死磕并发之CountDownLatch解析

CountDownLatch解析

CountDownLatch是什么

CountDownLatch是基于AQS的阻塞工具，阻塞一个或者多个线程，直到所有的线程都执行完成。

CountDownLatch解决了什么问题

当一个任务运算量比较大的时候，需要拆分为各种子任务，必须要所有子任务完成后才能汇总为总任务。
使用并发模拟的时候可以使用CountDownLatch.也可以设置超时等待时间，

CountDownLatch 用法

1.阻塞所有线程执行完成后再执行

@Slf4j
public class CountDownLatchExample {
    //线程数量
    private static final int THREAD_NUM = 10;

    // CountdownLatch阻塞模拟
    public static void main(String[] args) throws InterruptedException {

        // 创建线程池 用于执行线程
        ExecutorService executorService = Executors.newCachedThreadPool();
        //创建countDownLatch
        final CountDownLatch countDownLatch = new CountDownLatch(THREAD_NUM);
        long startTime = System.currentTimeMillis();
        //循环创建线程
        for (int i = 0; i < THREAD_NUM; i++) {
            final int a = i;

            executorService.execute(() -> {
                try {
                    test(a);
                } catch (Exception e) {
                    log.error("Exception", e);
                } finally {
                    countDownLatch.countDown();
                }
            });

        }

        countDownLatch.await();
        long endTime = System.currentTimeMillis();
        log.info("执行完毕,{}-{}",startTime,endTime);
        executorService.shutdown();
    }

    private static void test(int num) throws InterruptedException {
        Thread.sleep(100);
        log.info("{}-{}", num,System.currentTimeMillis());
        Thread.sleep(100);
    }

}

结果
10:56:02.544 [pool-1-thread-5] INFO AQSExample.CountDownLatchExampleTimeOutTest - 4-1559271362542
10:56:02.543 [pool-1-thread-2] INFO AQSExample.CountDownLatchExampleTimeOutTest - 1-1559271362541
10:56:02.548 [pool-1-thread-10] INFO AQSExample.CountDownLatchExampleTimeOutTest - 9-1559271362548
10:56:02.544 [pool-1-thread-7] INFO AQSExample.CountDownLatchExampleTimeOutTest - 6-1559271362543
10:56:02.543 [pool-1-thread-4] INFO AQSExample.CountDownLatchExampleTimeOutTest - 3-1559271362542
10:56:02.544 [pool-1-thread-3] INFO AQSExample.CountDownLatchExampleTimeOutTest - 2-1559271362541
10:56:02.544 [pool-1-thread-8] INFO AQSExample.CountDownLatchExampleTimeOutTest - 7-1559271362543
10:56:02.544 [pool-1-thread-6] INFO AQSExample.CountDownLatchExampleTimeOutTest - 5-1559271362543
10:56:02.543 [pool-1-thread-1] INFO AQSExample.CountDownLatchExampleTimeOutTest - 0-1559271362541
10:56:02.548 [pool-1-thread-9] INFO AQSExample.CountDownLatchExampleTimeOutTest - 8-1559271362548
10:56:02.548 [main] INFO AQSExample.CountDownLatchExampleTimeOutTest - 执行完毕,1559271362441-1559271362548

上述结果可以看到，所有的线程执行完毕后主线程才打印出“执行完毕”。

2.按照超时时间阻塞所有线程执行，到时间后直接释放。

如果我们设置超时时间之后

@Slf4j
public class CountDownLatchExampleTimeOutTest {


    //线程数量
    private static final int THREAD_NUM = 10;

    // CountdownLatch阻塞模拟
    public static void main(String[] args) throws InterruptedException {
        // 创建线程池 用于执行线程
        ExecutorService executorService = Executors.newCachedThreadPool();
        //创建countDownLatch
        final CountDownLatch countDownLatch = new CountDownLatch(THREAD_NUM);
        //循环创建线程
        long startTime = System.currentTimeMillis();
        for (int i = 0; i < THREAD_NUM; i++) {
            final int a = i;
            executorService.execute(() -> {
                try {
                    test(a);
                } catch (Exception e) {
                    log.error("Exception", e);
                } finally {
                    countDownLatch.countDown();
                }
            });

        }

        countDownLatch.await(10,TimeUnit.MILLISECONDS);
        long endTime = System.currentTimeMillis();
        log.info("执行完毕,{}-{}",startTime,endTime);
        executorService.shutdown();
    }

    private static void test(int num) throws InterruptedException {
        Thread.sleep(50);
        log.info("{}-{}", num,System.currentTimeMillis());
    }
}

由于每个线程延迟50毫秒之后再执行，count已经超时了所以优先打印出了执行完毕的结果。然后在继续执行线程中的内容。

结果
11:14:55.509 [main] INFO AQSExample.CountDownLatchExampleTimeOutTest - 执行完毕,1559272495373-1559272495506
11:14:55.542 [pool-1-thread-1] INFO AQSExample.CountDownLatchExampleTimeOutTest - 0-1559272495542
11:14:55.542 [pool-1-thread-2] INFO AQSExample.CountDownLatchExampleTimeOutTest - 1-1559272495542
11:14:55.543 [pool-1-thread-3] INFO AQSExample.CountDownLatchExampleTimeOutTest - 2-1559272495543
11:14:55.543 [pool-1-thread-4] INFO AQSExample.CountDownLatchExampleTimeOutTest - 3-1559272495543
11:14:55.543 [pool-1-thread-5] INFO AQSExample.CountDownLatchExampleTimeOutTest - 4-1559272495543
11:14:55.544 [pool-1-thread-6] INFO AQSExample.CountDownLatchExampleTimeOutTest - 5-1559272495544
11:14:55.544 [pool-1-thread-7] INFO AQSExample.CountDownLatchExampleTimeOutTest - 6-1559272495544
11:14:55.545 [pool-1-thread-9] INFO AQSExample.CountDownLatchExampleTimeOutTest - 8-1559272495545
11:14:55.545 [pool-1-thread-8] INFO AQSExample.CountDownLatchExampleTimeOutTest - 7-1559272495545
11:14:55.545 [pool-1-thread-10] INFO AQSExample.CountDownLatchExampleTimeOutTest - 9-1559272495545

CountDownLatch源码解析

CountDownLatch源码中的方法和属性并不多，下面我们来一一解析。

1.AQS框架以及构造方法

//当前对象中私有阻塞工具
private final Sync sync;
// 模板方法模式重写AQS工具
private static final class Sync extends AbstractQueuedSynchronizer {
   private static final long serialVersionUID = 4982264981922014374L;
   // 共享阻塞AQS
   Sync(int count) {
       setState(count);
   }
   // 获取当前还剩多少资源可以使用
   int getCount() {
       return getState();
   }

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

   protected boolean tryReleaseShared(int releases) {
       for (;;) {
           int c = getState();
           if (c == 0)
               return false;
           int nextc = c-1;
           if (compareAndSetState(c, nextc))
               return nextc == 0;
       }
   }
}
//构造方法创建一个锁对象
public CountDownLatch(int count) {
   if (count < 0) throw new IllegalArgumentException("count < 0");
   this.sync = new Sync(count);
}

2.countDown()方法解析

该方法用于线程执行完毕后减计统计数量，

// 该方法时释放一个共享锁。当所有锁都被释放完成后主线程就能继续执行了。
public void countDown() {
    sync.releaseShared(1);
}

3.await()方法解析

//拦截主线程的方法。主线程在这里等待条件达成后继续执行。
public void await() throws InterruptedException {
    //在这里阻塞线程的执行
    sync.acquireSharedInterruptibly(1);
}
public final void acquireSharedInterruptibly(int arg)
       throws InterruptedException {
   if (Thread.interrupted())
       throw new InterruptedException();
    //这里判断是否还有可以共享的资源
   // 如果有则返回-1 否则返回 1，重写AQS的方法参见（1.AQS框架以及构造方法）
   if (tryAcquireShared(arg) < 0)
       // 有资源则运行阻塞自旋等待所有线程执行完毕
       doAcquireSharedInterruptibly(arg);
   // 无资源可用就让线程继续执行
}

// 带延迟的减少数据拦截方法
// 返回的结果是没有跑完全部线程就继续执行下一步了。
public boolean await(long timeout, TimeUnit unit)
    throws InterruptedException {
    return sync.tryAcquireSharedNanos(1, unit.toNanos(timeout));
}

public final boolean tryAcquireSharedNanos(int arg, long nanosTimeout)
    throws InterruptedException {
    //线程如果被中断则抛出异常
    if (Thread.interrupted())
        throw new InterruptedException();
    // 表示如果线程被执行完了直接返回成功，如果没有执行完则看等待时间来决定是否要继续执行。
    return tryAcquireShared(arg) >= 0 ||
        doAcquireSharedNanos(arg, nanosTimeout);
}

CountDownLatch 总结

CountDownLatch这个类能够使一个线程等待其他线程完成各自的工作后再执行。 在分散计算统一合成结果，按某个流程加载资源的方面有着非诚好用的效果。下一篇我们讲解像蓄水池一样功能的Semphore。