Memory leak with dynamic clients

[Alan DeKok]

On Dec 27, 2018, at 5:19 PM, Jared Cantwell <jared.cantwell at jumpcloud.com> wrote:
> 
> Thanks for the quick reply.  I think this approach is sufficient.  My
> initial tests show that this change appears to have resolved our issue.  I
> will perform more extensive testing over the coming days/weeks and report
> back if I find any further issues.

  OK, thanks.

> Is there any reason these changes can't/shouldn't be made in client.c when
> client_delete is called?  listen.c is the only caller, but I wonder if its
> cleaner for client.c to handle the lifetime of the client after it's
> deleted.

  It doesn't make a huge difference either way.  Since listen.c creates the dynamic client, it seemed reasonable to have it free the client, too.

  Alan DeKok.