crypto/cipher: use SIMD to improve xor performance

[templexxx]

Please answer these questions before submitting your issue. Thanks!

What version of Go are you using (go version)?

go1.8

What operating system and processor architecture are you using (go env)?

arch=amd64
os=darwin

What did you expect to see?

SIMD ( SSE2 SSSE3 or AVX2) will help xor's performance a lot.

I think I can make it

[FiloSottile]

Hi @templexxx, thanks for offering to make this.

I'll be happy to review a CL. Performance changes usually come with a benchstat.

You might want to have a look at #20967, too.

[ALTree]

Implemented in 5168fcf. Closing here.