How do I parallelize a for loop through a C++ std::list using OpenMP?

Question

I would like to iterate through all elements in an std::list in parallel fashion using OpenMP. The loop should be able to alter the elements of the list. Is there a simple s

Answer 1

Here is a solution which allows inserting/removing new elements of a list in parallel.

For a list with N elements we first cut the list into nthreads lists with roughly N/nthreads elements. In a parallel region this can be done like this

int ithread = omp_get_thread_num();
int nthreads = omp_get_num_threads();
int t0 = (ithread+0)*N/nthreads;
int t1 = (ithread+1)*N/nthreads;

std::list l2;
#pragma omp for ordered schedule(static)
for(int i=0; i

Where l2 is the cut list for each thread.

Then we can act on each list in parallel. For example we can insert -1 every first position in the list like this

auto it = l2.begin();
for(int i=(t0+4)/5; i<(t1+4)/5; i++) {
    std::advance(it, 5*i-t0);
    l2.insert(it, -1);
}

Finally, after we are doing operating on the lists in parallel we splice the lists for each thread back to one list in order like this:

#pragma omp for ordered schedule(static)
for(int i=0; i

The algorithm is essentially.

1. Fast-forward through list sequential making cut lists.
2. Act on cut lists in parallel adding, modifying, or removing elements.
3. Splice the modified cut lists back together sequential.

---

Here is a working example

#include 
#include 
#include 
#include 

int main(void) {
  std::list l;
  for(int i=0; i<22; i++) {
    l.push_back(i);
  }
  for (auto it = l.begin(); it != l.end(); ++it) {
    std::cout << *it << " ";
  } std::cout << std::endl;

  int N = l.size();
  #pragma omp parallel
  {
    int ithread = omp_get_thread_num();
    int nthreads = omp_get_num_threads();
    int t0 = (ithread+0)*N/nthreads;
    int t1 = (ithread+1)*N/nthreads;

    //cut list into nthreads lists with size=N/nthreads
    std::list l2;
    #pragma omp for ordered schedule(static)
    for(int i=0; i

Result

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 
-1 0 1 2 3 4 -1 5 6 7 8 9 -1 10 11 12 13 14 -1 15 16 17 18 19 -1 20 21