Linux Kernel Internals

1 October 20262 min read

Community-written documentation on how the Linux kernel works — with a focus on why things are designed the way they are, rather than just API reference. It collects the rationale that is otherwise scattered across mailing list archives and tribal knowledge. Covers modern mainline Linux (io_uring, BPF, cgroup v2, sched_ext, MGLRU, …) on x86 and ARM64.

Docs are licensed CC BY-SA 4.0, code snippets GPL-2.0. Contributions go through the GitHub repo.

Sections

• Foundations — boot/init, architecture (x86, ARM64), syscalls, modules, timekeeping

• Memory — page allocator, SLUB, vmalloc, NUMA, address translation, page cache and reclaim, memory cgroups

• Scheduling & Concurrency — CFS/EEVDF, spinlocks and mutexes, RCU, interrupts, IPC

• Storage & I/O — VFS, filesystems (ext4, btrfs, XFS), block layer, io_uring

• Devices & Drivers — device model, USB, GPU/DRM, ALSA, input/HID, TTY

• Networking & BPF — packet flow, TCP/IP, netfilter, XDP, BPF verifier and maps

• Isolation & Security — cgroups, LSM, capabilities, seccomp, KVM, crypto, live patching

• Observability — ftrace, perf, kprobes, KGDB, kdump, power management

• Background / Evolution — history of Linux and hardware

Formats

• Deep dives — subsystem internals with the design rationale

• "Life of …" walkthroughs — e.g. Life of a malloc, Life of a packet receive

• War Stories — real kernel bugs and incidents

• Explainers and tuning guides

• References — glossary, /proc documentation