By the way usually a tunnel encapsulates a single packet. I'm not aware
of any other protocol multiplexing at the IP level. I would assume
Wireshark requires some replumbing to handle that. Something like TFS
being treated as a framing layer. Just food for thought.
On 21/01/20 14:46, João Valverde wrote:
On 21/01/20 14:33, Christian Hopps wrote:
So I've got a payload of packets in a single frame. I'm calling
dissector_try_uint_new() to dissect each payload (typically IPv4
packets). Some of these packets are considered "malformed" by
wireshark (e.g., created by scapy/trex with some bogus values).
The problem I'm hitting is that the first malformed inner packet
fails all the way out of my parent dissector, so it doesn't dissect
any of the other packets in the payload.
Another problem I'm having is that the IP sub-dissector is
overwriting my source and destination addresses in the pinfo/tree
(not sure which doesn't really matter).
Summary:
- How can I "catch" errors in a subdissector so I can call other
sub-dissectors?
Use TRY/CATCH (in epan/exceptions.h).
- How can I "block" sub-dissectors from overwriting my outer header
information?
I don't think you can. Maybe your IPTFS dissector can set it after the
sub-dissectors run.
Thanks,
Chris.
___________________________________________________________________________
Sent via: Wireshark-dev mailing list <wireshark-dev@xxxxxxxxxxxxx>
Archives: https://www.wireshark.org/lists/wireshark-dev
Unsubscribe: https://www.wireshark.org/mailman/options/wireshark-dev
mailto:wireshark-dev-request@xxxxxxxxxxxxx?subject=unsubscribe
___________________________________________________________________________
Sent via: Wireshark-dev mailing list <wireshark-dev@xxxxxxxxxxxxx>
Archives: https://www.wireshark.org/lists/wireshark-dev
Unsubscribe: https://www.wireshark.org/mailman/options/wireshark-dev
mailto:wireshark-dev-request@xxxxxxxxxxxxx?subject=unsubscribe
