[Bernd Becker <bb@xxxxxxxxxxxxxxx>]

Hi Craig,

your patch works fine with the GIOP captures I have tried (Visibroker 
traffic mainly).

A couple of changes though (mainly errors that were there before your 
patch):

- correct bitmask for hf_giop_iop_vscid and  hf_giop_iop_scid, reflecting 
the change
 to a 24bit vendor id and 8 bit service id
- set the length of the "Service Context List" tree correctly after 
dissecting. The
 length was just being set to the end of the tvb
- do not exit the loop through the Service Context List with return if the 
sequence length
 is 0, continue the loop instead. This should fix a problem reported by 
Mika Korpela.
 (see   http://www.ethereal.com/lists/ethereal-dev/200205/msg00234.html )
 I never really tested this though, because I don't have a capture with a 
Service Context
 List with more than one element.

Regards,
Bernd

--On Wednesday, February 12, 2003 19:07:33 -0500 Craig Rodrigues 
<rodrigc@xxxxxxxxx> wrote:

> Hi,
>
> I am working with TAO 1.3 ( http://www.cs.wustl.edu/~schmidt/TAO.html ),
> which is a CORBA compliant Object Request Broker (ORB).
>
> TAO has implementats some features of Real-time CORBA.  It encodes
> a priority value in the service context field.  I wanted to
> decode this value with Ethereal, so I had to make some enhancements
> to Ethereal.
>
> I added a few new methods:
> decode_RTCorbaPriority()
> decode_UnknownServiceContext()
>
> and I overhauled:
> decode_ServiceContextList() to make use of the above functions, and
> removed some code that was hidden behind #ifdef's.
>
> The OMG has standardized 16 ServiceContext ID's.  I didn't add decoding
> routines for all of those ServiceContext ID's, but it should be
> relatively easy to add them in at a later date.
>
> Right now I just added decoding for ServiceContext ID == 10, which is
> RTCorbaPriority.
>
> How is this patch?
>
> Thanks.
> --
> Craig Rodrigues
> http://home.attbi.com/~rodrigc
> rodrigc@xxxxxxxxx

Attachment: packet-giop.c.diff
Description: Binary data