matrix multiplication using parallel threads

I am writing a code for matrix multiplication using multi-threading.

The code works fine without multiple-threads. The code works when i add the line t.join(),but doing this won't parallelise the code.

All the threads will wait for their predecessor to finish.I want to execute them in parallel.

Run-time Error :

libc++abi.dylib: terminating
Abort trap: 6

When i looked up the error in SO mirrors,almost all of them where facing segmentation faults which gave this errror.

Note : I am using a mac machine with gcc(clang).

Here's my code:

void multiply (int a[][N],int b[][N],int n,int m,int p,int result[][N])
{
    int i,j,k;

    for (i=0;i<n;i++)
    {
        for (j=0;j<p;j++)
        {
            thread t(multiplyParallely,a,b,m,i,j,result);

            // t.join(); // Works When i add this line.But that doesnt parallelise.
            //multiplyParallely(a,b,m,i,j,result); // Obviously This Works
        }
    }

}