Anders Broman wrote:
Hi,
As far as I know the only change was to use range rather than a single
ssn value in the preferences of ANSI MAP, probably you got owerlaping ssn
definitions in your preferences ( CAMEL ,GSM MAP, RANAP ... ) what does
it say at the ssn entry in the SCCP part of the dissection?
Hmm, when I load that capture file the SCCP portion says:
SSN: 6
[Linked to TCAP, TCAP ssn linked to GSM_MAP]
which looks OK.
However, the bottom of the TCAP protocol tree says says "BER Error:
Wrong field in sequence [...]".
I suppose that's preventing the TCAP dissector from calling the GSM_MAP
dissector.
[If so should such BER errors fail an assertion or something so the
packet shows up as malformed/dissector bug/whatever??]
-----Original Message-----
From: Michael Lum [mailto:Michael.Lum@xxxxxxxxxx]
Sent: August 9, 2006 12:18 PM
To: 'wireshark-dev@xxxxxxxxxxxxx'
Subject: ANSI MAP / TCAP dissector hooks
Hello,
why was the way ANSI MAP hooks into TCAP changed?
I believe there is a problem with it.
I have attached a capture file with the following:
SCTP
M2UA
MTP3 (ITU)
SCCP
TCAP (ANSI)
ANSI MAP
My preferences were set to:
MTP3 (ANSI)
ANSI MAP (SSN RANGE 5-14)
Procedure:
1. Start Ethereal
2. Set filters
3. Load file
Frames show up as SCCP (ANSI), ERR and DT2
4. Edit preferences, change MTP3 -> ITU
With Ethereal 0.99.0 everything works fine the four frames come out decoded
as ANSI MAP
with:
UDT QueryWithPermInvoke(Last) Location Request UDT QueryWithPermInvoke(Last)
Routing Request UDT ResponseRetRes(Last) UDT ResponseRetRes(Last)
Following the same procedure with Wireshark 0.99.2 the decode goes as far as
TCAP but not ANSI MAP.
To get the decode to work properly I have to:
5. Edit preferences, change ANSI MAP SSN Range to something different,
apply
