256 KB of code to zero 64 KB of memory is the kind of optimization that makes you question every life choice that led to it.
I blame Intel. It took them 33 years (ERMSB) to finally standardize REP MOVSB as _the_ fast path. Another 10 years passed and someone discovered https://lock.cmpxchg8b.com/reptar.html
I blame Intel. It took them 33 years (ERMSB) to finally standardize REP MOVSB as _the_ fast path. Another 10 years passed and someone discovered https://lock.cmpxchg8b.com/reptar.html