Issue with HWaccel

[askpet]

Hi

I have searched here (and everywhere else it seems) and I can't get any further.

I keep getting "Unable to poll intel GPU stats: No device filter specified and no discrete/integrated i915 devices found"

My Frigate image: ghcr.io/blakeblackshear/frigate:dev-4dc6c93-tensorrt
Intel NUC i5 (5th gen) and HD Graphics 6000 (rev 9)
OS: Ubuntu 22.04 LTS

Command "lspci -v" gives
00:02.0 VGA compatible controller: Intel Corporation HD Graphics 6000 (rev 09) (prog-if 00 [VGA controller])
Subsystem: Intel Corporation HD Graphics 6000
Flags: bus master, fast devsel, latency 0, IRQ 49
Memory at a9000000 (64-bit, non-prefetchable) [size=16M]
Memory at b0000000 (64-bit, prefetchable) [size=256M]
I/O ports at 3000 [size=64]
Expansion ROM at 000c0000 [virtual] [disabled] [size=128K]
Capabilities:
Kernel driver in use: i915
Kernel modules: i915

docker-compose.yaml:

version: '3.9'
services:
  frigate:
    container_name: Myfrigate
    privileged: true # this may not be necessary for all setups
    restart: unless-stopped
    image: ghcr.io/blakeblackshear/frigate:dev-4dc6c93-tensorrt
    shm_size: "64mb"
    devices:
      - /dev/bus/usb:/dev/bus/usb
      - /dev/dri/renderD128 # for intel hwaccel, needs to be updated for your hardware
    volumes:
      - /etc/localtime:/etc/localtime:ro
      - /home/askpet/Docker/Frigate/frigate.yml:/config/config.yml:ro
      - /home/askpet/Docker/Frigate/media:/media/frigate
      - type: tmpfs # Optional: 1GB of mem~Docory, reduces SSD/SD Card wear
        target: /tmp/cache
        tmpfs:
          size: 1000000000
    ports:
      - "5000:5000"
      - "1935:1935"   # RTMP feeds
    environment:
      FRIGATE_RTSP_PASSWORD: "foo_foo"

the hwaccel command in frigate.yml
ffmpeg:
hwaccel_args: -hwaccel vaapi -hwaccel_device /dev/dri/renderD128 -hwaccel_output_format yuv420p

It seems as if my docker container does not have access to the GPU - but I am lost.
Maybe someone in here can point in the right direction?

[blakeblackshear]

If you don't have an nvidia GPU, there is no reason to use the tensorrt version.

What do you see when you run ls -lah /dev/dri/ on your host?

If frigate runs with the hwaccel_args that you have specified, then it is using the GPU. It's just the stats around GPU usage that can't be found.

[askpet]

ls -lah /dev/dri/
ls: cannot access '/dev/dri/': No such file or directory

OK - so it must be a rights issue.

[askpet]

I am running docker desktop and has placed my docker-compose.yaml file in my /Home/Docker/Frigate folder (created by me) - and running docker-compose up -d from there. Maybe thats the issue?

[blakeblackshear]

You could try moving the mapping in compose from a device to a volume mount. That could help. It should be in the container with the compose file you posted.

[askpet]

will update here when I am done. Thanks for helping me!

[askpet]

Ok - boom :-). Success. I ran "docker run" as root. That was effectfull.

VAINFO OUTPUT
Return Code: 0

Process Output:

vainfo: VA-API version: 1.17 (libva 2.10.0)
vainfo: Driver version: Intel i965 driver for Intel(R) Broadwell - 2.4.1
vainfo: Supported profile and entrypoints
VAProfileMPEG2Simple : VAEntrypointVLD
VAProfileMPEG2Simple : VAEntrypointEncSlice
VAProfileMPEG2Main : VAEntrypointVLD
VAProfileMPEG2Main : VAEntrypointEncSlice
VAProfileH264ConstrainedBaseline: VAEntrypointVLD
VAProfileH264ConstrainedBaseline: VAEntrypointEncSlice
VAProfileH264Main : VAEntrypointVLD
VAProfileH264Main : VAEntrypointEncSlice
VAProfileH264High : VAEntrypointVLD
VAProfileH264High : VAEntrypointEncSlice
VAProfileH264MultiviewHigh : VAEntrypointVLD
VAProfileH264MultiviewHigh : VAEntrypointEncSlice
VAProfileH264StereoHigh : VAEntrypointVLD
VAProfileH264StereoHigh : VAEntrypointEncSlice
VAProfileVC1Simple : VAEntrypointVLD
VAProfileVC1Main : VAEntrypointVLD
VAProfileVC1Advanced : VAEntrypointVLD
VAProfileNone : VAEntrypointVideoProc
VAProfileJPEGBaseline : VAEntrypointVLD
VAProfileVP8Version0_3 : VAEntrypointVLD

Massive improvement :-)

[askpet]

Problem solved