#include <mpi.h>
#include <stdio.h>
#define comm MPI_COMM_WORLD
int main(int argc, char* argv[])
{
        MPI_Init(&argc, &argv);
        MPI_Status status;
        MPI_Request req;
        int j, k, i, rank, size, N = 4;
        MPI_Comm_rank(comm, &rank);
        MPI_Comm_size(comm, &size);
        double a[N][N], b[N][N], c[N][N], d[N][N], t1 = 0, t2 = 0, t3 = 0;
        int del = N / size;
        int i0 = rank * del;
        int imax = i0 + del;
        for (i = 0; i < N; i++)
        {
                for (j = 0; j < N; j++)
                {
                        a[i][j] = i;
                        b[i][j] = j;
                        c[i][j] = 0;
                        d[i][j] = 0;
                }
        }
        if (rank == size - 1) { imax = N; }
        t1 = MPI_Wtime();
        for(i = i0; i < imax; ++i)
        {
                for(j = 0; j < N; ++j)
                {
                        for(k = 0; k < N; ++k)
                        {
                                c[i][j] = a[i][k] * b[k][j];
                        }
                }
        }
        t2 = MPI_Wtime();

        MPI_Reduce(&c, &d, N * N, MPI_DOUBLE, MPI_SUM, 0, comm);
        t3 = MPI_Wtime();
        printf("T1 = %f, t2 = %f, t3 = %f\n", t1, t2, t3);
        for (i = 0; i < N; ++i)
        {
                for (j = 0; j < N; ++j)
                {
                        printf("Rank = %i, Size = %i, A[%i, %i] = %f, B[%i, %i] = %f, C[%i, %i] = %f, D[%i, %i] = %f\n", rank, size, i, j, a[i][j], i, j,  b[i][j], i, j, c[i][j], i, j, d[i][j]);
                }
        }
        MPI_Finalize();
}