[17251] in Kerberos_V5_Development
Re: GSS memory allocation initial cut for review
daemon@ATHENA.MIT.EDU (Sam Hartman)
Wed Oct 5 13:16:02 2011
From: Sam Hartman <hartmans@painless-security.com>
To: Nico Williams <nico@cryptonector.com>
Date: Wed, 05 Oct 2011 13:15:56 -0400
In-Reply-To: <CAK3OfOgYXo0CkPh7R0yL0m2T-SzO_3VbzbHOiLUqYjzT+j8B3w@mail.gmail.com>
(Nico Williams's message of "Wed, 5 Oct 2011 11:02:52 -0500")
Message-ID: <tslbotvuzoz.fsf@mit.edu>
MIME-Version: 1.0
Cc: kevin.wasserman@painless-security.com, krbdev@mit.edu
Content-Type: text/plain; charset="iso-8859-1"
Errors-To: krbdev-bounces@mit.edu
Content-Transfer-Encoding: 8bit
>>>>> "Nico" == Nico Williams <nico@cryptonector.com> writes:
Nico> On Wed, Oct 5, 2011 at 10:23 AM, Sam Hartman
Nico> <hartmans@painless-security.com> wrote:
>> We've said that we will post on kitten. Our claim is that
>> having all allocations in MIT's mechglue go through one
>> allocation function is an MIT matter and changing what that
>> allocation function does on windows is not that hard to do.
Nico> For any one mechglue implementors are free to do as they wish.
Nico> But I thought you wanted all mechglues on a platform to share
Nico> an allocator, and *that* would require some consensus among
Nico> implementors.
I do, but my timely pressing concern is the changes to the MIT code.
For now we're calling heapalloc. But if the consensus is something else
and we've made changes to run allocations through one place it's easy to
match that consensus.
We presented the long-term plan only for context; the long-term
discussion needs to happen on kitten.
I've said this several times both on krbcore and krbdev.
Well, OK, I may have only said it once on krbdev, but that was in the
original message.
_______________________________________________
krbdev mailing list krbdev@mit.edu
https://mailman.mit.edu/mailman/listinfo/krbdev
