Wireshark-dev: Re: [Wireshark-dev] Patch to decode ERF type 5 record

From: Jeff Morriss <jeff.morriss@xxxxxxxxxxx>
Date: Sun, 18 Feb 2007 19:33:26 +0800


Florent.Drouin@xxxxxxxxxxxxxxxxx wrote:
No, the ERF type 5 record has a different header than the PCAP header, but
MTP2 part is not affected.
In fact, the MTP2 (FCS) is not specific to the ERF format, I would say,
MTP2 (FCS) is the standart MTP2, but the checksums are present in the 2
last bytes of the frame.

I could use a new DLT, but this would be to show the additional
informations given in the ERF type 5 header (like a kind of timeslot
information, etc)

In fact I didn't mean use a new DLT value but rather a different
WTAP_ENCAP value.  E.g., here:

+       case TYPE_MC_HDLC:
+               wtap_encap = WTAP_ENCAP_MTP2;
+               break;

it could be WTAP_ENCAP_MTP2_WITH_FCS (were that to exist) so the MTP2
dissector could automatically know if the FCS is there or not.

Anyway, I read a little bit about this file format and it seems to just
be a place holder for anything in HDLC, not just MTP2.

For ATM stored in ERF files wiretap uses 'getenv()' to tell which WTAP_ENCAP should be used. If/when there's a WTAP_ENCAP_MTP2_WITH_FCS, (or any other HDLC protocol stored in that file format) I guess that method could be used for HDLC, too.

Anyway, I checked in your patch (rev 20838).