[vfio-users] GPU Passthrough, 2 GPUs to different virtual machines

[Sascha Fröhlich]

Hi,

I am trying to get the following setup with a Xeon E5/ASUS X99-A II to work:

First slot:      GTX 1060     -> VM 1
Second slot: GTX 1070 -> VM 2
Third slot:     GT 710         -> Host

This is the latest ArchLinux. I had to use the propietary NVIDIA drivers
and an additional Xorg configuration for the host GPU, because nouveau
would flicker and crash. No need to blacklist anything.

IOMMU and VFIO works fine though, lspci -nn -k reports the two GTX GPUs
using vfio-pci:
https://gist.github.com/anonymous/89ebd0e18464cef3c5dde6897c39eb5f

So far I've managed to add the GTX 1070 too a VM and successfully installed
Windows 10, including NVIDIA drivers. This is a working QEMU xml:
https://gist.github.com/anonymous/f9de48a009e6bd8824e5f272da01e184

The second VM has a very similar configuration. The GTX 1060 doesn't work,
though. When I connect the screen, its just blank with the message "No
displayport output". Video output via SPICE works.

I have these messages in dmesg, which I find confusing. Do I need to use a
ROM for my GTX 1060?

[   40.368641] vfio-pci 0000:01:00.0: Invalid PCI ROM header signature:
expecting 0xaa55, got 0xffff
[   40.368691] vfio-pci 0000:01:00.0: Invalid PCI ROM header signature:
expecting 0xaa55, got 0xffff

Here is the link to the complete output:
https://gist.github.com/anonymous/f00ce9fefeaa7c7093cc6ee75ff42794

Do you have any idea how to debug this any further?

Kind regards, Sascha
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://listman.redhat.com/archives/vfio-users/attachments/20171231/8263f177/attachment.htm>