Look at increasing stacksize linking parameter on windows.
_____
From: lam-bounces_at_[hidden] [mailto:lam-bounces_at_[hidden]] On Behalf Of
Yu-Cheng Chou
Sent: Friday, December 16, 2005 2:55 PM
To: lam_at_[hidden]
Subject: LAM: execute a large-scale matrix multiplication program
hi,
I have a matrix multiplication program to be run on multiple machines.
The program works fine with small size matrices, but doesn't work with large
size matrices.
Besides, the working size of matrix is much smaller for windows than the one
for linux.
Can anyone tell me the possible reason that causes this error?
Below shows my program C source code.
#include <mpi.h>
#include <stdio.h>
#define NRA 3000 // number of rows in matrix A
#define NCA 2000 // number of columns in matrix A
#define NCB 400 // number of columns in matrix B
#define MASTER 0 // taskid of first task
#define FROM_MASTER 1 // setting a message type
#define FROM_WORKER 2 // setting a message type
int main(int argc, char *argv[]) {
int numtasks, // number of tasks
taskid, // task identifier
source, // task id of message source
dest, // task id of message destination
mtype, // message type
rows, // rows of matrix A sent to each worker
averow, extra, offset, // used to determine rows sent to each
worker
i, j, k; // misc
double a[NRA][NCA], // matrix A to be multiplied
b[NCA][NCB], // matrix B to be multiplied
c[NRA][NCB]; // result matrix C
MPI_Status status;
printf("debug_1\n"); // the execution did not come to this
point
MPI_Init(&argc, &argv);
MPI_Comm_size(MPI_COMM_WORLD, &numtasks);
MPI_Comm_rank(MPI_COMM_WORLD, &taskid);
if(taskid == MASTER) {
printf("Number of worker tasks = %d\n", numtasks-1);
// initialize matrix a and matrix b
for(i=0; i<NRA; i++) {
for(j=0; j<NCA; j++) {
a[i][j] = i+j;
}
}
for(i=0; i<NCA; i++) {
for(j=0; j<NCB; j++) {
b[i][j] = i*j;
}
}
// send matrix data to worker processes
averow = NRA/(numtasks-1);
extra = NRA%(numtasks-1);
offset = 0;
mtype = FROM_MASTER;
for(dest=1; dest<numtasks; dest++) {
rows = (dest <= extra) ? averow+1 : averow;
printf(" sending %d rows to task %d\n", rows, dest);
MPI_Send(&offset, 1, MPI_INT, dest, mtype, MPI_COMM_WORLD);
MPI_Send(&rows, 1, MPI_INT, dest, mtype, MPI_COMM_WORLD);
MPI_Send(&a[offset][0], rows*NCA, MPI_DOUBLE, dest, mtype,
MPI_COMM_WORLD);
MPI_Send(&b, NCA*NCB, MPI_DOUBLE, dest, mtype, MPI_COMM_WORLD);
offset = offset + rows;
}
}
else {
// receive matrix data from master process
mtype = FROM_MASTER;
MPI_Recv(&offset, 1, MPI_I! NT, MASTER, mtype, MPI_COMM_WORLD,
&status);
MPI_Recv(&rows, 1, MPI_INT, MASTER, mtype, MPI_COMM_WORLD, &status);
MPI_Recv(&a, rows*NCA, MPI_DOUBLE, MASTER, mtype, MPI_COMM_WORLD,
&status);
MPI_Recv(&b, NCA*NCB, MPI_DOUBLE, MASTER, mtype, MPI_COMM_WORLD,
&status);
}
if(taskid == MASTER) {
// receive results from worker processes
mtype = FROM_WORKER;
for(source=1; source<numtasks; source++) {
MPI_Recv(&offset, 1, MPI_INT, source, mtype, MPI_COMM_WORLD,
&status);
MPI_Recv(&rows! , 1, MPI_INT, source, mtype, MPI_COMM_WORLD,
&status);
MPI_Recv(&c[offset][0], rows*NCB, MPI_DOUBLE, source, mtype,
MPI_COMM_WORLD, &status);
}
}
else {
// matrix calculation
for(k=0; k<NCB; k++) {
for(i=0; i<rows; i++) {
c[i][k] = 0.0;
for(j=0; j<NCA; j++) {
c[i][k] = c[i][k] + a[i][j]*b[j][k];
}
}
}
// send results to master process
mtype = FROM_WORKER;
MPI_Send(&offset, 1, MPI_INT, MASTER, mtype, MPI_COMM_WORLD);
MPI_Send(&rows, 1, MPI_INT, MASTER, mtype, MPI_COMM_WORLD);
MPI_Send(&c, rows*NCB, MPI_DOUBLE, MASTER, mtype, MPI_COMM_WORLD);
}
if(taskid == MASTER) {
printf("Here is the result matrix\n\n");
/*
for(i=0; ! i<NRA; i++) {
for(j=0; j<NCB; j++) {
printf("%6.2f ", c[i][j]);
}
printf("\n");
}
*/
}
MPI_Finalize();
return 0;
}
|
