atomic 内存序

提高性能的一种方式  atomic 替换mutex

内存序跟原子操作本身并不冲突, 不论用那种内存序该原子还是原子操作

atomic内存序: 就是来指定顺序的

例子:

注 : 像类似 memory_order_relaxed 都是以函数为单位来重新排执行指令的 [这个注释后面再来看好了]

#include <atomic>
atomic<int> a (0);
atomic<int> b(0);
unsigned int __stdcall thread_write(void *ar){
    int hahah = 123;                             //1
    int balala = hahah;                         //2
    a.store(1,std::memory_order_relaxed);        //3
    b.store(666,std::memory_order_relaxed);        //4
    bool ooo = false;                         //7
    return 0;
}
unsigned int __stdcall thread_read(void * arg){

    while(b.load(std::memory_order_relaxed) != 666); //5
    int r = a.load(std::memory_order_relaxed);        //6
    cout << "a : " << r << endl;
    assert( r ==1 );

    return 0;
}
int m