CountDownLatch源码分析

CountDownLatch源码

这次来分析下CountDownLatch，CountDownLatch用于线程通信的一种方式，实现让一个或多个线程等待某些步骤完成之后再运行。也是比较典型的AQS的实现。

总览CountDownLatch方法

直接把源码全部拉下来看：

/**
 * A synchronization aid that allows one or more threads to wait until
 * a set of operations being performed in other threads completes.
 */
 public class CountDownLatch {
    //静态内部类实现AQS模板
    private static final class Sync extends AbstractQueuedSynchronizer {
        private static final long serialVersionUID = 4982264981922014374L;

        Sync(int count) {
            setState(count);
        }

        int getCount() {
            return getState();
        }

        protected int tryAcquireShared(int acquires) {
            return (getState() == 0) ? 1 : -1;
        }

        protected boolean tryReleaseShared(int releases) {
            // Decrement count; signal when transition to zero
            //CAS实现的释放锁过程
            for (;;) {
                int c = getState();
                if (c == 0)
                    return false;
                int nextc = c-1;
                if (compareAndSetState(c, nextc))
                    return nextc == 0;
            }
        }
    }

    private final Sync sync;//私有类，也就是具体的锁对象

    //创建锁对象，传入参数表示需要等待的线程数。
    public CountDownLatch(int count) {
        if (count < 0) throw new IllegalArgumentException("count < 0");
        this.sync = new Sync(count);
    }

    //关键方法：
    public void await() throws InterruptedException {
        sync.acquireSharedInterruptibly(1);
    }

    //同样的await方法，多提供了
    public boolean await(long timeout, TimeUnit unit)
        throws InterruptedException {
        return sync.tryAcquireSharedNanos(1, unit.toNanos(timeout));
    }

    
    public void countDown() {
        sync.releaseShared(1);
    }

    
    public long getCount() {
        return sync.getCount();
    }

    
    public String toString() {
        return super.toString() + "[Count = " + sync.getCount() + "]";
    }
}

加锁的实现

看下关键方法的实现：

public final void acquireSharedInterruptibly(int arg)
            throws InterruptedException {
        if (Thread.interrupted())//这里先判断当前线程是否被interrupted，底层是native方法
            throw new InterruptedException();
        if (tryAcquireShared(arg) < 0)//这里是核心，调用的是CountDownLatch实现的判断方法，Shared方式，也就是允许节点的nextWriter可写节点，允许state大于1，只有当state为0时返回1。
            doAcquireSharedInterruptibly(arg);
    }
    //带时间的等待，不超过设置的最大等待时间，超过了就被唤起。
    public boolean await(long timeout, TimeUnit unit)
        throws InterruptedException {
        return sync.tryAcquireSharedNanos(1, unit.toNanos(timeout));
    }

下面看看doAcquireSharedInterruptibly的实现：

private void doAcquireSharedInterruptibly(int arg)
        throws InterruptedException {
        final Node node = addWaiter(Node.SHARED);
        boolean failed = true;
        try {
            for (;;) {
                final Node p = node.predecessor();
                if (p == head) {
                //从末尾节点找到头节点，如果尝试获取锁是大于0（这里shared返回的只能是1或-1）的表明，如果是头节点再去尝试获取锁并释放共享相关的节点。
                    int r = tryAcquireShared(arg);
                    if (r >= 0) {
                        setHeadAndPropagate(node, r);
                        p.next = null; // help GC
                        failed = false;
                        return;
                    }
                }
                //获取锁失败，将对应节点挂起，park在parkAndCheckInterrupt方法里，后续被唤起也是从这里开始。
                if (shouldParkAfterFailedAcquire(p, node) &&
                    parkAndCheckInterrupt())
                    throw new InterruptedException();
            }
        } finally {
            if (failed)
                cancelAcquire(node);
        }
    }

解锁的实现

解锁的过程，看看countdown的实现：

public void countDown() {
        sync.releaseShared(1);
    }
    //这里的tryReleaseShared也是一样，共享锁的state--，只有state等于0的时候返回true
    public final boolean releaseShared(int arg) {
        if (tryReleaseShared(arg)) {
            doReleaseShared();
            return true;
        }
        return false;
    }

下面具体释放锁的流程：逻辑就是从队列头开始，找标志位位-1（表示后面还有节点需要释放，置为0）、标志位为0的节点（等待唤起的节点，置为-3）

private void doReleaseShared() {

        for (;;) {
            Node h = head;
            if (h != null && h != tail) {
                int ws = h.waitStatus;
                if (ws == Node.SIGNAL) {
                    if (!compareAndSetWaitStatus(h, Node.SIGNAL, 0))
                        continue;            // loop to recheck cases
                    unparkSuccessor(h);//unpark下一个节点（ws<=0），这里进来的是head节点，shared情况下头节点一般是空节点。
                }
                else if (ws == 0 &&
                         !compareAndSetWaitStatus(h, 0, Node.PROPAGATE))
                    continue;                // loop on failed CAS
            }
            //如果发现没有符合的释放了，就直接退出循环
            if (h == head)                   // loop if head changed
                break;
        }
    }

总结释放过程：

首先将count–，然后判断count是否是0，如果是0就去尝试释放（对于被countdownLatch给awit的节点，会被入队，然后-1表示后续还有共享节点（需要注意有头节点，ws就是Node.SIGNAL也就是-1，也就是整个阻塞队列的ws都在头节点来标识），0表示处理可以释放的节点，也就是队列中会是-1、-1、0来表示3个被阻塞的线程），释放从头开始，找-1、0。

至于为什么需要置为-3,可以看看注释的解释：

/*
         * Ensure that a release propagates, even if there are other
         * in-progress acquires/releases.  This proceeds in the usual
         * way of trying to unparkSuccessor of head if it needs
         * signal. But if it does not, status is set to PROPAGATE to
         * ensure that upon release, propagation continues.
         * Additionally, we must loop in case a new node is added
         * while we are doing this. Also, unlike other uses of
         * unparkSuccessor, we need to know if CAS to reset status
         * fails, if so rechecking.
         */

tips：park的底层

public static void park(Object blocker) {
        Thread t = Thread.currentThread();
        setBlocker(t, blocker);//被当前线程阻碍
        UNSAFE.park(false, 0L);//这里被park住，unpark后继续
        setBlocker(t, null);//释放障碍
    }