How Linux Works

This book is by the Japanese author Satoru Takeuchi (formerly of Fujitsu, where he worked on development and support for commercial Linux, especially the kernel), written for technical people who want to truly understand how Linux works. It does not list commands, nor does it dig deep into kernel source; it first "runs" system behavior through experiments, then explains the results with 198 diagrams and charts, turning abstract mechanisms into phenomena you can see for yourself. It covers the main core mechanisms of Linux: programs and processes, the kernel and system calls, the C standard library and the role of libraries; process creation (fork/execve), parent-child relationships, states and signals, zombie and orphan processes, shell job management and daemons; process scheduling, including scheduling under a single logical CPU and multiple logical CPUs, time slices, context switching, and performance; memory management, including reading memory information, memory reclamation, virtual memory, demand paging, and multilevel page tables; advanced process mechanisms such as copy-on-write, interprocess communication, mutexes, multiprocessing, and multithreading; and storage hierarchy, file systems, and external storage, including memory-mapped files, disk quotas, journaling and copy-on-write consistency, and Btrfs snapshots and volumes. The experiment programs are written in C and Python, with runnable accompanying code and verification methods. Readers run things by hand while checking the diagrams, working out one by one why Linux behaves as it does. The book suits application developers, system designers, operations and maintenance staff, and technical support personnel, as well as readers filling a gap in their knowledge of operating systems; only some basic Linux commands are needed to read it smoothly.