Arch Linux

jjcasmar

Member

Registered: 2019-11-24

Posts: 8

nvidia module used, but nvidia-smi shows nothing

I have a system with 2 GPUs, an AMD GPU and a nvidia GPU. I use plasma-wayland and I want to run the desktop and all the applications in the AMD GPU, and use only the nvidia one for selected applications.

The displays are connected to the AMD GPU and sddm is configured to use wayland. According to `nvidia-smi`, nothing is running on the nvidia card

+-----------------------------------------------------------------------------+
| NVIDIA-SMI 525.89.02    Driver Version: 525.89.02    CUDA Version: 12.0     |
|-------------------------------+----------------------+----------------------+
| 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 ...  Off  | 00000000:2D:00.0 Off |                  N/A |
|  0%   39C    P8    22W / 215W |      0MiB /  8192MiB |      0%      Default |
|                               |                      |                  N/A |
+-------------------------------+----------------------+----------------------+
                                                                               
+-----------------------------------------------------------------------------+
| Processes:                                                                  |
|  GPU   GI   CI        PID   Type   Process name                  GPU Memory |
|        ID   ID                                                   Usage      |
|=============================================================================|
|  No running processes found                                                 |
+-----------------------------------------------------------------------------+

However, `lsmod | grep nvidia` shows that the nvidia module is used

i2c_nvidia_gpu         16384  0
nvidia_drm             73728  0
nvidia_uvm           2908160  0
nvidia_modeset       1515520  1 nvidia_drm
nvidia              61444096  260 nvidia_uvm,nvidia_modeset
video                  65536  3 amdgpu,radeon,nvidia_modeset

If I logout from the plasma session, the nvidia module is not used anymore.

I want to be able to unload the nvidia module to be able to load vfio-pci on demand to do passthrough, however, since the nvidia module is used, its not possible to unloaded.

How can I know why applications are using the nvidia module? I guess it might be something related to XWayland, but when running `glxgears`, nvidia-smi still shows that nothing is running on the nvidia card.

Last edited by jjcasmar (2023-02-27 07:28:01)