在使用标准库中的加锁机制时,例如我们使用std::mutex,写了如下的代码(下面的代码使用condition_variable可能更合适)
std::mutex g_mtx; int g_resNum; // and how to get the resource ... // 线程1使用如下代码: { std::lock_guard<std::mutex> guard(g_mtx); // prepare the resource ... // set the resource num g_resNum = ...; } // 线程2使用如下的代码: { std::lock_guard<std::mutex> guard(g_mtx); // check the g_resNum if (g_resNum > 0) { ... } }
我们知道如果先运行线程1的代码,也就是说线程1获得了锁,进行了准备资源的操作。然后运行线程2,也就是说线程2之后获得了锁,可以知道线程1加锁区域的代码已经执行完毕,而且对线程2可见,我们可以放心的使用线程1中设置的资源。这里面的原因是
标准库在实现std::mutex过程中使用了memory_order。http://en.cppreference.com/w/cpp/atomic/memory_order 中的描述如下:
Acquire operation
Atomic load with memory_order_acquire or stronger is an acquire operation. The lock() operation on a Mutex is also an acquire operation.
Release operation
Atomic store with memory_order_release or stronger is a release operation. The unlock() operation on a Mutex is also a release operation.
C++ atomic中对Acquire- Release ordering的简单介绍如下:
If an atomic store in thread A is tagged memory_order_release and an atomic load in thread B from the same variable is tagged memory_order_acquire, all memory writes (non-atomic and relaxed atomic) that happened-before the atomic store from the point of view of thread A, become visible side-effects in thread B, that is, once the atomic load is completed, thread B is guaranteed to see everything thread A wrote to memory.
关于memory_order的简单介绍就到这里,下面我用一个很简单的spinlock的实现来简单说明一下memory_order的使用:
class spinlock_mutex { std::atomic_flag flag; public: spinlock_mutex() : flag(ATOMIC_FLAG_INIT) {} void lock() { while(flag.test_and_set(std::memory_order_acquire)); } bool try_lock() { if (flag.test_and_set(std::memory_order_acquire)) return false; return true; } void unlock() { flag.clear(std::memory_order_release); } };
上述spinlock实现存在性能上的问题,网上有关于这个实现问题的讨论,参考网址:https://www.zhihu.com/question/55764216
希望能给初学C++标准库多线程的程序员有所启示。