Booting, halting and power management

Booting and halting

Boot sequence

Firmware/POST – hardware initialization, device detection

Bootloader (GRUB, systemd-boot, EFI stub_doc) – loads kernel image into memory

Kernel – decompresses, initializes, starts man 1 init

Kernel booting

This is loaded in two stages - in the first stage the kernel (as a compressed image file) is loaded into memory and decompressed, and a few fundamental functions such as essential hardware and basic memory management (memory paging) are set up. Control is then switched one final time to the main kernel start process calling start_kernel_id, which then performs the majority of system setup (interrupts, the rest of memory management, device and driver initialization, etc.) before spawning separately, the idle process and scheduler, and the init process (which is executed in user space).

Kernel loading stage

The kernel as loaded is typically an image file, compressed into either zImage or bzImage formats with zlib. A routine at the head of it does a minimal amount of hardware setup, decompresses the image fully into high memory, and takes note of any RAM disk if configured. It then executes kernel startup via startup_64 (for x86_64 architecture).

arch/x86/boot_src – boot protocol, CPU feature detection (x86 Feature Flags_doc)

arch/x86/boot/compressed/vmlinux.lds.S_src - linker script defines entry startup_64_id in

arch/x86/boot/compressed/head_64.S_src - assembly of extractor

extract_kernel_id - extractor in language C

prints

Decompressing Linux... done.
Booting the kernel.

Kernel startup stage

The startup function for the kernel (also called the swapper or process 0) establishes memory management (paging tables and memory paging), detects the type of CPU and any additional functionality such as floating point capabilities, and then switches to non-architecture specific Linux kernel functionality via a call to start_kernel_id.

↯ Startup call hierarchy:

arch/x86/kernel/vmlinux.lds.S_src – linker script

arch/x86/kernel/head_64.S_src – assembly of uncompressed startup code

arch/x86/kernel/head64.c_src – platform dependent startup:

x86_64_start_kernel_id

x86_64_start_reservations_id

init/main.c_src – main initialization code

init/init_task.c_src – initial task (PID 0, swapper), init_task_id

init/do_mounts.c_src – root device mounting (mount_root_id, prepare_namespace_id)

init/do_mounts_initrd.c_src – initrd (initial ramdisk) unpacking and pivot to real root

init/do_mounts_rd.c_src – RAM disk image mounting (legacy rd= parameter)

init/initramfs.c_src – built-in initramfs (cpio) unpacking (populate_rootfs_id)

init/calibrate.c_src – BogoMIPS delay loop calibration (calibrate_delay_id)

init/version.c_src – kernel version string and build timestamp (linux_banner_id)

init/noinitramfs.c_src – default root mount when no initramfs is configured

start_kernel_id 200 SLOC

rcu_init_id – Read-copy-update

rest_init_id

kernel_init_id - deferred kernel thread #1

kernel_init_freeable_id This and following functions are defined with attribute __init_id

prepare_namespace_id

initrd_load_id

mount_root_id

run_init_process_id obviously runs the first process man 1 init

kthreadd_id – deferred kernel thread #2

cpu_startup_entry_id

do_idle_id

🛠️ Utilities

man 8 dracut – generate initramfs images; dracut --list-modules shows available modules

man 1 lsinitrd – list contents of an initramfs image

start_kernel_id executes a wide range of initialization functions. It sets up interrupt handling (IRQs), further configures memory, starts the man 1 init process (the first user-space process), and then starts the idle task via cpu_startup_entry_id. Notably, the kernel startup process also mounts the initial ramdisk (initrd) that was loaded previously as the temporary root file system during the boot phase. The initrd allows driver modules to be loaded directly from memory, without reliance upon other devices (e.g. a hard disk) and the drivers that are needed to access them (e.g. a SATA driver). This split of some drivers statically compiled into the kernel and other drivers loaded from initrd allows for a smaller kernel. The root file system is later switched via a call to man 8 pivot_root / man 2 pivot_root which unmounts the temporary root file system and replaces it with the use of the real one, once the latter is accessible. The memory used by the temporary root file system is then reclaimed.