XCP-ng
    • Categories
    • Recent
    • Tags
    • Popular
    • Users
    • Groups
    • Register
    • Login

    RDNA 4 GPU Passthrough

    Scheduled Pinned Locked Moved Compute
    3 Posts 2 Posters 136 Views 2 Watching
    Loading More Posts
    • Oldest to Newest
    • Newest to Oldest
    • Most Votes
    Reply
    • Reply as topic
    Log in to reply
    This topic has been deleted. Only users with topic management privileges can see it.
    • P Offline
      PessimistTech
      last edited by

      Hi all,
      I am trying to pass through 2 AMD Radeon Ai Pro 9700 GPUs to a ubuntu VM. I have gottent to the point where I have the GPUs passed through, I can see them fine in the VM (Ie rocm-info/rocm-smi/nvtop all show the GPUs and current status without isssue), and I can even start workloads against them and see load start on the GPUs, but the workloads always time out or fail as they don't seem to get a response back or something along those lines.

      In my digging so far I have seen that no interrupts are being registered, and as far as I can tell this seems to be an issue with the XCP-NG -> VM layer in the repsponse (as I cannot fine any issues, errors or misconfigs on the VM other than the timeouts).

      Has anyone tried/had success with newer RDNA 4 GPUs? Any suggestions on things to try next? I'm at a bit of a loss here...

      Y 2 Replies Last reply Reply Quote 0
      • Y Offline
        yannsionneau Vates 🪐 XCP-ng Team @PessimistTech
        last edited by

        Hi @pessimisttech let me ping @Team-Hypervisor-Kernel here 🙂

        1 Reply Last reply Reply Quote 0
        • Y Offline
          yannsionneau Vates 🪐 XCP-ng Team @PessimistTech
          last edited by

          @PessimistTech Let's try to see if we have errors in the logs.

          Run ls /boot/xen* and note the name of the xen file that ends with -d.gz

          Then, can you reboot and , while in the grub menu, edit the XCP-ng menu entry (the first one) and make the following modifications:

          • modify the Xen ELF loaded by multiboot2 to load the debug Xen build by specifying the file that ends with -d.gz instead of /boot/xen.gz
          • on the same line, at the end, add iommu=debug

          Then boot (by doing ctrl+X) and once booted and the VM with PCI passthrough started, please issue xl dmesg command in dom0 shell and report the content here.

          Also please paste the result of dom0 dmesg.

          then also an lspci -vvv -nn -s SBDF from both dom0 and the VM

          replace SBDF with the segment bus device function address of the device (beware it won't be the same in dom0 and inside the VM).

          Regards,

          Yann

          1 Reply Last reply Reply Quote 0

          Hello! It looks like you're interested in this conversation, but you don't have an account yet.

          Getting fed up of having to scroll through the same posts each visit? When you register for an account, you'll always come back to exactly where you were before, and choose to be notified of new replies (either via email, or push notification). You'll also be able to save bookmarks and upvote posts to show your appreciation to other community members.

          With your input, this post could be even better 💗

          Register Login
          • First post
            Last post