Date: Sat, 4 Aug 2012 17:32:13 -0500 From: "jfoug" <jfoug@....net> To: <john-dev@...ts.openwall.com> Subject: RE: sunmd5 >From: Solar Designer [mailto:solar@...nwall.com] > > Here's an obvious optimization, removing the modulo division. There are a ton of tweaks to be made in the coin flip. I have not touched that code at all, it is 100% original to what Bartavelle did (which was original Moffet code, I think). The gains on this one, likely will be clawing and grabbing a few extra cracks per sec, unless we find better optimizations in either crypt reloading, or better ways to optimize, 'knowing' that we have a large array of SSE_PARA buffers, where the code as it is, may only be optimal for a single buffer.
Powered by blists - more mailing lists
Confused about mailing lists and their use? Read about mailing lists on Wikipedia and check out these guidelines on proper formatting of your messages.