Setup¶
This document describes the Linux for Tegra (L4T) installation process for a custom Jetson Nano based carrier board. The packages used are derived from NVIDIA’s official releases and adapted to support board specific hardware components.
Note
The steps are written for Jetson Nano. The same procedure applies to other Jetson models; only the Ubuntu version and package names differ.
Requirements¶
- Jetson-210_Linux_R32.7.6_aarch64.tbz2
- Tegra_Linux_Sample-Root-Filesystem_R32.7.6_aarch64.tbz2
- Host system:
- Ubuntu 18.04
- Or Ubuntu 18.04 running inside Docker
Info
By default, the installation is performed under a workspace directory in the user’s home path. A custom directory may be used if preferred
-
Extracting the L4T Package: Create the workspace and extract the L4T package (This will create the Linux_for_Tegra directory.)
-
Installing the Root Filesystem: Extract the root filesystem into the rootfs directory:
-
Applying NVIDIA Binary Files: Run the following script to install NVIDIA-specific binaries into the root filesystem
-
Verifying the Installation: Verify that the base system was created successfully. (If the file exists, the setup is valid.)
-
Creating a Default User: To predefine the default user credentials (
-u: Username,-p: Password,-n: Hostname) -
Entering Force Recovery Mode (RCM): To flash the device, Jetson Nano must be placed into Force Recovery Mode (RCM).
- Press and hold Recovery and Reset pins together
- Wait 15–30 seconds
- Release Reset first
- Then release Recovery
- Connect the board to the host PC via USB Type-B and verify:
lsusbIf NVIDIA Corp. appears, the device is ready for flashing.
-
Flashing the Device: