OpenMP常用指令释义

来源：互联网发布：同城团购用什么软件编辑：程序博客网时间：2024/05/28 06:04

1 #pragma omp for

int j = 0;#pragma omp forfor ( j = 0; j < 4; j++ ){         printf(“j = %d, ThreadId = %d\n”, j, omp_get_thread_num());     }执行以上代码后打印出以下结果j = 0, ThreadId = 0j = 1, ThreadId = 0j = 2, ThreadId = 0j = 3, ThreadId = 0

单独使用for,后续循环在一个线程中执行。可以看出没有形成多线程的效果，所以一般不推荐使用。

2 #pragma omp parallel

int j = 0;#pragma omp parallel forfor ( j = 0; j < 4; j++ ){         printf(“j = %d, ThreadId = %d\n”, j, omp_get_thread_num());     }执行后会打印出以下结果：j = 0, ThreadId = 0j = 2, ThreadId = 2j = 1, ThreadId = 1j = 3, ThreadId = 3

区域并行化指导语句，说明后续语句按多线程方式运行，线程数默认为CPU线程数，也可以用num_threads()自行指定。

3 #pragma omp parallel for

#pragma omp parallel{#pragma omp forfor ( j = 0; j < 4; j++ ){         printf(“j = %d, ThreadId = %d\n”, j, omp_get_thread_num());     }}执行以上代码会打印出以下结果：j = 1, ThreadId = 1j = 3, ThreadId = 3j = 2, ThreadId = 2j = 0, ThreadId = 0

循环并行化指导语句，说明后续循环按多线程执行，线程数与循环次数相等。

0 0