Follow @Openwall on Twitter for new release announcements and other news
[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Date: Tue, 15 Sep 2015 04:03:11 +0300
From: Solar Designer <>
Subject: Re: Judy array

Fred, magnum -

On Sun, Sep 13, 2015 at 08:28:08PM -0700, Fred Wang wrote:
> I use a 10-year-old Dell 2950 as my test environment, precisely because it uses slower memory, and more easily shows improvements.  For my "standard" test case (MD5, 29 million hashes, a ~13 million entry dictionary, and best64 rules, yielding about 1 billion hash attempts to find about 1.7 million solutions)
> hashcat	3 minute 54 seconds
> mdxfind	1 minute 15 seconds  (Judy only)
> mdxfind	47 seconds  (Current code, Bloom filter + Judy)

Fred - is this with the original best64 rules you e-mailed me, or with
the cut-down version with only JtR-compatible rules left in it (14 rules
removed)?  (BTW, this best64 is currently misnamed since it contains
more than 64 rules either way.)

magnum - testing this stuff, I see that pot sync is a major bottleneck.
Since this is your feature, you might want to benchmark and optimize it
some more, or/and maybe we just want it disabled by default when cracking
saltless hashes.  As it is, it's just not able to handle Fred's "about
1.7 million solutions" in a minute - it takes several minutes to process
those guesses, so it becomes the primary bottleneck for the whole run.

> Its important to note that this includes the time to read 29M hashes, and store them - this takes about 22 seconds on the test box.  The box uses dual E5410  @ 2.33GHz, so 8 cores.

I am now running john with your test case on 2x E5420, which is similar
(2.5 GHz, though).  I'll post more on this soon.


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.