Nvidia driver installation

[mgrenier25]

I can't get the nvidia driver to install on Proxmox 8.0.2

I followed the instructions here
The script builds the module up to like 90% then fails.

/var/log/nvidia-installer.log doesn't seem to have any relevant info about why it failed

I used this driver, Nvidia lists my gpu as supported(Quadro m4000)

 

[dcsapak]

can you post the nvidia install log?
what is the exact output?

why wouldn't the nvidia driver from the debian repository be good enough?
in the non-free bookworm repository there is the 'nvidia-driver' package: https://packages.debian.org/bookworm/nvidia-driver
in version 525.125.06

 

[LnxBil]

The driver you tried is an original nvidia driver, so any problems with them needs to be discussed with them. There is nothing PVE can do about it.

Have you tried the Debian way? Installing nvidia-detect and follow the instructions. Maybe you need to add the backports to get a working environment. I did this a while ago with a quadro K2000 and it worked.

 

[mgrenier25]

if I use apt to install it and it doesn't seem to build the nvidia kernel modules.
I tried with stable and backports, both on a fresh install, same result. systemd-modules-load spits out somme errors because it cannot find nvidia modules.

 

[Uruk]

Same problem here the only version working for now is NVIDIA-Linux-x86_64-535.113.01 but all of the above version is not working,

tried to install the latest one for my RTX 3060 and here are the log output from the installer at the end

-> Kernel module compilation complete.
-> Kernel messages:
[  127.777035] nvidia_uvm: Unknown symbol nvUvmInterfaceUnregisterGpu (err -2)
[  127.777040] nvidia_uvm: Unknown symbol nvUvmInterfaceAddressSpaceDestroy (err -2)
[  127.777044] nvidia_uvm: Unknown symbol nvUvmInterfaceRegisterUvmCallbacks (err -2)
[  127.777047] nvidia_uvm: Unknown symbol nvUvmInterfacePmaPinPages (err -2)
[  127.777053] nvidia_uvm: Unknown symbol nvUvmInterfaceChannelAllocate (err -2)
[  127.777056] nvidia_uvm: Unknown symbol nvUvmInterfaceGetP2PCaps (err -2)
[  127.777059] nvidia_uvm: Unknown symbol nvUvmInterfaceEnableAccessCntr (err -2)
[  127.777068] nvidia_uvm: Unknown symbol nvUvmInterfacePmaUnregisterEvictionCallbacks (err -2)
[  127.777073] nvidia_uvm: Unknown symbol nvUvmInterfaceDupAllocation (err -2)
[  127.777078] nvidia_uvm: Unknown symbol nvUvmInterfacePmaAllocPages (err -2)
[  127.777088] nvidia_uvm: Unknown symbol nvUvmInterfaceSessionCreate (err -2)
[  127.867591] nvidia-nvlink: Nvlink Core is being initialized, major device number 236

[  127.868429] nvidia 0000:01:00.0: vgaarb: changed VGA decodes: olddecodes=none,decodes=none:owns=none
[  127.917253] NVRM: loading NVIDIA UNIX x86_64 Kernel Module  535.113.01  Tue Sep 12 19:41:24 UTC 2023
[  128.791474] nvidia_uvm: module uses symbols nvUvmInterfaceDisableAccessCntr from proprietary module nvidia, inheriting taint.
[  128.813267] nvidia-uvm: Loaded the UVM driver, major device number 508.
[  153.968824] VFIO - User Level meta-driver version: 0.3
[  153.993156] nvidia-uvm: Unloaded the UVM driver.
[  154.014228] nvidia-nvlink: Unregistered Nvlink Core, major device number 236
[  154.131662] nvidia-nvlink: Nvlink Core is being initialized, major device number 236

[  154.132689] nvidia 0000:01:00.0: vgaarb: changed VGA decodes: olddecodes=none,decodes=none:owns=none
[  154.179205] NVRM: loading NVIDIA UNIX x86_64 Kernel Module  535.113.01  Tue Sep 12 19:41:24 UTC 2023
[  154.189203] nvidia_uvm: module uses symbols nvUvmInterfaceDisableAccessCntr from proprietary module nvidia, inheriting taint.
-> One or more kernel modules were already loaded before the module test load; trying again with the udev event queue paused.
-> Kernel messages:
[  127.777040] nvidia_uvm: Unknown symbol nvUvmInterfaceAddressSpaceDestroy (err -2)
[  127.777044] nvidia_uvm: Unknown symbol nvUvmInterfaceRegisterUvmCallbacks (err -2)
[  127.777047] nvidia_uvm: Unknown symbol nvUvmInterfacePmaPinPages (err -2)
[  127.777053] nvidia_uvm: Unknown symbol nvUvmInterfaceChannelAllocate (err -2)
[  127.777056] nvidia_uvm: Unknown symbol nvUvmInterfaceGetP2PCaps (err -2)
[  127.777059] nvidia_uvm: Unknown symbol nvUvmInterfaceEnableAccessCntr (err -2)
[  127.777068] nvidia_uvm: Unknown symbol nvUvmInterfacePmaUnregisterEvictionCallbacks (err -2)
[  127.777073] nvidia_uvm: Unknown symbol nvUvmInterfaceDupAllocation (err -2)
[  127.777078] nvidia_uvm: Unknown symbol nvUvmInterfacePmaAllocPages (err -2)
[  127.777088] nvidia_uvm: Unknown symbol nvUvmInterfaceSessionCreate (err -2)
[  127.867591] nvidia-nvlink: Nvlink Core is being initialized, major device number 236

[  127.868429] nvidia 0000:01:00.0: vgaarb: changed VGA decodes: olddecodes=none,decodes=none:owns=none
[  127.917253] NVRM: loading NVIDIA UNIX x86_64 Kernel Module  535.113.01  Tue Sep 12 19:41:24 UTC 2023
[  128.791474] nvidia_uvm: module uses symbols nvUvmInterfaceDisableAccessCntr from proprietary module nvidia, inheriting taint.
[  128.813267] nvidia-uvm: Loaded the UVM driver, major device number 508.
[  153.968824] VFIO - User Level meta-driver version: 0.3
[  153.993156] nvidia-uvm: Unloaded the UVM driver.
[  154.014228] nvidia-nvlink: Unregistered Nvlink Core, major device number 236
[  154.131662] nvidia-nvlink: Nvlink Core is being initialized, major device number 236

[  154.132689] nvidia 0000:01:00.0: vgaarb: changed VGA decodes: olddecodes=none,decodes=none:owns=none
[  154.179205] NVRM: loading NVIDIA UNIX x86_64 Kernel Module  535.113.01  Tue Sep 12 19:41:24 UTC 2023
[  154.189203] nvidia_uvm: module uses symbols nvUvmInterfaceDisableAccessCntr from proprietary module nvidia, inheriting taint.
[  154.467738] VFIO - User Level meta-driver version: 0.3
ERROR: Installation has failed.  Please see the file '/var/log/nvidia-installer.log' for details.  You may find suggestions on fixing installation problems in the README available on the Linux driver download page at www.nvidia.com.

 

[Uruk]

Problem as been fixed switch to newer pve-kernel-6.5