Scheduling
C
The structure omp_sched_t represents a schedule that can be passed to omp_set_schedule, or retrieved from omp_get_schedule, as well as the modifier that can be applied to that schedule. Its possible values are omp_sched_static, omp_sched_dynamic, omp_sched_auto, omp_sched_guided and omp_sched_monotonic. The first four correspond go the static, dynamic, auto and guided schedules respectively. The last one corresponds to the monotonic modifier.
#include <stdio.h>
#include <stdlib.h>
#include <omp.h>
/**
* @brief Illustrates the use of the omp_sched_t.
* @details A for loop is parallelised across 4 threads using the runtime
* schedule policy. The schedule to apply is defined by omp_set_schedule as
* being a dynamic scheduling with chunks made of 4 iterations.
**/
int main(int argc, char* argv[])
{
// Use 4 threads when creating OpenMP parallel regions
omp_set_num_threads(4);
// Tell OpenMP which scheduling policy and chunk size use for runtime schedule clauses
omp_sched_t kind = omp_sched_dynamic;
omp_set_schedule(kind, 2);
// Parallelise the for loop using the runtime schedule
#pragma omp parallel for schedule(runtime)
for(int i = 0; i < 10; i++)
{
printf("Thread %d processes iteration %d.\n", omp_get_thread_num(), i);
}
return EXIT_SUCCESS;
}