crypto/md5: optimize amd64 assembly
* Use two ADDL instead of LEAL
* Keep ones in R11
* Use XORL with lower latency instead of NOTL
* Remove loads and load the correct value in the previous round
* Reduce dependency chain in round 2.
* Remove MOVL in round 3.
name old time/op new time/op delta
Hash8Bytes-32 104ns ± 0% 96ns ± 1% -7.83% (p=0.000 n=9+10)
Hash64-32 169ns ± 0% 155ns ± 0% -7.97% (p=0.000 n=10+10)
Hash128-32 244ns ± 0% 224ns ± 0% -8.16% (p=0.000 n=9+10)
Hash256-32 396ns ± 0% 360ns ± 1% -9.01% (p=0.000 n=10+10)
Hash512-32 700ns ± 1% 634ns ± 1% -9.43% (p=0.000 n=10+10)
Hash1K-32 1.30µs ± 0% 1.18µs ± 1% -9.32% (p=0.000 n=9+10)
Hash8K-32 9.77µs ± 0% 8.81µs ± 0% -9.78% (p=0.000 n=9+10)
Hash1M-32 1.24ms ± 1% 1.12ms ± 1% -9.54% (p=0.000 n=10+10)
Hash8M-32 10.0ms ± 1% 9.0ms ± 1% -10.04% (p=0.000 n=10+10)
Hash8BytesUnaligned-32 104ns ± 0% 96ns ± 0% -7.50% (p=0.000 n=10+10)
Hash1KUnaligned-32 1.32µs ± 1% 1.18µs ± 1% -10.42% (p=0.000 n=10+10)
Hash8KUnaligned-32 9.80µs ± 0% 8.79µs ± 1% -10.29% (p=0.000 n=10+10)
name old speed new speed delta
Hash8Bytes-32 77.1MB/s ± 0% 83.6MB/s ± 1% +8.49% (p=0.000 n=9+10)
Hash64-32 379MB/s ± 0% 412MB/s ± 0% +8.66% (p=0.000 n=10+10)
Hash128-32 525MB/s ± 0% 572MB/s ± 0% +8.89% (p=0.000 n=9+10)
Hash256-32 646MB/s ± 0% 710MB/s ± 1% +9.90% (p=0.000 n=10+10)
Hash512-32 732MB/s ± 1% 808MB/s ± 1% +10.41% (p=0.000 n=10+10)
Hash1K-32 786MB/s ± 0% 866MB/s ± 1% +10.30% (p=0.000 n=9+10)
Hash8K-32 839MB/s ± 0% 930MB/s ± 0% +10.79% (p=0.000 n=10+10)
Hash1M-32 849MB/s ± 1% 938MB/s ± 1% +10.54% (p=0.000 n=10+10)
Hash8M-32 841MB/s ± 1% 935MB/s ± 1% +11.16% (p=0.000 n=10+10)
Hash8BytesUnaligned-32 77.1MB/s ± 0% 83.4MB/s ± 0% +8.12% (p=0.000 n=10+10)
Hash1KUnaligned-32 778MB/s ± 1% 869MB/s ± 1% +11.64% (p=0.000 n=10+10)
Hash8KUnaligned-32 836MB/s ± 0% 932MB/s ± 1% +11.47% (p=0.000 n=10+10)
Change-Id: I02b31229b857e9257dc9d36538883eb3af4ad993
This PR will be imported into Gerrit with the title and first
comment (this text) used to generate the subject and body of
the Gerrit change.
Change-Id: I02b31229b857e9257dc9d36538883eb3af4ad993
GitHub-Last-Rev:
ec8b15d789181d0dac57bf0ba5041ee7aeb305c9
GitHub-Pull-Request: golang/go#43690
Reviewed-on: https://go-review.googlesource.com/c/go/+/283538
Run-TryBot: Joel Sing <joel@sing.id.au>
Reviewed-by: Matthew Dempsky <mdempsky@google.com>
Reviewed-by: David Chase <drchase@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Joel Sing <joel@sing.id.au>