crypto/sha512: speed-up for very small blocks

crypto/sha512: speed-up for very small blocks

Similar to https://golang.org/cl/54391, but for sha512
name          old time/op    new time/op    delta
Hash8Bytes-8     289ns ± 1%     253ns ± 2%  -12.59%  (p=0.000 n=10+10)
Hash1K-8        1.85µs ± 1%    1.82µs ± 1%   -1.77%  (p=0.000 n=9+10)
Hash8K-8        12.7µs ± 2%    12.5µs ± 1%     ~     (p=0.075 n=10+10)

name          old speed      new speed      delta
Hash8Bytes-8  27.6MB/s ± 1%  31.6MB/s ± 2%  +14.43%  (p=0.000 n=10+10)
Hash1K-8       554MB/s ± 1%   564MB/s ± 1%   +1.81%  (p=0.000 n=9+10)
Hash8K-8       647MB/s ± 2%   653MB/s ± 1%     ~     (p=0.075 n=10+10)

Change-Id: I437668c96ad55f8dbb62c89c8fc3f433453b5330
Reviewed-on: https://go-review.googlesource.com/82996
Run-TryBot: Ilya Tocar <ilya.tocar@intel.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Filippo Valsorda <hi@filippo.io>