After a lot of fiddling around with settings and hardware, I finally have a stable Proxmox 5.1 ATI GPU pass-through system.   What helped me was this helpful article to finally get all the bugs ironed out.   I did have to make several tweaks for my system.   I’m running an Intel system with ATI Radeon GPU.

1. Ensure VT-d is supported and enabled in the BIOS
2. Enable IOMMU on the host
   1. append the following to the GRUB_CMDLINE_LINUX_DEFAULT line in /etc/default/grub

      intel_iommu=on

   2. Save your changes by running

      update-grub

3. Blacklist NVIDIA & Nouveau kernel modules so they don’t get loaded at boot

   1. echo "blacklist nouveau" >> /etc/modprobe.d/blacklist.conf
      echo "blacklist nvidia" >> /etc/modprobe.d/blacklist.conf
      echo "blacklist radeon" >>
      /etc/modprobe.d/blacklist.conf

   2. Save your changes by running

      update-initramfs -u

4. Add the following lines to /etc/modules

   vfio
   vfio_iommu_type1
   vfio_pci
   vfio_virqfd

5. Reboot the host
6. Create your Windows VM using the UEFI bios hardware option (not the deafoult seabios) but do not start it yet.  Use VirtIO.  Modify /etc/pve/qemu-server/<vmid>.conf and ensure the following are in the file. Create / modify existing entries as necessary.

   bios: ovmf
   machine: q35
   cpu: host,hidden=1
   numa: 1

7. Install Windows
   1. Mount second ISO  (virtio-win*.iso)
   2. Load IO driver from d:\viostore\w10\amd64\viostore.inf
   3. After install be sure to enable Remote desktop.
8. Pass through the GPU.
   1. Modify /etc/pve/qemu-server/<vmid>.conf and add
      hostpci0: <device address>,x-vga=on,pcie=1. Example

      hostpci0: 01:00,x-vga=on,pcie=1

9. Passthrough USB keyboard and mouse
   1. I find it best to passthrough specific USB ports rather than device IDs.  That way I can hotplug different devices to specific ports later without having to reboot.
10. Done.

Troubleshooting

Blue screening when launching certain applications

AMD drivers setup application and/or Windows boot would consistently blue screen on me with the following error:

kmode_exception_not_handled

The fix as outlined here was to create /etc/modprobe.d/kvm.conf and add the parameter “options kvm ignore_msrs=1”

echo "options kvm ignore_msrs=1" > /etc/modprobe.d/kvm.conf

Same fix can be applied at runtime with

echo 1 > /sys/module/kvm/parameters/ignore_msrs

Update 4/9/18: Blue screening happens to Windows 10 1803 as well with the error

System Thread Exception Not Handled

The fix for this is the same – ignore_msrs=1

Frezing keyboard/mouse:

Device Manager -> Human Interface Devices -> Microsoft Hardware USB Keyboard -> Power Management -> Uncheck “Allow the computer to turn off this device to save power”

Notes:

Other guides require setting up vfio.conf.  With my hardware it was not required.  It’s probably needed for a nVidia card though.

1. Determine the PCI address of your GPU
   1. Run

      lspci -v

      and look for your card.  Usually 01:00.0 & 01:00.1. You can omit the part after the decimal to include them both in one go – so in that case it would be 01:00

   2. Run lspci -n -s <PCI address> to obtain vendor IDs. Example :

      lspci -n -s 01:00
      01:00.0 0300: 10de:1b81 (rev a1)
      01:00.1 0403: 10de:10f0 (rev a1)

2. Assign your GPU to vfio driver using the IDs obtained above. Example:

   echo "options vfio-pci ids=10de:1b81,10de:10f0" > /etc/modprobe.d/vfio.conf