PWR024: Loop can be rewritten in OpenMP canonical form
Issue
The loop is currently not in OpenMP canonical form but it can be made OpenMP compliant through refactoring.
Actions
Rewrite the loop to make it an OpenMP canonical form using single statements involving only the iterator variable in the loop header.
Relevance
OpenMP requires a loop to fulfill some requirements: it must be a for loop,
have single statements in its loop headers, use only one iterator variable which
must be an integer and use only integer increments. Failing to fulfill any of
these requirements inhibits the loop parallelization with OpenMP. In many cases,
a simple rewrite can fix this.
Code example
The following loop is not in OpenMP canonical form because the iterator is initialized outside the loop:
void example(int **A, int n, int m) {
int i = 0;
for (; i < n; i++) {
A[i] = m;
}
}
This can be easily fixed by moving the iterator initializer inside the for statement:
void example(int **A, int n, int m) {
for (int i = 0; i < n; i++) {
A[i] = m;
}
}