talloc destructor in brlock.c?

[Volker Lendecke]

Hi, Jeremy!

Log message from r15082 (trunk):

> Using talloc with destructors is nice and all, but in this
> case it's in a performace critical path and it *hurts* us.
> Go back to plain malloc/free with an explicit destructor
> call.
> Jeremy.

Does that really hurt that much? I'm asking because we're
trying to advance the cluster stuff, and having a nice
straight API to the tdb that essentially matches the API to
locking.tdb would make the work a bit easier.

What would persuade you to get it in again? How much loss
in what test would you accept? :-)

Volker
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 189 bytes
Desc: not available
Url : http://lists.samba.org/archive/samba-technical/attachments/20060613/a50e9d06/attachment.bin