Low-Level Linux: Memory, Syscalls, and Performance Tuning from the Metal Up
暫譯: 低階 Linux：從底層到性能調優的記憶體與系統呼叫

Modern applications live or die by performance, reliability, and efficiency - yet most developers never truly understand what happens after their code hits the kernel. Low-Level Linux takes you beneath abstractions and frameworks and into the machinery that actually runs your software.

This book is a practical, deeply technical guide to how Linux works from the metal up - from virtual memory and page tables to syscalls, scheduling, and performance tuning at scale.

Whether you're optimizing high-throughput systems, debugging impossible latency spikes, or preparing for kernel-level work, this book gives you the mental models used by elite systems engineers.

You will learn how Linux really behaves under load - not just how it's supposed to.

Inside you'll master:

• How Linux virtual memory actually works
  (paging, MMUs, TLBs, NUMA, and page faults demystified)

• The full syscall lifecycle
  from user space → kernel → hardware → back again

• How context switching, scheduling, and interrupts impact latency

• Memory allocation internals
  (slab, slub, vmalloc, brk, mmap)

• CPU caches, false sharing, and cache-line-level optimization

• Diagnosing real performance problems using
  perf, ftrace, strace, bcc, and eBPF

• Why "fast code" is often slow - and how to prove it

• How to reason about performance using first principles, not guesswork

Who this book is for:

• Software engineers ready to go beyond frameworks

• Linux users who want to understand what their system is truly doing

• Backend, infrastructure, and systems engineers

• Performance engineers and SREs

• Anyone who wants to think like the kernel does

No fluff. No hand-waving. Just real Linux internals explained clearly, visually, and practically.