[英]Segmentation fault in matrix multiplication using multiple threads
#include <stdio.h>
#include <pthread.h>
int arr[1000][1000];
int brr[1000][1000];
int h;
int f;
void *BMM(void *arg)
{
int* neo = (int*) arg;
int ne = *neo;
int sum = 0;
for(int i = 0; i < n; ++i)
{
sum += arr[x][i]*brr[x][f];
++f;
}
printf("%d\n", sum);
crr[x][h] = sum;
pthread_exit(NULL);
}
int main()
{
pthread_t* ar = malloc(3*sizeof(*ar));
printf("Enter the value of m and n\n");
scanf("%d %d",&m,&n);
for(int i = 0; i < m; ++i)
{
for(int j = 0; j < n; ++j)
{
scanf("%d",&arr[i][j]);
}
}
printf("Enter the value of p and q\n");
scanf("%d %d",&p,&q);
if(p != n)
{
printf("The matrix multiplication is not possible\n");
return 0;
}
int* id;
id = (int *)malloc(4*sizeof(int));
for(int i = 0; i < p; ++i)
{
for(int j = 0; j < q; ++j)
{
scanf("%d",&brr[i][j]);
}
}
for(x = 0; x < m; ++x)
{
for(z = 0; z < q; z+=4)
{
f = z;
h = z;
for(int k = 0; k < 3; ++k)
{
pthread_create(&ar[k],NULL,BMM,NULL);
}
for(int k = 0; k < 3; ++k)
{
pthread_join(ar[k],NULL);
}
}
}
for (int i = 0; i < m; ++i)
{
for(int j = 0; j < q; ++j)
{
printf("%d ",crr[i][j]);
}
printf("\n");
}
}
The above program is supposed to multiply two matrix by multiplying row one of matrix by all the columns of other matrix using 3 threads and then row two by all the other columns and so on and then store the respective values int another matrix but it is giving segmentation fault. 上面的程序假设通过使用3个线程将矩阵的第一行与其他矩阵的所有列相乘,然后将第二行与所有其他列相乘,以此类推,然后将相应的值存储到另一个矩阵中,从而将两个矩阵相乘,然后给出分段故障。 Where am I going wrong? 我要去哪里错了?
I think your problem is here: 我认为您的问题在这里:
pthread_create(&ar[k],NULL,BMM,NULL);
^^^^
void *arg is NULL
and then: 接着:
void *BMM(void *arg)
{
int* neo = (int*) arg;
int ne = *neo; // Dereference NULL --> segmentation fault
Further this looks strange: 此外,这看起来很奇怪:
void *BMM(void *arg)
{
int* neo = (int*) arg;
int ne = *neo; // ne is never used !!
int sum = 0;
for(int i = 0; i < n; ++i) // Where does n come from ?
Perhaps it should be n
instead of ne
? 也许应该是n
而不是ne
?
If n
, x
, f
and h
are global variables you are into trouble as all threads will work on the same variables. 如果n
, x
, f
和h
是全局变量,那么您会遇到麻烦,因为所有线程都将在相同的变量上工作。 That would be real bad. 那真是太糟糕了。 Each thread needs it own variables. 每个线程都需要自己的变量。
BTW: 顺便说一句:
Always check the value returned by scanf
- something like: 始终检查scanf
返回的值-类似于:
if (scanf("%d %d",&m,&n) != 2)
{
// Add error handling here
}
and 和
if (scanf("%d",&arr[i][j]) != 1)
{
// Add error handling here
}
声明:本站的技术帖子网页,遵循CC BY-SA 4.0协议,如果您需要转载,请注明本站网址或者原文地址。任何问题请咨询:yoyou2525@163.com.