[<prev] [next>] [<thread-prev] [day] [month] [year] [list]
Message-ID: <e74ae2a4383a1c1117f6c3eb35f9f693@smtp.hushmail.com>
Date: Sat, 31 May 2014 23:44:09 +0200
From: magnum <john.magnum@...hmail.com>
To: john-dev@...ts.openwall.com
Subject: Re: Different c/s speed after compiling with new autoconf
on 32-bit system (64bit system not affected)
On 2014-05-31 22:09, [Jano] wrote:
> I have also a 32-bit machine and today when pull the new
> bleeding-jumbo I see this difference:
> Will run 2 OpenMP threads
> Benchmarking: wpapsk, WPA/WPA2 PSK [PBKDF2-SHA1 128/128 SSE2 4x]...
> (2xOMP) DONE
> Raw: 376 c/s real, 188 c/s virtual
> jano:~$ john --test --format=wpapsk
> Will run 2 OpenMP threads
> Benchmarking: wpapsk, WPA/WPA2 PSK [PBKDF2-SHA1 128/128 SSE2 8x]...
> (2xOMP) DONE
> Raw: 345 c/s real, 176 c/s virtual
This is supposedly not caused by autoconf - the "4x" or "8x" is chosen
from your version of gcc, regardless of building w/ autoconf or not. But
it's not the sharpest of swords.
Try "./configure --with-icc-asm && make -s clean default" and see if it
ends up faster.
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.
