Re: Conversation tracking

[Stephen Fisher <steve () stephen-fisher com>]

---- On Thu, 10 May 2012 10:26:46 -0600 Tobias Weiss  wrote ---- 

> I'm using conversation tracking and ran into an issue. My code looks like this (as suggested in the README.developer): 
>
> conv = find_conversation(pinfo->fd->num, 
>    &pinfo->src, 
>    &pinfo->dst, 
>    pinfo->ptype, 
>    pinfo->srcport, 
>    pinfo->destport, 
>    0); 
>
> if(conv != NULL) { get conversation data } 
> else { create conversation data with se_alloc() and add it to the conversation } 
>
> When the dissectors main function is called for the first time, find_conversation() returns a non-null pointer! As far 
> as I understood it should return 0 as conversation_new() was never called for this address/port combination yet. What 
> am I missing here? 

Using a dissector I wrote, packet-exec.c, as an example, I first called find_or_create_conversation(), then 
conversation_get_proto_data() and then check the return of that second function to see if I need to create the data:

        conversation = find_or_create_conversation(pinfo);

        /* Retrieve information from conversation
         * or add it if it isn't there yet
         */
        hash_info = conversation_get_proto_data(conversation, proto_exec);
        if(!hash_info){
                hash_info = se_alloc(sizeof(exec_hash_entry_t));

<<< set the variables here >>>

                conversation_add_proto_data(conversation, proto_exec, hash_info);
        }

___________________________________________________________________________
Sent via:    Wireshark-dev mailing list <wireshark-dev () wireshark org>
Archives:    http://www.wireshark.org/lists/wireshark-dev
Unsubscribe: https://wireshark.org/mailman/options/wireshark-dev
             mailto:wireshark-dev-request () wireshark org?subject=unsubscribe