Update: Kerberos Ticket Forwarding Patch/Update [3.2]

[Derrick Schommer]

The latest one I just sent (passed in transmission) removes the struct in total. It's been awhile, I didn't know the krb5_mk_1cred was non-portable though. Hmm..

I also saw your use of the SMB_MALLOC, didn't know you had that :) The code I wrote came out of a in-house CIFS client I wrote a few years ago for internal testing, it only had to run in this one environment, I tried to port it as cleanly as possible but I see I've been pretty unsuccessful :-)

Derrick


-----Original Message-----
From: Jeremy Allison [mailto:jra at samba.org] 
Sent: Friday, August 01, 2008 15:45
To: Derrick Schommer
Cc: Love Hörnquist Åstrand; samba-technical at lists.samba.org
Subject: Re: Update: Kerberos Ticket Forwarding Patch/Update [3.2]

On Fri, Aug 01, 2008 at 03:27:03PM -0400, Derrick Schommer wrote:
> Okay, I was going to ask about the structure, and if it was "cool" to use "magic numbers" in order to get the size right. 

Actually I'm in the process of rewriting this patch, as 
it's nearly "close enough" :-).

Here's what I have so far. Not quite finished, but if
you're going to rewrite it I thought you might was well
start with what I'd do to it anyway. I added several
checks for return calls that were missing and re-arranged
the termination logic to come first, which is what
we prefer.

NB. You're using a non-portable krb5 interface
krb5_mk_1cred(). As far as I can tell this is not
defined as an external entry point. We do that in
other places, but we have to create wrappers for
other kinds of krb5 (Heimdal etc.).

If you can finish this up I'd appreciate it.
Anyone got any good ideas on what to do about
krb5_mk_1cred() ?

Jeremy