The Linux boot process includes five key stages. The first stage is the BIOS/UEFI self-test and select the startup device, which is responsible for detecting the hardware and finding the bootable device in the set order; the second stage is the boot loader (such as GRUB) running, loading the kernel and passing the startup parameters. If GRUB is corrupted or overwritten, the system will not be able to start; the third stage is kernel initialization and hardware detection, loading initramfs to access the root file system, and creating the first process (systemd or init); the fourth stage is user space startup and service initialization, and systemd starts various services such as network, logs, login interfaces, etc. according to the configuration; the fifth stage is the basis for troubleshooting startup problems, and can be gradually checked through GRUB, kernel loading, and service status.

The Linux boot process is actually not that mysterious, but it involves many links. The system needs to go through several critical stages from pressing the power supply to being able to use, and each step is linked together. Understanding these steps will be helpful for troubleshooting startup problems and optimizing system performance.

BIOS/UEFI self-test and boot selection

The first thing to do when powering on is hardware level self-test. BIOS or UEFI will check whether the basic hardware exists and works normally, such as memory, hard disk, graphics card, etc. Then it will look for bootable devices in the set order, such as hard disk, USB disk, CD, etc.

• If the boot device is not specified correctly in the BIOS settings, the system cannot continue
• UEFI has stronger functions than traditional BIOS, supports larger hard disks and faster boot mode
• Some motherboards can press specific keys (such as F12, Del) to enter the startup menu

At this stage, what users can see is usually the manufacturer's logo or short-term hardware detection information. If there is a problem, you may see error messages, such as "Missing operating system".

Bootloader runs

After the boot device is found, the system will load the Bootloader, such as GRUB (Grand Unified Bootloader). GRUB is responsible for loading the core part of the operating system (kernel) and passing startup parameters.

• GRUB is usually installed in the MBR (main boot record) or EFI partition of the hard disk
• If GRUB is damaged, it will enter the "grub rescue>" prompt, and the boot needs to be repaired.
• The GRUB menu lists the available kernel versions and supports entering recovery mode.

Common problems such as Windows overwrites GRUB after dual-system installation, or kernel files are corrupted after system upgrade, which will lead to inability to enter the system.

Kernel initialization and hardware detection

After GRUB loads the kernel, the kernel begins to initialize itself and detects the system hardware. Initramfs (initial RAM disk) is then loaded to prepare to access the real root file system.

• initramfs contains the basic drivers and programs required to mount the root partition
• The kernel creates the first process (PID 1), usually systemd or old version of init
• This stage will output a lot of log information, and some distributions are hidden by default. You can press Esc to view

If you see stuck in "Started GNOME Display Manager" or similar prompts, the problem may be in the graphical interface or service startup process.

Userspace startup and service initialization

After the system switches to the root file system, the first user space process (usually systemd) will be started, and various services will be started according to the configuration, such as network, logs, login interface, etc.

• systemd controls service startup order and dependencies through unit files (.service)
• You can use systemctl list-units --type=service to view the current service status
• The service that failed to start can be viewed using journalctl -b

Some services fail not prevent the system from starting, but may affect subsequent use. For example, if the network service does not start, it cannot be connected to the Internet; if the graphical interface is not started, it can only be logged in using the command line.

Basically that's it. The entire process may have problems at every step from hardware self-test to service startup, but as long as you know the process, there will be directions when troubleshooting. When the startup fails, first check whether GRUB can come out, then check whether the kernel is loaded, and finally check whether the service has reported an error, just take it step by step.

The above is the detailed content of linux boot process explained. For more information, please follow other related articles on the PHP Chinese website!