NVENC Error: unsupported device in Docker with Nvidia Container Runtime

NVIDIA Open GPU Kernel Modules Version

570.86.16

Please confirm this issue does not happen with the proprietary driver (of the same version). This issue tracker is only for bugs specific to the open kernel driver.

• I confirm that this does not happen with the proprietary driver package.

Operating System and Version

Ubuntu 22.04.5 LTS

Kernel Release

5.15.0-113-generic

Please confirm you are running a stable release kernel (e.g. not a -rc). We do not accept bug reports for unreleased kernels.

• I am running on a stable kernel release.

Hardware: GPU

NVIDIA GeForce RTX 4090

Describe the bug

After installing version 570.86.16 of the open GPU kernel modules, I encountered an error when using the NVENC functionality inside a container with Nvidia Container Runtime. The error indicates that the device is unsupported. However, when running the application directly on the host machine, the NVENC feature works correctly.

I confirm that this does not happen with the proprietary driver package.

To Reproduce

docker run --runtime=nvidia --gpus '"device=0,1"' jrottenberg/ffmpeg:4.1-nvidia -report -loglevel debug -f lavfi -i testsrc=duration=5:size=1280x720:rate=30 -c:v h264_nvenc -preset fast -y /tmp/test_output.mp4

Bug Incidence

Always

nvidia-bug-report.log.gz

nvidia-bug-report.log.gz

More Info

Similar issues were #104 and #378, but the nvenc problem occurs again in the new version of the driver

Start with the provided Docker reproduction command and compare its behavior with NVENC on the host, using the attached nvidia-bug-report.log.gz for driver details. Investigate the open kernel module and Nvidia Container Runtime interaction; done means the same container command no longer reports an unsupported device with the open driver.