svn commit: samba r22020 - in branches/SAMBA_3_0/source: auth include lib smbd

[Gerald (Jerry) Carter]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Andrew Bartlett wrote:

> I think the getpwnam cache is now cleaner 
> avoiding talloc_reference, and doing a tcopy_passwd instead.
> Having a magic cache, using talloc_reference() I think
> caused confusing behaviours.  Because callers
> are not expecting a cache, they would be entirely 
> within their rights to talloc_steal() parts of the the
> results, but this would break the cache.

It's the stealing members from a structure that concerns me.

> The bug I was actually trying to fix is that I suspected 
> that not everything was being cleaned up when a session
> terminated.  I was hoping to make this simpler, with talloc
> taking over from the manual cleanup system.

Has Jeremy chatted with you about this?  He was going to
follow up.  Like I said,  I'll review it again but the
talloc_steal() calls make it more confusing for me, and
not cleaner.  Maybe another read through will change my mind.

It will take me about a day to get to it though.





cheers, jerry
=====================================================================
Samba                                    ------- http://www.samba.org
Centeris                         -----------  http://www.centeris.com
"What man is a man who does not make the world better?"      --Balian
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.3 (GNU/Linux)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org

iD8DBQFGEwz7IR7qMdg1EfYRAj8AAJ4mmT1P6GFH7G4ygT+qH+inwNemTQCgjp7J
tNdINCLeBV+6Yn4wqcBBu7w=
=yPLG
-----END PGP SIGNATURE-----