SW installations on NVIDIA DRIVE

Please provide the following info (check/uncheck the boxes after creating this topic):
Software Version
DRIVE OS Linux 5.2.6
DRIVE OS Linux 5.2.6 and DriveWorks 4.0
DRIVE OS Linux 5.2.0
DRIVE OS Linux 5.2.0 and DriveWorks 3.5
[*] NVIDIA DRIVE™ Software 10.0 (Linux)
NVIDIA DRIVE™ Software 9.0 (Linux)
other DRIVE OS version
other

Target Operating System
[*] Linux
QNX
other

Hardware Platform
[*] NVIDIA DRIVE™ AGX Xavier DevKit (E3550)
NVIDIA DRIVE™ AGX Pegasus DevKit (E3550)
other

SDK Manager Version
1.8.0.10363
other

Host Machine Version
[*] native Ubuntu 18.04
other

Hi everyone,

We would like to perform some SW installations on NVIDIA DRIVE AGX Xavier. Is there any precaution we need to take into account before deploying installations on DRIVE?

We would like to install C++ boost library (ex with apt get install libboost-all-dev) and we also have a few peripheral devices we would like to use that can be interfaced with python, so we may install via pip a few python packages as well (ex ‘libusb1 == 1.6.6’, ‘hexdump >= 3.3’, ‘pycrypto >= 2.6.1’, ‘tqdm >= 4.14.0’, ‘requests’, etc)

We are concerned about possible conflicts or system corruption if some package installation conflicts in general with what is already install, so please advise us for any guidelines from your side upon installing anything on the Target PC.

Are you tryng to install on the target PC or on NVIDIA DRIVE AGX Xavier?

Customer can refer to DRIVE OS 5.2.6 Installation Guide to install SW on host PC or DRIVE AGX Xavier device.

You can also find some webinars at Training | NVIDIA Developer to get more undrstanding how to setup your environment and developing project.

Hello,

We are trying to install SW on NVIDIA DRIVE AGX Xavier device.

The first link you have provided is the DRIVE Platform Installation Guide for NVIDIA SW. We are about to install extra libraries and SW on NVIDIA DRIVE AGX, on top of NVIDIA DRIVE SW itself and we though of asking if there might be a conflict in doing that.

What is generally the recommendation from your side when performing “sudo apt-get” installations or “pip” type installations for any kind of SW package on the NVIDIA DRIVE platform? Should we just go through with it and see if it works or should we consider anything in advance, any compatibility issue for example?

1 Like

In general, at Linux OS, the “sudo apt-get” installations or “pip” type installations should work.
For those SW which are not provided by NVIDIA DRIVE SW package, that will be case by case if any issue met due to we’e not possible to verify all of them until customer reported issues.

I’m trying to install numpy onto the target filesystem in one of the nvidia provided docker containers, I think pip allows this with the --platform and --target flags the package I’m trying to install is numpy the host container is x86_64 and the target is aarch64 (orin).

This is the command I’m trying but when cross-compiling I’m unable to find numpy on the target.

pip install numpy --platform manylinux_2_17_aarch64 --target=/drive/drive-linux/filesystem/targetfs/usr/python3.8 --implementation cp --python 3.8 --only-binary=:all:

Has anyone done this or is there something I’m missing?

Please create a new topic for your issue. Thanks.