Openwall GNU/*/Linux - a small security-enhanced Linux distro for servers
[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Date: Fri, 22 Jun 2012 02:38:20 +0200
From: Frank Dittrich <frank_dittrich@...mail.com>
To: john-dev@...ts.openwall.com
Subject: Re: relbench and changed format names

On 06/22/2012 12:08 AM, magnum wrote:
> On 2012-06-22 00:04, Frank Dittrich wrote:
>> For now, I just changed relbench in a way that the jumbo-5 dynamic
>> formats are recognized as the same as those in current git.
>> This alone increased the number of matching formats by 36.
>> See attached patch.
> 
> Cool. Should I commit this?

Unless Alexander objects, I think you should.

Replacing the square brackets in format names affects these jumbo-5 formats:

$ fgrep '] [' j5-sse2i-omp.txt
Benchmarking: dynamic_1008 md5($p.$s) [RADIUS User-Password] [SSE2i
10x4x3]... DONE
Benchmarking: dynamic_1009 md5($s.$p) [RADIUS Responses] [SSE2i
10x4x3]... DONE
Benchmarking: NTLMv1 C/R MD4 DES [ESS MD5] [netntlm]... (2xOMP) DONE


And these are the corresponding jumbo-6 lines:

$ grep -E '(dynamic_1008|dynamic_1009|NTLMv1)' j6-sse2i-omp.txt
Benchmarking: dynamic_1008 md5($p.$s) (RADIUS User-Password) [128/128
SSE2 intrinsics 10x4x3]... DONE
Benchmarking: dynamic_1009 md5($s.$p) (RADIUS Responses) [128/128 SSE2
intrinsics 10x4x3]... DONE
Benchmarking: NTLMv1 C/R MD4 DES [ESS MD5] [32/32]... (2xOMP) DONE

The NTLMv1 line is the only one in the upcoming jumbo-6 which uses [...]
in the format name, so no harm done, since it will be treated similar to
the jumbo-5 output.

These are the jumbo-5 format names that use additional spaces at the end
of the name or in the middle of the name:

$ grep "Benchmarking: .*  " j5-sse2i-omp.txt
Benchmarking: dynamic_0: md5($p)  (raw-md5)  [SSE2i 10x4x3]... DONE
Benchmarking: dynamic_1: md5($p.$s)  (joomla)  [SSE2i 10x4x3]... DONE
Benchmarking: dynamic_2: md5(md5($p))  (e107)  [SSE2i 10x4x3]... DONE
Benchmarking: dynamic_3: md5(md5(md5($p)))  [SSE2i 10x4x3]... DONE
Benchmarking: dynamic_4: md5($s.$p)  (OSC)  [SSE2i 10x4x3]... DONE
Benchmarking: dynamic_5: md5($s.$p.$s)  [SSE2i 10x4x3]... DONE
Benchmarking: dynamic_6: md5(md5($p).$s)  [SSE2i 10x4x3]... DONE
Benchmarking: dynamic_8: md5(md5($s).$p)  [SSE2i 10x4x3]... DONE
Benchmarking: dynamic_9: md5($s.md5($p))  [SSE2i 10x4x3]... DONE
Benchmarking: dynamic_10: md5($s.md5($s.$p))  [SSE2i 10x4x3]... DONE
Benchmarking: dynamic_11: md5($s.md5($p.$s))  [SSE2i 10x4x3]... DONE
Benchmarking: dynamic_12: md5(md5($s).md5($p))  (IPB)  [SSE2i 10x4x3]...
DONE
Benchmarking: dynamic_13: md5(md5($p).md5($s))  [SSE2i 10x4x3]... DONE
Benchmarking: dynamic_14: md5($s.md5($p).$s)  [SSE2i 10x4x3]... DONE
Benchmarking: dynamic_15: md5($u.md5($p).$s)  [128x1 (MD5_Body)]... DONE
Benchmarking: dynamic_16: md5(md5(md5($p).$s).$s2)  [128x1
(MD5_Body)]... DONE
Benchmarking: dynamic_17: phpass ($P$ or $H$)  [SSE2i 2x4x3]... DONE
Benchmarking: dynamic_18: md5($s.Y.$p.0xF7.$s) (Post.Office MD5)  [128x1
(MD5_Body)]... DONE

For the upcoming jumbo-6, no such format names exist.

These are the new format names that have to be matched with the jumbo-5
format names listed above:

$ grep -E "Benchmarking: dynamic_([0-689]|1[0-8]):" j6-sse2i-omp.txt
Benchmarking: dynamic_0: md5($p) (raw-md5) [128/128 SSE2 intrinsics
10x4x3]... DONE
Benchmarking: dynamic_1: md5($p.$s) (joomla) [128/128 SSE2 intrinsics
10x4x3]... DONE
Benchmarking: dynamic_2: md5(md5($p)) (e107) [128/128 SSE2 intrinsics
10x4x3]... DONE
Benchmarking: dynamic_3: md5(md5(md5($p))) [128/128 SSE2 intrinsics
10x4x3]... DONE
Benchmarking: dynamic_4: md5($s.$p) (OSC) [128/128 SSE2 intrinsics
10x4x3]... DONE
Benchmarking: dynamic_5: md5($s.$p.$s) [128/128 SSE2 intrinsics
10x4x3]... DONE
Benchmarking: dynamic_6: md5(md5($p).$s) [128/128 SSE2 intrinsics
10x4x3]... DONE
Benchmarking: dynamic_8: md5(md5($s).$p) [128/128 SSE2 intrinsics
10x4x3]... DONE
Benchmarking: dynamic_9: md5($s.md5($p)) [128/128 SSE2 intrinsics
10x4x3]... DONE
Benchmarking: dynamic_10: md5($s.md5($s.$p)) [128/128 SSE2 intrinsics
10x4x3]... DONE
Benchmarking: dynamic_11: md5($s.md5($p.$s)) [128/128 SSE2 intrinsics
10x4x3]... DONE
Benchmarking: dynamic_12: md5(md5($s).md5($p)) (IPB) [128/128 SSE2
intrinsics 10x4x3]... DONE
Benchmarking: dynamic_13: md5(md5($p).md5($s)) [128/128 SSE2 intrinsics
10x4x3]... DONE
Benchmarking: dynamic_14: md5($s.md5($p).$s) [128/128 SSE2 intrinsics
10x4x3]... DONE
Benchmarking: dynamic_15: md5($u.md5($p).$s) [32/32 128x1 (MD5_Body)]...
DONE
Benchmarking: dynamic_16: md5(md5(md5($p).$s).$s2) [32/32 128x1
(MD5_Body)]... DONE
Benchmarking: dynamic_17: phpass ($P$ or $H$) [SSE2 intrinsics 2x4x3]...
DONE
Benchmarking: dynamic_18: md5($s.Y.$p.0xF7.$s) (Post.Office MD5) [32/32
128x1 (MD5_Body)]... DONE


So I think the patch can be applied even if I (or someone else) will
come up with some way to map old and new format names, because the list
of mappings required will be shorter.

Frank

Powered by blists - more mailing lists

Your e-mail address:

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