Java嵌套同步

Question

我无法理解嵌套同步的工作原理。我这里举两个例子做个对比:

//example 1
public class Account {
    int balance = 1000;
    Object lock = new Object();

    public int getBalance() throws InterruptedException {
        synchronized(lock) {
            // step 1- Thread 1 enter here and sleep
            Thread.sleep(20);
            return balance;
        }
    }
    public void setBalance(int amount) {
        // step 2 - Thread 2 have to wait until thread 1 release the lock.
        synchronized(lock) {
            this.balance = amount;
        }
    }
}

上面的例子很清楚也很合乎逻辑。
现在看例子2:

public class Account {
    int balance = 1000;
    Object lock = new Object();

    public int getBalance() {
        synchronized(lock) {
            // step 1 - Thread 1 enter here and lock the object.
            synchronized(lock) { //same lock 
                //step 2 - Thread 1 can enter here also if the lock is already locked why????

                //...do something...
                return balance;
            }
        }
    }
}

在例子2中我不明白，如果外层锁已经被锁定，为什么同一个线程可以进入同一个锁2次...

Answer 1

在这种情况下，内部锁不起作用。 Java 使用 recursive mutexes ，因此在给定互斥锁上持有锁的线程可以再次锁定它，并保持锁定计数。只有当最后一个 synchronized block 退出时，互斥体才会真正解锁。