通用线程池

通用线程池

1. 架构模型

2. 核心参数

3. 继承体系

  • Executor: 顶级接口,任务执行器
  • ExecutorService:即Executor Service,跟我们正常写方法比较类似,定义了线程池的通用方法
  • AbstractExecutorService: 典型的模版方法模式实现,主流程有抽象类实现,提供钩子方法,由子类实现。

4. AbstractExecutorService实现

4.1 submit

将Runable和Callable包装成RunnableFuture对象,调用子类实现的execute(RunableFuture)防范】

    /**
     * @throws RejectedExecutionException {@inheritDoc}
     * @throws NullPointerException       {@inheritDoc}
     */
    public Future<?> submit(Runnable task) {
        if (task == null) throw new NullPointerException();
        RunnableFuture<Void> ftask = newTaskFor(task, null);
        execute(ftask);
        return ftask;
    }

    /**
     * @throws RejectedExecutionException {@inheritDoc}
     * @throws NullPointerException       {@inheritDoc}
     */
    public <T> Future<T> submit(Runnable task, T result) {
        if (task == null) throw new NullPointerException();
        RunnableFuture<T> ftask = newTaskFor(task, result);
       // 钩子函数,由子类实现具体的调度逻辑
        execute(ftask);
        return ftask;
    }

    /**
     * @throws RejectedExecutionException {@inheritDoc}
     * @throws NullPointerException       {@inheritDoc}
     */
    public <T> Future<T> submit(Callable<T> task) {
        if (task == null) throw new NullPointerException();
        RunnableFuture<T> ftask = newTaskFor(task);
        execute(ftask);
        return ftask;
    }

4.2 invokeAll

  • 首先将集合中的所有Callable包装成RunnableFuture,并调用execute(Runable)方法
  • 依次调用Future.get(),注意主动cancel和执行异常会被吞掉
  • 超时后,会尝试中断未执行完的线程

4.3 invokeAny

  • 有任何一个执行完成就结束
  • 完成后,会尝试中断正在执行的任务(不一定能取消掉)
  • 超时会中断所有正在执行的任务

5. ThreadPoolExecutorService实现

  • 线程池状态:

  • 运行状态
    • RUNNING:接收新的任务,处理队列中的任务
    • SHUTDOWN:不接受新任务,但是处理队列中的任务
    • STOP:不接受新任务,不处理队列中的任务,同时打断队列中的任务
    • TIDYING:所有任务都终止,工作线程数量为0,在转换为TIDYING状态后会执行钩子函数terminated()
    • TERMINATED: terminated() 执行结束
// 用高3位表示线程池的状态, 总共5个状态,3位正好可以表示
private static final int COUNT_BITS = Integer.SIZE - 3;
private static final int COUNT_MASK = (1 << COUNT_BITS) - 1;

// runState is stored in the high-order bits
private static final int RUNNING    = -1 << COUNT_BITS;
private static final int SHUTDOWN   =  0 << COUNT_BITS;
private static final int STOP       =  1 << COUNT_BITS;
private static final int TIDYING    =  2 << COUNT_BITS;
private static final int TERMINATED =  3 << COUNT_BITS;

5.1 shutdown()

public void shutdown() {
    final ReentrantLock mainLock = this.mainLock;
    mainLock.lock();
    try {
        checkShutdownAccess();
        // CAS 设置保证执行状态为SHUTDOWN
        advanceRunState(SHUTDOWN);
        // 中断空闲线程
        interruptIdleWorkers();
        onShutdown(); // hook for ScheduledThreadPoolExecutor
    } finally {
        mainLock.unlock();
    }
    // 这里会尝试终止,实际不一定能终止,最后一个线程会调用终止
    tryTerminate();
}

5.2 shutdownNow()

public List<Runnable> shutdownNow() {
    List<Runnable> tasks;
    final ReentrantLock mainLock = this.mainLock;
    mainLock.lock();
    try {
        checkShutdownAccess();
        // CAS 设置保证执行状态为STOP
        advanceRunState(STOP);
        // 中断所有的执行线程
        interruptWorkers();
        // 取出所有未执行的任务返回,给业务线程机会是否处理该线程
        tasks = drainQueue();
    } finally {
        mainLock.unlock();
    }
    // 这里会尝试终止,实际不一定能终止,最后一个线程会调用终止
    tryTerminate();
    return tasks;
}

5.3 awaitTerminated() & tryTerminate()

    public boolean awaitTermination(long timeout, TimeUnit unit)
        throws InterruptedException {
        long nanos = unit.toNanos(timeout);
        final ReentrantLock mainLock = this.mainLock;
        mainLock.lock();
        try {
            for (;;) {
              // 线程池为Terminated才会正常结束
                if (runStateAtLeast(ctl.get(), TERMINATED))
                    return true;
                if (nanos <= 0)
                    return false;
               // 利用条件变量,类似wait notify,但是这里支持等待时长
                nanos = termination.awaitNanos(nanos);
            }
        } finally {
            mainLock.unlock();
        }
    }

    final void tryTerminate() {
        for (;;) {
            int c = ctl.get();
            if (isRunning(c) ||
                runStateAtLeast(c, TIDYING) ||
                (runStateOf(c) == SHUTDOWN && ! workQueue.isEmpty()))
                return;
            if (workerCountOf(c) != 0) { // Eligible to terminate
                interruptIdleWorkers(ONLY_ONE);
                return;
            }

         	  // 只有一个线程会执行到下面的代码,其他线程在上面return了
            final ReentrantLock mainLock = this.mainLock;
            mainLock.lock();
            try {
                // tidying状态才能终止,线程数为0,队列是空
                if (ctl.compareAndSet(c, ctlOf(TIDYING, 0))) {
                    try {
                        // 钩子方法
                        terminated();
                    } finally {
                        // 状态标记为终止
                        ctl.set(ctlOf(TERMINATED, 0));
                        // 条件标量通知等待结束的线程可以放行了,之所以是signall->多个线程等待都会被放行
                        termination.signalAll();
                    }
                    return;
                }
            } finally {
                mainLock.unlock();
            }
            // else retry on failed CAS
        }
    }

5.4 核心方法:execute

执行下面的操作:

  • 一言以蔽之:先添加核心线程,然后添加到队列,队列满了后创建非核心线程。最后执行拒绝策略。
    public void execute(Runnable command) {
        if (command == null)
            throw new NullPointerException();
      
        int c = ctl.get();
        // 小于核心线程数,需要启动新任务
        if (workerCountOf(c) < corePoolSize) {
            // 会自动检测runState和workerCount, 
            // 如果添加失败,如果返回false,要么线程数超过核心线程数,要么runState已经变更,执行后续的处理
            if (addWorker(command, true))
                return;
            c = ctl.get();
        }
      
        // 如果是运行状态说明,添加失败的原因是超过核心线线程数,先添加到队列中
        if (isRunning(c) && workQueue.offer(command)) {
            // 多线程场景,double-check
            int recheck = ctl.get();
            // 不在运行态,直接回滚
            if (! isRunning(recheck) && remove(command))
              // 执行拒绝策略  
              reject(command);
            // 运行态,运行线程数等于0
            else if (workerCountOf(recheck) == 0)
              // 第一个任务为null, 会从队列中取1个任务作为第一个任务执行  
              addWorker(null, false);
        }
        // 添加到队列失败,创建非核心线程,执行任务
        else if (!addWorker(command, false))
            // 执行拒绝策略
            reject(command);
    }

5.5 核心方法addWorker

Worker(Runnable firstTask) {
   setState(-1); // inhibit interrupts until runWorker
   this.firstTask = firstTask;
   // 创建线程的runable对象传的是this,即worker对象,t.start会执行worker的run方法,调用runWorker(this)
   this.thread = getThreadFactory().newThread(this);
}
/**
 * firstTask: 第一个需要执行的任务
 * core: 是否创建核心线程数
 **/
private boolean addWorker(Runnable firstTask, boolean core) {
    retry:
    for (int c = ctl.get();;) {
        // 如果线程池runState 为Stop状态,直接返回false
        // shutDown状态,会执行队列中的任务,但不会执行新的任务,所以不需要创建新线程
        if (runStateAtLeast(c, SHUTDOWN)
            && (runStateAtLeast(c, STOP)
                || firstTask != null
                || workQueue.isEmpty()))
            return false;

        for (;;) {
            // 大于核心线程数还是大雨最大线程数,取决于core参数,超过了就不能创建新线程了,返回false,外层调用者会执行拒绝策略
            if (workerCountOf(c)
                >= ((core ? corePoolSize : maximumPoolSize) & COUNT_MASK))
                return false;
            // CAS保证线程安全+1,即workerCount+1
            if (compareAndIncrementWorkerCount(c))
                // break跳出循环,执行循环后面的内容,continue不会跳出循环
                break retry;
            // CAS失败,说明workerCount已经被其他线程变更, 重新取值判断
            c = ctl.get();  // Re-read ctl
            // 运行状态到了SHUT DOWN以后(STOP, TIDYing)重新跳出到外层循环
            if (runStateAtLeast(c, SHUTDOWN))
                continue retry;
            // 其他情况运行状态不变,只需要重新执行下内层循环判断数量
        }
    }
    
    // 工作线程数已经+1, 如果真正启动失败,会回滚 
    boolean workerStarted = false;
    boolean workerAdded = false;
    Worker w = null;
    try {
        w = new Worker(firstTask);
        final Thread t = w.thread;
        if (t != null) {
            final ReentrantLock mainLock = this.mainLock;
            mainLock.lock();
            try {
                // Recheck while holding lock.
                // Back out on ThreadFactory failure or if
                // shut down before lock acquired.
                int c = ctl.get();

                if (isRunning(c) ||
                    (runStateLessThan(c, STOP) && firstTask == null)) {
                    if (t.getState() != Thread.State.NEW)
                        throw new IllegalThreadStateException();
                    // 向容器中添加工作对象
                    workers.add(w);
                    workerAdded = true;
                    int s = workers.size();
                    if (s > largestPoolSize)
                        largestPoolSize = s;
                }
            } finally {
                mainLock.unlock();
            }
            if (workerAdded) {
                // 启动的时候,会执行Worker对象的run方法
                t.start();
                workerStarted = true;
            }
        }
    } finally {
        if (! workerStarted)
            // 添加失败,这里会回滚线程数
            addWorkerFailed(w);
    }
    return workerStarted;
}

5.6 Woker.runWork(Worker w)方法

final void runWorker(Worker w) {
    Thread wt = Thread.currentThread();
    Runnable task = w.firstTask;
    w.firstTask = null;
    w.unlock(); // allow interrupts
    boolean completedAbruptly = true;
    try {
       // keepAliveTime, getTask会调用阻塞队列的poll方法一直到取到为止,
       // 如果超时(keepAliveTime)未取到,会抛出中断异常,processWorkerExit会执行,删除工作线程,由GC回收
        while (task != null || (task = getTask()) != null) {
            w.lock();
            // shutdown的时候需要清除中断标志位,因为当前线程还要执行线程中的任务
            // shutDownNow, 需要确保处于中断状态, 所以在任务中调用中断后,下一次任务会清除中断标志位
            // 中断的时候,join,wait, notify等都可以响应中断标志位
            if ((runStateAtLeast(ctl.get(), STOP) ||
                 (Thread.interrupted() &&
                  runStateAtLeast(ctl.get(), STOP))) &&
                !wt.isInterrupted())
                wt.interrupt();
            try {
                //子类实现钩子函数
                beforeExecute(wt, task);
                try {
                    // 执行任务
                    task.run();
                    // 子类实现
                    afterExecute(task, null);
                } catch (Throwable ex) {
                    afterExecute(task, ex);
                    throw ex;
                }
            } finally {
                task = null;
                w.completedTasks++;
                w.unlock();
            }
        }
        // 有异常的情况下,该值位true
        completedAbruptly = false;
    } finally {
        // 有异常的情况下,该值位true, 会减少workerCount, 
        // 以便能重新创建线程,所有抛出异常并不会导致没有线程可用
        processWorkerExit(w, completedAbruptly);
    }
}

private void processWorkerExit(Worker w, boolean completedAbruptly) {
    if (completedAbruptly) // If abrupt, then workerCount wasn't adjusted
        decrementWorkerCount();

    final ReentrantLock mainLock = this.mainLock;
    mainLock.lock();
    try {
        completedTaskCount += w.completedTasks;
        workers.remove(w);
    } finally {
        mainLock.unlock();  
    }

    tryTerminate();

    int c = ctl.get();
    if (runStateLessThan(c, STOP)) {
        if (!completedAbruptly) {
            int min = allowCoreThreadTimeOut ? 0 : corePoolSize;
            // 保证至少一个线程运行
            if (min == 0 && ! workQueue.isEmpty())
                min = 1;
            if (workerCountOf(c) >= min)
                return; // replacement not needed
        }
        addWorker(null, false);
    }
}
posted @ 2022-01-11 22:19  dragonfei  阅读(171)  评论(0编辑  收藏  举报