crypto/des: improve the throughput of DES and 3DES

For detailed explanation of the adopted (Eric Young's) algorithm,
see http://ftp.nluug.nl/security/coast/libs/libdes/ALGORITHM

benchmark                   old ns/op     new ns/op     delta
BenchmarkEncrypt-16         649           164           -74.73%
BenchmarkDecrypt-16         546           156           -71.43%
BenchmarkTDESEncrypt-16     1651          385           -76.68%
BenchmarkTDESDecrypt-16     1645          378           -77.02%

benchmark                   old MB/s     new MB/s     speedup
BenchmarkEncrypt-16         12.31        48.76        3.96x
BenchmarkDecrypt-16         14.64        51.03        3.49x
BenchmarkTDESEncrypt-16     4.84         20.74        4.29x
BenchmarkTDESDecrypt-16     4.86         21.16        4.35x

Change-Id: Ic3e1fe3340419ec5a0e6379434911eb41e0246f6
Reviewed-on: https://go-review.googlesource.com/36490


Run-TryBot: Minux Ma <minux@golang.org>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Russ Cox <rsc@golang.org>