本文共 2160 字，大约阅读时间需要 7 分钟。
基本思想：OpenMP进行加减乘除运算的规约处理 reduction 可以支持的运算符号 +、-、*、/、&&、|、^、|| 符号
#pragma omp parallel for reduction(+:sum) for(int i=0;i
测试代码
#include #include #include using namespace std;using namespace chrono;void sequentialProgram(int num){ int sum=0; for(int i=0;i (end_time-start_time).count()<<" seconds"< (end_time-start_time).count()<<" seconds"<
测试结果i 可以看出 在增加归约之后，结果和串行程序保持一致，随着叠加次数增加，其时间明显比串行时间的少
F:\OpenMP\cmake-build-debug\OpenMP.exesum=0 the current thread id: 0sum=1 the current thread id: 0sum=3 the current thread id: 0sum=6 the current thread id: 0sum=10 the current thread id: 0sum=15 the current thread id: 0sum=21 the current thread id: 0sum=28 the current thread id: 0sum=36 the current thread id: 0sum=45 the current thread id: 0sum=55 the current thread id: 0sum=66 the current thread id: 0sum=66sequentialProgram elapse time: 0.0376821 secondsi=1 sum=1 the current thread id: 1i=8 sum=8 the current thread id: 8i=9 sum=9 the current thread id: 9i=4 sum=4 the current thread id: 4i=2 sum=2 the current thread id: 2i=7 sum=7 the current thread id: 7i=10 sum=10 the current thread id: 10i=6 sum=6 the current thread id: 6i=5 sum=5 the current thread id: 5i=0 sum=0 the current thread id: 0i=3 sum=3 the current thread id: 3i=11 sum=11 the current thread id: 11 sum=66parallelProgram elapse time: 0.0163177 secondsProcess finished with exit code 0
同样的逻辑代码
shared 对一个变量的同一个内存进行读写
#pragma omp parallel for shared(sum) for(int i=0;i
simd 编译器可以忽略矢量依赖关系，使循环尽可能实现矢量友好，并尊重用户对同时执行多个循环迭代的意图。
#pragma omp simdfor(int i=0;i
测试代码
#include #include #include using namespace std;using namespace chrono;static omp_lock_t m_lock;void sequentialProgram(int num){ int sum=0; for(int i=0;i (end_time-start_time).count()<<" seconds"< (end_time-start_time).count()<<" seconds"<
测试结果
F:\OpenMP\cmake-build-debug\OpenMP.exesum=66sequentialProgram elapse time: 0.0011415 seconds sum=66 sum=66 sum=66 sum=66 sum=66 sum=66parallelProgram elapse time: 0.0028554 secondsProcess finished with exit code 0
转载地址：http://ntyai.baihongyu.com/