Date: Sun, 17 Jun 2012 16:24:43 +0200 From: magnum <john.magnum@...hmail.com> To: john-dev@...ts.openwall.com Subject: Re: Re: [patch] optional new raw sha1 implemetation Nice, you boosted it by 13% on my core2 laptop. You need to re-add some #ifdef (apart from __GNUC__) for building this, now that the __SSE4_1__ is gone. I'm not sure what is the requiremtent now, just __SSE2__ perhaps? magnum On 2012-06-17 14:23, Tavis Ormandy wrote: > Hey magnum, I implemented one of Simon's suggestions, as well as your > fixes, and some other minor performance adjustments like an XMM bswap > and using a lookup table instead of a branch in set_key(). > > Simon suggested requesting more than 4 passwords at a time, then > buffering them to reduce the overhead of multiple function calls. That sounds > like a nice easy performance win, and from testing around 128 seems to be the > right number to request on my machine (i tried powers of two up to 8192, and > 256 wins). > > Current HEAD: > > $ time ../run/john --format=rawsha1_sse4 passwords > Loaded 1 password hash (Raw SHA-1 [taviso sse4]) > madmda16 (?) > guesses: 1 time: 0:00:01:30 DONE (Sun Jun 17 13:41:17 2012) c/s: 9429K trying: madmda11 - madmda16 > Use the "--show" option to display all of the cracked passwords reliably > > real 1m30.985s > user 1m30.596s > sys 0m0.064s > > With Simon's idea @256: > > $ rm ../run/john.pot; time ../run/john --format=rawsha1_sse4 passwords > Loaded 1 password hash (Raw SHA-1 [taviso sse4]) > madmda16 (?) > guesses: 1 time: 0:00:01:19 DONE (Sun Jun 17 13:56:51 2012) c/s: 10843K trying: madmduft - madmdr93 > Use the "--show" option to display all of the cracked passwords reliably > > real 1m19.122s > user 1m18.554s > sys 0m0.091s > > Not bad! > > I've also written an alternative to the SSE4 compare, it's slower on my > machine, but not by much, so you can use it even if you don't have SSE4. > > I'll try to send you a pull request, but I've never used github before, > so please forgive me if I get it wrong. > > Tavis. > > On Sun, Jun 17, 2012 at 11:06:26AM +0200, magnum wrote: >> Blimey, you are correct. How long has it been this way?? I'll fix them. >> >> magnum >> >> >> On 2012-06-17 03:56, jfoug wrote: >>> Dyna 12 is NOT correct. Any raw-sha1 format doing that is wrong. It should >>> be dyna_26 >>> >>>> >>>> I now replaced it with strrchr() as it's working on a string anyway. I >>>> also added the $dynamic_12$ tag other raw SHA1 formats use. And the >>>> format announced FMT_SPLIT_UNIFIES_CASE which was not correct until now. >>>> >>>> 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.