I had some problems configuring VGPU and Vgpu_UNLOCK

simplePve

New Member
Jun 2, 2022
3
0
1
Hello, there are some problems in configuring the virtual graphics card for PVE.
My machine configuration is i9-7940X and RTX2080ti 12GB

Now I have installed the graphics driver
1654139669284.png

and mdevctl types shows
1654139854819.png
And configure VM 100
etc/pve/qemu-server/100.conf
args: -uuid 00000000-0000-0000-0000-000000000100 boot: order=ide0;ide2 cores: 1 hostpci0: 0000:65:00.0,mdev=nvidia-257,pcie=1,x-vga=1 ide0: local:iso/zh-cn_windows_10_consumer_editions_version_21h2_x64_dvd_cf28838e.iso,media=cdrom,size=5722610K ide2: none,media=cdrom machine: q35 memory: 10240 meta: creation-qemu=6.1.0,ctime=1654078209 name: 123 numa: 0 ostype: l26 sata0: lvm-1:vm-100-disk-0,size=200G scsihw: virtio-scsi-pci smbios1: uuid=ed647fc2-b48c-4c03-b3c9-7ce98c2bc3fa sockets: 1 vga: virtio vmgenid: 580ad737-523a-478f-887e-d7b9fdbe61af
qm start 100 err



root@pve-10:~# qm start 100 mdev instance '00000000-0000-0000-0000-000000000100' already existed, using it. kvm: -device vfio-pci,sysfsdev=/sys/bus/pci/devices/0000:65:00.0/00000000-0000-0000-0000-000000000100,id=hostpci0,bus=ich9-pcie-port-1,addr=0x0,x-vga=on: vfio 00000000-0000-0000-0000-000000000100: error getting device from group 81: Input/output error Verify all devices in group 81 are bound to vfio-<bus> or pci-stub and not already in use start failed: QEMU exited with code 1
1654141913736.png
please help me ,thanks
 
Hi,

can you post the complete output of 'dmesg' here?
also could you try to remove the mdev instances before starting the vms?

Code:
echo 1 > /sys/bus/mdev/devices/$mdev_UUID/remove
 
XXXX - VMID
pci_device_id - your id
pci_id - your id

nano /etc/vgpu_unlock/profile_override.toml
Bash:
[profile.nvidia-256]
num_displays = 1
display_width = 1920
display_height = 1080
max_pixels = 2073600
cuda_enabled = 1
frl_enabled = 1
framebuffer = 0x76000000
pci_device_id = 0x1E30
pci_id = 0x1E3012BA

[mdev.00000000-0000-0000-0000-000000000XXXX]
frl_enabled = 0


nano /etc/pve/qemu-server/XXXX.conf
add line:
Bash:
args: -uuid 00000000-0000-0000-0000-00000000XXXX
 
Last edited:
你好,

你能在这里发布'dmesg'的完整输出吗?
您也可以尝试在启动虚拟机之前删除 mdev 实例吗?

[代码]
回声 1 > /sys/bus/mdev/devices/$mdev_UUID/remove
[/代码]
○oh I'm so stupid, I should have uploaded the attachment to you
 

Attachments

XXXX - VMID
pci_device_id - your id
pci_id - your id

nano /etc/vgpu_unlock/profile_override.toml
Bash:
[profile.nvidia-256]
num_displays = 1
display_width = 1920
display_height = 1080
max_pixels = 2073600
cuda_enabled = 1
frl_enabled = 1
framebuffer = 0x76000000
pci_device_id = 0x1E30
pci_id = 0x1E3012BA

[mdev.00000000-0000-0000-0000-000000000XXXX]
frl_enabled = 0


nano /etc/pve/qemu-server/XXXX.conf
add line:
Bash:
args: -uuid 00000000-0000-0000-0000-00000000XXXX
please this profile_override.toml files
[mdev.00000000-0000-0000-0000-000000000XXXX] frl_enabled = 0

xxxx is my vmid?
or just xxxx