Openwall GNU/*/Linux - a small security-enhanced Linux distro for servers
[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Date: Sun, 17 Jul 2011 19:54:51 -0500
From: "JimF" <jfoug@....net>
To: <john-users@...ts.openwall.com>
Subject: Re: Re: sha1(md5($pass)

All of this wedding stuff, and having to run around all day never seems to 
end.  I have not gotten a chance to fully get the md5_gen format working 
properly on all builds 'yet', but this is my top priority.  I hope to get 
some time tonight, after I rest up a bit.

.....  Email written earlier today, but not sent ----- 

Until I get the problems found and fixed, the current version simply
compiles 'out' the md5_gen 12/13/21-26 formats for ANY x86_64 build.  During
the later redesign of the code, these builds started failing (even coring),
so I simply wrote them 'out' for those builds.   The 32 bit or true generic
should still work.  I am not sure if true generic on built with a x86 64 bit
compiler would work, but I think not.

Like I said, until the bugs are found, and the code is changed, those
formats simply do not work.   This is the way it was for many versions of
md5_gen.  If a format went over 55 bytes, then for ANY sse build, it was
off.  I worked around some of those issues, and have a 'hybrid' SSE/x86 set
of functions, which do some of the crypt work in SSE (if they can). This
helped open up some of the formats to SSE builds, however, on the 64 bit
systems (especially 64 bit with SSE2int) it was not working right.  Also,
the SHA1 additions was very new code (added the last couple of days before I
released it), and thus is not nearly as optimized, nor as well working.  I
think all of the SHA1 is commented out in x86_64 builds right now.

I will get to it later today (I hope).  I have been very busy the last few
days (last couple of weeks actually). Just got the stepson married off on
Friday, so things are still a little hectic.  But it is coming to an end,
and I should be able to get some coding done later this afternoon.   Getting
md5_gen fully working is high on my todo list.  Having the formats commented
out on the 64 bit builds was supposed to be very temporary, and simply due
to lack of time to find the bugs.

Jim.
----- Original Message ----- 
From: "Donovan" <contact.newangels@...il.com>
To: <john-users@...ts.openwall.com>
Sent: Sunday, July 17, 2011 10:59 AM
Subject: [john-users] Re: sha1(md5($pass)


> Hi,
>
> Ok, i compiled an new one ( X86_SSE2 for macOS).
>
> Work great with sha 1 (md5($pass)
>
> new-host-4:run xxx$ mpirun -np 4 ./john --format='md5_gen(23)' --test
> Benchmarking:md5_gen(23): sha1(md5($p)) SSE2 [SSE2 32x4 (.S)]..(4xMPI)DONE
> Raw: 12135K c/s real, 12135K c/s virtual
>
> new-host-4:run xxx$ ./john --format='md5_gen(23)' --test
> Benchmarking:md5_gen(23): sha1(md5($p)) SSE2 [SSE2 32x4 (.S)]..DONE
> Raw: 5218K c/s real, 5218K c/s virtual
>
> Best Regards,
>
> Donovan
>

Powered by blists - more mailing lists

Your e-mail address:

Confused about mailing lists and their use? Read about mailing lists on Wikipedia and check out these guidelines on proper formatting of your messages.