https://bugs.wireshark.org/bugzilla/show_bug.cgi?id=7066
--- Comment #15 from Cristian Constantin <const.crist@xxxxxxxxxxxxxx> 2012-04-13 03:11:36 PDT ---
(In reply to comment #11)
> Yes, it should take into account the most recent setup frame.
> I was looking at
> http://anonsvn.wireshark.org/viewvc/trunk/epan/conversation.c?0=39484&1=40837&pathrev=40837
> and don't see anything particularly shocking there, but then i never worked
> much with conversations.
> As i understand it, there can be multiple conversations for the same key,
> sorted by setup-frame. Did you maintain that scheme?
cristian: for the sip traffic, as far as I can tell, the design seems to be the
following:
1. SDP is dissected
2. ip address and port used for the media traffic are extracted from the
SDP payload
3. a session is either created or looked up based on the ip address and port
(and having as 0 the second ip address and port)
so basically one half of the session is stored in the session hashtable.
they can overlap with other half sessions that use the same ip address and port
(on the receiving or transmitting side)
>
> What about if you debug the calls to conversation_new for the various SDP
> setup-frames? Maybe you could see something wrong there.
--
Configure bugmail: https://bugs.wireshark.org/bugzilla/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are watching all bug changes.
