An explanation of how Linux handles system calls on x86-64 and why they show up as expensive operations in performance profiles

Code Confessions is a platform or blog where developers share their experiences, challenges, and insights from their journey in software development. Through personal anecdotes, lessons learned, and reflections on coding practices, Code Confessions offers a candid and relatable perspective on the joys and struggles of being a programmer. Readers can find inspiration, empathy, and  lessons from the stories shared by fellow developers.

Confessions of a Code Addict

System calls are expensive operations that go beyond their kernel code execution cost. When transitioning between user and kernel space on Linux x86-64, the CPU must drain instruction pipelines, switch page tables and stacks, clear branch predictor buffers, and apply security mitigations against speculative execution attacks like Spectre. These microarchitectural disruptions force the CPU to rebuild its optimization state, making system calls significantly more costly than they appear. The article explores the Linux kernel's syscall handler, measures direct overhead using clock_gettime benchmarks, and explains indirect costs from pipeline draining and branch predictor clearing. Practical optimization strategies include using vDSO, caching values, batching I/O operations, leveraging io_uring, and utilizing eBPF to reduce kernel crossings.

What Makes System Calls Expensive: A Linux Internals Deep Dive

<p>Very interesting, despite knowing my faire share of x86 assembly, I never knew about vDSOs until now.</p>
