OpenMP* Examples

The following examples show how to use the OpenMP* features.

A Simple Difference Operator

This example shows a simple parallel loop where the amount of work in each iteration is different. Dynamic scheduling is used to get good load balancing. The for has a nowait because there is an implicit barrier at the end of the parallel region.

Example

void for1(float a[], float b[], int n)

{

  int i, j;

  #pragma omp parallel shared(a,b,n) private(i,j)

  {

    #pragma omp for schedule(dynamic,1) nowait

    for (i = 1; i < n; i++)

      for (j = 0; j <= i; j++)

        b[j + n*i] = (a[j + n*i] + a[j + n*(i-1)]) / 2.0;

  }

}

Two Difference Operators

This example uses two parallel loops fused to reduce fork/join overhead. The first for has a nowait because all the data used in the second loop is different than all the data used in the first loop.

Example

void for2(float a[], float b[], float c[], float d[],

          int n, int m)

{

  int i, j;

  #pragma omp parallel shared(a,b,c,d,n,m) private(i,j)

  {

    #pragma omp for schedule(dynamic,1) nowait

    for (i = 1; i < n; i++)

      for (j = 0; j <= i; j++)

        b[j + n*i] = ( a[j + n*i] + a[j + n*(i-1)] )/2.0;

    #pragma omp for schedule(dynamic,1) nowait

    for (i = 1; i < m; i++)

      for (j = 0; j <= i; j++)

        d[j + m*i] = ( c[j + m*i] + c[j + m*(i-1)] )/2.0;

  }

}

See more examples in the OpenMP C/C++ version 2.5 specifications (http://www.openmp.org/specs).