5.3.2. Send-only Partitioning Example with Tasks and OpenMP version 4.0 or later

The previous example is tailored specifically for send-side partitioning using threads. This is an example where parallel task producers produce input to part of an overall buffer; they complete in any order and contribute to the overall buffer.
Example Parallel task producers for partitioned communication using threads.

#include <stdlib.h> 
#include "mpi.h" 
#define NUM_THREADS 8 
#define NUM_TASKS 64 
#define PARTITIONS NUM_TASKS 
#define PARTLENGTH 16 
#define MESSAGE_LENGTH PARTITIONS*PARTLENGTH 
int main(int argc, char *argv[]) /* send-side partitioning */ 
{ 
  double message[MESSAGE_LENGTH]; 
  int send_partitions = PARTITIONS, 
      send_partlength = PARTLENGTH, 
      recv_partitions = 1, 
      recv_partlength = PARTITIONS*PARTLENGTH; 
  int count = 1, source = 0, dest = 1, tag = 1, flag = 0; 
  int myrank; 
  int provided; 
  MPI_Request request; 
  MPI_Info info = MPI_INFO_NULL; 
  MPI_Datatype send_type; 
  MPI_Init_thread(&argc, &argv, MPI_THREAD_MULTIPLE, &provided); 
  if (provided < MPI_THREAD_MULTIPLE) 
     MPI_Abort(MPI_COMM_WORLD, EXIT_FAILURE); 
  MPI_Comm_rank(MPI_COMM_WORLD, &myrank); 
  MPI_Type_contiguous(send_partlength, MPI_DOUBLE, &send_type); 
  MPI_Type_commit(&send_type); 
 
  if (myrank == 0) 
  { 
     MPI_Psend_init(message, send_partitions, count, send_type, dest, tag, 
                    MPI_COMM_WORLD, info, &request); 
     MPI_Start(&request); 
 
     #pragma omp parallel shared(request) num_threads(NUM_THREADS) 
     { 
        #pragma omp single 
        { 
           /* single thread creates 64 tasks to be executed by 8 threads */ 
           for (int partition_num=0;partition_num<NUM_TASKS;partition_num++) 
           { 
              #pragma omp task firstprivate(partition_num) 
              { 
                 /* compute and fill partition #partition_num, then mark 
                 ready: */ 
                 /* buffer is filled in arbitrary order from each task */ 
                 MPI_Pready(partition_num, request); 
              } /*end task*/ 
           } /* end for */ 
        } /* end single */ 
     } /* end parallel */ 
     while(!flag) 
     { 
        /* Do useful work */ 
        MPI_Test(&request, &flag, MPI_STATUS_IGNORE); 
        /* Do useful work */ 
     } 
     MPI_Request_free(&request); 
  } 
  else if (myrank == 1) 
  { 
     MPI_Precv_init(message, recv_partitions, recv_partlength, MPI_DOUBLE, 
                    source, tag, MPI_COMM_WORLD, info, &request); 
 
     MPI_Start(&request); 
     while(!flag) 
     { 
        /* Do useful work */ 
        MPI_Test(&request, &flag, MPI_STATUS_IGNORE); 
        /* Do useful work */ 
     } 
     MPI_Request_free(&request); 
  } 
  MPI_Finalize(); 
  return 0; 
}