Ronnie,
> Please see the CMIP (which uses ROS) FTAM and MMS which are all
> implemented ontop of ACSE and which shows how some data is passed from
> the PRES layer to the dissector using the SESSION structure.
>
> Since this structure is already used for these things, this is i think
> where you would add your extra variable.
Thanks - I'll have a look at this and see how this works.
> As for ROS, CMIP implements it by hand in the conformance file since
> we dont have a ROS dissector yet.
> Cutnpaste from that file would probably do the trick until we have ROS
> in ethereal.
Given there is data passing, I may try and call a ROS dissector from
the RTSE dissector - and then from ROS into X.411.
> Please also note that ACSE was replaced wholesale with an asn2eth
> generated dissector a few weeks ago which also replaced the way you
> register a subdissector for use with ACSE.
That is the one I am using.
> I dont think i have seen a real SET yet in a real packet, so if you
> can donate a capture file and the corresponding ASN1 definition
> maybe Tomas can see whatever needs to be added to asn2eth and I can
> add the support to packet-ber.c
I have one that I have been using for testing but can't donate it for
various reasons.
I will try and generate a distributable one for you. X.400 is probably
going to exercise the ASN.1 handling quite a bit.
I've added basic support for SETs to packet-ber.c (basically a
cut/paste of dissect_ber_sequence). I want to check that all the
non-OPTIONAL elements have been seen after parsing the SET, but the
ber_sequence_t being const means I can't modify the flags to indicate
it has been seen. Another approach is required.
asn2eth was just fine with SETs.
Graeme
