION
The backtrace provided is unfortunately without complete debugging symbols. I followed the instructions provided at the following links to obtain a more useful backtrace, but had a lot of trouble doing so:
- https://community.kde.org/KWin/Debugging
- https://community.kde.org/Guidelines_and_HOWTOs/Debugging/How_to_create_useful_crash_reports#Install_debugging_packages
- https://wiki.archlinux.org/title/Debugging/Getting_traces

As well as other linked sources at those pages. They all point to Arch utilizing `debuginfod` for obtaining debug symbols in `gdb`, however I have been unable to run the recommended commands:
```
echo 0 | sudo tee /proc/sys/kernel/yama/ptrace_scope
sudo gdb -pid $(pidof kwin_wayland) -batch -ex "set logging file kwin_wayland.gdb" -ex "set logging on" -ex "continue" -ex "thread apply all backtrace" -ex "quit"
```

while also having `debuginfod` enabled and working. I've set the correct environment variable ahead of time as well as tried running the command with the environment variable set immediately after `sudo` however in doing so the output to the console shows that debuginfod is automatically disabled. Runing the command as-is also does not offer any prompt about using debuginfod either. 

Am willing to try and get a good backtrace, however I am lost as to how to do so as the additional information at the various wiki pages and alternate methods are very confusing. Hopefully the one provided for now is sufficient to start a bug report on this at least. I am relatively new to using Linux and am unfamiliar with using `gdb` and other debugging tools for more technical bug reports, so I appreciate any help and guidance in how to obtain the necessary info for anyone who ends up looking into this crash. 

Additionally, the graphics tablet in question is a Wacom Mobile Studio Pro 13 (1st Gen), connected via the Wacom Link peripheral (Thunderbolt 3 USB-C to HDMI/microUSB adaptor, essentially). The device itself can also be used as a standalone computer, however connecting it to another PC in this way allows for it to be used as a secondary display, acting as a more common Wacom Cintiq drawing tablet. I'm aware this hardware stack is a bit niche, however I've found since switching to Linux, it is seemingly well supported all things considered. Worth mentioning at least as I do not know if this crash affects more traditional display drawing tablets, however I can also try to reproduce the issue using a Huion display tablet instead, if that would help at all. Needless to say in either case I first need to be able to get a better backtrace from the crash either way. 

Finally, I have a few other bad backtraces saved, as well as related coredumps, info from journalctl, and general notes I've taken in trying to troubleshoot the issue. I can provide those as well if requested, however as noted several times now, I don't know if that will be any helpful to diagnosing the problem due to missing symbols, and don't want to clutter the initial report.