I'm going to attempt to duplicate the problem. It looks like the actual issue with the VM object is occuring a while before the actual crash so the backtraces aren't enough to diagnose the bug. -Matt Matthew Dillon <dillon at xxxxxxxxxxxxx>