What is MPI_Start function?

MPI library

MPI (Message Passing Interface) is a library that allows you to write parallel programs in C or Fortran77. The library uses commonly available operating system services to create parallel processes and exchange information among these processes.

MPI_Start

This method starts a communication channel with a persistent request handle. The request handle is activated when the MPI_Start function returns.

Syntax

int MPI_Start(MPI_Request *request)

Parameters

request is a communication request handle.

Return value

The function returns an error if it is unsuccessful. By default, the error aborts the MPI job.

In case of success, it returns MPI_SUCCESS- the value returned upon successful termination of any MPI routine.

#include <stdio.h>
#include <stdlib.h>
#include <mpi.h>
int main(int argc, char* argv[])
{
    MPI_Init(&argc, &argv);
    // Get the number of processes and check only 2 processes are used
    int size;
    MPI_Comm_size(MPI_COMM_WORLD, &size);
    if(size != 2)
    {
        printf("This application is meant to be run with 2 processes.\n");
        MPI_Abort(MPI_COMM_WORLD, EXIT_FAILURE);
    }
    // Get my rank and do the corresponding job
    enum role_ranks { SENDER, RECEIVER };
    int my_rank;
    MPI_Comm_rank(MPI_COMM_WORLD, &my_rank);
    switch(my_rank)
    {
        case SENDER:
        {
            int data_sent;
            MPI_Request request;
            // Prepare the send request handle
            MPI_Send_init(&data_sent, 1, MPI_INT, RECEIVER, 0, MPI_COMM_WORLD, &request);
            data_sent = 12345;
            // Launch the send
            MPI_Start(&request);
            printf("MPI process %d sends value %d.\n", my_rank, buffer_sent);
            break;
        }
        case RECEIVER:
        {
            int received;
            MPI_Recv(&received, 1, MPI_INT, SENDER, 0, MPI_COMM_WORLD, MPI_STATUS_IGNORE);
            printf("MPI process %d received value %d.\n", my_rank, received);
            break;
        }
    }
    MPI_Finalize();
    return EXIT_SUCCESS;
}

In the program above, we perform the following steps:

Include the mpi.h library to use MPI functions.
Compute the number of processes using the MPI_Comm_size function and check that the application is run with 2 processes, i.e., a sender and a receiver (line 9).
For the sender, we create a request handle (line 24), prepare the send using the MPI_Send_Init function (line 26), and launch the send using the MPI_Start function (line 29).
The receiver process receives the data sent by the sender in line 36 using the MPI_Recv function.