Summary

This blog post describes my experience following the first tutorial for Linux Kernel Development. The main goal is to create a safe and practical environment for kernel development using Virtual Machines (VMs), avoiding the risks of testing directly on the host machine. Instead of rebooting the physical system after every kernel change, we can compile, boot, debug, and test everything inside an isolated VM.

The tutorial uses QEMU as the virtualization backend and libvirt as the management layer. QEMU provides the actual virtual machine execution, while libvirt simplifies tasks such as VM creation, networking, storage handling, and repeated launches. This makes the workflow much cleaner compared to manually invoking long QEMU commands every time. For kernel work, especially when debugging drivers or experimenting with system internals, this setup is far more convenient and significantly reduces friction.

My Experience

Overall, the process was smoother than I expected. Since my main machine runs Ubuntu, I decided to use the Debian-based path suggested by the tutorial. Package installation, configuring the VM and creating the activate.sh script worked without major issues.

The only issue I encountered was that I needed to install SSH inside the VM, which was not explicit in the tutorial. For reference, my final activate.sh file looked like this.

activate.sh

#!/bin/bash

#!/usr/bin/env bash

# environment variables
export LK_DEV_DIR='/home/lk_dev' # path to testing environment directory
export VM_DIR="${LK_DEV_DIR}/vm"
export BOOT_DIR="${VM_DIR}/arm64_boot" # path to boot artifacts

# Launches a VM with a custom kernel and `initrd`
function launch_vm_qemu() {
    sudo qemu-system-aarch64 \
        -M virt,gic-version=3 \
        -m 2G -cpu cortex-a57 \
        -smp 2 \
        -netdev user,id=net0 -device virtio-net-device,netdev=net0 \
        -initrd "${BOOT_DIR}/initrd.img-6.1.0-43-arm64" \
        -kernel "${IIO_TREE}/arch/arm64/boot/Image" \
        -append "loglevel=8 root=/dev/vda2 rootwait" \
        -device virtio-blk-pci,drive=hd \
        -drive if=none,file="${VM_DIR}/arm64_img.qcow2",format=qcow2,id=hd \
        -nographic
}

# Registers and starts a VM with `libvirt`
function create_vm_virsh() {
    sudo virt-install \
      --name "arm64" \
      --memory 2048 \
      --arch aarch64 --machine virt \
      --import \
      --features acpi=off \
      --disk path="${VM_DIR}/arm64_img.qcow2" \
      --boot kernel=${IIO_TREE}/arch/arm64/boot/Image,initrd=${BOOT_DIR}/initrd.img-6.1.0-43-arm64,kernel_args="loglevel=8 root=/dev/vda2 rootwait" \
      --network bridge:virbr0 \
      --graphics none
}


# export functions so they persist in the new Bash shell session
export -f launch_vm_qemu
export -f create_vm_virsh

# prompt preamble
prompt_preamble='(LK-DEV)'

# colors
GREEN="\e[32m"
PINK="\e[35m"
BOLD="\e[1m"
RESET="\e[0m"

# launch Bash shell session w/ env vars defined
echo -e "${GREEN}Entering shell session for Linux Kernel Dev${RESET}"
echo -e "${GREEN}To exit, type 'exit' or press Ctrl+D.${RESET}"
exec bash --rcfile <(echo "source ~/.bashrc; PS1=\"\[${BOLD}\]\[${PINK}\]${prompt_preamble}\[${RESET}\] \$PS1\"")