Follow @Openwall on Twitter for new release announcements and other news
[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Date: Sat, 2 Feb 2013 00:12:04 +0100
From: magnum <>
Subject: Re: Password Cracking a DMG...

On 29 Jan, 2013, at 22:10 , magnum <> wrote:
> On 29 Jan, 2013, at 21:01 , Jeremiah Grossman <> wrote:
>> On Jan 29, 2013, at 11:18 AM, magnum <> wrote:
>>> On 29 Jan, 2013, at 19:09 , Jeremiah Grossman <> wrote:
>>>> from run/
>>>>> $ dmg2john aes_256.dmg 
>>>> Segmentation fault: 11
>>>> Same issue as yesterday. "aes_256.dmg" is a newly created 15GB DMG encrypted with AES-256 (OS X 10.8.2). No data contained within. 
>>> That should be "./dmg2john". Maybe you just didn't copy it verbatim? Otherwise, maybe you actually did not run the newly built ./dmg2john but an old bad one from somewhere in your path.
>>> magnum
>> Positive I got it right. I was just snipping the command line for brevity sake. The dmg2john I ran was in the run/ directory automatically built during compile of JtR.
> I figured so, just checking. I will try to reproduce the problem and debug it.

For people not subsribed to john-dev: This is resolved in latest git. The dmg2john bug was fixed, and then we realized both dmg2john and the format blatantly ignored the iterations count - which is bumped a lot in later OSX versions (it was hard-coded to 1000 while newer Macs produce files with over 200,000 iterations, and seemingly depending on available CPU power at creation time). Finally, some known-plain stuff was tweaked.

So the good news is everything hopefully works now if you check out a Git Jumbo. The bad news is with this high iteration count, you get about 5-10 c/s per core on CPU. Using OpenCL and GPU we can get a little more but this is the toughest format I know of right now.


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.