Follow @Openwall on Twitter for new release announcements and other news
[<prev] [next>] [<thread-prev] [day] [month] [year] [list]
Date: Sun, 12 Jul 2015 17:34:47 +0300
From: Solar Designer <solar@...nwall.com>
To: john-dev@...ts.openwall.com
Subject: Re: JtR on Power

On Fri, Jul 10, 2015 at 05:34:34PM +0800, Lei Zhang wrote:
> Speedup of each format:
> 
> -------------------
> md5crypt	6.4
> phpass		7.2
> sha256crypt	2.1
> sha512crypt	3.4
> pbkdf2-md5	9.6
> pbkdf2-sha256	4.8
> pbkdf2-sha512	2.6
> -------------------
> 
> Well, some of the speedups seem too high. I once doubted if I got other load when benchmarking those non-AltiVec figures, so I retried the 'before' part. But the figures didn't get higher. Perhaps something other than vectorization is playing a role here, but I don't know.

Yes, right, something else must be playing a role.

For example, it is expected that the speedup for sha256crypt would be
higher than for sha512crypt (since it's 4x vs. 2x increase in number of
concurrent instances), but you have it the other way around.  Perhaps
there was something crippling the performance of non-SIMD code way
beyond of what's essential to that type of code.  You'd need to review
the generated assembly code to find out.

Alexander

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.