Hey everyone, I have researched and tried multiple different kernel flags to get my display working properly on my laptop. I found that amdgpu.dcdebugmask=0x10 seems to make all issues go away with my AMD 7840U and 780M GPU.
I updated to kernel 6.13 in hopes that I could take advantage of the better performance being reported on AMD systems. However it seems that a regression has been introduced somewhere. Upon rebooting with the new kernel installed, I am greeted with a login screen that flashes entirely white seemingly each time the cursor blinks in the password field. My wife descibed this behavior as my laptop starting a rave. This is similar to the issue that I had before adding the aforementioned kernel flag. Also it’s important to note that the issue seems to go away once logged in, and only reappears when I full screen some sort of media such as a video regardless of my adaptive sync setting in KDE’s display configuration.
I consider myself relatively experienced with Linux in general but if a post doesn’t already exist online about an issue I’m having, I don’t really know where to start troubleshooting. I have zero issues out of kernel 6.12.15 but I’d like to maybe help someone else and submit a bug report if I can figure out where the issue lies. This is an HP Pavilion Aero 14 with OLED display. The specs are 2880x1800@120hz.
Any help is appreciated.
EDIT: I have also tried amdgpu.sg_display=0 and it causes my particular laptop to completely freeze and crash.
Boot to the 6.12.15 kernel and post the output of inxi -Fzxx please.
The issue is likely graphics driver + kernel and we need to see exactly what GPUs are in use and the related drivers.
Please also post dnf list --installed \*nvidia\* kernel
It seems that @lbolan243 only has one GPU. I asked for the nvidia output just in case the problem may have been related to an nvidia GPU, but since that does not exist the problem must exist between the various drivers (amdgpu, radeonsi) and wayland. I note that the Display: does not show a DE and that the Monitor-1: (API: EGL) shows both wayland and X11 as inactive.
I’m running the same and not seeing an issue on my AMD GPU, but its not the same type of GPU.
After you have booted under 6.13 use Ctrl-Alt-F3 to get to a console and login.
From there you can get access to the system journal and scan the logs for errors.
For example sudo journalctl -b 0 -e.
You may have errors in the user journal journalctl --user -b 0 -e.
Do you see any errors?
Out of curiousity, what happens if you completely get rid of that kernel parameter? In my case (AMD Ryzen 7 8845HS + AMD Radeon 780M) kernel 6.13.5 seems to have solved all my amdgpu problems… no flickering, no glitches, no artifacts or freezes with no additional kernel parameters in two days of intense work.
I tried that by editing my command list before booting. I still get the flashing screen but it doesn’t freeze and lockup as it did before the flag. So I’m fighting two different issues with different kernels
Yes at 120hz. My entire screen actually goes solid white if I change the refresh rate and I have to wait for the timeout so the display settings revert. Other settings under display config me are set as follows:
Adaptive Sync - Auto
Color profile - built in
Color accuracy - prefer efficiency as this impacts performance for me.
I’ve tested changing all of these settings individually but none of them seem to correct the flashing issue
I haven’t got any solution and I like both Plasma and Gnome for different reasons. Having say that, my Radeon 780M was a nightmare with Plasma Wayland and now works great with Gnome Wayland and kernel 6.13.
Funny that you mention that, this is actually a fresh install. I was running Silverblue and rebased to 42 beta to try the new kernel and the issue was present there as well.
