Wireshark-bugs: [Wireshark-bugs] [Bug 2681] glithes in current ssl decrypt code
Date: Thu, 10 Jul 2008 03:40:52 -0700 (PDT)
https://bugs.wireshark.org/bugzilla/show_bug.cgi?id=2681





--- Comment #10 from Paolo Abeni <paolo.abeni@xxxxxxxx>  2008-07-10 03:40:51 PDT ---
(In reply to comment #9)
> Created an attachment (id=1990)
 --> (https://bugs.wireshark.org/bugzilla/attachment.cgi?id=1990) [details]
> Additional SSL dissector registration
> 
> While working on your patch I came across the registration of two ports as SSL
> equivalents to the plain service. I'm no expert in this field, but should this
> patch be applied as well? It adds HTTP and LDAP on their respective SSL ports.
> I've no means of testing this, and I'm not sure if this creates conflicts.

It seams that the ssl_association_add() calls in 
proto_reg_handoff_ssl() are some kind of legacy. The proper associations for
well known protocols are cretated via ssl_dissector_add() in other dissector
proto_reg_handoff(). See for example: epan/dissectors/packet-sip.c.

Imap and pop dissector does not call ssl_dissector_add (yet), so they must
explicitly initialized in packet-ssl.c.

Perhaps proper ssl_dissector_add calls() should me placed in packet-pop.c and
packet-imap.c and the corresponding calls to ssl_association_add() should be
removed from packet-ssl.c. What do you think ?

BTW can we close this bug ?


-- 
Configure bugmail: https://bugs.wireshark.org/bugzilla/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are the assignee for the bug.