Follow @Openwall on Twitter for new release announcements and other news
[<prev] [next>] [thread-next>] [day] [month] [year] [list]
Date: Mon, 23 May 2011 14:37:05 +0200
From: magnum <rawsmooth@...dband.net>
To: john-dev@...ts.openwall.com
Subject: BSDI poor in OMP4/OMP7 when doing real work

This is not Jumbo specific but I found it during regression testing of 
jumbo-5-rc6:

Using the OMP4 *or* OMP7 patch on 1.7.7, the BSDI format performs very 
poorly in real cracking (test suite) even though --test show a good 
speedup. Actually the reported figure from real cracking is also high 
but the real time is awful. It does find all 1320 passwords.

Plain 1.7.7:
--test     77004 c/s
testsuite  2 seconds, reported c/s 78586

1.7.7 + OMP4
--test     128098 c/s
testsuite  1 minute, 11 seconds, reported c/s 123860

(btw the testsuite can't be used as-is for plain jumbo as it lacks some 
command line options used by the test suite)

Notes:
1. This happens with OMP_NUM_THREADS=1 too.
2. It even happens if OMP is disabled in Makefile (but patch applied).
3. Compiling with icc make no difference, problem remains the same.
4. DES and LM does not seem to have any problem.
5. If you find a fix for this, please don't forget the FMT_OMP flags :)

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.