[sr-dev] fix_nated_contact() then encode_contact() causes core dump

Nick Lemberger Nick.Lemberger at lkfd.net
Sat Oct 17 11:48:59 CEST 2009


When one runs fix_nated_contact() from the nathelper module for it's intended purpose (it must complete successfully) and then run encode_contact() from the siputils module (it also must complete) Kamailio always core-dumps with:
CRITICAL:core:del_lump: offset exceeds message size (202688 > 602) aborting...

I'm guessing this has something to do with re-writing the contact header twice and not fixing the message size the second time but I can't say for sure.

I have a core file, but it seems rather large (44M). It can be uploaded if need be. This is occurring with both the release 1.5.2tls & 1.6.0notls dev trunk.

I posted this bug to the SourceForge Bug Tracker - if this is the wrong list for bug reports on 1.5/1.6 let me know.  If this should be put into a different bug tracker, I can do that as well.

Regards,
Nicholas Lemberger
Lakefield Communications




More information about the sr-dev mailing list