Question

I have a c program as below:

int a[10];
int b;

for(int i = 0; i < 10; i++)

     function1(a[i]);
function1(b);

Now I want to parallelize all these 11 calls of function function1(). How can I do this using openmp?

I have tried

#pragma omp parallel sections
{
#pragma omp section
#pragmal omp parallel for
for(int i = 0; i < 10; i++)
     function1(a[i]);
#pragma omp section
function1(b);
}

But the above code doesn’t seem to work.
EDIT: Please read function1(b) as some different function, ie function2(b).

Answer 1

A simple way, that doesn’t depend on OpemMP, is to add b to the a array.
This way, you have a single loop to parallelize.
Just make a 11 ints long, and put the value of b in the last one.

In a more general case (assuming the members of a are not integers, but something larger), you may want to change function1 to get a pointer. Then build another array, of 11 pointers. Set 10 to point to cells of a, the last to b.

In an even more general case, the function called for b is a different one (possibly with entirely different parameters). In this case, you can still use one loop:

for (i=0; i<11; i++) {
    if (i<10) {
        function1(a[i]);
    } else {
        function2(b);
    }
}