#include <stdio.h>
#include <stdlib.h>
#include <math.h>
#include <mpi.h>

//#include "serial.h"

#define D 64

//float ** matrix;
float matrix[D][D];

void createAndInitMatrix()
{
    int i = 0, j = 0;
    /*
    matrix = (float **) malloc(D * sizeof(float *));

       if (matrix == NULL)
       {
      printf("NULL\n");
      return;
       }

    int i = 0, j = 0;
    for (i = 0; i < D; i++)
    {
    	matrix[i] = (float *) malloc(D * sizeof(float));
    	if (matrix == NULL)
      {
      printf("NULL in for 2\n");
      return;
      }
    }
    */

    for (i = 0; i < D; i++)
    {
        for (j = 0; j < D; j++)
        {
            matrix[i][j] = i * j;
            //matrix[i][j] = (float) sqrt(i * j);
            //printf("%7.3f | ", array[i][j]);
        }
        //printf("\n");
    }
}



int main(int argc, char** argv)
{
    //serial_calculator();

    MPI_Init(NULL, NULL);
    int world_rank;
    MPI_Comm_rank(MPI_COMM_WORLD, &world_rank);
    int world_size;
    MPI_Comm_size(MPI_COMM_WORLD, &world_size);

    if (world_size < 2)
    {
        fprintf(stderr, "World size must be greater than 0 for %s\n", argv[0]);
        MPI_Abort(MPI_COMM_WORLD, 1);
    }



    //float matrix_recieve2[D][D]; //temporary buffer matrix

    //int number;
    //Initialize  matrix in master and scat it
    if (world_rank == 0)
    {
        createAndInitMatrix();

        //MPI_Send(&number, 1, MPI_INT, 1, 0, MPI_COMM_WORLD);
    }

    //
    int row = 0;
    int send_count = (int) (D / (world_size - 1)); //e.g 64/2 = 32
    int receive_count = send_count;

    for (row = 0; row < D; row++)
    {
        float matrix_recieve[D]; //temporary buffer matrix
        //printf("before scatter %d = \n", world_rank);

        MPI_Scatter(matrix[row], send_count,
                    MPI_FLOAT, matrix_recieve, receive_count, MPI_FLOAT,
                    0, MPI_COMM_WORLD);

        //printf("after scatter %d = \n", world_rank);

        if (world_rank != 0)
        {

            int col = 0;
            /*for (col = ((world_rank - 1) * (send_count));
                    col < ((world_rank) * (send_count)); col++)
            {
                matrix_recieve[row] = sqrtf(col * row);
            }*/

            for(col=0; col < send_count; col++)
            {
               //printf("%7.2f|", matrix_recieve[row]);
               matrix_recieve[col] = (float)sqrtf(matrix_recieve[col]);
               //matrix_recieve[row] = (float)sqrt();
            }

        }


        MPI_Gather(matrix_recieve, send_count, MPI_FLOAT,
                   matrix[row], receive_count, MPI_FLOAT,
                   0, MPI_COMM_WORLD);

        //print result in master

    }

    if (world_rank == 0)
    {
        int row2 = 0, col2 = 0;
        for (row2 = 0; row2 < D; row2++)
        {
            for (col2 = 0; col2 < D; col2++)
            {

                printf("%7.2f|", matrix[row2][col2]);
            }
            printf("\n");
        }
    }

    //MPI_Recv(&number, 1, MPI_INT, 0, 0, MPI_COMM_WORLD, MPI_STATUS_IGNORE);
    //free(matrix);
    MPI_Finalize();

    return 0;
}