Freezing and white screen
This started happening over the weekend. Happens in multiple titles. Some do it on launch. Others randomly after some time. Each time, the fans go nuts in my machine and the system is completely frozen and non responsive to input.
Video attached.
23 Replies
Did you put gamescope on the command options for your games?
That kinda looks like Gamescope failing to run FSR
I haven’t modified any launch options for any title.
Is there general guidance for picking up the stack traces of what’s happening here? I’m an OSX guy but certainly know my way around a terminal
I attempted a
ujust update and still run into it. Interestingly, I don’t seem to need to enter a game in order to trigger it; I just did so by sitting on the Big Picture UI (or whatever it’s called these days).
🤯 I switched from a DP -> HDMI cable… to directly placing the HDMI cable and forcing the output in ~/.config/environment.d/10-gamescope-session.conf.
So far so good (as in I’ve launched CODBO3) but I’m genuinely confused why any of this should matter.LOL no way that just worked
thats so weird
that probably just is a gamescope-specific problem then!
you probably should just use this workaround for now since theres really no chance we can fix that. if you can reproduce this problem consistently you can try reporting your problem to the gamescope team
For sure. I’m happy to do so and report this to the chimera team (?)
Thanks for the help, @tulip
chimera?
i mean the actual gamescope peeps
Isn’t gamescope forked from chimera?
gamescope isnt forked from anything
gamescope session is forked from chimeraOS
that isnt a gamescope session issue i think thoOh I see — https://github.com/ChimeraOS/gamescope is what I was thinking of, but that’s a fork of Valve’s, so the report should likely go to Valve
GitHub
GitHub - ChimeraOS/gamescope: SteamOS session compositing window ma...
SteamOS session compositing window manager. Contribute to ChimeraOS/gamescope development by creating an account on GitHub.
you probably should check if that issue happens on nested gamescope too btw
but yeah! its straight from valve
(AFAIK)
no
we use a fork of gamescope
gamescope session is still somewhat shared with chimeraos although the bazzite one is significantly different
we make minimal changes, however they could be interfering
do not expect a response if you open an issue there
TIL. Can you point me to the correct repo? I am happy to report there, read up on any docs for collecting logs / stack traces.
journalctl -b | fpaste
drop this hereCan do once I’m at my machine. Should I reproduce prior to doing so?
if you reproduce and have to reboot use
-b -1
it does not look like gamescope, this looks like a kernel driver crashCan do. Appreciate the assist.
or a hardware failure
somewhere in that realm. If you can rollback and see if a previous version does it it would help a lot
I was using Bazzite for weeks prior with no issue. Happy to try a few rollbacks as well.
Frustrating: I cannot repro at the moment despite being able to do so at will yesterday. If I manage a repro, I’ll reach back out with the output.
@antheas https://paste.centos.org/view/a698f494
A lot of:
But that just seems like noise to me.
Found this report: https://gitlab.freedesktop.org/drm/amd/-/issues/3725
Will attempt a rollback to 40 if the standard rollback doesn’t do the trick
GitLab
AMD GPU screen blanking for seconds with a warning (#3725) · Issues...
Hello, I run Fedora 40 on a ThinkPad T14 Gen3 - comes with AMD Ryzen 7 PRO 6850U with Radeon Graphics. I have my monitor connected...
No dice on standard rollback. Seems possibly kernel related — I notice that any ongoing signals to devices seem to be continuous during one of these freezes
Seems I'm on 41-stable:
Rebasing to stable-40.20241020.0
(as an aside, gosh I never realized how baked in OSX shortcuts and commands are at this point - esp. text manipulation commands)
Managed a crash in desktop mode after the rebase command completed but before I managed a restart. Interesting. I think this at least confirms it has nothing to do with game scope.
Still crashing on stable-40.20241020.0
Aaaand I’m getting a motherboard indicator light on boot now — so something seems wrong here.
I may have found the culprit and it’s… ugh.
My retaining bracket for the GPU seems to have snapped.
So likely the GPU wasn’t making full contact. I’ll reseat after cleaning it with IPA.
I don’t have another machine to test this in but it may honestly be the board.
Oof yeah — the more I think on it… fans running full blast and no display atm = major component failure. Rip 7900 XTX, I hardly knew ye
RMA sent out. New hardware in a few weeks.