News for hackers. Stuff that matters.

A comment on Popcount CPU instruction. Parent comment

Re:Popcount CPU instruction by vnorilo on Monday September 09, 2019 @04:26AM (#20916199)
It could be a bug; I don't see an immediate reason to zero eax in this fragment. Sometimes it's easy to miss the reasons why compilers zero apparently unused things though: it can be about breaking data dependencies or making sure flags are in a known state.
[ Parent | on HN ]
Re:Popcount CPU instruction by gpderetta on Monday September 09, 2019 @04:49AM (#20916314)
It is a bug, but in the CPU. Until recently, on intel cpu, the popcnt instruction had an implicit dependency on its output register, which can be a performance issue. Explicitly clearing the register (which is very fast and almost free), is an easy way to workaround it.
[ Parent | on HN ]
Re:Popcount CPU instruction by account42 on Monday September 09, 2019 @07:28AM (#20917135)
Oh, nice info. It's gone when using e.g. -march=znver1.
[ Parent | on HN ]