Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

feat: add nvidia-open images #69

Merged
merged 2 commits into from
Jan 2, 2025
Merged

feat: add nvidia-open images #69

merged 2 commits into from
Jan 2, 2025

Conversation

ledif
Copy link
Collaborator

@ledif ledif commented Jan 2, 2025

Pulling in changes by @m2Giles from ublue-os/bluefin#2092.

I confirmed that just build aurora latest nvidia-open produces an image without errors, but did not verify on hardware.

@ledif ledif marked this pull request as ready for review January 2, 2025 05:45
@ledif ledif requested a review from castrojo as a code owner January 2, 2025 05:45
@dosubot dosubot bot added size:M This PR changes 30-99 lines, ignoring generated files. enhancement New feature or request github_actions Pull requests that update GitHub Actions code labels Jan 2, 2025
Copy link
Collaborator

@RealVishy RealVishy left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

should we test this on :beta with hardware and then push to :stable and :latest ?

@NiHaiden
Copy link
Collaborator

NiHaiden commented Jan 2, 2025

should we test this on :beta with hardware and then push to :stable and :latest ?

Yeah i think we should do this with beta for now and then push it out when people who have nvidia hardware can use it. I sadly can't test nvidia stuff since I don't have any hardware from them anymore 😅

@ledif
Copy link
Collaborator Author

ledif commented Jan 2, 2025

I have a machine with an RTX 4080 and a spare SSD that I can use to test this on. I'll try to mess around with it in the next few days to confirm that it's working as expected on hardware.

@ledif
Copy link
Collaborator Author

ledif commented Jan 2, 2025

I was able to test it out today and it appears to be working, from what I can tell.

Driver Identification

The system log shows a distinction in the how the driver from the kernel module identifies itself.

With aurora-nvidia

❯ sudo dmesg | grep NVIDIA -i -A 10
[   11.398242] nvidia-nvlink: Nvlink Core is being initialized, major device number 511
[   11.399120] nvidia 0000:01:00.0: vgaarb: VGA decodes changed: olddecodes=io+mem,decodes=none:owns=none
[   11.446509] NVRM: loading NVIDIA UNIX x86_64 Kernel Module  565.77  Wed Nov 27 23:33:08 UTC 2024
[   11.538970] nvidia_uvm: module uses symbols nvUvmInterfaceDisableAccessCntr from proprietary module nvidia, inheriting taint.
[   11.639839] nvidia-uvm: Loaded the UVM driver, major device number 509.
[   11.698898] nvidia-modeset: Loading NVIDIA Kernel Mode Setting Driver for UNIX platforms  565.77  Wed Nov 27 22:53:48 UTC 2024

With aurora-nvidia-open

❯ sudo dmesg | grep NVIDIA -i -A 10
[    9.887388] nvidia-nvlink: Nvlink Core is being initialized, major device number 511
[    9.887989] nvidia 0000:01:00.0: vgaarb: VGA decodes changed: olddecodes=io+mem,decodes=none:owns=none
[    9.935547] NVRM: loading NVIDIA UNIX Open Kernel Module for x86_64  565.77  Release Build  (dvs-builder@U16-A21-31-3)  Wed Nov 27 23:09:12 UTC 2024
[   10.017230] nvidia-modeset: Loading NVIDIA UNIX Open Kernel Mode Setting Driver for x86_64  565.77  Release Build  (dvs-builder@U16-A21-31-3)  Wed Nov 27 22:57:13 UTC 2024
[   10.022840] [drm] [nvidia-drm] [GPU ID 0x00000100] Loading driver
[   11.215420] [drm] Initialized nvidia-drm 0.0.0 for 0000:01:00.0 on minor 1
[   11.215440] nvidia 0000:01:00.0: vgaarb: deactivate vga console
[   11.245966] fbcon: nvidia-drmdrmfb (fb1) is primary device
[   11.245967] fbcon: Deferring console take-over
[   11.245969] nvidia 0000:01:00.0: [drm] fb1: nvidia-drmdrmfb frame buffer device
[   11.496343] nvidia-uvm: Loaded the UVM driver, major device number 509.

Notice that the non-open version identifies itself as NVIDIA UNIX x86_64 Kernel Module 565.77 whereas the open version identifies itself as NVIDIA UNIX Open Kernel Module for x86_64 565.77 Release Build (dvs-builder@U16-A21-31-3).

NVIDIA Tooling

I was able to confirm that tools like nvidia-smi still work as expected with the new nvidia-open image:

❯ nvidia-smi
Thu Jan  2 12:49:26 2025
+-----------------------------------------------------------------------------------------+
| NVIDIA-SMI 565.77                 Driver Version: 565.77         CUDA Version: 12.7     |
|-----------------------------------------+------------------------+----------------------+
| GPU  Name                 Persistence-M | Bus-Id          Disp.A | Volatile Uncorr. ECC |
| Fan  Temp   Perf          Pwr:Usage/Cap |           Memory-Usage | GPU-Util  Compute M. |
|                                         |                        |               MIG M. |
|=========================================+========================+======================|
|   0  NVIDIA GeForce RTX 4080        Off |   00000000:01:00.0  On |                  N/A |
|  0%   42C    P8             16W /  320W |    1195MiB /  16376MiB |      0%      Default |
|                                         |                        |                  N/A |
+-----------------------------------------+------------------------+----------------------+

Based on this, it seems to be working as expected.

@NiHaiden
Copy link
Collaborator

NiHaiden commented Jan 2, 2025

I think it's safe to say we can push this into stable and latest and maybe put something in the forums or on the website to be aware of the new open modules. but they seem to work fine for bluefin and bazzite too.

@dosubot dosubot bot added the lgtm This PR has been approved by a maintainer label Jan 2, 2025
@RealVishy RealVishy added this pull request to the merge queue Jan 2, 2025
Merged via the queue into ublue-os:main with commit 75c473b Jan 2, 2025
35 checks passed
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
enhancement New feature or request github_actions Pull requests that update GitHub Actions code lgtm This PR has been approved by a maintainer size:M This PR changes 30-99 lines, ignoring generated files.
Projects
None yet
Development

Successfully merging this pull request may close these issues.

3 participants