我已经在单个节点上成功编程了矩阵矩阵乘法，现在我的目标是链接该程序以在集群节点上并行执行。
主要工作是用我的程序（dgemm_来部分地计算矩阵矩阵乘法（在这种情况下为mydgemm）中，更改Scalapack Netlib源代码中的代码，并更改原始代码（ScaLAPACK）。 ）。
在这里，原始代码是C程序，但是该程序中的所有例程都调用Fortran例程（例如dgemm_是Fortran语言），而我的程序（mydgemm）是C程序。
修改后，我可以使用任意大小的矩阵的单个节点成功执行，但是当我使用4个节点（矩阵的大小大于200）运行时->关于节点之间的通信数据有错误（MPI）。
这是一个错误：
*BAD TERMINATION OF ONE OF YOUR APPLICATION PROCESSES
PID 69754 RUNNING AT localhost.localdomain
EXIT CODE: 11
CLEANING UP REMAINING PROCESSES
YOU CAN IGNORE THE BELOW CLEANUP MESSAGES*
我只在主函数中使用MPI在每个节点上创建随机矩阵（后面附有）-例程称为new_pdgemm (...)。 （我在new-pdgemm内修改了代码）。
在mydgemm.c内部，我使用OMP进行并行处理，并且此代码在内核上执行。
能给我一个指导或想法来解决我的问题吗？
您是否认为此问题是因为Fortran是列主行，而C是行主行？
还是我需要将mydgemm.c更改为mydgemm.f（这真的很难，也许我做不到）？
我的代码：
int main(int argc, char **argv) {
int i, j, k;
/************ MPI ***************************/
int myrank_mpi, nprocs_mpi;
MPI_Init( &argc, &argv);
MPI_Comm_rank(MPI_COMM_WORLD, &myrank_mpi);
MPI_Comm_size(MPI_COMM_WORLD, &nprocs_mpi);
/************ BLACS ***************************/
int ictxt, nprow, npcol, myrow, mycol,nb;
int info,itemp;
int _ZERO=0,_ONE=1;
int M=20000;
int K=20000;
int N=20000;
nprow = 2; npcol = 2;
nb=1200;
Cblacs_pinfo( &myrank_mpi, &nprocs_mpi ) ;
Cblacs_get( -1, 0, &ictxt );
Cblacs_gridinit( &ictxt, "Row", nprow, npcol );
Cblacs_gridinfo( ictxt, &nprow, &npcol, &myrow, &mycol );
//printf("myrank = %d\n",myrank_mpi);
int rA = numroc_( &M, &nb, &myrow, &_ZERO, &nprow );
int cA = numroc_( &K, &nb, &mycol, &_ZERO, &npcol );
int rB = numroc_( &K, &nb, &myrow, &_ZERO, &nprow );
int cB = numroc_( &N, &nb, &mycol, &_ZERO, &npcol );
int rC = numroc_( &M, &nb, &myrow, &_ZERO, &nprow );
int cC = numroc_( &N, &nb, &mycol, &_ZERO, &npcol );
double *A = (double*) malloc(rA*cA*sizeof(double));
double *B = (double*) malloc(rB*cB*sizeof(double));
double *C = (double*) malloc(rC*cC*sizeof(double));
int descA[9],descB[9],descC[9];
descinit_(descA, &M, &K, &nb, &nb, &_ZERO, &_ZERO, &ictxt, &rA, &info);
descinit_(descB, &K, &N, &nb, &nb, &_ZERO, &_ZERO, &ictxt, &rB, &info);
descinit_(descC, &M, &N, &nb, &nb, &_ZERO, &_ZERO, &ictxt, &rC, &info);
double alpha = 1.0; double beta = 1.0;
double start, end, flops;
srand(time(NULL)*myrow+mycol);
#pragma simd
for (j=0; j<rA*cA; j++)
{
A[j]=((double)rand()-(double)(RAND_MAX)*0.5)/(double)(RAND_MAX);
// printf("A in myrank: %d\n",myrank_mpi);
}
// printf("A: %d\n",myrank_mpi);
#pragma simd
for (j=0; j<rB*cB; j++)
{
B[j]=((double)rand()-(double)(RAND_MAX)*0.5)/(double)(RAND_MAX);
}
#pragma simd
for (j=0; j<rC*cC; j++)
{
C[j]=((double)rand()-(double)(RAND_MAX)*0.5)/(double)(RAND_MAX);
}
MPI_Barrier(MPI_COMM_WORLD);
start=MPI_Wtime();
new_pdgemm ("N", "N", &M , &N , &K , &alpha, A , &_ONE, &_ONE , descA , B , &_ONE, &_ONE , descB , &beta , C , &_ONE, &_ONE , descC );
MPI_Barrier(MPI_COMM_WORLD);
end=MPI_Wtime();
if (myrow==0 && mycol==0)
{
flops = 2 * (double) M * (double) N * (double) K / (end-start) / 1e9;
/* printf("This is value: %d\t%d\t%d\t%d\t%d\t%d\t\n",rA,cA,rB,cB,rC,cC);
printf("%f\t%f\t%f\n", A[4], B[6], C[3]);*/
printf("%f Gflops\n", flops);
}
Cblacs_gridexit( 0 );
MPI_Finalize();
free(A);
free(B);
free(C);
return 0;
}
答案 0 :(得分：0)
好的，这并不是真正的答案，但是评论太长了，我希望答案能为您提供格式。
因此，我使用注释中提到的blacs_gridexit修复了该错误，即根据例程描述的要求设置了ictxt参数。然后，我用标准的pdgemm替换了您的例程。一旦进行了这些更改，并将矩阵尺寸减小到2,000 * 2,000，以适合我的笔记本电脑。这样，代码就可以成功运行，至少在某种意义上说它没有报告错误，并且给出了一种合理的GFlopage。所以这向我暗示了
因此，我将重新安装您正在使用的库，确保它们与您正在使用的编译器一致，并运行库提供的测试，并且还包括在代码中省略的头文件（DON 'T，这些非常重要！）。如果这些工作正常，我建议这是由于您的代码中的错误所致。您无法显示此图片的原因是什么？
无论如何，下面是我成功运行的代码。如果我在自己的代码中正确执行了此操作，那么我肯定会通过在调用函数时确保适当的原型在范围内来修复所有这些编译器警告。
ian-admin@agon ~/work/stack/mpi $ cat stack.c
#include <stdio.h>
#include <stdlib.h>
#include <time.h>
#include "mpi.h"
int main(void) {
int i, j, k;
/************ MPI ***************************/
int myrank_mpi, nprocs_mpi;
MPI_Init( NULL, NULL);
MPI_Comm_rank(MPI_COMM_WORLD, &myrank_mpi);
MPI_Comm_size(MPI_COMM_WORLD, &nprocs_mpi);
/************ BLACS ***************************/
int ictxt, nprow, npcol, myrow, mycol,nb;
int info,itemp;
int _ZERO=0,_ONE=1;
int M=2000;
int K=2000;
int N=2000;
nprow = 2; npcol = 2;
nb=1200;
Cblacs_pinfo( &myrank_mpi, &nprocs_mpi ) ;
Cblacs_get( -1, 0, &ictxt );
Cblacs_gridinit( &ictxt, "Row", nprow, npcol );
Cblacs_gridinfo( ictxt, &nprow, &npcol, &myrow, &mycol );
//printf("myrank = %d\n",myrank_mpi);
int rA = numroc_( &M, &nb, &myrow, &_ZERO, &nprow );
int cA = numroc_( &K, &nb, &mycol, &_ZERO, &npcol );
int rB = numroc_( &K, &nb, &myrow, &_ZERO, &nprow );
int cB = numroc_( &N, &nb, &mycol, &_ZERO, &npcol );
int rC = numroc_( &M, &nb, &myrow, &_ZERO, &nprow );
int cC = numroc_( &N, &nb, &mycol, &_ZERO, &npcol );
double *A = (double*) malloc(rA*cA*sizeof(double));
double *B = (double*) malloc(rB*cB*sizeof(double));
double *C = (double*) malloc(rC*cC*sizeof(double));
int descA[9],descB[9],descC[9];
descinit_(descA, &M, &K, &nb, &nb, &_ZERO, &_ZERO, &ictxt, &rA, &info);
descinit_(descB, &K, &N, &nb, &nb, &_ZERO, &_ZERO, &ictxt, &rB, &info);
descinit_(descC, &M, &N, &nb, &nb, &_ZERO, &_ZERO, &ictxt, &rC, &info);
double alpha = 1.0; double beta = 1.0;
double start, end, flops;
srand(time(NULL)*myrow+mycol);
#pragma simd
for (j=0; j<rA*cA; j++)
{
A[j]=((double)rand()-(double)(RAND_MAX)*0.5)/(double)(RAND_MAX);
// printf("A in myrank: %d\n",myrank_mpi);
}
// printf("A: %d\n",myrank_mpi);
#pragma simd
for (j=0; j<rB*cB; j++)
{
B[j]=((double)rand()-(double)(RAND_MAX)*0.5)/(double)(RAND_MAX);
}
#pragma simd
for (j=0; j<rC*cC; j++)
{
C[j]=((double)rand()-(double)(RAND_MAX)*0.5)/(double)(RAND_MAX);
}
MPI_Barrier(MPI_COMM_WORLD);
start=MPI_Wtime();
pdgemm_ ("N", "N", &M , &N , &K , &alpha, A , &_ONE, &_ONE , descA , B , &_ONE, &_ONE , descB , &beta , C , &_ONE, &_ONE , descC );
MPI_Barrier(MPI_COMM_WORLD);
end=MPI_Wtime();
if (myrow==0 && mycol==0)
{
flops = 2 * (double) M * (double) N * (double) K / (end-start) / 1e9;
/* printf("This is value: %d\t%d\t%d\t%d\t%d\t%d\t\n",rA,cA,rB,cB,rC,cC);
printf("%f\t%f\t%f\n", A[4], B[6], C[3]);*/
printf("%f Gflops\n", flops);
}
Cblacs_gridexit( ictxt );
MPI_Finalize();
free(A);
free(B);
free(C);
return 0;
}
ian-admin@agon ~/work/stack/mpi $ mpicc -g stack.c /home/ian-admin/Downloads/scalapack-2.0.2/libscalapack.a -llapack -lblas -lgfortran
stack.c: In function ‘main’:
stack.c:24:4: warning: implicit declaration of function ‘Cblacs_pinfo’ [-Wimplicit-function-declaration]
Cblacs_pinfo( &myrank_mpi, &nprocs_mpi ) ;
^~~~~~~~~~~~
stack.c:25:4: warning: implicit declaration of function ‘Cblacs_get’ [-Wimplicit-function-declaration]
Cblacs_get( -1, 0, &ictxt );
^~~~~~~~~~
stack.c:26:4: warning: implicit declaration of function ‘Cblacs_gridinit’ [-Wimplicit-function-declaration]
Cblacs_gridinit( &ictxt, "Row", nprow, npcol );
^~~~~~~~~~~~~~~
stack.c:27:4: warning: implicit declaration of function ‘Cblacs_gridinfo’ [-Wimplicit-function-declaration]
Cblacs_gridinfo( ictxt, &nprow, &npcol, &myrow, &mycol );
^~~~~~~~~~~~~~~
stack.c:31:13: warning: implicit declaration of function ‘numroc_’ [-Wimplicit-function-declaration]
int rA = numroc_( &M, &nb, &myrow, &_ZERO, &nprow );
^~~~~~~
stack.c:44:6: warning: implicit declaration of function ‘descinit_’ [-Wimplicit-function-declaration]
descinit_(descA, &M, &K, &nb, &nb, &_ZERO, &_ZERO, &ictxt, &rA, &info);
^~~~~~~~~
stack.c:72:5: warning: implicit declaration of function ‘pdgemm_’ [-Wimplicit-function-declaration]
pdgemm_ ("N", "N", &M , &N , &K , &alpha, A , &_ONE, &_ONE , descA , B , &_ONE, &_ONE , descB , &beta , C , &_ONE, &_ONE , descC );
^~~~~~~
stack.c:83:4: warning: implicit declaration of function ‘Cblacs_gridexit’ [-Wimplicit-function-declaration]
Cblacs_gridexit( ictxt );
^~~~~~~~~~~~~~~
/usr/bin/ld: warning: libgfortran.so.3, needed by //usr/lib/liblapack.so, may conflict with libgfortran.so.5
ian-admin@agon ~/work/stack/mpi $ mpirun -np 4 --oversubscribe ./a.out
9.424291 Gflops