Follow @Openwall on Twitter for new release announcements and other news
[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Date: Mon, 17 Dec 2012 00:09:13 +0100
From: john@...jenski.de
To: john-users@...ts.openwall.com
Subject: Re: gpg2john -> false positive -> how to exclude?

Hi magnum,

thanks for your fast and detailed answer, that helps in understanding 
how it works.
Supplying the key would be no real problem since - as you guessed 
correctly - this key is in no
productive use anymore.
Here some facts about my setup: 'make linux-x86-64-native', enabled the 
two "gcc with openMP" lines in the Makefile.

the converted gpg2john hash is:
secring.gpg:$gpg$*17*24*1024*2fd8c6834db06ddfe073fd944b6bd8dbd268163e6374ef6f*3*255*2*3*8*bf07a2f4faafa916*65536*6c7784ea65895667

the one false positive i got is in clear-text: bortaloo
(which is not my phrase | and thus does not work for unlocking)

output of gpg --list-secret-keys is:
---------------------------
sec   1024D/615B53E0 2001-01-29
uid                  Donny
ssb   1024g/8646A815 2001-01-29

john-version used: unstable-jumbo (from .git as zip-download via http)

If you need the real secring.pgp and a ciphertext file, private email 
would be great :)

Greets,
Seb

On 16.12.2012 23:15, magnum wrote:
> On 16 Dec, 2012, at 21:04 , john@...jenski.de wrote:
>> is it possible to let john go on with brute-forcing in incremental mode, after
>> an obvious false positive is found?
> Some formats are expected to have collisions, like CRC32 for instance. You can make any format behave like such by adding FMT_NOT_EXACT to the format flags in the end (normally) of the source file. In this case, this line:
>
> 	FMT_CASE | FMT_8_BIT | FMT_OMP,
>
> of gpg_fmt_plug.c (or opencl_gpg_fmt.c) would be
>
> 	FMT_CASE | FMT_8_BIT | FMT_OMP | FMT_NOT_EXACT,
>
> However, see below.
>
>> I recently did 'gpg2john' and finally ended up with incremental mode and got an false positive.
>> I forgot my gpg-passphrase (from>10 years ago), and I know it's not the found word, besides:
>> it simply does not work :)
> As far as I understand from the source, it should really not emit false positives. Maybe there is a bug in the format. At worst that means once Dhiru fixes it you'd have to re-start from scratch.
>
> We might need more information. Were you running the CPU format or the OpenCL one? And are you using Jumbo-7 or the unstable-jumbo from git? Would you by any chance reveal the gpg key or input file (perhaps privately, to Dhiru) so we can reproduce the bug? I take it you do not use that passphrase anymore anyway :)  Failing that, I guess we'd be helped by knowing what exact time key it is.
>
> magnum

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.