Hi Pablo,
I've now submitted the 0001-abis_oml.patch (after removing some dead code) to wireshark, see https://bugs.wireshark.org/bugzilla/show_bug.cgi?id=5784 for details (I suggest you add yourself to Cc).
On Sun, Mar 06, 2011 at 02:16:43PM +0100, Pablo Neira Ayuso wrote:
18:14:15 Warn Dissector bug, protocol A-bis OML, in packet 13: packet-gsm_abis_oml.c:940: failed assertion "DISSECTOR_ASSERT_NOT_REACHED"
It seems that some TLVs that appear in the body of packets are unknown. I'll debug which are the complaining tags and get back you.
Did you enable the 'use ip.access nanoBTS' flag in the protocol preferences for the OML dissector? If you don't it will try to use the Siemens definitions on an ip.access protocol trace (which will break).
Yes, the nanoBTS mode is enabled.
I've been debugging this a bit. The problem seems to be related with "Activate Software" and "Activate Software ACK" messages. They contain an attribute whose tag is 0x42 which the dissector does not know how to interpret.
Any clue on this?
I don't know yet, but I think it is not a blocker for submitting it. I really want to get it included. We can always fix those kind of minor bugs later.
I think the next should be 0004-rsl_ipaccess.c, but maybe make it a (default: on) preference whether or not to enable ip.access support. It has the same C99 array initializer problems.