一、LinkedBlockingQueue

1.1 简介

LinkedBlockingQueue是一个由链表结构组成的有界阻塞队列，此队列是FIFO(先进先出)的顺序来访问的，它由队尾插入后再从队头取出或移除，其中队列的头部是在队列中时间最长的元素，队列的尾部是在队列中时间最短的元素。在LinkedBlockingQueue类中分别用2个不同的锁takeLock、putLock来保护队头和队尾操作。如下图所示：

1.2 类图

1.3 源码分析

1.3.1 属性与链表节点类

//链表节点类，next指向下一个节点。如果下一个节点时null表示没有节点了。
static class Node<E> {
    E item;

    Node<E> next;

    Node(E x) { item = x; }
}

// 最大容量上限，默认是 Integer.MAX_VALUE
private final int capacity;

// 当前元素数量，这是个原子类。
private final AtomicInteger count = new AtomicInteger(0);

// 头结点
private transient Node<E> head;

// 尾结点
private transient Node<E> last;

// 队头访问锁
private final ReentrantLock takeLock = new ReentrantLock();

// 队头访问等待条件、队列
private final Condition notEmpty = takeLock.newCondition();

// 队尾访问锁
private final ReentrantLock putLock = new ReentrantLock();

// 队尾访问等待条件、队列
private final Condition notFull = putLock.newCondition();

使用原子类AtomicInteger是因为读写分别使用了不同的锁，但都会访问这个属性来计算队列中元素的数量，所以它需要是线程安全的。关AtomicInteger详细请看我的这一篇文章：【Java并发编程】深入分析AtomicInteger(二)

1.3.2 offer操作

public boolean offer(E e) {
	if (e == null) throw new NullPointerException();
	final AtomicInteger count = this.count;
	//当队列满时，直接返回了false，没有被阻塞等待元素插入
	if (count.get() == capacity)
		return false;
	int c = -1;
	Node<E> node = new Node(e);
	//开启队尾保护锁
	final ReentrantLock putLock = this.putLock;
	putLock.lock();
	try {
		if (count.get() < capacity) {
			enqueue(node);
			//原则计数类
			c = count.getAndIncrement();
			if (c + 1 < capacity)
				notFull.signal();
		}
	} finally {
		//释放锁
		putLock.unlock();
	}
	if (c == 0)
		signalNotEmpty();
	return c >= 0;
}

//在持有锁下指向下一个节点
private void enqueue(Node<E> node) {
	// assert putLock.isHeldByCurrentThread();
	// assert last.next == null;
	last = last.next = node;
}

1.3.3 put操作

//put 操作把指定元素添加到队尾，如果没有空间则一直等待。
public void put(E e) throws InterruptedException {
	if (e == null) throw new NullPointerException();
	// Note: convention in all put/take/etc is to preset local var
	// holding count negative to indicate failure unless set.
	//注释：在所有的 put/take/etc等操作中变量c为负数表示失败，>=0表示成功。
	int c = -1;
	Node<E> node = new Node(e);
	final ReentrantLock putLock = this.putLock;
	final AtomicInteger count = this.count;
	putLock.lockInterruptibly();
	try {
		/*
		 * Note that count is used in wait guard even though it is
		 * not protected by lock. This works because count can
		 * only decrease at this point (all other puts are shut
		 * out by lock), and we (or some other waiting put) are
		 * signalled if it ever changes from capacity. Similarly
		 * for all other uses of count in other wait guards.
		 */
		/*
		 * 注意，count用于等待监视，即使它没有用锁保护。这个可行是因为
		 * count 只能在此刻（持有putLock）减小（其他put线程都被锁拒之门外），
		 * 当count对capacity发生变化时，当前线程（或其他put等待线程）将被通知。
		 * 在其他等待监视的使用中也类似。
		 */
		while (count.get() == capacity) {
			notFull.await();
		}
		enqueue(node);
		c = count.getAndIncrement();
		// 还有可添加空间则唤醒put等待线程。
		if (c + 1 < capacity)
			notFull.signal();
	} finally {
		putLock.unlock();
	}
	if (c == 0)
		signalNotEmpty();
}

1.3.4 take操作

//弹出队头元素，如果没有会被阻塞直到元素返回
public E take() throws InterruptedException {
	E x;
	int c = -1;
	final AtomicInteger count = this.count;

	final ReentrantLock takeLock = this.takeLock;
	takeLock.lockInterruptibly();
	try {
		while (count.get() == 0) {
			notEmpty.await();//没有元素一直阻塞
		}
		x = dequeue();
		c = count.getAndDecrement();
		if (c > 1)//如果还有可获取元素，唤醒等待获取的线程。
		  notEmpty.signal();
	} finally {
		//拿到元素后释放锁
		takeLock.unlock();
	}
	if (c == capacity)
		signalNotFull();
	return x;
}

//在持有锁下返回队列队头第一个节点
private E dequeue() {
    // assert takeLock.isHeldByCurrentThread();
    // assert head.item == null;
    Node<E> h = head;
    Node<E> first = h.next;
    h.next = h; // help GC
    //出队后的节点作为头节点并将元素置空
    head = first;
    E x = first.item;
    first.item = null;
    return x;
}

1.3.5 remove操作

//移除指定元素。
public boolean remove(Object o) {
	if (o == null) return false;
	//对两把锁加锁
	fullyLock();
	try {
		for (Node<E> trail = head, p = trail.next;
				p != null;
				trail = p, p = p.next) {
			if (o.equals(p.item)) {
				unlink(p, trail);
				return true;
			}
		}
		return false;
	} finally {
		fullyUnlock();
	}
}

//p是移除元素所在节点，trail是移除元素的上一个节点
void unlink(Node<E> p, Node<E> trail) {
	// assert isFullyLocked();
	// p.next is not changed, to allow iterators that are
	// traversing p to maintain their weak-consistency guarantee.
	p.item = null;
	//将trail下一个节点指向p的下一个节点
	trail.next = p.next;
	if (last == p)
		last = trail;
	if (count.getAndDecrement() == capacity)
		notFull.signal();
}

void fullyLock() {
	putLock.lock();
	takeLock.lock();
}

//释放锁时确保和加锁顺序一致
void fullyUnlock() {
	takeLock.unlock();
	putLock.unlock();
}

注意，锁的释放顺序与加锁顺序是相反的。

参考资料
http://ifeve.com/juc-linkedblockingqueue/