On Tue, Apr 19, 2011 at 02:48:32PM +0200, Raphael Coeffic wrote: > # ulimit -c unlimited > # /usr/sbin/sems -f /etc/sems/sems.conf > > hopefully you should get something saying 'core dumped'. Then, just > look for a file called core.XXXX, where XXXX is the process ID of > the sems instance which just crashed. Once you have the file (should > be in /var/cores/; if not, check this page: > http://aplawrence.com/Linux/limit_core_files.html), open it with gdb > and then use the backtrace command. If you have some way of > uploading the file somewhere (>100MB), I could also inspect it > myself. > > For example:
Thanks a lot, I fixed the problem, it was that this little dsm script get invoked throught a di_diaoult call, but di_dial.so it's not compiled by default, so it was using the old one of sems 1.2 and that caused the segfault ... :-P Thanks for you time and help -- _______________________________________________ Semsdev mailing list [email protected] http://lists.iptel.org/mailman/listinfo/semsdev
