On 08/29/2011 06:00 PM, Bart Van Assche wrote: > On Fri, Aug 26, 2011 at 1:40 PM, Jan Safranek <[email protected]> wrote: >> I debugged bug 1633670 as far as I could and I have even a >> proof-of-concept patch, but I am not so happy with it. Could someone >> with better AgentX knowledge look at it? These parts look a bit fragile >> and I don't want to break them apart with wrong patch. > > That patch is probably fine but looks suboptimal to me. IMHO it > doesn't make sense to dequeue new requests while closing a session. > The (untested) patch I added to the patch tracker should also fix the > crash but without dequeueing new requests while closing a session.
Problem is that you *must* dequeue all requests which are related to the closing AgentX session - somewhere deep in these requests there is pointer to the AgentX session and the session is being freed when closing. Now, how to dequeue them quicker than walking AgentX subtree? Jan ------------------------------------------------------------------------------ Special Offer -- Download ArcSight Logger for FREE! Finally, a world-class log management solution at an even better price-free! And you'll get a free "Love Thy Logs" t-shirt when you download Logger. Secure your free ArcSight Logger TODAY! http://p.sf.net/sfu/arcsisghtdev2dev _______________________________________________ Net-snmp-coders mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/net-snmp-coders
