(new users can only have one link per post, so I’ve hosted the question as a gist here with links: Jetson-l4t-question.md · GitHub)
On my Jetson Xavier NX board, I do
$ cat /etc/nv_tegra_release
# R32 (release), REVISION: 7.2, GCID: 30192233, BOARD: t186ref, EABI: aarch64, DATE: Sun Apr 17 09:53:50 UTC 2022
If I’ve understood correctly, that means that I should build docker images off e.g.l4t-pytorch:r32.7.2
, but I only see JetPack 4.6.1 (L4T R32.7.1) at https://catalog.ngc.nvidia.com/orgs/nvidia/containers/l4t-pytorch
.
That’s probably stemming from the fact that there is no R32.7.2 at https://catalog.ngc.nvidia.com/orgs/nvidia/containers/l4t-base/tags
I believe the Dockerfiles for those l4t-base images are at https://gitlab.com/nvidia/container-images/l4t-base/-/blob/master/Dockerfile.l4t_r32
Ultimately, I’d like to use an image like dustynv/ros:humble-pytorch-l4t-r35.1.0
(at https://hub.docker.com/layers/dustynv/ros/humble-pytorch-l4t-r35.1.0/images/sha256-538fe2976ac24289422e20aea579ded571148c7c6f5e1fb0df8d31b2a4c11c93
, from https://github.com/dusty-nv/jetson-containers
).
What goes wrong if I run an r35/r34 image on a Jetson with r32 installed?