[Post-Install] Randomly kicked out to lock screen then freeze after entering password

[rediskrad]

I just upgraded today from version 25 through a notification offering an upgrade. The upgrade process went smoothly, after reboot i logged in and everything looked fine. I launched a browser and roughly after 1 miunte i got kicked out back to lock screen.
When i entered my password the screen froze, still on lock screen, not even mouse responding. So i rebooted through ctrl+alt+del and logged in again. Same thing happened, again roughly after 1 min.
I rebooted again, expecting the same thing, but it has been over 3 hours and issue did not return (yet).
I have no idea if it was just some post install problems and it wont come back or i just got lucky the 3rd time. I can post logs or other stuff but i don't know where should i look.

I didn't do any major modifications to the os, just using apt packages from default repositories and the disover store.

My HW:
AMD Ryzen 7 9700X 8-Core Processor
32 GiB of RAM
NVIDIA GeForce RTX 3080

Nvidia driver: 595.58.03

[rediskrad]

The issue came back after rebooting so I did some more troubleshooting with AI (gemini).

We found these error logs:
kvě 15 06:32:53 fanda-b650eagleax kwin_wayland_wrapper[2763]: kwin_core: Failed to authenticate the drm magic token. path: "/dev/dri/card2" error: Operace zamítnuta
kvě 15 06:32:53 fanda-b650eagleax kwin_wayland_wrapper[2763]: kwin_wayland_drm: drmModeListLessees() failed: Operace zamítnuta
kvě 15 06:32:53 fanda-b650eagleax kwin_wayland_wrapper[2763]: kwin_wayland_drm: Atomic modeset test failed! Operace zamítnuta

Gemini deduced that this is likely OS trying to use the wrong GPU and recommended setting priority for nvida by adding:
KWIN_DRM_DEVICES=/dev/dri/card1:/dev/dri/card2 to /etc/environment

after listing which is the correct one with:

ls -l /sys/class/drm/card*/device/driver
lrwxrwxrwx 1 root root 0 May 15 17:40 /sys/class/drm/card1/device/driver -> ../../../../bus/pci/drivers/nvidia
lrwxrwxrwx 1 root root 0 May 15 17:40 /sys/class/drm/card2/device/driver -> ../../../../bus/pci/drivers/amdgpu
​

Also clearing kwin/plasma configs and cache with:

mv ~/.config/plasma-org.kde.plasma.desktop-appletsrc ~/.config/plasma-org.kde.plasma.desktop-appletsrc.bak
mv ~/.config/kwinrc ~/.config/kwinrc.bak
rm -rf ~/.cache/plasmashell* ~/.cache/kwin*

I have now tried 3 more reboots to check if the problem comes back and so far it seems to be solved.

Although command:

journalctl -b -1 | grep -i kwin | grep -E "error|failed|critical"​

is still returning some errors:

kvě 15 17:40:11 fanda-b650eagleax kwin_wayland_wrapper[2632]: kwin_wayland_drm: Atomic modeset test failed! Operace zamítnuta
kvě 15 17:40:11 fanda-b650eagleax kwin_wayland_wrapper[2632]: kwin_core: Applying output configuration failed!
kvě 15 17:40:11 fanda-b650eagleax kwin_wayland_wrapper[2632]: kwin_wayland_drm: Atomic modeset test failed! Operace zamítnuta
kvě 15 17:40:11 fanda-b650eagleax kwin_wayland_wrapper[2632]: kwin_core: Applying output configuration failed!
kvě 15 17:40:11 fanda-b650eagleax kwin_wayland_wrapper[2632]: kwin_wayland_drm: Atomic modeset test failed! Operace zamítnuta
kvě 15 17:40:11 fanda-b650eagleax kwin_wayland_wrapper[2632]: kwin_core: Applying output configuration failed!
kvě 15 17:40:11 fanda-b650eagleax kwin_wayland_wrapper[2632]: kwin_wayland_drm: atomic commit failed: Operace zamítnuta
kvě 15 17:40:11 fanda-b650eagleax kwin_wayland_wrapper[2632]: kwin_screencast: PipeWire remote error:  connection error
​