Assembly version of AndUnsafeInplace?

[shenwei356]

I need a faster AND(x, y []byte function (stackoverflow post, and benchmark repo) and just find you provide one.

func AndUnsafeInplace(main, arg []byte) {}

It's so fast by converting every 8 bytes to uint64,

I'm wondering if it could be faster to use golang assembly code utilizing SIMD instructions.

I just begin to learn assembly language, hope I can help one day.

[shenwei356]

I wrote one https://github.com/shenwei356/pand

It works, and the performance is OK.