Guidelines for Compiling and Building the Linux Kernel

Example 1

The following uses the Hi3516D V300 board and Ubuntu x86 server as an example.

Scenario 1: Building the Native Kernel at the Version Level

Perform a full build for the project to generate the uImage kernel image.

./build.sh --product-name Hi3516DV300 # Build the uImage kernel image of the Hi3516D V300 board.

Scenario 2: Building the Modified Kernel Separately

  1. Set up the build environment.

    1. Merge the required patch by referring to guidelines for using patches on development boards.
    2. Prepare for the build environment. You can use the Arm Clang or GCC compiler.

    Enter the root directory of the project and configure environment variables:

    export PATH=`pwd`/prebuilts/clang/host/linux-x86/clang-r353983c/bin:`pwd`/prebuilts/gcc/linux-x86/arm/gcc-linaro-7.5.0-arm-linux-gnueabi/bin/:$PATH # Configure the build environment.
MAKE_OPTIONES="ARCH=arm CROSS_COMPILE=arm-linux-gnueabi- CC=clang HOSTCC=clang" # Use Clang provided by the project.

  2. Modify the kernel code or kernel configuration (defconfig file provided by OpenHarmony can be used for reference).

  3. Create a build directory and generate the .config file of the kernel.

    make ${MAKE_OPTIONES} hi3516dv300_emmc_smp_hos_l2_defconfig # Use the defconfig file to build the kernel.

  4. Build the kernel image.

    make ${MAKE_OPTIONES} -j32 uImage # Build the uImage kernel image.