Your code is fast – if you're lucky

## Your code is fast - if you’re lucky Recently, while I was working on an optimized Quicksort implementation, I came across a rather interesting quirk. Modern compilers (especially Clang) optimize loops using fast, branch-free instructions - provided you use the right programming style. sort.h -

Your code is fast - if you’re lucky

Recently, while I was working on an optimized Quicksort implementation, I came across a rather interesting quirk. Modern compilers (especially Clang) optimize loops using fast, branch-free instructions - provided you use the right programming style.

sort.h - a quicksort with sorting networks

A few cosmetic changes

It is already micro-optimized using sorting networks and loop unrolling. Only a few cosmetic changes remain.

We rewrite this beginner‑friendly style, which explicitly shows how the pointers are moved:

But what actually happened?

This “small cosmetic” change causes Clang to replace branches with csel.

With branches

GCC does not exhibit this “quirk” (different code generation for logically equivalent source). It consistently generates the slower branch-based version.

Links

blqsort - Fast Quicksort with C and C++ Interface

When ‘if’ slows you down, avoid it

Interactive sorting demo

christof.kaser@gmail.com