[golang-dev] Re: code review 6820096: crypto/sha1: Make sha-1 do block mixup in place (issue 6820096)

[golang-dev] Re: code review 6820096: crypto/sha1: Make sha-1 do block mixup in place (issue 6820096)

	Notcarl
	Nov 6, 2012 at 7:39 pm

Hello dave@cheney.net, rsc@golang.org (cc: golang-dev@googlegroups.com),

Please take another look.

http://codereview.appspot.com/6820096/

Search Discussions

Notcarl: I adding this change, and got a small increase in speed. at The fear is that unrolling the loop means potentially more i-cache misses, so I agree it should be in a future CL. Hoisting i out of the loop in the meantime showed a little extra gain. https://codereview.appspot.com/6820096/

Follow ups

Notcarl: Thanks for the quick review! https://codereview.appspot.com/6820096/diff/6001/src/pkg/crypto/sha1/sha1block.go File src/pkg/crypto/sha1/sha1block.go (right): https://codereview.appspot.com/6820096/diff/6001/src/pkg/crypto/sha1/sha1block.go#newcode39 src/pkg/crypto/sha1/sha1block.go:39: t := a5 + f + e + w[i] + _K0 I am not sure I understand. Since the loop only goes from [0, 15], anding with 0xf should be a no-op
Notcarl: Done. https://codereview.appspot.com/6820096/
Dave: hmmm, i can't replicate your benchmark results. linux/amd64: Intel(R) Core(TM) i5-2520M CPU @ 2.50GHz benchmark old ns/op new ns/op delta BenchmarkHash8Bytes 866 824 -4.85% BenchmarkHash1K 10058 9520 -5.35% BenchmarkHash8K 74567 70953 -4.85% benchmark old MB/s new MB/s speedup BenchmarkHash8Bytes 9.24 9.71 1.05x BenchmarkHash1K 101.81 107.56 1.06x BenchmarkHash8K 109.86 115.46 1.05x https://codereview.appspot.com/6820096/

Discussion Navigation

view	thread \| post
posts	‹ prev \| 9 of 20 \| next ›

Discussion Overview

font	variable	Hotkey: f
user style	avatars	Hotkey: a