AMD Renoir drivers (4650G)
- Thread starter j1a2o
- Start date
I was able to figure this out. Renoir GPUs are still experimental in the current Proxmox kernel. They're standard in 5.5 kernel. So adding amdgpu.exp_hw_support=1 to kernel command line made this work.
Never tried this, but it could be helpful: https://www.reddit.com/r/Proxmox/comments/l0cjve/solved_ryzen_apu_vaapi_decoding_on_proxmox/
The easier way now is to upgrade to the newly released 5.11 kernel: https://forum.proxmox.com/threads/kernel-5-11.86225/
Don't need to enable that pvetest repository. If you're already on pve-no-subscription, just run:
and reboot.
Don't need to enable that pvetest repository. If you're already on pve-no-subscription, just run:
Code:
apt install pve-kernel-5.11sadly it only worked partially...
I hav /dev/dri/* now and amdgpu is loaded and used
however VA-API doesn't work
I checked the problem with "unsuported power profile" and found that some people seem to have problems shutting down their laptops but I'm pretty sure this is not the problem here...
I hav /dev/dri/* now and amdgpu is loaded and used
Code:
root@proxmix:~# dmesg | grep -i amdgpu
[ 5.210878] [drm] amdgpu kernel modesetting enabled.
[ 5.211021] amdgpu: Topology: Add CPU node
[ 5.211118] fb0: switching to amdgpudrmfb from EFI VGA
[ 5.211253] amdgpu 0000:07:00.0: vgaarb: deactivate vga console
[ 5.211369] amdgpu 0000:07:00.0: amdgpu: Trusted Memory Zone (TMZ) feature disabled as experimental (default)
[ 5.215818] amdgpu 0000:07:00.0: amdgpu: Fetched VBIOS from ROM BAR
[ 5.215821] amdgpu: ATOM BIOS: 113-RENOIR-033
[ 5.217375] amdgpu 0000:07:00.0: amdgpu: VRAM: 512M 0x000000F400000000 - 0x000000F41FFFFFFF (512M used)
[ 5.217378] amdgpu 0000:07:00.0: amdgpu: GART: 1024M 0x0000000000000000 - 0x000000003FFFFFFF
[ 5.217379] amdgpu 0000:07:00.0: amdgpu: AGP: 267419648M 0x000000F800000000 - 0x0000FFFFFFFFFFFF
[ 5.217532] [drm] amdgpu: 512M of VRAM memory ready
[ 5.217536] [drm] amdgpu: 3072M of GTT memory ready.
[ 6.109090] amdgpu 0000:07:00.0: amdgpu: RAS: optional ras ta ucode is not available
[ 6.132879] amdgpu 0000:07:00.0: amdgpu: RAP: optional rap ta ucode is not available
[ 6.133129] amdgpu 0000:07:00.0: amdgpu: SMU is initialized successfully!
[ 6.179347] snd_hda_intel 0000:07:00.1: bound 0000:07:00.0 (ops amdgpu_dm_audio_component_bind_ops [amdgpu])
[ 6.224751] amdgpu: Topology: Add dGPU node [0x1636:0x1002]
[ 6.224760] amdgpu 0000:07:00.0: amdgpu: SE 1, SH per SE 2, CU per SH 18, active_cu_number 27
[ 6.226232] fbcon: amdgpudrmfb (fb0) is primary device
[ 6.320353] amdgpu 0000:07:00.0: [drm] fb0: amdgpudrmfb frame buffer device
[ 6.357272] amdgpu 0000:07:00.0: amdgpu: ring gfx uses VM inv eng 0 on hub 0
[ 6.357280] amdgpu 0000:07:00.0: amdgpu: ring comp_1.0.0 uses VM inv eng 1 on hub 0
[ 6.357282] amdgpu 0000:07:00.0: amdgpu: ring comp_1.1.0 uses VM inv eng 4 on hub 0
[ 6.357283] amdgpu 0000:07:00.0: amdgpu: ring comp_1.2.0 uses VM inv eng 5 on hub 0
[ 6.357285] amdgpu 0000:07:00.0: amdgpu: ring comp_1.3.0 uses VM inv eng 6 on hub 0
[ 6.357286] amdgpu 0000:07:00.0: amdgpu: ring comp_1.0.1 uses VM inv eng 7 on hub 0
[ 6.357287] amdgpu 0000:07:00.0: amdgpu: ring comp_1.1.1 uses VM inv eng 8 on hub 0
[ 6.357288] amdgpu 0000:07:00.0: amdgpu: ring comp_1.2.1 uses VM inv eng 9 on hub 0
[ 6.357290] amdgpu 0000:07:00.0: amdgpu: ring comp_1.3.1 uses VM inv eng 10 on hub 0
[ 6.357291] amdgpu 0000:07:00.0: amdgpu: ring kiq_2.1.0 uses VM inv eng 11 on hub 0
[ 6.357293] amdgpu 0000:07:00.0: amdgpu: ring sdma0 uses VM inv eng 0 on hub 1
[ 6.357295] amdgpu 0000:07:00.0: amdgpu: ring vcn_dec uses VM inv eng 1 on hub 1
[ 6.357296] amdgpu 0000:07:00.0: amdgpu: ring vcn_enc0 uses VM inv eng 4 on hub 1
[ 6.357298] amdgpu 0000:07:00.0: amdgpu: ring vcn_enc1 uses VM inv eng 5 on hub 1
[ 6.357299] amdgpu 0000:07:00.0: amdgpu: ring jpeg_dec uses VM inv eng 6 on hub 1
[ 6.359017] [drm] Initialized amdgpu 3.40.0 20150101 for 0000:07:00.0 on minor 0
[ 7.237503] amdgpu 0000:07:00.0: amdgpu: Unsupported power profile mode 0 on RENOIRhowever VA-API doesn't work
Code:
root@proxmix:~# vainfo
error: can't connect to X server!
libva info: VA-API version 1.4.0
libva info: va_getDriverName() returns 0
libva info: Trying to open /usr/lib/x86_64-linux-gnu/dri/radeonsi_drv_video.so
libva info: Found init function __vaDriverInit_1_4
pci id for fd 4: 1002:1636, driver (null)
libva error: /usr/lib/x86_64-linux-gnu/dri/radeonsi_drv_video.so init failed
libva info: va_openDriver() returns 2
vaInitialize failed with error code 2 (resource allocation failed),exitI checked the problem with "unsuported power profile" and found that some people seem to have problems shutting down their laptops but I'm pretty sure this is not the problem here...
That vainfo version is pretty old by now and might not support this GPU. Try a newer version.
Last edited:
@j1a2o @thex I have a 4650g as above with exactly the same problem. Did you make any progress with this?
I've found some different information in other places, specifically that AMD's drivers are needed. I tried to install the latest AMD drivers but the DKMS build fails on kernel 5.11 and the drivers seem to only support Kernel 5.4. Although I have /dev/dri, VAAPI fails and my vainfo gives exactly same as you got above.
I've found some different information in other places, specifically that AMD's drivers are needed. I tried to install the latest AMD drivers but the DKMS build fails on kernel 5.11 and the drivers seem to only support Kernel 5.4. Although I have /dev/dri, VAAPI fails and my vainfo gives exactly same as you got above.
I got it working in a Jellyfin container with the ANMD mod configured so I guess its related entirely to what is in the docker image rather than libraries installed on the proxmox host or lxc container.