43 lines
1.2 KiB
C++
43 lines
1.2 KiB
C++
|
#include "sort.hpp"
|
||
|
|
#include <omp.h>
|
||
|
|
|
||
|
|
// OMP parallel sort
|
||
|
|
// https://cw.fel.cvut.cz/old/_media/courses/b4m35pag/lab6_slides_advanced_openmp.pdf
|
||
|
|
template <typename V>
|
||
|
|
void mergeSortRecursive(V& v, size_t left, size_t right) {
|
||
|
|
if (left < right) {
|
||
|
|
if (right-left >= 32) {
|
||
|
|
size_t mid = (left+right)/2;
|
||
|
|
#pragma omp taskgroup
|
||
|
|
{
|
||
|
|
#pragma omp task shared(v) untied if(right-left >= (1<<14))
|
||
|
|
mergeSortRecursive(v, left, mid);
|
||
|
|
#pragma omp task shared(v) untied if(right-left >= (1<<14))
|
||
|
|
mergeSortRecursive(v, mid+1, right);
|
||
|
|
#pragma omp taskyield
|
||
|
|
}
|
||
|
|
std::inplace_merge(v.begin()+left, v.begin()+mid+1, v.begin()+right+1);
|
||
|
|
}
|
||
|
|
else {
|
||
|
|
std::sort(v.begin()+left, v.begin()+right+1);
|
||
|
|
}
|
||
|
|
}
|
||
|
|
}
|
||
|
|
|
||
|
|
// sort_omp
|
||
|
|
void sort_omp(size_t nthreads) {
|
||
|
|
#pragma omp parallel num_threads(nthreads)
|
||
|
|
{
|
||
|
|
#pragma omp single
|
||
|
|
mergeSortRecursive(vec, 0, vec.size()-1);
|
||
|
|
}
|
||
|
|
}
|
||
|
|
|
||
|
|
std::chrono::microseconds measure_time_omp(size_t num_threads) {
|
||
|
|
auto beg = std::chrono::high_resolution_clock::now();
|
||
|
|
sort_omp(num_threads);
|
||
|
|
auto end = std::chrono::high_resolution_clock::now();
|
||
|
|
return std::chrono::duration_cast<std::chrono::microseconds>(end - beg);
|
||
|
|
}
|
||
|
|
|