I have the same Issues in Deep Rock Galatic. With driver 575.51.02:
kernel: NVRM: Xid (PCI:0000:01:00): 109, pid=7758, name=FSD-Win64-Shipp, Ch 00000021, errorString CTX SWITCH TIMEOUT, Info 0x3c034
The issue appears here after have installed the last driver 575.57.08.
It is interesting to note that my precedent driver was working smooth without this issue.
So I have used without any issue for 6 weeks the driver 575.51.02 (beta).
“Avatar: Frontiers of Pandora” also being bombarded with this, especially on photo mode. Tried a plethora of nvidia drivers - same results. Different version of Protons - same results.
kernel: [ 6139.157212] NVRM: Xid (PCI:0000:01:00): 109, pid=19460, name=afop.exe, Ch 00000031, errorString CTX SWITCH TIMEOUT, Info 0x3c02e
There is another related thread that I posted in for Clair Obscur Expedition 33:
I got XID 109 errors with:
- RTX 3080Ti
- nvidia driver 575.57.08
- Kernel 6.14.8 (cachyos)
- application: Steam + Proton 9.0-4
- game: Stalker 2 (DLSS+Frame Gen)
- behaviour: game freeze when game start or after few minutes while monitor display a 10Hz FPS (monitor integrated function) and while there is no problem with other application that use GPU (terminal, browser, picom, Xorg)
Rollback on nvidia beta driver 575.51.02 fixes it for me.
You know i’ve checked everything thats public, tested driver after driver and i’m at the end of the rope. This just doesn’t work - nvidia software i mean. I could say its linux-only issue, that big company doesn’t give a dime about it because why should they with free-of-charge os’s, but that would be lying. Im getting more and more reports on different support channels that nvidia doesn’t work even on commercial platforms like windows 11; people are getting ctds all over and are forced to rollback their drivers way before 570 branch happened. First it was that bs scamming with prices, bitcoins and whatnots and now this? Im seeing this topic, its not new, months passed by and its just a waiting game. But who really has time to wait? My costumers would devour me alive and no bone would be left if my work would make them wait.
They seem to have a lot more people working on the hardware side than on the software side, unfortunately
Wondering if this will be fixed in 580 or so. @ifrexalis if you’re still there, what GPU are you using?
Fair warning: you might be getting these now because of Steam’s new overlay.
It’s happened without steam open, and I’ve disabled the overlay a while ago, so that shouldn’t be it
I’ve recently upgraded from a 3080 to a 5070TI so I had to switch to nvidia open here is some of my SMI info
| NVIDIA-SMI 575.64.03 Driver Version: 575.64.03 CUDA Version: 12.9 |
0 NVIDIA GeForce RTX 5070 Ti
Kernel Version: 6.15.6-arch1-1
I’m using nvidia-open-beta-dkms
this is the error kernel: NVRM: Xid (PCI:0000:08:00): 109, pid=127019, name=WoW.exe, Ch 00000019, errorString CTX SWITCH TIMEOUT, Info 0x9c01c
I can’t find anything else going on besides this error my game seems to hang or randomly revert back to a scene from mins prior when then hang occurs.
It never happened before switching gpus. also it dont matter if I try vanilla wine with dxvk or ge-proton 10-9. I tried a lot of different stuff from switching API in game DX11-DX12 no difference. I tried wow64 wine but I still get this error.
Any help would be much appreciated or how I can help just let me know :)
After months struggling with this issue while playing Steam games on Proton like Street Fighter 6 and Frostpunk 2, I was finally able to solve it by using the latest NTSYNC feature in Linux kernel 6.15 and glorious eggroll proton version 10.9, which allows proton to use it.
Sadly I tried using ntsync and even wow64 and older proton/wine/dxvk just what for whatever reason this area I was in was just instantly making my system get the ctx timeout. it was as soon as I loaded in but other characters were fine in starting zones so I think its whatever operations is being called in the new area as my gpu usage is still low.
I even swapped to cachyOS and tried using that but even still I get the same issue though cachyOS is still based on arch so that could be it. I had endeavourOS before
For World of Warcraft, this appears to only happen to me with ‘compute effects’ turned on. Disabling it eliminates the CTX SWITCH TIMEOUT related crashes. Setting it to low sometimes eliminates the crashes.
It also does appear to be zone and environment dependent. Old zones have no problems. Newer zones will work fine until your camera turns to look at, say, a torch or fire.
1 Like
My reproducible crashes in World of Warcraft appear to have stopped after following @pphop 's suggestion to disable “compute effects” within the game’s graphics settings. My issue is also zone-dependent.
In WoW, the setting has a description: Controls the quality of Compute-based effects such as Volumetric Fog and some particle effects.
It has a description for the disabled setting as well: Volume Fog disabled. Compute-based particle collision disabled.
I am also having this issue in Metal Garden on driver 580.76.05
NVRM: Xid (PCI:0000:01:00): 109, pid=7878, name=GameThread, channel 0x000000bc, errorString CTX SWITCH TIMEOUT, Info 0x3ac064
Sadly it didn’t fixed the problem for me. Still reproducible 100% of times, even with Compute Effects disabled.
My workaround was to enable Wine-Wayland, no problems there.
Edit: the Compute Effects disabled seems to work if you choose DirectX11 as the Graphics API instead of DirectX12. I have been switching windows for a few minutes now and the error didn’t occur, it happened 100% of times 3 seconds after switching the window with DirectX12.
Edit2: after switching to DirectX11, I can enabled Compute Effects again and the issue still doesn’t happen, seems to be a DirectX12 problem.
TLDR (on Gnome 48 + Wayland):
On Xwayland:
- DirectX12 with or without Compute Effects: broken
- DirectX11 with or without Compute Effects: works
On Wine-Wayland:
- DirectX11 or DirectX12 with or without Compute Effects: works
Using Fsync/Esync/Ntsync doesn’t have any impact on the issue.
Still happens on 580.76.05, CTX timeouts get handled more gracefully though it seems like
NVRM: Xid (PCI:0000:01:00): 109, pid=3313, name=Emuthread - Sta, channel 0x00000008, errorString CTX SWITCH TIMEOUT, Info 0x14000
It’s fixed for me on driver 580.82.07
Not for me. Issue still persist (already about 8 months).
I am on rtx4090 with gnome48 and wayland.
I have this issue, freezing system while audio continues playing, also without any game open. Just browsing, vscodium and terminal.