Initialization and Termination All processes must initialize MPI by calling the MPI_Init function and must finalize MPI processing by calling MPI_Finalize. These calls use no [IN] arguments and return an int error number (0 = no error). The calling syntax for these functions is: ierr= MPI_Init() ierr= MPI_Finalize()

where:

Parameter	Description	Status
ierr (int)	MPI error number (0 = no error)	[OUT]

The following example program illustrates MPI initialization and termination. Each processor prints "Hello world.".

#include <stdio.h>
#include <mpi.h>
main(int argc, char **argv){
  MPI_Init(&argc, &argv);
  printf(" Hello World.\n");
  MPI_Finalize();
}


OUTPUT (4 PEs):
 Hello World.
 Hello World.
 Hello World.
 Hello World.