Openwall GNU/*/Linux - a small security-enhanced Linux distro for servers
[<prev] [next>] [<thread-prev] [day] [month] [year] [list]
Date: Fri, 3 Feb 2012 13:28:52 +0400
From: Solar Designer <>
Subject: Re: Bit slice implementation of DES based hashes


I just took a look at the code.  It turns out that you haven't moved to
1.7.9+ yet, although I thought you said you did.  If what you did was
actually move your old 1.7.8'ish files into a 1.7.9+ tree, that may be
causing extra problems.

On Tue, Jan 24, 2012 at 09:53:19PM +0530, Piyush Mittal wrote:
> Here I am enclosing my code, please check whether key initialisation is
> done right or not?

It depends on what you're referring to.  The changes to DES_bs_init()
look OK to me (as I suggested, but for 1.7.8's code).  The call to
DES_bs_set_key() is OK, although the key string is probably wrong (I see
that you have the one we discussed commented out, so perhaps you were
experimenting with that).  However, when we consider Oracle specific key
setup, things get weird - but perhaps you don't actually expect that to
work yet?

Somehow temp_key[] is not NUL-terminated (but may work fine anyway when
you have just one test vector), temp_salt[] is initialized in get_salt()
instead of in set_salt(), which you don't even have (you should).  But
these are just a few obvious things.  It is difficult to comment on the
rest because I don't see whether and how it was supposed to work, or
probably it was not (yet).

It looks like initially your goal was to get things working for just one
bit layer - correct?  And you haven't completed that implementation yet -
correct?  What/how are you testing, then?  Are you only testing the
"Key values.." printout from DES_bs_crypt_Oracle()?


Powered by blists - more mailing lists

Your e-mail address:

Powered by Openwall GNU/*/Linux - Powered by OpenVZ