Openwall GNU/*/Linux - a small security-enhanced Linux distro for servers
[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Date: Thu, 17 Sep 2015 13:38:00 +0200
From: magnum <john.magnum@...hmail.com>
To: john-dev@...ts.openwall.com
Subject: Re: prince mode and rules

On 2015-09-17 01:39, magnum wrote:
> On 2015-09-17 01:34, Frank Dittrich wrote:
>> On 09/17/2015 01:17 AM, Frank Dittrich wrote:
>>> Yes, you do call rpp_init() and rules_init() twice in do_prince_crack,
>>> apparently once in a loop.
>>> I am sure that has to be avoided.

I agree, but I'm not sure how to "reset" the ctx for each word, I see no 
mechanism for that. I now tried copying it prior to the candidate 
generation and reset to the copy each time and it's 2x faster (which is 
still slow) but I seem to get different results (will investigate more).

>> You also test whether a rule should be applied or rejected once per
>> candidate:
>>
>> $ grep -c "Rule #1:" testprince.log
>> 3084
>>
>> With all that logging, it is no wonder it takes ages.

The logging is bad and should be dropped, but it doesn't affect speed much.

>> You should process rules_reject() just once per rule, outside of the
>> loop processing the candidates.

I see no mechanism for doing this (and ending up with a pruned list of 
rules to use). If there is one we should use it in wordlist mode too.

magnum

Powered by blists - more mailing lists

Your e-mail address:

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