Same kernel and blob, but 3 monitors. 1x 144 Hz gsync and 2x 60 Hz non-gsync. No flickering at all.
It must be something in my configs then, thanks.
Honestly this issue is unbelievable annoying, seems like the 535.xx drivers completely broke Starfield aswell.
Well, 525 can launch the game, but with an absolutely abysmal performance. Pick your poison, I guessā¦
I think it also broke cyberpunk 2077 ā¦ Well , maybe itās not all nvidia cards , but thereās definitely a problem with the rtx 20xx and causing a xid 31.
Itās been the case since May.
[ 9232.991334] [drm:drm_new_set_master] ERROR [nvidia-drm] [GPU ID 0x00000100] Failed to grab modeset ownership
[45128.855944] Process accounting resumed
[95925.650671] NVRM: GPU at PCI:0000:01:00: GPU-3440fcd9-ad72-5684-052e-87619260bcbf
[95925.650674] NVRM: Xid (PCI:0000:01:00): 109, pid=67885, name=Warframe.x64.ex, Ch 000000ce, errorString CTX SWITCH TIMEOUT, Info 0x1c067
Nvidia module 535.104.05, Arch kernel 6.5.2-zen1-1-zen. First occurrence in a few days for this game.
nvidia-bug-report.log.gz (2.0 MB)
Has anyone tried the newly released driver version 535.113.01? They didnāt address the issue in the patch notes though
@amrits is there any new update on this issue?
still broken.
This is still being investigated and I am following up internally about the progress.
1 Like
Been following this thread for awhile, only started noticing freezes due to the109 errors about a month or so ago. All games affected will either freeze while loading in, or the moment it has finished loading. Iāve tried the 520.56.06 driver workaround mentioned in this thread and everything seemed fine until about 20-ish minutes in I abruptly froze again, after which it relapsed to its previous behavior. I have since switched kernels (from the normal 22.04 LTS Kubuntu to 22.04 LTS Kubuntu Focus) in an attempt to fix the issue, so far it seems to have done nothing. While I am currently on a more dated 535 driver, I had previously tried the more up to date 535.104.xx drivers with no luck. Might try switching to 23.04ā¦ though at this point I doubt itāll do anything.
Running driver 535.86.05, 6.2.0-31-generic, i5-12600k, 3080, and 32Gb Ram on 1 2560x1080 144Hz monitor.
nvidia-bug-report.log.gz (552.1 KB)
We have a separate bug [4304454] filed for Horizon Zero Dawn freeze issue with xid 31 errors.
Shall keep updated on the progress.
1 Like
Is there anything you can tell us about how far along they are in terms of developing a fix for this? Have they found any sort of pattern in terms of hardware combinations that might cause this issue? I find it odd more users arenāt mentioning experiencing these issues, especially considering that (at least for me) 95% of my steam library has been wrecked.
1 Like
RoboCob Rogue City (Demo)
Proton: 8.0-3
VKD3D: 2.9
Driver: 535.113.01
Vulkan: 1.3.242
NVRM: Xid (PCI:0000:26:00): 31, pid=18091, name=GameThread, Ch 00000069, intr 00000000. MMU Fault: ENGINE HOST0 HUBCLIENT_ESC faulted @ 0x1_03fb0000. Fault is of type FAULT_PTE ACCESS_TYPE_VIRT_READ
nvidia-bug-report.log.gz (604.1 KB)
Edit: Using Proton Experimental with VKD3D 2.10 seems to at least circumvent this issue.
Also, I saw this Xid 109 error in GRID 16.1 vib in ESXi. The guest VMs are crashing whenever we load up Nuke. Able to reproduce the issue in dual 4K monitors over PCoIP and start Nuke. The VM will hang and the ESXi will have the following error.
2023-10-09T04:20:18.418Z cpu91:2113650)NVRM: Xid (PCI:0000:21:00): 109, pid=2110318, name=, Ch 00000602, errorString CTX SWITCH TIMEOUT, Info 0x12c035
Is this related to the reported issue?
This is on CentOS 7.9 BTW with A40 running 12Q.
Sorry for the delayed response on this issue. We posted a new build from our Vulkan developer branch with a possible fix for the āXid 109 CTX SWITCH TIMEOUTā errors, available at https://developer.nvidia.com/vulkan-driver (link to the Linux build).
For those able to test this build, could you please let us know if it helps with the Xid 109 errors observed in some games and applications?
Thank you.
Hello,
Was able to test 535.43.11 with my ampere rtx 3090, the Xid now shows in games that it didnāt previously - mainly Starfield and RoboCop Rogue City Demo(after the last update yesterday).
1 Like
@dleone Thank you for taking the time to address our concerns. Iām afraid, however, that the driver that youāre referring to, 535.43.11, has been the main topic in the vastly popular thread about Starfield for the past few days. Many people have tested it and shared already their reports.
The good news is that it does seem to fix most ā though not all ā problems with Starfield on the Turing architecture. The bad news is that on Ampere, compared to 535.43.10, it reintroduces the Xid 109 errors, known from the stable 535 series. This seems like a clear regression compared to 535.43.10. The latter did have its own problems, mainly performance-related, but at least the Xid 109 seemed to be fixed also on Ampere.
In fact, even the developer of vkd3d-proton noticed the reintroduced crashes and had reported them to you. Other people in the Starfield thread seem to be noticing Xid 109s in other games too, which may be even more concerning.
Iām sure everyone here appreciates highly the hard work of Nvidiaās Linux team and I believe that each one of you is accomplishing well beyond what can be expected from the undoubtedly limited resources that you have. I personally have always liked your drivers: sometimes slow to catch up, but dependable and performing.
But the situation with Starfield ā and Cyberpunkās Xid 31 errors, for that matter ā isnāt that reassuring, Iām afraid. While I have no doubt that your team is very dedicated and giving your very best, in my humble opinion your management does need to seriously consider allocating more resources to the Linux development.
Linux gaming might not be that big, but itās certainly showing a growing trend. Most importantly perhaps, many of us are not only gamers. Nvidiaās reputation in fields like AI is also relevant to many of us. And the experience we have with Nvidiaās Linux drivers can influence our future purchasing decisions, even in these broader domains. It may not be the primary factor, but it will be an important factor nonetheless. Itās a matter of trust.
4 Likes
Iām currently using Arch linux with a rtx 2060 mobile. Iāve been receiving Xid 109 errors for a while. I was able to identify it after playing Diablo 4 and lately I observed it also in Cyberpunk 2077.
Iām currently using the 535.113.01 driver, and itās happening from time to time. Iāve observed that depending on the configuration Iām using the errors becomes more or less frequent.
When playing Cyberpunk 2077 I observed that when playing with some browser windows open in the background the errors are more likely to happen. So what Iāve been doing is letting everything closed but Steam and then starting the game. The error may happen after 20 minutes gaming. After reopening the game it usually stays fine until the end of a gaming session.
Itās frustrating to see an error such as that for such a long time. Itās just one of several issues Iāve been having with Nvidia in Linux. It makes me wonder if I should really keep up with Nvidia GPUās when building my next deskop.
2 Likes
So :
- Starfield crashes with a Xid 109 using the 535.113.01 and 535.43.11 drivers on all cards except 20xx series .
- Cyberpunk 2077 crashes with a Xid 31 using pretty much every driver that has DLSS implemented on 20xx series. (Maybe release a patch allowing to disable DLSS compilation in the meantime ? )
1 Like
I have tested the 535-43-11.
Starfield runs on amazing 20FPS on low with 2060s.
CP77 As always Xid 31.
have you tried building vkd3d-proton by hand ?
latest commits give you a nice boost of performance , compared to the previous state which was at the level of a PowerPoint slides.