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

    Intel Core Ultra iGPU passthrough

    Scheduled Pinned Locked Moved Hardware
    10 Posts 4 Posters 205 Views 3 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.
    • A Offline
      antest
      last edited by olivierlambert

      Hi everybody,

      First of all I would like to thank all the contributors for the hard effort put into building and maintaining this project and for the vast amount of support that this forum has been giving!

      Specs:
      DELL Tower
      CPU: Intel Core Ultra 5 235
      XCP-ng 8.3

      I have been trying to passthrough the iGPU to a VM running Debian 13 but keep getting errors.
      I have successfully passed older intel iGPUs and have always worked perfectly.

      PCI pass through enabled:
      61c0db70-3708-4d9e-a580-d1b0d1207df1-image.png

      I don't know if this is normal but if I attach the GPU using xcp-ng center it shows up in XOA:
      c791357a-b8e7-48d3-8f20-59ead8f5890f-image.png

      But not as a pci device:
      00453442-af40-4383-bec2-61a0e9a20737-image.png

      Regardless, I have also attached as pci:
      Halted VM:
      3091b95d-f5bf-4dd2-8790-0f0c1d51b676-image.png

      Running VM:
      89de623a-6e21-4999-912f-91f26db91851-image.png

      Upon boot console gets filled:

      [   24.283971] i915 0000:00:08.0: [drm] *ERROR* GT1: GSC proxy component didn't bind within the expected timeout
      [   24.283993] i915 0000:00:08.0: [drm] *ERROR* GT1: GSC proxy handler failed to init
      [   48.271474] i915 0000:00:08.0: [drm] *ERROR* GT0: GUC: TLB invalidation response timed out for seqno 60
      [   50.282144] i915 0000:00:08.0: [drm] *ERROR* GT0: GUC: TLB invalidation response timed out for seqno 61
      [   52.294046] i915 0000:00:08.0: [drm] *ERROR* GT1: GUC: TLB invalidation response timed out for seqno 41
      [   54.306843] i915 0000:00:08.0: [drm] *ERROR* GT1: GUC: TLB invalidation response timed out for seqno 42
      

      lspci detects the gpu, but it's not usable.

      Also tried changing to XE drivers instead of i915 and didn't help.

      Because I don't know enough to diagnose this, I tried asking AI which prompted me to perform "xen-pciback.hide" for the Intel Management Engine (ME) and still didn't help.
      Also tried with Ubuntu 25.10.

      Has anyone been able to pass through a similar iGPU?

      Sorry for the image spamming.

      Thank you.

      TeddyAstieT 1 Reply Last reply Reply Quote 0
      • olivierlambertO Offline
        olivierlambert Vates 🪐 Co-Founder CEO
        last edited by

        I'm not sure Intel iGPU passthrough can work. Let me ping @teddyastie maybe

        1 Reply Last reply Reply Quote 0
        • TeddyAstieT Offline
          TeddyAstie Vates 🪐 XCP-ng Team Xen Guru
          last edited by

          Do you get anything special written to xl dmesg in Dom0 ?

          A 1 Reply Last reply Reply Quote 0
          • A Offline
            antest @TeddyAstie
            last edited by

            @TeddyAstie

            I don't think so, it seems IOMMU is enabled:

            (XEN) [ 0.183949] Intel VT-d iommu 0 supported page sizes: 4kB, 2MB, 1GB
            (XEN) [ 0.191524] I/O virtualisation enabled
            

            I'll leave the full log attached.

            xl_dmesg.txt

            1 Reply Last reply Reply Quote 0
            • TeddyAstieT Offline
              TeddyAstie Vates 🪐 XCP-ng Team Xen Guru @antest
              last edited by

              @antest can you retry after doing /opt/xensource/libexec/xen-cmdline --set-xen iommu=debug then rebooting.
              And also reporting full DomU and Dom0 dmesg (in addition to xl dmesg) ?

              A 1 Reply Last reply Reply Quote 1
              • A Offline
                antest @TeddyAstie
                last edited by

                @TeddyAstie

                After rebooting, with pci attached:

                xl_dmesg.txt
                Dom0_dmesg.txt
                Dom1_dmesg.txt

                1 Reply Last reply Reply Quote 1
                • O Offline
                  ovicz
                  last edited by

                  Kernel command line: BOOT_IMAGE=/boot/vmlinuz-6.12.57+deb13-amd64 root=UUID=d706ed91-c2dd-4c8f-92cb-d2a595e24fe4 ro quiet splash i915.force_probe=7d67 i915.enable_guc=3 i915.max_vfs=7

                  Try to remove the i915 options and rebuild grub then reboot. see if that errors regarding guc pops up again,

                  I don't see in modinfo i915 that value of enable_guc=3

                  parm: enable_guc:Enable GuC load for GuC submission and/or HuC load. Required functionality can be selected using bitmask values. (-1=auto [default], 0=disable, 1=GuC submission, 2=HuC load) (int)

                  A 1 Reply Last reply Reply Quote 0
                  • A Offline
                    antest @ovicz
                    last edited by

                    @ovicz Tried it, didn't change.

                    But thanks anyway.

                    TeddyAstieT 1 Reply Last reply Reply Quote 0
                    • TeddyAstieT Offline
                      TeddyAstie Vates 🪐 XCP-ng Team Xen Guru @antest
                      last edited by

                      @antest can you try running the VM with only 1 vCPU ?

                      A 1 Reply Last reply Reply Quote 0
                      • A Offline
                        antest @TeddyAstie
                        last edited by

                        @TeddyAstie Sure thing:

                        xl_dmesg.txt
                        Dom0_dmesg.txt
                        DomU_dmesg.txt

                        1 Reply Last reply Reply Quote 0
                        • First post
                          Last post