Date: Thu, 09 Feb 2012 10:20:31 +0100 From: Simon Marechal <simon@...quise.net> To: john-dev@...ts.openwall.com Subject: Re: cryptmd5 optimizations On 09/02/2012 07:54, Lukas Odzioba wrote: > Thank you Simon I digged though code one made changes but > unfortunatelly 8*64 bytes for each thread is still to much memory (42 > from MD5_std.c was an overkill). Actually you should be able to just keep the first hash and the current hash. If will require some creative specialized unrolling though. Indeed, for the same salt you could share those buffers for all threads. Not sure if this is the way to go.
Powered by blists - more mailing lists
Powered by Openwall GNU/*/Linux - Powered by OpenVZ