Can't run any program via NVIDIA card

nikolasumus · February 22, 2021, 8:01pm

I installed GPU drivers from NVIDIA repository (CUDA Toolkit 11.2 Update 1 Downloads | NVIDIA Developer) by using “sudo zypper install -y cuda” command.

If I run this “NV_PRIME_RENDER_OFFLOAD=1 GLX_VENDOR_LIBRARY_NAME=nvidia glxinfo | grep vendor” command I get an error:

user@localhost:~> __NV_PRIME_RENDER_OFFLOAD=1 __GLX_VENDOPreformatted textR_LIBRARY_NAME=nvidia glxinfo | grep vendor
X Error of failed request:  BadValue (integer parameter out of range for operation)
  Major opcode of failed request:  152 (GLX)
  Minor opcode of failed request:  24 (X_GLXCreateNewContext)
  Value in failed request:  0x0
  Serial number of failed request:  39
  Current serial number in output stream:  40.

But if I run “lsmod | grep nvidia” command I can see that driver is seems to be running. Output:

user@localhost:~> lsmod | grep nvidia
nvidia_drm             65536  0
nvidia_modeset       1232896  1 nvidia_drm
nvidia_uvm           1155072  0
nvidia              34082816  2 nvidia_uvm,nvidia_modeset
drm_kms_helper        274432  2 amdgpu,nvidia_drm
drm                   561152  16 gpu_sched,drm_kms_helper,amdgpu,nvidia_drm,ttm

Also here is the output of the “nvidia-smi” command:

user@localhost:~> nvidia-smi
Mon Feb 22 19:49:01 2021       
+-----------------------------------------------------------------------------+
| NVIDIA-SMI 460.32.03    Driver Version: 460.32.03    CUDA Version: 11.2     |
|-------------------------------+----------------------+----------------------+
| 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  GeForce GTX 1650    Off  | 00000000:01:00.0 Off |                  N/A |
| N/A   49C    P0    16W /  N/A |      0MiB /  3911MiB |      0%      Default |
|                               |                      |                  N/A |
+-------------------------------+----------------------+----------------------+
                                                                               
+-----------------------------------------------------------------------------+
| Processes:                                                                  |
|  GPU   GI   CI        PID   Type   Process name                  GPU Memory |
|        ID   ID                                                   Usage      |
|=============================================================================|
|  No running processes found                                                 |
+-----------------------------------------------------------------------------+

nvidia-bug-report.log.gz (1.5 MB)

generix · February 22, 2021, 10:31pm

Please run nvidia-bug-report.sh as root and attach the resulting nvidia-bug-report.log.gz file to your post.

nikolasumus · February 23, 2021, 9:59am

Attached.

generix · February 23, 2021, 10:07am

The driver loads too late, please try embedding it into the initrd using the add_drivers+ option in /etc/dracut.conf and recreating the initrd.
https://man7.org/linux/man-pages/man5/dracut.conf.5.html

nikolasumus · February 23, 2021, 1:01pm

Still don’t work
Now my 60-nvidia-default.conf in /etc/dracut.conf.d looks like this:

#line omit_drivers+=" nvidia nvidia-drm nvidia-modeset nvidia-uvm " is a default one
omit_drivers+=" nvidia nvidia-drm nvidia-modeset nvidia-uvm "
add_drivers+=" nvidia "
add_drivers+=" nvidia_uvm "
add_drivers+=" nvidia_modeset "
add_drivers+=" drm "
add_drivers+=" drm_kms_helper "
add_drivers+=" nvidia_drm "

Did I do something wrong ?

generix · February 23, 2021, 1:07pm

The omit_drivers line has to be commented, it disables the nvidia drivers from being added to the initrd. so the file should look like

#omit_drivers+=" nvidia nvidia-drm nvidia-modeset nvidia-uvm "
add_drivers+=" nvidia nvidia-drm nvidia-modeset nvidia-uvm "

and then run sudo dracut -f

nikolasumus · February 24, 2021, 11:43am

It works ! It fixed the problem.
Thank you so much!

dr.wes.snyder.v · February 24, 2021, 10:02pm

Hello,

±----------------------------------------------------------------------------+
| Processes: |
| GPU GI CI PID Type Process name GPU Memory |
| ID ID Usage |
|=============================================================================|
| 0 N/A N/A 1631 G /usr/lib/xorg/Xorg 50MiB |
| 0 N/A N/A 5872 G /usr/lib/xorg/Xorg 258MiB |
| 0 N/A N/A 6115 G /usr/bin/gnome-shell 62MiB |
| 0 N/A N/A 11192 G …gAAAAAAAAA --shared-files 53MiB |
| 0 N/A N/A 19001 G …AAAAAAAAA= --shared-files 70MiB |
±----------------------------------------------------------------------------+

xxx@xxxxx:~$ lsmod | grep nvidia
nvidia_uvm 1011712 0
nvidia_drm 57344 9
nvidia_modeset 1228800 15 nvidia_drm
nvidia 34041856 754 nvidia_uvm,nvidia_modeset
drm_kms_helper 217088 1 nvidia_drm
drm 552960 12 drm_kms_helper,nvidia_drm

Error:
xxx@xxxxxx:~$ steam
Running Steam on ubuntu 20.04 64-bit
STEAM_RUNTIME is enabled automatically
Pins up-to-date!
Steam client’s requirements are satisfied
/home/wes/.local/share/Steam/ubuntu12_32/steam -nominidumps -nobreakpad
[2021-02-24 17:01:40] Startup - updater built Feb 12 2021 23:21:33
[2021-02-24 17:01:40] Opted in to client beta ‘publicbeta’ via beta file
You are in the ‘publicbeta’ client beta.
SteamUpdateUI: An X Error occurred
X Error of failed request: GLXBadContext
SteamUpdateUI: An X Error occurred
X Error of failed request: BadValue (integer parameter out of range for operation)
Major opcode of failed request: 151 (GLX)
Minor opcode of failed request: 3 (X_GLXCreateContext)
Value in failed request: 0x0
Serial number of failed request: 45
xerror_handler: X failed, continuing
Major opcode of failed request: 151 (GLX)
Minor opcode of failed request: 6 (X_GLXIsDirect)
Serial number of failed request: 46
xerror_handler: X failed, continuing

generix · February 24, 2021, 11:03pm

The Xserver is running correctly on the nvidia gpu. Steam is 32bit, so you need to have the 32bit compatibility libs installed. How did you install the driver?

dr.wes.snyder.v · February 25, 2021, 12:19am

wget -O /etc/apt/preferences.d/cuda-repository-pin-600 https://developer.download.nvidia.com/compute/cuda/repos/ubuntu1804/x86_64/cuda-ubuntu1804.pin
sudo apt-key adv --fetch-keys https://developer.download.nvidia.com/compute/cuda/repos/ubuntu1804/x86_64/7fa2af80.pub sudo add-apt-repository “deb Index of /compute/cuda/repos/ubuntu1804/x86_64 /”
$ sudo apt install cuda

I see, only the 64bit was installed. Now to find the 32bit installer.

dr.wes.snyder.v · February 25, 2021, 1:39am

If no easy way exists to have both 32 and 64 bit installed, I believe I found flatpak has Steam. Going to try that in a VM.

The flatpak gave error. :(

Software & Update:

I work on this later today.

Thanks @generix . I followed this and now all is working: Installing the NVIDIA driver, CUDA and cuDNN on Linux · GitHub. I did not install cuDNN.

generix · February 25, 2021, 7:50am

By installing the full cuda bundle, you installed the compute driver, which doesn’t have the 32bit compatibility libs. You need to install the graphics driver from Ubuntu’s software&Updates.

nadeemm · October 12, 2021, 6:30pm