Boot Process Overview

Boot is a chain of trust and control transfers that turns platform reset into a running user-space environment. A common PC path can be summarized as firmware, boot manager or loader, kernel with optional early user space, and the PID 1 init system. Architectures, virtual machines, embedded systems, and containers can use different paths.

Firmware Initialization

Platform firmware initializes enough CPU, memory, and device state to choose a boot target. Traditional PCs use BIOS conventions; current PCs commonly use UEFI. Firmware settings, boot order, platform verification, and Secure Boot policy can determine which next-stage executable is allowed to run.

Firmware does not necessarily understand the installed Linux root filesystem. It locates a boot path according to its interface—for example, BIOS boot code on a selected disk or a UEFI boot entry pointing to an EFI executable on an EFI System Partition.

Boot Loader or Boot Manager

A loader such as GRUB can present entries, load a selected Linux kernel and initial RAM filesystem into memory, construct the kernel command line, and transfer control. UEFI can also load a kernel built as an EFI executable directly, so a separate multi-stage loader is common rather than universal.

The selected artifacts must agree: kernel version, initramfs content, root identifier, security signatures, and command-line options all affect whether the next handoff succeeds.

Kernel and Early User Space

The kernel decompresses or relocates as required, initializes core subsystems, parses its command line, and discovers available hardware. An initramfs can supply modules and early tools needed for storage discovery, RAID, encryption, LVM, networking, or other work required to assemble the real root filesystem.

After the intended root is available, early user space switches to it and the kernel executes the configured first user-space program. Details such as who performs filesystem checks or read-write remounting belong to the distribution's boot design rather than one universal sequence.

PID 1 and System Readiness

The first user-space process receives PID 1. On many distributions it is systemd; other systems use sysvinit, OpenRC, runit, BusyBox init, or a specialized program. PID 1 establishes the user-space service environment, reaps orphaned children, and handles shutdown responsibilities.

Reaching PID 1 does not mean the system is fully ready. Services can still be starting, storage can be mounting, network configuration can be pending, and a graphical or console login is only one possible target state.

The Customize the GRUB2 Boot Menu lab demonstrates one loader configuration path. Apply changes only in a recovery-capable lab system.

Summary

You can now trace the major Linux boot handoffs without treating them as universal implementation details.

1. Start with firmware initialization and target selection.
2. Relate the loader to kernel, initramfs, and command-line selection.
3. Use early user space to understand complex root assembly.
4. Treat PID 1 as the beginning of service initialization, not proof of readiness.