[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Date: Fri, 05 Mar 2010 20:11:43 +0100
From: "Magnum, P.I." <rawsmooth@...dband.net>
To: john-users@...ts.openwall.com
Subject: Re: Feedback on the generic crypt(3) patch
> http://www.openwall.com/lists/john-users/2009/11/05/1
Some more thoughts regarding the generic-crypt(3) patch. I always
wondered what was actually measured when benchmarking this "format".
Just DES or a mixture of formats? The code revealed it was DES only.
In case anyone share my curiousity, I enclose a little patch making it
easy to choose which format is tested/benchmarked, as a compile time
option. It's supposed to be applied after the crypt-1 patch.
Here's some example figures, including figures from the optimized
versions of DES and MD5 for comparison:
---8<---------8<---------8<---------8<------
Benchmarking: Traditional DES [128/128 BS SSE2-16]... DONE
Many salts: 4504K c/s real, 4563K c/s virtual
Only one salt: 3884K c/s real, 3884K c/s virtual
Benchmarking: generic crypt(3) OS call (using DES) [?/64]... DONE
Many salts: 421162 c/s real, 425416 c/s virtual
Only one salt: 417396 c/s real, 417396 c/s virtual
Benchmarking: FreeBSD MD5 [32/64 X2]... DONE
Raw: 20248 c/s real, 20640 c/s virtual
Benchmarking: generic crypt(3) OS call (using MD5) [?/64]... DONE
Many salts: 8061 c/s real, 8118 c/s virtual
Only one salt: 8087 c/s real, 8087 c/s virtual
Benchmarking: generic crypt(3) OS call (using SHA-256) [?/64]... DONE
Many salts: 385 c/s real, 387 c/s virtual
Only one salt: 463 c/s real, 463 c/s virtual
Benchmarking: generic crypt(3) OS call (using SHA-512) [?/64]... DONE
Many salts: 383 c/s real, 384 c/s virtual
Only one salt: 461 c/s real, 461 c/s virtual
---8<---------8<---------8<---------8<------
View attachment "crypt-benchmark.diff" of type "text/x-patch" (2134 bytes)
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.
