LU 分解，采用行连续划分方式下的 MPI 实现

#include "stdio.h"
#include "stdlib.h"
#include "mpi.h"
#define a(x,y) a[x*M+y]
/*A为M*M矩阵*/
#define A(x,y) A[x*M+y]
#define l(x,y) l[x*M+y]
#define u(x,y) u[x*M+y]
#define floatsize sizeof(float)
#define intsize sizeof(int)

int M,N;
int m;
float *A;
int my_rank;
int p;
MPI_Status status;

void fatal(char *message)
{
    printf("%s\n",message);
    exit(1);
}

void Environment_Finalize(float *a,float *f)
{
    free(a);
    free(f);
}

int main(int argc, char **argv)
{
    int i,j,k,my_rank,group_size;
    int i1,i2;
    int v,w;
    float *a,*f,*l,*u;
    FILE *fdA;

    MPI_Init(&argc,&argv);
    MPI_Comm_size(MPI_COMM_WORLD,&group_size);
    MPI_Comm_rank(MPI_COMM_WORLD,&my_rank);

    p=group_size;

    if (my_rank==0)
    {
        fdA=fopen("dataIn.txt","r");
        fscanf(fdA,"%d %d", &M, &N);
        if(M != N)
        {