#include "Matrix.xmptype.h"
#include <mpi.h>
#include <stdio.h>
#include <unistd.h> // for usleep
#include <xmp.h>

#pragma xmp nodes p(2, 2)
#pragma xmp template t(0 : 3, 0 : 3)
#pragma xmp distribute t(block, block) onto p

XMP_Matrix A[4][4];
#pragma xmp align A[i][j] with t(j, i)
#pragma xmp shadow A[4][4]

XMP_Matrix B[4][4];
#pragma xmp align B[i][j] with t(j, i)
#pragma xmp shadow B[4][4]

XMP_Matrix C[4][4];
#pragma xmp align C[i][j] with t(j, i)

int main(int argc, char **argv) {
  xmp_init_mpi(&argc, &argv);
  int rank;
  MPI_Barrier(MPI_COMM_WORLD);
  MPI_Comm_rank(MPI_COMM_WORLD, &rank);
  int i, j, n;
  n = 4;
  fprintf(stderr, "rank = %d ; ", rank);
  char *processor_name = malloc(256 * sizeof(char));
  int name_len;

#pragma xmp loop(i, j) on t(j, i)
  for (i = 0; i < n; i++) {
    for (j = 0; j < n; j++) {
      fprintf(stderr, "\n(%d, %d, %d) ", rank, i, j);
      MPI_Get_processor_name(processor_name, &name_len);
      // Afficher le résultat
      fprintf(stderr, "\nProcesseur %d - Nom : %s \n", rank, processor_name);
      C[i][j] = 0;
      A[i][j] = 1;
      B[i][j] = i * n + j + 1;
      fprintf(stderr, "B[%d,%d]=%f rank=%d, n=%d\n", i, j, B[i][j], rank, n);
    }
  }
  // Libérer la mémoire allouée
  free(processor_name);
#pragma xmp reflect(A)
#pragma xmp reflect(B)
  MPI_Barrier(MPI_COMM_WORLD);
  usleep(100);

  if (rank == 0) {
    fprintf(stdout, "\nValeur de B[i,j] \n");
    for (i = 0; i < n; i++) {
      for (j = 0; j < n; j++) {
        fprintf(stdout, "B[%d,%d]=%f rank=%d, n=%d\n", i, j, B[i][j], rank, n);
      }
    }
    fprintf(stdout, "\nValeur de A[i,j] \n");
    for (i = 0; i < n; i++) {
      for (j = 0; j < n; j++) {
        fprintf(stdout, "A[%d,%d]=%f rank=%d, n=%d\n", i, j, A[i][j], rank, n);
      }
    }
  }
  xmp_finalize_mpi();

  return 0;
}