News for hackers. Stuff that matters.
Re:tolower() in bulk at speed by rep_movsd on Tuesday June 28, 2022 @04:43AM (#31905545)
Would it not be simpler to bitwise OR every 5th bit to one based on a byte mask for uppercase chars?

Also, use SIMD - I think even AVX can do byte level paralellism for 256 and 512 bits at a time?

There are no comments on this comment yet.