gpu passthrough operation not permitted error

SyntaxError210

New Member
Apr 10, 2021
7
1
3
27
Hi all. Earlier today I posted a thread because I was having this issue with an HP Proliant dl360 g7 which turned out to be because of hp itself. So, I moved the gpu (radeon pro wx 5100) to a dell poweredge r710 and followed the guide again. I ended up getting the same error and after spending three days last time trying to fix it and ending up with the result that I couldn't because hp sucks, I would really like to not find out that dell has the same issue. I think it might be a problem with the gpu itself but I hope that's also wrong because its the only one we have. If anyone can help at all it would be greatly appreciated.
Grub settings:
Code:
GRUB_CMDLINE_LINUX_DEFAULT="quiet intel_iommu=on"
dmesg | grep -e DMAR -e IOMMU output:
Code:
[    0.016479] ACPI: DMAR 0x000000009F3B3674 000240 (v01 DELL   PE_SC3   00000001 DELL 00000001)
[    0.352640] DMAR: IOMMU enabled
[    0.669178] DMAR-IR: This system BIOS has enabled interrupt remapping
[    1.800072] DMAR: Host address width 40
[    1.800074] DMAR: DRHD base: 0x000000af000000 flags: 0x0
[    1.800095] DMAR: dmar0: reg_base_addr af000000 ver 1:0 cap c90780106f0462 ecap f020fe
[    1.800096] DMAR: DRHD base: 0x000000fed90000 flags: 0x1
[    1.800103] DMAR: dmar1: reg_base_addr fed90000 ver 1:0 cap c90780106f0462 ecap f020fe
[    1.800104] DMAR: RMRR base: 0x0000009f4c8000 end: 0x0000009f4dffff
[    1.800105] DMAR: RMRR base: 0x0000009f4b1000 end: 0x0000009f4bffff
[    1.800106] DMAR: RMRR base: 0x0000009f4a1000 end: 0x0000009f4a1fff
[    1.800107] DMAR: RMRR base: 0x0000009f4a3000 end: 0x0000009f4a3fff
[    1.800108] DMAR: RMRR base: 0x0000009f4a5000 end: 0x0000009f4a5fff
[    1.800109] DMAR: RMRR base: 0x0000009f4a7000 end: 0x0000009f4a7fff
[    1.800110] DMAR: RMRR base: 0x0000009f4a9000 end: 0x0000009f4a9fff
[    1.800110] DMAR: RMRR base: 0x0000009f4c0000 end: 0x0000009f4c0fff
[    1.800111] DMAR: RMRR base: 0x0000009f4c2000 end: 0x0000009f4c2fff
[    1.800112] DMAR: ATSR flags: 0x0
[    1.800437] DMAR: dmar0: Using Queued invalidation
[    1.800453] DMAR: dmar1: Using Queued invalidation
[    1.814212] DMAR: Intel(R) Virtualization Technology for Directed I/O
[    9.538835] AMD-Vi: AMD IOMMUv2 driver by Joerg Roedel <jroedel@suse.de>
[    9.538837] AMD-Vi: AMD IOMMUv2 functionality not available on this system
[  352.806986] vfio_iommu_type1_attach_group: No interrupt remapping support.  Use the module param "allow_unsafe_interrupts" to enable VFIO IOMMU support on this platform
[  593.049423] vfio_iommu_type1_attach_group: No interrupt remapping support.  Use the module param "allow_unsafe_interrupts" to enable VFIO IOMMU support on this platform
/etc/modules settings:
Code:
# /etc/modules: kernel modules to load at boot time.
#
# This file contains the names of kernel modules that should be loaded
# at boot time, one per line. Lines beginning with "#" are ignored.
vfio
vfio_iommu_type1
vfio_pci
vfio_virqfd
I followed the allow unsafe interrupts part of the guide with this command and it seemed to do what it was supposed to:
Code:
echo "options vfio_iommu_type1 allow_unsafe_interrupts=1" > /etc/modprobe.d/iommu_unsafe_interrupts.conf
Finally, this is my 126.conf file (126 is the vmid):
Code:
bios: ovmf
bootdisk: ide0
cores: 3
efidisk0: HDDPool:126/vm-126-disk-1.qcow2,size=128K
hostpci0: 05:00,pcie=1,x-vga=1
ide0: HDDPool:126/vm-126-disk-0.qcow2,size=32G
ide2: SSDPool:iso/Windows.iso,media=cdrom
machine: q35
memory: 10272
name: windows-vm
net0: e1000=52:81:9E:A5:ED:89,bridge=vmbr1,firewall=1,tag=10
numa: 0
ostype: win10
scsihw: virtio-scsi-pci
smbios1: uuid=a931b443-8243-4b3b-b9ce-19ef8b8edcbb
sockets: 2
vmgenid: 527be331-79ff-4833-9d15-5d2cd804877e

This is the error that occurs when I attempt to start the vm:
Code:
kvm: -device vfio-pci,host=0000:05:00.0,id=hostpci0.0,bus=ich9-pcie-port-1,addr=0x0.0,multifunction=on: vfio 0000:05:00.0: failed to setup container for group 14: Failed to set iommu for container: Operation not permitted
TASK ERROR: start failed: QEMU exited with code 1
 
Have you tried adding vfio_iommu_type1.allow_unsafe_interrupts=1 to the kernel parameters (after intel_iommu=on)? The error message in the logs suggests that this might help with this issue.
 
I have no idea what to do. There's just no vnc connection.
You added x-vga=1, which is necessary for NVidia card but also disables the virtual display, I think. If all went well, you should have output from the GPU and no VNC/Spice/virtual display. Therefore, this is to be expected, I think. However, your reaction seems to imply that you also have no output from the GPU? Maybe you can try with x-vga=0 to check if the GPU is seen by Windows (but installing the NVidia drivers might not work)?
EDIT: At least the passthrough appears to be working now for your system. I have no experience with Nvidia cards. Maybe search the forum or maybe someone else can help you here.
 
Last edited:
It works! Your right, the gpu is outputting the display. I had to bike a good bit to get to the physical servers so it took a bit to check but it is outputting through the gpu after 5 days of trying to get it to work. Thanks for all your help.
 
  • Like
Reactions: leesteken

About

The Proxmox community has been around for many years and offers help and support for Proxmox VE, Proxmox Backup Server, and Proxmox Mail Gateway.
We think our community is one of the best thanks to people like you!

Get your subscription!

The Proxmox team works very hard to make sure you are running the best software and getting stable updates and security enhancements, as well as quick enterprise support. Tens of thousands of happy customers have a Proxmox subscription. Get yours easily in our online shop.

Buy now!