Follow @Openwall on Twitter for new release announcements and other news
[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Date: Fri, 29 Jun 2012 01:10:51 +0200
From: Frank Dittrich <frank_dittrich@...mail.com>
To: john-dev@...ts.openwall.com
Subject: Re: Shouldn't the TS work with a pw.dic in reversed order
 as well?

On 06/29/2012 12:59 AM, magnum wrote:
> I do see problems on 64-bit after reversing *.dic files. It's so many, I
> suspect it's a red herring and something else is amiss:
> 
> $ ./jtrts.pl -q full
> -------------------------------------------------------------------------------
> - JtR-TestSuite (jtrts). Version 1.12.4, June 22, 2012.  By, Jim
> Fougeron & others
> - Testing:  John the Ripper password cracker, ver: 1.7.9-jumbo-6-RC0
> [linux-x86-64i]
> --------------------------------------------------------------------------------
> form=dynamic_17                   guesses:    0 time: 0:00:00:14 :
> Expected count(s) (1500)  [!!!FAILED!!!]
> form=dynamic_27                   guesses:    0 time: 0:00:00:12 :
> Expected count(s) (1500)  [!!!FAILED!!!]
> form=dynamic_28                   guesses:    0 time: 0:00:00:12 :
> Expected count(s) (1500)  [!!!FAILED!!!]
> form=md5                          guesses:    0 time: 0:00:00:10 :
> Expected count(s) (1500)  [!!!FAILED!!!]
> form=md5_apr1                     guesses:    0 time: 0:00:00:10 :
> Expected count(s) (1500)  [!!!FAILED!!!]
> form=BSDI                         guesses:    0 time: 0:00:00:04 :
> Expected count(s) (1500)  [!!!FAILED!!!]
> form=BF                           guesses:    0 time: 0:00:01:25 :

If even the core formats fail completely, are you sure you didn't empty
o otherwise destroy your pw.dic?

> form=crc32_dups2                  guesses: 1498 time: 0:00:00:00 :
> Expected count(s) (1500)  [!!!FAILED!!!]
> form=crc32_dups2_read_file        guesses: 1498 time: 0:00:00:00 :

OK, these use another dictionary, and I don't know what the duplicates
suppression does here.
But even so, usually, the sequence of candidates should only matter for
collisions and FMT_NOT_EXACT not set. But CRC32 has this flag set.

> Expected count(s) (1500)  [!!!FAILED!!!]
> form=mscash-utf8                  guesses: 1502 time: 0:00:00:00 :

More guesses than expected?

Frank

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.