Iommu. GPU Passthrough problem because of identical GPU's


first: Sry for my English! I am native German.

I have this System:

Mainboard ASUS ROG Zenith Extreme
CPU AMD Threadripper 1950x
RAM 64GB Corsair Dominator Platinum
2x NVME's

And as GPU (in this Order build in):
1x ASUS Strix 1080ti
1x AMD Radeon Pro WX3100
1x ASUS Strix 1080ti

i buyed last week the AMD Radeon Pro, because i want to use Linux again as a Desktop System. I tried it with one of the Nvidia Card, but with the properity Driver i had a big problem with lagging Windows if resizing. Everything else was fast. Only resizing was extremly slow. I dont tried a lot to fix it, because i want it anyway to use the GPU's in another way. So i buyed the AMD Card and built it in.

Im using only the AMD Card. I want to use one of the Nvidia GPU's for a VM. But i cant passthrough it because the two nvidia cards have the same ID.

I read already the wiki article But that dont help me, because that isnt working.

Because this Script


for i in /sys/bus/pci/devices/*/boot_vga; do
	if [ $(cat "$i") -eq 0 ]; then
		AUDIO="$(echo "$GPU" | sed -e "s/0$/1/")"
		echo "vfio-pci" > "$GPU/driver_override"
		if [ -d "$AUDIO" ]; then
			echo "vfio-pci" > "$AUDIO/driver_override"

modprobe -i vfio-pci

gives me a "Permission Denied" at Bootscreen. I tried chmod 777 and chmod +x. Even than it dont work.

I hope somebody can give me a step for step guide or help.

After work, i can send few logs and infos like lspci and others.

Im Home now. Here the detailed Specs:

  Host: colossus Kernel: 5.1.0-mainline x86_64 bits: 64 compiler: gcc 
  v: 8.3.0 Console: tty 0 Distro: Manjaro Linux 
  Type: Desktop Mobo: ASUSTeK model: ROG ZENITH EXTREME v: Rev 1.xx 
  serial: <filter> UEFI: American Megatrends v: 1701 date: 01/09/2019 
  Topology: 16-Core (2-Die) model: AMD Ryzen Threadripper 1950X bits: 64 
  type: MT MCP MCM arch: Zen rev: 1 L2 cache: 8192 KiB 
  flags: lm nx pae sse sse2 sse3 sse4_1 sse4_2 sse4a ssse3 svm 
  bogomips: 256070 
  Speed: 2000 MHz min/max: 2200/4000 MHz Core speeds (MHz): 1: 3120 2: 2000 
  3: 2200 4: 2200 5: 2000 6: 2000 7: 2000 8: 3999 9: 3223 10: 2000 11: 3999 
  12: 2000 13: 3999 14: 2200 15: 2200 16: 2200 17: 2200 18: 2200 19: 2200 
  20: 2200 21: 2200 22: 2000 23: 2000 24: 2000 25: 2000 26: 3999 27: 2200 
  28: 2200 29: 2200 30: 2200 31: 2200 32: 2200 
  Device-1: NVIDIA GP102 [GeForce GTX 1080 Ti] vendor: ASUSTeK 
  driver: nouveau v: kernel bus ID: 08:00.0 
  Device-2: AMD Lexa XT [Radeon PRO WX 3100] driver: amdgpu v: kernel 
  bus ID: 42:00.0 
  Device-3: NVIDIA GP102 [GeForce GTX 1080 Ti] vendor: ASUSTeK 
  driver: nouveau v: kernel bus ID: 43:00.0 
  Display: server: 1.20.4 driver: nouveau 
  resolution: <xdpyinfo missing> 
  OpenGL: renderer: AMD Radeon Pro WX3100 (POLARIS12 DRM 3.30.0 
  5.1.0-mainline LLVM 8.0.0) 
  v: 4.5 Mesa 19.0.3 direct render: Yes 
  Device-1: NVIDIA GP102 HDMI Audio vendor: ASUSTeK driver: snd_hda_intel 
  v: kernel bus ID: 08:00.1 
  Device-2: AMD Family 17h HD Audio vendor: ASUSTeK driver: snd_hda_intel 
  v: kernel bus ID: 0a:00.3 
  Device-3: AMD Baffin HDMI/DP Audio [Radeon RX 550 640SP / RX 560/560X] 
  driver: snd_hda_intel v: kernel bus ID: 42:00.1 
  Device-4: NVIDIA GP102 HDMI Audio vendor: ASUSTeK driver: snd_hda_intel 
  v: kernel bus ID: 43:00.1 
  Sound Server: ALSA v: k5.1.0-mainline 
  Device-1: Intel I211 Gigabit Network vendor: ASUSTeK driver: igb 
  v: 5.6.0-k port: 2000 bus ID: 03:00.0 
  IF: enp3s0 state: up speed: 1000 Mbps duplex: full mac: <filter> 
  Device-2: Aquantia AQC107 NBase-T/IEEE 802.3bz Ethernet [AQtion] 
  vendor: ASUSTeK driver: atlantic v: port: 2000 
  bus ID: 05:00.0 
  IF: enp5s0 state: up speed: 10000 Mbps duplex: full mac: <filter> 
  Local Storage: total: 2.73 TiB used: 85.24 GiB (3.1%) 
  ID-1: /dev/nvme0n1 vendor: Samsung model: SSD 970 EVO 1TB size: 931.51 GiB 
  ID-2: /dev/nvme1n1 vendor: Samsung model: SSD 970 EVO 1TB size: 931.51 GiB 
  ID-3: /dev/sda vendor: Western Digital model: WD10EADS-42P6B0 
  size: 931.51 GiB 
  ID-4: /dev/sdb vendor: Apple model: HDD WD10EZES-40UFAA0 size: 931.51 GiB 
  ID-1: / size: 915.60 GiB used: 85.24 GiB (9.3%) fs: ext4 
  dev: /dev/nvme1n1p2 
  System Temperatures: cpu: 30.9 C mobo: N/A 
  Fan Speeds (RPM): cpu: 0 
  GPU: device: nouveau temp: 24 C fan: 0 device: nouveau temp: 23 C fan: 0 
  device: amdgpu temp: 41 C fan: 1916 
  Processes: 499 Uptime: 5m Memory: 62.83 GiB used: 4.42 GiB (7.0%) 
  Init: systemd Compilers: gcc: 8.3.0 clang: 8.0.0 Shell: bash v: 5.0.3 
  inxi: 3.0.33 

Have nobody some ideas?

This dont work too :frowning:

Run sudo su, then input your root password, then run the script.

Its do nothing. It makes only a new empty line, and the Cursor is pulsating. Can only abort with ctrl+c to get a new commandline.

Sry, have following Error:

Somehow i got it working.

i still get the last posted Error in the Screenshot. But it works anyway.

But the start of the VM is very slow. Even the Windows Start Circle loops very laggy.

This Thread can be closed. KVM performance is still slow (only Windows 10 as guest). But this is another problem.

I reinstalled already and switched to Archlinux.

The Solution for my specific passthrough Problem: Simply do nothing. Dont activate iommu with Kernel parameters. Dont do anything. Only(!!) blacklist all Drivers for the nvidia cards. After a restart no driver use the GPU's and i can add them without problem into virt-manager...

This topic was automatically closed 30 days after the last reply. New replies are no longer allowed.