Linux brutessl client
Hi. I've been able to compile a working brutessl 1.02 cllient under linux and gcc 2.7.0, but I get abysmal search speeds. My system is a pentium 60, which, according to the docs, ought to have a speed of something over 14100 kps (that's what a 486dx2/66 with no rotate left macro gets). brutessl -t reports that my system can do 11200 keys per second. Is there any obvious reason this number is so much lower than expected? I'm reasonably certain gcc implemets a rotl macro, but I don't know how to make use of it -- has anyone out there gotten assembly.c to compile under gcc? Finally, are there any additional optimizations for pentium machines available? thanks, -sq
-----BEGIN PGP SIGNED MESSAGE----- On Wed, 23 Aug 1995, Sam Quigley wrote:
Date: Wed, 23 Aug 1995 21:48:33 -0700 From: Sam Quigley <poodge@econ.Berkeley.EDU> To: cypherpunks@toad.com Subject: Linux brutessl client
Hi.
I've been able to compile a working brutessl 1.02 cllient under linux and gcc 2.7.0, but I get abysmal search speeds.
My system is a pentium 60, which, according to the docs, ought to have a speed of something over 14100 kps (that's what a 486dx2/66 with no rotate left macro gets).
brutessl -t reports that my system can do 11200 keys per second.
i'm running linux 1.2.11 on my DX4-100 (might be faster than your P60) and i get these results.. zifi:~/crypto/ssl> brutessl -t BruteSSL 1.0 Self test... OK 17 minutes and 8 seconds per segment, 16300 keys per second. zifi:~/crypto/ssl> file brutessl brutessl: ELF 32-bit LSB executable i386 (386 and up) Version 1 the perl code doesn't seem to work quite right for me though, giving errors when i try to use more than one option at a time, ie. Identifier "main::dkltrc4" used only once: possible typo. anyway, i'm not sure i'll be participating this time as i won't be around too much. ciao, - -pjf patrick finerty = zinc@zifi.genetics.utah.edu = pfinerty@nyx.cs.du.edu U of Utah biochem grad student in the Bass lab - zinc fingers + dsRNA! ** FINGER zinc-pgp@zifi.genetics.utah.edu for pgp public key - CRYPTO! zifi runs LINUX 1.2.11 -=-=-=WEB=-=-=-> http://zifi.genetics.utah.edu -----BEGIN PGP SIGNATURE----- Version: 2.6.2 iQCVAwUBMDwSA03Qo/lG0AH5AQG1jQP/ec+MO0BTFGF53eWohIlCBxpdaXnCnFJe 9TvhbRPX7kjuVyVKcuT6KdHrZ/+qrsLAXCrlml5UuAemUmyAE4NeR3NrbJ7rgXWg g+UfzIC46799/MNu0XQeCZdip2aq+3VojakoPcd/VThW7BUKxdJeNyP/C8XhjxMg +tLRnuiVrjA= =xGxM -----END PGP SIGNATURE-----
the perl code doesn't seem to work quite right for me though, giving errors when i try to use more than one option at a time, ie. Identifier "main::dkltrc4" used only once: possible typo.
Sorry about that .... I discovered that there was a problem with "require" under perl 5.001 (if one does a require on getopts.pl, the perl CRASHES with some flag combinations) so I inlined getopts.pl, and changed the comments, but left the first part of the sentance -- the "if you have no getopts" was omitted. SO: in brief do *NOT* add the "s" flag to the first line of brclient ! There are new versions of brloop and brclient which might cope better with problems with the transatlantic link -- you should use brc0.09 and brl0.03
anyway, i'm not sure i'll be participating this time as i won't be around too much.
All the better -- your machine will have more cycles for CRACKing ! The purpose of SKSP is to make it work when you are not in ... There is a "how to" page referenced from http://www.brute.cl.cam.ac.uk/brute/ which includes some "fault finding" tips. If you have problems, look there first. If that doesn't help, email me, and I'll add any "common" problems to the "how to".
On Thu, 24 Aug 1995, Piete Brooks wrote:
Date: Thu, 24 Aug 1995 07:50:32 +0100 From: Piete Brooks <Piete.Brooks@cl.cam.ac.uk> To: zinc <zinc@zifi.genetics.utah.edu> Cc: Sam Quigley <poodge@econ.Berkeley.EDU>, cypherpunks@toad.com Subject: Re: Linux brutessl client
I discovered that there was a problem with "require" under perl 5.001 (if one does a require on getopts.pl, the perl CRASHES with some flag combinations) so I inlined getopts.pl, and changed the comments, but left the first part of the sentance -- the "if you have no getopts" was omitted.
SO: in brief do *NOT* add the "s" flag to the first line of brclient !
i just want to mention that i obtained the newest code from the web site, compiled and everything is working fine. i started brloop as a background process, hope that's ok... anyway, seems to be doing ok. i guess i'll know tomorrow when the cracking starts. i think i'll try to get things working on an SGI i have access to as well. -pjf patrick finerty = zinc@zifi.genetics.utah.edu = pfinerty@nyx.cs.du.edu U of Utah biochem grad student in the Bass lab - zinc fingers + dsRNA! ** FINGER zinc-pgp@zifi.genetics.utah.edu for pgp public key - CRYPTO! zifi runs LINUX 1.2.11 -=-=-=WEB=-=-=-> http://zifi.genetics.utah.edu
participants (4)
-
cg@bofh.lake.de -
Piete Brooks -
Sam Quigley -
zinc