How to build NVIDIA Jetson Nano kernel

From RidgeRun Developer Wiki


Previous: Development/Getting_into_the_Board Index Next: GStreamer/Accelerated_Elements






Introduction

This wiki page contains instructions to download and build kernel source code for Jetson Nano, several parts of this wiki were based in the document: NVIDIA Tegra Linux Driver Package Development Guide 32.3.1.

L4T 32.3.1 is used by JetPack 4.3 (You can get more information from jetpack-archive).

Previous versions:

Build Kernel

Follow these instructions to build and install the kernel image and device tree.

1. Download and install the Toolchain

NVIDIA recommends using the Linaro 7.3.1 2018.05 toolchain for L4T 32.3 (Toolchain)

Download the pre-built toolchain binaries from: http://releases.linaro.org/components/toolchain/binaries/7.3-2018.05/aarch64-linux-gnu/gcc-linaro-7.3.1-2018.05-x86_64_aarch64-linux-gnu.tar.xz

wget http://releases.linaro.org/components/toolchain/binaries/7.3-2018.05/aarch64-linux-gnu/gcc-linaro-7.3.1-2018.05-x86_64_aarch64-linux-gnu.tar.xz

Execute the following commands to extract the toolchain:

mkdir $HOME/l4t-gcc
cd $HOME/l4t-gcc
tar xf gcc-linaro-7.3.1-2018.05-x86_64_aarch64-linux-gnu.tar.xz

2. Download the kernel sources

You can download the kernel source files and then manually extract them. it is recommended to instead sync with git.

In a browser, navigate to: https://developer.nvidia.com/embedded/downloads. Locate and download the L4T Sources for your release. (L4T Sources 32.3.1 2019/12/17) or run the command below:

wget https://developer.download.nvidia.com/embedded/L4T/r32-3-1_Release_v1.0/Sources/T210/public_sources.tbz2

Execute the following commands to extract the kernel:

tar -xvf public_sources.tbz2
cd Linux_for_Tegra/source/public
JETSON_NANO_KERNEL_SOURCE=$(pwd)
tar -xf kernel_src.tbz2

3. Compile kernel and dtb

Follow the steps:

cd $JETSON_NANO_KERNEL_SOURCE
TOOLCHAIN_PREFIX=$HOME/l4t-gcc/gcc-linaro-7.3.1-2018.05-x86_64_aarch64-linux-gnu/bin/aarch64-linux-gnu-
TEGRA_KERNEL_OUT=$JETSON_NANO_KERNEL_SOURCE/build
KERNEL_MODULES_OUT=$JETSON_NANO_KERNEL_SOURCE/modules
make -C kernel/kernel-4.9/ ARCH=arm64 O=$TEGRA_KERNEL_OUT LOCALVERSION=-tegra CROSS_COMPILE=${TOOLCHAIN_PREFIX} tegra_defconfig
make -C kernel/kernel-4.9/ ARCH=arm64 O=$TEGRA_KERNEL_OUT LOCALVERSION=-tegra CROSS_COMPILE=${TOOLCHAIN_PREFIX} -j8 --output-sync=target zImage
make -C kernel/kernel-4.9/ ARCH=arm64 O=$TEGRA_KERNEL_OUT LOCALVERSION=-tegra CROSS_COMPILE=${TOOLCHAIN_PREFIX} -j8 --output-sync=target modules
make -C kernel/kernel-4.9/ ARCH=arm64 O=$TEGRA_KERNEL_OUT LOCALVERSION=-tegra CROSS_COMPILE=${TOOLCHAIN_PREFIX} -j8 --output-sync=target dtbs
make -C kernel/kernel-4.9/ ARCH=arm64 O=$TEGRA_KERNEL_OUT LOCALVERSION=-tegra INSTALL_MOD_PATH=$KERNEL_MODULES_OUT modules_install

Flash Jetson NANO

This guide assumes that the user already has sdk-manager installed. This link contains details about how to install sdk-manager: https://developer.nvidia.com/nvidia-sdk-manager https://docs.nvidia.com/sdk-manager/download-run-sdkm/index.html

NVIDIA_SDK_MANAGER contains the directory where NVIDIA SDK manager was installed. For example.

export NVIDIA_SDK_MANAGER=$HOME/nvidia/nvidia_sdk/

Select DTB and directory based on Jetson Nano module type

Production module (P3448-0020)

export DTB=tegra210-p3448-0002-p3449-0000-b00.dtb
JETPACK_4_3_P3448=${NVIDIA_SDK_MANAGER}/JetPack_4.3_Linux_P3448-0020/Linux_for_Tegra

Devkit module (P3448)

export DTB=tegra210-p3448-0000-p3449-0000-a02.dtb
JETPACK_4_3_P3448=${NVIDIA_SDK_MANAGER}/JetPack_4.3_Linux_P3448/Linux_for_Tegra

Make sure the Jetson Nano is in recovery mode

Copy kernel, device tree and modules into jetpack

cd ${JETPACK_4_3_P3448}
# Copy kernel generated
cp $JETSON_NANO_KERNEL_SOURCE/build/arch/arm64/boot/Image kernel/
# Copy device tree generated
cp $JETSON_NANO_KERNEL_SOURCE/build/arch/arm64/boot/dts/${DTB} kernel/dtb/

Apply root filesystem patch

With the default root file system, the changes in the device tree are lost when the board is powered off. This is a known issue (https://devtalk.nvidia.com/default/topic/1072091/jetson-nano/how-to-modify-the-dts-file-for-jetson-nano-board/) Apply this patch to avoid losing the changes added to the device tree.

Create patch file

Patch: fix_dtb_lost.patch

diff --git a/rootfs/opt/nvidia/l4t-bootloader-config/nv-l4t-bootloader-config.sh b/rootfs/opt/nvidia/l4t-bootloader-config/nv-l4t-bootloader-config.sh
index 921efc2..fe60093 100755
--- a/rootfs/opt/nvidia/l4t-bootloader-config/nv-l4t-bootloader-config.sh
+++ b/rootfs/opt/nvidia/l4t-bootloader-config/nv-l4t-bootloader-config.sh
@@ -123,7 +123,7 @@ t210ref_update_qspi_check () {
                        # need to update QSPI
                        return 0
                elif (( "${sys_maj_rev}" == "${qspi_maj_rev}" )); then
-                       if (( "${sys_min_rev}" >= "${qspi_min_rev}" )); then
+                       if (( "${sys_min_rev}" > "${qspi_min_rev}" )); then
                                # sys_rel == qspi_rel
                                # sys_maj_rev == qspi_maj_rev
                                # sys_min_rev >= qspi_min_rev

Apply the patch

cd ${JETPACK_4_3_P3448}
sudo patch -p 1 < fix_dtb_lost.patch

Flash memory

Make sure the Jetson NANO is in recovery mode.

Flash production module

sudo ./flash.sh jetson-nano-emmc mmcblk0p1

Flash Devkit module

sudo ./flash.sh jetson-nano-qspi-sd mmcblk0p1

Flash device tree only

This command flashes only the device tree, excluding the kernel image and filesystem. Make sure the Jetson NANO is in recovery mode.

cd ${JETPACK_4_3_P3448}
# Copy device tree generated
cp $JETSON_NANO_KERNEL_SOURCE/build/arch/arm64/boot/dts/${DTB} kernel/dtb/
Flash production module DTB
sudo ./flash.sh -r -k DTB jetson-nano-emmc mmcblk0p1
Flash Devkit module DTB
sudo ./flash.sh -r -k DTB jetson-nano-qspi-sd mmcblk0p1


Previous: Development/Getting_into_the_Board Index Next: GStreamer/Accelerated_Elements