#Fortran

Copy this code into fixed_loops.f.

      program main
      implicit none
      include 'mpif.h'

      integer i, rank, nprocs, count, start, stop, nloops
      integer total_nloops,err
      integer status(MPI_STATUS_SIZE)

      call MPI_Init(err)

c     get the number of processes, and the id of this process
      call MPI_Comm_rank(MPI_COMM_WORLD, rank, err)
      call MPI_Comm_size(MPI_COMM_WORLD, nprocs, err)

c     we want to perform 1000 iterations in total. Work out the 
c     number of iterations to perform per process...
      count = 1000 / nprocs

c     we use the rank of this process to work out which
c     iterations to perform.
      start = rank * count
      stop = start + count - 1

c     now perform the loop
      nloops = 0

      do i=start,stop
        nloops = nloops + 1
      enddo

      print *,"Process ",rank," performed ",nloops,
     .        " iterations of the loop."

      if (rank .ne. 0) then
c        if we are not the master process (rank 0) then we need to
c        tell that process how many iterations we performed
         call MPI_Send( nloops, 1, MPI_INTEGER, 0, 0, MPI_COMM_WORLD, 
     .                  err )
      else
c        if we are the master process, then we need to wait to be
c        told how many iterations were performed by each other
c        process
         total_nloops = nloops;

         do i=1,nprocs-1
            call MPI_Recv( nloops, 1, MPI_INTEGER, i, 0, MPI_COMM_WORLD,
     .                     status, err )

            total_nloops = total_nloops + nloops
         enddo

c        ok - are there any more loops to run?
         nloops = 0

         if (total_nloops .lt. 1000) then
           do i=total_nloops,1000-1
              nloops = nloops + 1
           enddo
         endif

         print *,"Process 0 performed the remaining ",nloops,
     .           " iterations of the loop"
      endif

      call MPI_Finalize()

      end

This example uses two new functions;

• MPI_Send(void message, integer size, MPI_INTEGER, integer process, 0, MPI_COMM_WORLD, err). This function sends the message in the memory pointed to by message to the process whose rank is given in process. The message is an array of integers, where the number of integers is given in size.
• MPI_Recv(void message, integer size, MPI_INTEGER, integer process, 0, MPI_COMM_WORLD, status, err). This function receives the message from the process whose rank is given in process. The received message is placed into the memory pointed to by message. The message is an array of integers, where the number of integers is given in size. status is an array of integers that gives the status of the transfer. This must be an array declared as integer status(MPI_STATUS_SIZE).

It is possible to send messages of other data types, e.g.

• MPI_INTEGER - the message is a single integer, or an array of integers
• MPI_REAL - the message is a single real number or an array of real numbers
• MPI_DOUBLE_PRECISION - the message is a single double precision number or an array of double precision numbers

Just replace MPI_INTEGER in MPI_Send and MPI_Recv with the correct data type for the message.

The full list of MPI data types is given here.

Note that it is also possible for you to define your own data types, although this is beyond the scope of this basic introductory course.

In this example, MPI_Send is used to send the number of completed loops from each worker process (processes with ranks 1 to N-1) to the master process (process with rank 0). MPI_Recv is called N-1 times on the master process to receive the N-1 messages sent by the worker processes.

You can compile this example using;

mpif77 fixed_loops.f -o fixed_loops

This will produce the executable fixed_loops.

Return to the previous page.