Libargus + Rivermax + CUDA

[janderson]

Hi,
I have been looking into dockerizing my application that uses the following features on a Jetson Xavier:
LibArgus for image capture
CUDA
Rivermax

I have seen examples of how to run CUDA in a docker container, but was wondering if there are any more resources / examples of how to run LibArgus and the Rivermax card in a docker container.

Thanks!

[dusty_nv]

Hi @janderson, on JetPack 5 the libargus libraries are automatically mounted into the containers from the device when --runtime nvidia is used, and in l4t-jetpack container the Multimedia API and Argus developer headers/samples can be found under /usr/src/jetson_multimedia_api inside the container:

sudo docker run --runtime nvidia -it --rm --network host \
  --volume /tmp/argus_socket:/tmp/argus_socket \
  nvcr.io/nvidia/l4t-jetpack:r35.4.1 \
  /bin/bash -c 'find /usr -name "*argus*"'

Note that to use Argus inside container, you should include --volume /tmp/argus_socket:/tmp/argus_socket and --runtime nvidia in your docker run command when you start the container.

I’m not personally familiar with Rivermax or running that in a container, hopefully others will be able to share their experience if any special steps are needed.