Hi All, I'm trying to restart a couple of transient systems but I have found some problems that I can't really understand.
Running in debug mode I find a failing assert in System::read_header(), line 220, assert(this->_vectors.count(vec_name) == 0); the problem is that both _old_local_solution and _older_local_solution, which I'm trying to read, are initialized by the TransientSystem constructor. Anyway it seems that the function add_vector() takes into account the possibility of adding an already added vector so I've removed this assert. All works fine running in serial but in parallel (with serial mesh) I get this PETSC message [0]PETSC ERROR: Object is in wrong state! [0]PETSC ERROR: You cannot call this after you have called VecSetValues() but before you have called VecAssemblyBegin/End()! I have figured out that this happens in the function ImplicitSystem::assemble() when we do rhs->zero(); This could be due to the fact that rhs's values has been read from file, but I don't know why this happens only in parallel. I've tried setting rhs->close() before rhs->zero() but then in debug mode I get Assertion `((i >= this->first_local_index()) && (i < this->last_local_index()))' failed. [1] /home/lorenzo/src2/libmesh-cvs-new/include/numerics/petsc_vector.h, line 705, compiled Apr 22 2008 at 11:58:39 terminate called after throwing an instance of 'libMesh::LogicError' what(): Error in libMesh internal logic and running in optimized mode i find that my solution is wrong near the mesh partitions boundaries. Any idea? Thanks Lorenzo ------------------------------------------------------------------------- This SF.net email is sponsored by the 2008 JavaOne(SM) Conference Don't miss this year's exciting event. There's still time to save $100. Use priority code J8TL2D2. http://ad.doubleclick.net/clk;198757673;13503038;p?http://java.sun.com/javaone _______________________________________________ Libmesh-users mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/libmesh-users
