strings: use SSE4.2 in strings.Index on AMD64

strings: use SSE4.2 in strings.Index on AMD64

Use PCMPESTRI instruction if available.

Index-4              21.1ns ± 0%  21.1ns ± 0%     ~     (all samples are equal)
IndexHard1-4          395µs ± 0%   105µs ± 0%  -73.53%        (p=0.000 n=19+20)
IndexHard2-4          300µs ± 0%   147µs ± 0%  -51.11%        (p=0.000 n=19+20)
IndexHard3-4          665µs ± 0%   665µs ± 0%     ~           (p=0.942 n=16+19)

Change-Id: I4f66794164740a2b939eb1c78934e2390b489064
Reviewed-on: https://go-review.googlesource.com/22337
Run-TryBot: Ilya Tocar <ilya.tocar@intel.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Russ Cox <rsc@golang.org>