strings, bytes, strconv, runtime: use SSE4.2 string instructions on amd64

[gopherbot]

by rayneolivetti:

Chapter 10 of Intel Architectures Optimization Reference Manual describes use cases of
several SSE4 instructions that can speed up various routines in strings, bytes, strconv
and runtime packages.

The instructions are missing from 6l/6a by the way.

[rsc]

Comment 1:

Labels changed: added go1.3maybe.

[rsc]

Comment 2:

Labels changed: added release-none, removed go1.3maybe.

[rsc]

Comment 3:

Labels changed: added repo-main.

[agnivade]

@rsc - Is this still valid ? We already have SSE and AVX2 supported instructions in asm_amd64.s.

[agnivade]

ping @martisch / @quasilyte - Does this still need to remain open ?

[randall77]

@TocarIP