john-users - Re: bitslice DES parallelization with OpenMP

Openwall wordlists collection for password cracking (20+ languages)
[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Date: Sat, 3 Jul 2010 23:19:32 -0400
From: Erik Winkler <ewinkler@...ls.com>
To: john-users@...ts.openwall.com
Subject: Re: bitslice DES parallelization with OpenMP

Alexander,

Both patches compile under MacOS 10.6 using gcc-4.5 I downloaded from this site (http://hpc.sourceforge.net/).  Here are the results on a 2.4 Ghz Core2:

Without OpenMP:
./john -test
Benchmarking: Traditional DES [128/128 BS SSE2-16]... DONE
Many salts:	2031K c/s real, 2035K c/s virtual
Only one salt:	1972K c/s real, 1976K c/s virtual

Benchmarking: BSDI DES (x725) [128/128 BS SSE2-16]... DONE
Many salts:	67763 c/s real, 67763 c/s virtual
Only one salt:	66457 c/s real, 66457 c/s virtual

Benchmarking: FreeBSD MD5 [32/64 X2]... DONE
Raw:	10267 c/s real, 10267 c/s virtual

Benchmarking: OpenBSD Blowfish (x32) [32/64 X2]... DONE
Raw:	622 c/s real, 620 c/s virtual

Benchmarking: Kerberos AFS DES [48/64 4K]... DONE
Short:	328499 c/s real, 328499 c/s virtual
Long:	1047K c/s real, 1047K c/s virtual

Benchmarking: LM DES [128/128 BS SSE2-16]... DONE
Raw:	29621K c/s real, 29680K c/s virtual


With OpenMP:
./john -test
Benchmarking: Traditional DES [128/128 BS SSE2-16]... DONE
Many salts:	3639K c/s real, 1866K c/s virtual
Only one salt:	3188K c/s real, 1677K c/s virtual

Benchmarking: BSDI DES (x725) [128/128 BS SSE2-16]... DONE
Many salts:	124314 c/s real, 63453 c/s virtual
Only one salt:	120334 c/s real, 62278 c/s virtual

Benchmarking: FreeBSD MD5 [32/64 X2]... DONE
Raw:	10266 c/s real, 10266 c/s virtual

Benchmarking: OpenBSD Blowfish (x32) [32/64 X2]... DONE
Raw:	1219 c/s real, 620 c/s virtual

Benchmarking: Kerberos AFS DES [48/64 4K]... DONE
Short:	327424 c/s real, 326770 c/s virtual
Long:	1071K c/s real, 1071K c/s virtual

Benchmarking: LM DES [128/128 BS SSE2-16]... DONE
Raw:	29795K c/s real, 19398K c/s virtual

Without patching, the relevant results for DES and LM are shown below:

Benchmarking: Traditional DES [128/128 BS SSE2-16]... DONE
Many salts:	2279K c/s real, 2284K c/s virtual
Only one salt:	1971K c/s real, 1975K c/s virtual

Benchmarking: LM DES [128/128 BS SSE2-16]... DONE
Raw:	13027K c/s real, 13053K c/s virtual

Erik

On Jul 3, 2010, at 4:47 PM, Solar Designer wrote:

> 
> As usual, feedback is welcome.  I'd like this stuff to become a bit more
> mature before I approach integrating it.
> 
> Alexander