Follow @Openwall on Twitter for new release announcements and other news
[<prev] [next>] [<thread-prev] [day] [month] [year] [list]
Date: Wed, 20 Jun 2012 11:23:43 -0500
From: "jfoug" <jfoug@....net>
To: <john-dev@...ts.openwall.com>
Subject: RE: Record Loaded hashes/salts with current interface

>> i.e. as explained above, 'which' pointer, there are 50 or so of them
>> ;)
>
>Yet I think you can store the pointer in init() and use it in reset(),
>without having it as a parameter to reset().  The "thin" formats will
>have their (shared) init() called one by one, so the global pointer
>should be right at the time of reset().  No?

I believe I have been able to work around the issues in the format. I was
calling init myself in many places (prepare() in the dynamic format, and
several places within the thin formats).

I have re-done prepare() in dynamic_fmt.c, so that it does not require
init() to be called.  I have changed the THIN_LINKING code, so it can be
called twice.  Once to simply get the underlying format pointer (but not to
call init), and then it init is called on the thin format the linkage is
called again, but this time, the init() is called within the linkage call,
to properly init the format.

I think with these changes, the problem I was seeing, with the wrong pointer
being used, should go away. I was getting the proper format pointer, but
within that format pointer, the 'private->data' was not always setup
properly.  Now, I think it is.

Jim.

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.