Re: Dissector lifecycle and per-file data

[Pascal Quantin <pascal () wireshark org>]

Hi Vincent,

Le lun. 29 mars 2021 à 18:24, Vincent Mallet <vmallet () gmail com> a écrit :

> I am hoping this is the right place to ask this question but if not please
> redirect me.

That's the right place :)


> I am working on a custom C dissector and need to keep some data around for
> the lifetime of the current file. At the moment I allocate my structures
> using various wmem_ functions with a wmem_file_scope() scope and keep them
> in conversations with conversation_get_proto_data() and
> conversation_add_proto_data(). It works great for conversation-related data
> but is not ideal for file-related data.

The wmem file scope is definitely the right memory pool.


> Is there a way for a dissector to know a new file is being worked on and
> it’s time to reinitialize data structures? Or is there an equivalent to
> conversation_get_proto_data() but for a file?

Did you look at the register_cleanup_routine() found in epan/packet.h? It
seems to be what you are looking for. To quote the comments:
/**
 * Allows protocols to register "cleanup" routines which are called
 * after closing a capture file (or when preferences are changed, in
 * that case these routines are called before the init routines are
 * executed). It can be used to release resources that are allocated in
 * register_init_routine.
 */
You can find multiple dissectors using it in our code base that should give
you plenty of examples.

Best regards,
Pascal.

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