No /usr/lib/aarch64-linux-gnu/tegra in nvcr.io/nvidia/l4t-jetpack:r36.4.0

When I was installing FFmpeg inmnvcr.io/nvidia/l4t-jetpack:r36.4.0 Docker, I used the GitHub - Keylost/jetson-ffmpeg: ffmpeg support on nvidia jetson library. An error occurred during the CMake process.
root@ls-031:/data/jetson-ffmpeg# mkdir build
root@ls-031:/data/jetson-ffmpeg# cd build/
root@ls-031:/data/jetson-ffmpeg/build# ls
root@ls-031:/data/jetson-ffmpeg/build# cmake ..
-- The C compiler identification is GNU 11.4.0
-- The CXX compiler identification is GNU 11.4.0
-- Detecting C compiler ABI info
-- Detecting C compiler ABI info - done
-- Check for working C compiler: /usr/bin/cc - skipped
-- Detecting C compile features
-- Detecting C compile features - done
-- Detecting CXX compiler ABI info
-- Detecting CXX compiler ABI info - done
-- Check for working CXX compiler: /usr/bin/c++ - skipped
-- Detecting CXX compile features
-- Detecting CXX compile features - done
-- Looking for pthread.h
-- Looking for pthread.h - found
-- Performing Test CMAKE_HAVE_LIBC_PTHREAD
-- Performing Test CMAKE_HAVE_LIBC_PTHREAD - Success
-- Found Threads: TRUE
-- Using NvUtils API.
-- Configuring done
CMake Error: The following variables are used in this project, but they are set to NOTFOUND.
Please set them or make sure they are set and tested correctly in the CMake files:
LIB_NVBUFSURFACE
linked by target "nvmpi" in directory /data/jetson-ffmpeg
LIB_NVBUFSURFTRANSFORM
linked by target "nvmpi" in directory /data/jetson-ffmpeg
LIB_NVJPEG
linked by target "nvmpi" in directory /data/jetson-ffmpeg
LIB_V4L2
linked by target "nvmpi" in directory /data/jetson-ffmpeg
-- Generating done
CMake Generate step failed. Build files cannot be regenerated correctly.
I use the following startup command

docker run -d \
  --name build \
  --gpus all \
  --network=host \
  --device /dev/snd \
  --group-add audio \
  -e DISPLAY=$DISPLAY \
  -e PULSE_SERVER=unix:/run/user/1000/pulse/native \
  -e XAUTHORITY=/root/.Xauthority \
  -v ~/.Xauthority:/root/.Xauthority \
  -v /tmp/.X11-unix:/tmp/.X11-unix:rw \
  -v /dev:/dev \
  -v /mnt/data:/data/ \
  -v /var/run/dbus:/var/run/dbus \
  l4t-jetpack:20260120 \
  sleep infinity

i ask ai and find that i don’t have /usr/lib/aarch64-linux-gnu/tegra in docker , and i wander if it’s normal or not

As long as your nvidia-l4t-apt-source.list looks like this with your nv_tegra_release in place of r36.4

head -n 1 /etc/nv_tegra_release 
# R36 (release), REVISION: 4.7
cat /etc/apt/sources.list.d/nvidia-l4t-apt-source.list 
deb https://repo.download.nvidia.com/jetson/common r36.4 main
deb https://repo.download.nvidia.com/jetson/t234 r36.4 main
deb https://repo.download.nvidia.com/jetson/ffmpeg r36.4 main

Then install ffmpeg and its dependencies with

sudo apt update
sudo apt install ffmpeg

i’m trying to build ffmpeg 7 in jetpack, and source deb ffmpeg vision is 4.
while i use -v /usr/lib/aarch64-linux-gnu/tegra:/usr/lib/aarch64-linux-gnu/tegra i could build, so i’m wandering if is my docker start command unright.

Hi,

Please try to launch the container with --runtime nvidia.
This will mount some hardware-specific library from /usr/lib/aarch64-linux-gnu/tegra.

Thanks.

I tried using --runtime nvidia, but it seems that /usr/lib/aarch64-linux-gnu/tegra was not mounted either. Currently, it appears that I have to mount it manually. Could this be an issue with my Docker settings?

Maybe try jetson-containers as it appears to have ffmpeg 7.1 as a version you could have it install.

package = [
  # ffmpeg('apt', default=True),
  ffmpeg('git', version='7.1', alias=['ffmpeg:7.1'], default=False),
  ffmpeg('git', version='8.0', alias=['ffmpeg:8'], default=True),
  ffmpeg('jetpack', requires='==36.*'),
]

Hi,

When running the container with --runtime nvidia, it will automatically mount the driver listed in the below csv:

/etc/nvidia-container-runtime/host-files-for-container.d/drivers.csv

The libraries you need are included in the csv:

sym, /usr/lib/aarch64-linux-gnu/nvidia/libnvbufsurface.so
sym, /usr/lib/aarch64-linux-gnu/nvidia/libnvbufsurftransform.so
...

Thanks.