Writing better code with help from Qt and the compiler
summary
Shows how C/C++ intrinsics and compiler features (GCC, Clang, Intel, MSVC) replace hand assembly for real speedups: uses intrinsics like the crc32 instruction and popcount, combines compile-time macros and runtime dispatching (Qt’s cpu-detection and function-target macros) and linker multi-version libraries. Reports concrete wins: an existing loop drops from ~5M to ~3M cycles with a hand-tuned version and gains another ~3x with lookup-table tricks; those optimizations make it into Qt 5.3/5.4 and behave correctly when tested.
AI-written summary · about
chapters (7)
- 0:00Introduction
- 4:11Bit counting strategies
- 10:40Using compiler intrinsics
- 13:33Targeting specific hardware
- 18:34Qt code examples
- 29:01Runtime dispatching
- 37:11Conclusion and Q&A
