[<prev] [next>] [thread-next>] [day] [month] [year] [list]
Message-ID: <loom.20081007T003643-40@post.gmane.org>
Date: Tue, 7 Oct 2008 00:45:26 +0000 (UTC)
From: Marc Bevand <m.bevand@...il.com>
To: john-users@...ts.openwall.com
Subject: Breaking UNIX crypt() on the PlayStation 3
I have seen people in the past, on this mailing list, inquiring about
a bitslice DES implementation optimized for the Cell B.E. processor.
I have created one, as well as a UNIX crypt() bruteforcer based on it,
and it is indeed pretty fast. The 128 128-bit registers per SPU and
the logical instructions available on this architecture (selb, eqv,
andc, orc, nand, nor, etc) are just perfect for bitslicing. The
implementation I came up with averages 45.5 gates per S-box and is
capable of testing 11.5 million password/sec on a PS3 (1.9 Mpw/sec
per SPU core).
Slides and source code:
http://www.epitech.eu/v4/perso/~bevand_m/talks/breaking-unix-crypt.pdf
http://www.epitech.eu/v4/perso/~bevand_m/cell-bf/cell-bf-1.tar.gz
The source code also includes an "S-box circuit bruteforcer" I wrote
based on Matthew Kwan's ideas (he made the S-boxes used in John the
Ripper itself). I ran it for 2-3 hours to find good S-box circuits
leveraging all the logical instructions available to the SPU.
-marc
--
To unsubscribe, e-mail john-users-unsubscribe@...ts.openwall.com and reply
to the automated confirmation request that will be sent to you.
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.
