Closed wangyi-fudan closed 5 years ago
you missed a inline keyword which make wyrng slower.
Have you checked the assembly? With all compilers I can try on my test machine, the function gets inlined as it is obviously desirable.
Added the inline keyword. No change either to the binary or to the performance.
Dear Lemire: In your wyrng simplification code, you missed a inline keyword which make wyrng slower.
inline uint64_t wyhash64_stateless(uint64_t *seed) {
adding this inline, wyrng is the fastest :-)