Nvidia Quadro P400 not working on Ubuntu server via GPU/PCIe passthrough
-
@olivierlambert Yes indeed and it got no further reactions...
For Proxmox I followed this guide: https://www.reddit.com/r/homelab/comments/b5xpua/the_ultimate_beginners_guide_to_gpu_passthrough/
And it worked out of the box very solid, of course this can be different for XCP-NG.
EDIT: I also tried different NVidia drivers with no luck...
-
Nice to see a bit of action in this thread after a few weeks. It seems both of use are using Ivy Bridge CPU's, I wonder if the issue is related to our CPU's or chipsets?
I notice in the Reddit thread you posted it mentions something about the E3-12xx CPU's:
IMPORTANT ADDITIONAL COMMANDS
You might need to add additional commands to this line, if the passthrough ends up failing. For example, if you're using a similar CPU as I am (Xeon E3-12xx series), which has horrible IOMMU grouping capabilities, and/or you are trying to passthrough a single GPU.
These additional commands essentially tell Proxmox not to utilize the GPUs present for itself, as well as helping to split each PCI device into its own IOMMU group. This is important because, if you try to use a GPU in say, IOMMU group 1, and group 1 also has your CPU grouped together for example, then your GPU passthrough will fail.Is it possible that same GRUB config is required here? Sorry, I'm probably not much help when it comes to hardware troubleshooting on Linux.
GRUB_CMDLINE_LINUX_DEFAULT="quiet intel_iommu=on iommu=pt pcie_acs_override=downstream,multifunction nofb nomodeset video=vesafb:off,efifb:off"I don't have the faintest idea of what any of this does or if it could help.
-
Yeah I am having this problem for almost 2 years now, as I have been trying to migrate to XCP-NG on and off to see if I can get this to work. So far I couldnt and I was pushed back to Proxmox.
The part of the steps that you are referring to are the steps that need to be done on the Proxmox host itself, although in this case the GPU is actually seen by the VM so the passthrough steps provided in this documentation: https://xcp-ng.org/docs/compute.html#pci-passthrough are working. Also the documentation part you are referring to are for E3-12xx CPU's, mine is an E5, so I dont have to do these steps and havent done those when on Proxmox.
But when you installed the Nvidia drivers on lets say in my case an ubuntu 20.04 VM, it does install it succesfully but for some reason nvidia-smi says 'no devices were found' while the VM sees the GPU (Quadro P400) and does confirm that the nvidia drivers have been installed.
So I am a bit clueless as where this error could be, as I have searched alot of forms right now about this particular problem but without succes. Some even say that the card could be 'dead' while it does work on Proxmox and even VMware ESXi.
-
I truly hope the community can assist in there, I have also no idea
-
Well community bring it on!
Curious on what you guys think about this and how to resolve this issue... -
Okay so did some test/troubleshooting again, did not do anything. Tried multiple nvidia drivers but no luck still the nvidia-smi error 'no devices were found'.
And still this error occurs:
[ 3.244947] NVRM: loading NVIDIA UNIX x86_64 Kernel Module 495.29.05 Thu Sep 30 16:00:29 UTC 2021 [ 4.061597] NVRM: GPU 0000:00:09.0: RmInitAdapter failed! (0x22:0x56:751) [ 4.064473] NVRM: GPU 0000:00:09.0: rm_init_adapter failed, device minor number 0 [ 4.074551] NVRM: GPU 0000:00:09.0: RmInitAdapter failed! (0x22:0x56:751) [ 4.074650] NVRM: GPU 0000:00:09.0: rm_init_adapter failed, device minor number 0 [ 28.791682] NVRM: GPU 0000:00:09.0: RmInitAdapter failed! (0x22:0x56:751) [ 28.791750] NVRM: GPU 0000:00:09.0: rm_init_adapter failed, device minor number 0 [ 28.799413] NVRM: GPU 0000:00:09.0: RmInitAdapter failed! (0x22:0x56:751) [ 28.799476] NVRM: GPU 0000:00:09.0: rm_init_adapter failed, device minor number 0 [ 37.904451] NVRM: GPU 0000:00:09.0: RmInitAdapter failed! (0x22:0x56:751) [ 37.904532] NVRM: GPU 0000:00:09.0: rm_init_adapter failed, device minor number 0 [ 37.913443] NVRM: GPU 0000:00:09.0: RmInitAdapter failed! (0x22:0x56:751) [ 37.913527] NVRM: GPU 0000:00:09.0: rm_init_adapter failed, device minor number 0I have no idea on what this is and how to resolve this. I only have this issue with XCP-NG. ESXi and Proxmox work just fine without any errors. Could it be a kernel parameter which I am missing in the hypervisor?
-
CentOS8/RockyLinux 8.5 apparently has the same issue. I believe it has something to do with the passthrough handled by XCP-NG for some reason...
-
I'm having the exact same problem on my R630 with a E5-2620V3. Tested on Ubuntu 20.04. Tried using the latest drivers (470) as well as the future build (495) with the same results of no device being found when running nvidia-smi. Card has been confirmed working on a baremetal windows machine as well as inside a proxmox VM. Happy to do anything needed to help get this figured out.
-
Same error message?
-
@olivierlambert yes I'm getting this
[ 5.091321] NVRM: loading NVIDIA UNIX x86_64 Kernel Module 470.82.00 Thu Oct 14 10:24:40 UTC 2021 [ 46.870183] NVRM: GPU 0000:00:05.0: RmInitAdapter failed! (0x22:0x56:667) [ 46.870225] NVRM: GPU 0000:00:05.0: rm_init_adapter failed, device minor number 0 [ 46.877175] NVRM: GPU 0000:00:05.0: RmInitAdapter failed! (0x22:0x56:667) [ 46.877214] NVRM: GPU 0000:00:05.0: rm_init_adapter failed, device minor number 0 [ 53.426603] NVRM: GPU 0000:00:05.0: RmInitAdapter failed! (0x22:0x56:667) [ 53.426701] NVRM: GPU 0000:00:05.0: rm_init_adapter failed, device minor number 0 [ 53.433752] NVRM: GPU 0000:00:05.0: RmInitAdapter failed! (0x22:0x56:667) [ 53.433830] NVRM: GPU 0000:00:05.0: rm_init_adapter failed, device minor number 0 [ 412.663599] NVRM: GPU 0000:00:05.0: RmInitAdapter failed! (0x22:0x56:667) [ 412.663716] NVRM: GPU 0000:00:05.0: rm_init_adapter failed, device minor number 0 [ 412.671452] NVRM: GPU 0000:00:05.0: RmInitAdapter failed! (0x22:0x56:667) [ 412.671554] NVRM: GPU 0000:00:05.0: rm_init_adapter failed, device minor number 0 -
Okay thanks for the feedback.
-
Well I went back to Proxmox now and it works just fine on a Ubuntu VM. The passthrough works as expected... So I have no idea in what is going on within XCP-NG, as it should work as well?
-
At this point, you best option is to ask on Xen devel mailing list. I have the feeling it's a very low level problem on Xen itself.
In the meantime, I'll try to get my hand on a P400 for our lab.
-
@olivierlambert Did you manage to replicate the issue or have found similar issues?
Because I am actually wanting to upgrade again to XCP-NG as I am a bit tired of having 'quorum' in proxmox which forces me to have 3 hosts up a time when using only 1 or 2 ...
-
I still need to get the card, I have not managed yet.
Not sure to understand your quorum issue on ProxMox?
-
@olivierlambert Well what I have with the quorum issue is that I have 3 machines to have quorum, while I believe in XCP-NG I can even have one machine turned on with the VM's on that particular machine without having to have quorum and be able to use the VM's.
In Proxmox when I don't have quorum I cannot use the VM's except when I do this on one host 'pvecm expected 1' which is actually a dangerous command if I am not mistaken...
So this is why I would like to migrate to XCP-NG only the PCI GPU passthrough keeps me back...
-
Could my issue be that I have done this on a R620 where the PCI riser has a max of 75Watt? That the P400 draws more power?
because now I have the P400 in a different machine where the card can draw more power without issues and it also works without issues. -
So you mean in a different hardware, now it works with XCP-ng?
What's the other machine?
-
Well not at the moment, I have not tried it on a different hardware machine with XCP-NG.
I have however currently been using the 'different' hardware machine with Proxmox and the card works great with PCI passthrough'ing the Quadro P400. So I might try this tomorrow with XCP-NG on this machine...
If this does not work, I am really wondering what makes this not working on XCP-NG...
-
That's the information I need indeed (to be able to know if it's a hardware problem vs software)
