I have an NVIDIA Quadro K2000 that will work for a brief moment and then stop working via PCI passthrough, either through Windows or Linux. My latest attempts have been in Ubuntu 20.04 and 22.04 to get more detailed logging. This graphics card is not the primary one PVE used during install.
If I boot the Ubuntu VM and try
But if I type it again, it will say no devices found. If I tail the syslog I see this.
The NVIDIA persistenced service is still running.
On the Windows side, I followed the guide I believe on these forums for getting it set up, I can never get the driver to not show an error in Device Manager, but if I plug a monitor into the card I can watch Windows boot up to the login screen.
I'm fairly certain I have PCI Passthrough working properly because in another PCI slot I have a SATA controller passed through to another VM running TrueNAS with no issues, but I understand graphic cards can be a whole different thing.
I'm running VE 8.1.10. Happy to share additional info or logs for any guidance. I have watched multiple videos and read many threads, and after about 20 attempts I'm asking for help
If I boot the Ubuntu VM and try
nvidia-smi I see the graphics card and the driver and all the info.
Code:
timw@ubuntu:~$ nvidia-smi
Wed Apr 24 12:11:52 2024
+-----------------------------------------------------------------------------+
| NVIDIA-SMI 470.239.06 Driver Version: 470.239.06 CUDA Version: 11.4 |
|-------------------------------+----------------------+----------------------+
| 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 Quadro K2000 Off | 00000000:01:00.0 Off | N/A |
| 30% 45C P0 N/A / N/A | 0MiB / 2000MiB | 0% Default |
| | | N/A |
+-------------------------------+----------------------+----------------------+
+-----------------------------------------------------------------------------+
| Processes: |
| GPU GI CI PID Type Process name GPU Memory |
| ID ID Usage |
|=============================================================================|
| No running processes found |
+-----------------------------------------------------------------------------+But if I type it again, it will say no devices found. If I tail the syslog I see this.
Code:
May 12 22:28:17 ubuntu2004 kernel: [ 41.130802] NVRM: GPU 0000:01:00.0: RmInitAdapter failed! (0x25:0x65:1241)
May 12 22:28:17 ubuntu2004 kernel: [ 41.131167] NVRM: GPU 0000:01:00.0: rm_init_adapter failed, device minor number 0
May 12 22:28:19 ubuntu2004 kernel: [ 42.744861] NVRM: GPU 0000:01:00.0: RmInitAdapter failed! (0x24:0xffff:1211)
May 12 22:28:19 ubuntu2004 kernel: [ 42.745228] NVRM: GPU 0000:01:00.0: rm_init_adapter failed, device minor number 0The NVIDIA persistenced service is still running.
On the Windows side, I followed the guide I believe on these forums for getting it set up, I can never get the driver to not show an error in Device Manager, but if I plug a monitor into the card I can watch Windows boot up to the login screen.
I'm fairly certain I have PCI Passthrough working properly because in another PCI slot I have a SATA controller passed through to another VM running TrueNAS with no issues, but I understand graphic cards can be a whole different thing.
I'm running VE 8.1.10. Happy to share additional info or logs for any guidance. I have watched multiple videos and read many threads, and after about 20 attempts I'm asking for help