Openwall GNU/*/Linux - a small security-enhanced Linux distro for servers
[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Date: Fri, 15 May 2015 21:47:05 +0800
From: Lei Zhang <zhanglei.april@...il.com>
To: john-dev@...ts.openwall.com
Subject: Re: lotus85

Sorry it's a bit late when I read through this message.  I guess there isn't enough time to do it...


Lei

> On May 15, 2015, at 8:58 AM, Solar Designer <solar@...nwall.com> wrote:
> 
> Agnieszka, Lei -
> 
> Want to practice optimizing a format where interleaving is actually
> expected to be beneficial?  lotus85_fmt_plug.c is in need of
> optimizations very similar to these:
> 
> git show 57a0e272bbba7187b3240c9c605281100b97515f lotus5_fmt_plug.c
> git show f64b42fee9e368cd85cf546f08b694510824fea2 lotus5_fmt_plug.c
> 
> In short, we need proper get_hash*() and binary_hash*() functions, and
> 3x interleaving.  The former should provide huge speedup when cracking a
> large number of hashes per salt (unusual?), and the latter a further
> 2x-ish speedup.  Oh, and OMP_SCALE needs to be tuned, and
> BENCHMARK_LENGTH should be changed from -1 to 0 because this is a salted
> format (unlike lotus5).  See also: DOMINOSEC_fmt_plug.c (another related
> format, for reference).
> 
> These hashes are expected to be seen in the Hash Runner contest starting
> later today, so ideally you'd implement the optimizations today (or
> perhaps someone else will, and you'll miss the opportunity to practice).
> 
> Thanks,
> 
> Alexander

Powered by blists - more mailing lists

Your e-mail address:

Powered by Openwall GNU/*/Linux - Powered by OpenVZ