parallel-processing - OpenMP，使用所有内核并行

Question

我有 4 个内核的计算机和带有 2 个重要任务的 OMP 应用程序。

int main()
{
    #pragma omp parallel sections
    {
        #pragma omp section
        WeightyTask1();

        #pragma omp section
        WeightyTask2();
    }

    return 0;
}

每个任务都有这么重要的部分:

#omp pragma parallel for
for (int i = 0; i < N; i++)
{
    ...
}

我用 -fopenmp 编译程序参数，制作 export OMP_NUM_THREADS=4 .
问题是 仅加载两个内核 .如何在我的任务中使用所有核心？

Answer 1

我最初的 react 是:你必须声明更多的并行性。

您已经定义了两个可以并行运行的任务。 OpenMP 尝试在两个以上的内核上运行它会减慢您的速度(因为缓存局部性和可能的​​错误共享)。

编辑 如果并行 for 循环的数量很大(例如，不低于 8 次迭代)，并且您没有看到使用了 2 个以上的内核，请查看

omp_set_nested()

OMP_NESTED = TRUE | FALSE environment variable

This environment variable enables or disables nested parallelism. The setting of this environment variable can be overridden by calling the omp_set_nested() runtime library function.

If nested parallelism is disabled, nested parallel regions are serialized and run in the current thread.

In the current implementation, nested parallel regions are always serialized. As a result, OMP_SET_NESTED does not have any effect, and omp_get_nested() always returns 0. If -qsmp=nested_par option is on (only in non-strict OMP mode), nested parallel regions may employ additional threads as available. However, no new team will be created to run nested parallel regions. The default value for OMP_NESTED is FALSE.