Avoid memcpy in overlapped memory regions
Fix a problem introduced when memory reduction was implemented
Add some checks
Check we are making different diferent computation in every process
Initialize every process with a different seed
Remove config.h
ADD Barriers to help localize segFaults in multiprocess executions
Reorder pmap actions in a sense way. Not having a SegFault with 1 MPI process
Modify the region executed only by master process in main
Prevent all processes enter in save_coarse function
