Trying to set up a VM with a passed through L4 GPU and nvidia-smi returns “No devices found”. There is 2 L4’s in an Oracle X9-2 server. One is being passed through to the VM. The server OS is Ubuntu 22.04 and the VM OS is Ubuntu 20.04. Nvidia driver is the latest 535.
I’m stuck at what to do next to get this VM working with the pass through L4 GPU.
Any help is greatly appreciated.
Looking at the dmesg log I see this with the “PCI I/O region assigned to your NVIDIA device is invalid”
In grub I have set the pci=realloc and also tried pci=realloc=off and still the same outcome.
nvidia-bug-report.log.gz (239.5 KB)
Part of dmesg log.
[ 4.395365] nvidia: loading out-of-tree module taints kernel.
[ 4.395379] nvidia: module license ‘NVIDIA’ taints kernel.
[ 4.395380] Disabling lock debugging due to kernel taint
[ 4.446312] nvidia: module verification failed: signature and/or required key missing - tainting kernel
[ 4.483973] nvidia-nvlink: Nvlink Core is being initialized, major device number 235
[ 4.485503] nvidia 0000:05:00.0: enabling device (0140 → 0142)
[ 4.488937] NVRM: This PCI I/O region assigned to your NVIDIA device is invalid:
NVRM: BAR1 is 0M @ 0x0 (PCI:0000:05:00.0)
[ 4.488942] NVRM: This PCI I/O region assigned to your NVIDIA device is invalid:
NVRM: BAR2 is 0M @ 0x0 (PCI:0000:05:00.0)
[ 4.488944] NVRM: This PCI I/O region assigned to your NVIDIA device is invalid:
NVRM: BAR3 is 0M @ 0x0 (PCI:0000:05:00.0)
[ 4.488945] NVRM: This PCI I/O region assigned to your NVIDIA device is invalid:
NVRM: BAR4 is 0M @ 0x0 (PCI:0000:05:00.0)
[ 4.488947] NVRM: This PCI I/O region assigned to your NVIDIA device is invalid:
NVRM: BAR5 is 0M @ 0x0 (PCI:0000:05:00.0)
[ 4.537667] NVRM: loading NVIDIA UNIX x86_64 Kernel Module 535.161.07 Sat Feb 17 22:55:48 UTC 2024
[ 4.585273] nvidia-modeset: Loading NVIDIA Kernel Mode Setting Driver for UNIX platforms 535.
Full dmesg
dmsg.txt (80.6 KB)
lspci of the VM
vm-lspci.txt (4.0 KB)
Server info
Oracle X9-2, 2 processor Xeon
server-lspci.txt (35.7 KB)