On 04/14/2010 06:43 PM, Simo Sorce wrote:
On Tue, 13 Apr 2010 12:51:39 -0400
Simo Sorce<ssorce(a)redhat.com> wrote:
>
> This patch attempts to address the problem Eugene found with
> sdap_handle_release()
>
Attached a follow-up patch to address Eugene's legitimate concern about
restrictions put on _send() functions. We never free gsh in a _send()
either but only in the _recv() function of the call that actually
establish a new connection. This guarantees that gsh will never be
freed during sdap_handle_release() or any other error code path.
I can squash the 2 patches in a single one if people feels that is
better. Comments welcome.
I would appreciate very much if during this review you compare the
solution suggested by me with the solution suggested by Simo.
My strong opinion is that the solution suggested by me is more reliable,
robust to future changes and has less impact on the way already released
code operates.
Well, I am definitely a biased judge on this question. :-)
Regards, Eugene