Both Wayland and X11 dumps core in gaming after random time (1h+) - Desktop PC AMD Radeon card
I've recently updated to Radeon 9070xt, rebased the system from nvidia-open to bazzite. Everything looked great, started extensively using the PC and ... ye every so often i get a crash while gaming. On Wayland it just hangs the system and i need to power off the PC to restart. So I thought it's Wayland issue. I've layered on X11, same thing happened - luckly this just boots me back to log-in screen. I've attached journal logs from X11 and Xwayland.
I've also disconnected my second monitor (as i thought it might be the issue) - so that's not the case 🙁
Any suggestions? This might be a faulty card (as it's the new thing in the system) - if so, how do i check for that?
Solution:Jump to solution
Tried to replicate the issue yesterday and today. Looks like it's gone, so it's either:
- bazzite 43 update
- unlayering X11
- swithing to different monitor (from old 144 to new 180Hz)
- disabling vencord hardware acceleration...
25 Replies
That looks like it's inside the AMD driver (_ZL30amdgpu), that's unfortunate
Now if this were nVidia the answer would be "don't use nVidia" so let's hope the answer isn't "don't use AMD" 😅
Is your crash resulting in no video output but the system keeps running (audio still works) ?
yes
even discord is still running, and i can talk with ppl
on x11 it just logs me out after ~ 20 seconds
on wayland it's just neverending no-signal on monitor but the system runs (discord works and spotify plays music)
I may have the same issue on nvidia. I'm kinda new to Linux, care to explain to me how to get the logs so I can check if they are similar ? I just had the issue and haven't restarted the computer yet
i used journal browser app - i believe it's part of the install
it's pretty tricky to copy from (you need to swith mode to select using icon on the top right of the window) but you can browse all crashes. You are looking for critical type of entries
I had monitor going blank on nvidia which i fixed by stopping using any type of adaptive sync - maybe that will help you @Nitokie
Okay thanks, I'll check right now
I have adaptive sync set to never for my display
them you might be facing the same stuff as I am, though i'm on AMD
try getting the logs and post them
Please unlayer X11 before posting any more logs
We will not support your system in that state
You can get back to standards with
rpm-ostree resetuderstood, so only wayland, right ?
Yep
should i then create a new post or just update this one ?
Update is fine
okies, let me produce a crash report (hopefully not) 😉
Well, my latest journal logs appears to be from last week 🤔
In a terminal, try
journalctl -S 2025-10-28
you will likely need to scroll up or down
or redirect to a fileujust logs-this-boot
ujust logs-last-boot
Either can be piped to a file and uploaded hereThanks, got the logs from before the crash
am I totally crazy or are NVRM / nvAssertOkFailedNoLog nVidia things
on your AMD system 👀
I wonder if the rebase process did not fully rebase or something?
I'm running bazzite nvidia-open, I have a nvidia gpu
I confused you with the OP, sorry
OP's crash is on an AMD system rebased from nVidia
@Nitokie do you monitor RAM during use at all, e.g. with system monitor or a HUD during gaming? I've seen someone here mention growing RAM usage due to a Plasma (dekstop) theme / module, and I've read Overwatch 2 has a massive memory leak if you use DX11 with Proton, just to give 2 examples.
I ask this because the log mentions "Out of memory", so there's a chance it's related to some sort of memory leak.
I sometimes monitor with mangohug but never noticed anything alarming
It would be good to know if you see memory usage maxing out just before a crash, but it also sounds like you might want your own thread since your crash is different than the OP on an AMD card
But I was never monitoring when I had a crash, so maybe
Solution
Tried to replicate the issue yesterday and today. Looks like it's gone, so it's either:
- bazzite 43 update
- unlayering X11
- swithing to different monitor (from old 144 to new 180Hz)
- disabling vencord hardware acceleration
- swithing from 2k-144Hz + HD-60hz monitor combo to 2x2k:180Hz monitors (matching refresh rate rather than unmatched resolutions and ref-rates) - this was the reason i layered X11 to begin with
- turning off LACT overdrive flag (enabled it and hadn't change a thing, but still rolled back the flag also)
I'll re-enable the hardware acc in other parallel apps and see whether this happens again. I'll close this ticket so it won't stay open marking this message as solution.
Thank you for help!