All the errors I've sent you have been from devel mode (including the trace).
I'm running it through GDB now (using XCode on my Mac)... to get a better picture. It's still pretty unclear to me at this point. I don't understand why build_cube would work... but not reading xda/ exodus.... Derek On Jun 2, 2008, at 4:16 PM, Roy Stogner wrote: > > On Mon, 2 Jun 2008, Derek Gaston wrote: > >> I've attached the tracefile. It doesn't look all that helpful to me. >> Hopefully it will be more useful to you. > > It's at least a start - it's one of the dof_indices calls in > add_neighbors_to_send_list that's triggering the failure. So > somehow either the active_elements iterators or an > active_family_tree_by_neighbor call is returning a remote_elem? > >> I'm going to try to get it loaded up in the debugger so I can watch >> it. > > More importantly, recompile and run with METHOD=dbg or at least > METHOD=devel, not opt mode. If we're lucky, it'll trip over an > libmesh_assert() near the source of the bug, which would be a lot more > helpful than hitting a libmesh_error() that's triggered by the bug. > --- > Roy ------------------------------------------------------------------------- This SF.net email is sponsored by: Microsoft Defy all challenges. Microsoft(R) Visual Studio 2008. http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/ _______________________________________________ Libmesh-users mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/libmesh-users
