nvidia open: driver won't start after update 570.124.04

[LachCrafter]

Nixpkgs version

• Unstable (25.05)

Describe the bug

The operating system will load but the drivers won't on startup.
When I switch to a tty, the nvidia drm gives the an error shown down below.

Steps to reproduce

1. install the latest open nvidia driver package on nixos-unstable
2. restart

Expected behaviour

The drivers load

Screenshots

No response

Relevant log output

[drm:nv_drm_atomic_commit [nvidia_drm]] *ERROR* [nvidia-drm] [GPU ID 0x00000100] Flip Event timout on head 1

Additional context

No response

System metadata

• system: "x86_64-linux"
• host os: Linux 6.13.5, NixOS, 25.05 (Warbler), 25.05beta762848.d69ab0d71b22
• multi-user?: yes
• sandbox: yes
• version: nix-env (Nix) 2.24.12
• channels(root): "nixos"
• nixpkgs: /nix/var/nix/profiles/per-user/root/channels/nixos

Notify maintainers

No response

I assert that this issue is relevant for Nixpkgs

• I assert that this is a bug and not a support request.
• I assert that this is not a duplicate of an existing issue.
• I assert that I have read the NixOS Code of Conduct and agree to abide by it.

Is this issue important to you?

Add a 👍 reaction to issues you find important.

[gtebbutt]

I'm getting a similar issue with a slightly different error on the same driver version:

[    6.599434] [drm] [nvidia-drm] [GPU ID 0x00000300] Loading driver
[    7.459290] [drm:nv_drm_load [nvidia_drm]] *ERROR* [nvidia-drm] [GPU ID 0x00000300] Failed to allocate NvKmsKapiDevice
[    7.459999] [drm:nv_drm_register_drm_device [nvidia_drm]] *ERROR* [nvidia-drm] [GPU ID 0x00000300] Failed to register device

Looks like there are one or two reports of similar problems in the NVIDIA forum, but I couldn't see any helpful answers there.

[LachCrafter]

Fixed for me in latest kernel release.

[ajgon]

@LachCrafter Can you say which driver and which kernel did you use? I tried both nvidia stable and latest with most recent pkgs.linuxPackages_6_12 (which is 6.12.19) and pkgs.linuxPackages_6_13 (which is 6.13.7) from nix "unstable" branch, and still getting this error.

[gtebbutt]

Not OP, but I'm no longer getting the error I mentioned above, or any dmesg error I can see - I'm now using driver 570.124.04 on kernel 6.12.17 (clean install from nixos-unstable about a week ago). I'm also not getting any display output at all after the scroll of boot text, so there's definitely still something wrong on my install, though; same config was working fine using nixpkgs 24.11 on an older GPU.

I'm using it for CUDA work, and running that over SSH seems fine even without the display working, so I haven't done any real debugging beyond that yet. Was planning to try reinstalling once NVIDIA release the next iteration of the driver.

[LachCrafter]

This seems to start happening too after a restart when changing the display refresh rate.
It seems to be an issue mostly with the driver.

[LachCrafter]

Nvidia released the 570.133.07 driver yesterday, it provides a few fixes. I will give it a try.

[ajgon]

570.133.07 worked for me without any issues, so looks like it was a bug in a driver 🥳

[gtebbutt]

Also now working on 570.133.07, but I did initially need to set

services.xserver.resolutions = [
    {
      x = 1024;
      y = 786;
    }
  ];

in order to get any display output. Changing it to 1920x1080 after gnome had loaded was then fine, but trying 4K hard crashed the system (couldn't even SSH in to reboot, needed to physically power off). That last issue could plausibly be a flaky HDMI cable given that it's now stable at 1080p with the new driver, where previously I wasn't able to get a display output at all.

[LachCrafter]

The new release fixed my initial issue too. I'm closing the issue.