Hacker Newsnew | past | comments | ask | show | jobs | submitlogin

Unfortunately __builtin_popcnt isn't emitting a popcnt instruction with the GCC I've got here, even using -msse4.2. I believe that very recent GCC does get this right.


GCC 4.4 isn't very recent, but it generates a popcnt with -msse4.2.

GCC 4.5, using popcnt on my Core-i7 860 takes the trivial loop mentioned using "Complement and Compare" from ~10.5s to ~7.5s




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: