ThreadPoolExecutor解析-主要源码研究

注：本文的分析和源码基于jdk1.7;

一、ThreadPoolExecutor创建

ThreadPoolExecutor作为java.util.concurrent包中核心的类，先看下类型的结构：

 

最顶级的接口都是Executor，而ThreadPoolExecutor继承于抽象类AbstractExecutorService，提供以下4个构造函数用于创建：

public ThreadPoolExecutor(int corePoolSize,int maximumPoolSize,ong keepAliveTime,TimeUnit unit,BlockingQueue<Runnable> workQueue);

public ThreadPoolExecutor(int corePoolSize,int maximumPoolSize,long keepAliveTime,TimeUnit unit,BlockingQueue<Runnable> workQueue,ThreadFactory threadFactory)

 public ThreadPoolExecutor(int corePoolSize, int maximumPoolSize,long keepAliveTime,TimeUnit unit,BlockingQueue<Runnable> workQueue,RejectedExecutionHandler handler)

public ThreadPoolExecutor(int corePoolSize,int maximumPoolSize,long keepAliveTime,TimeUnit unit,BlockingQueue<Runnable> workQueue,ThreadFactory threadFactory,RejectedExecutionHandler handler)

 

前面的3个方法都是使用通过this调用最后一个方法，没有指定的构造参数使用默认参数，参数解析：

1、

    /**
     * Core pool size is the minimum number of workers to keep alive
     * (and not allow to time out etc) unless allowCoreThreadTimeOut
     * is set, in which case the minimum is zero.
     */
    private volatile int corePoolSize;

线程池核心线程数大小，初始化时核心线程数也是0，除非先调用prestartCoreThread或者prestartAllCoreThreads先创建核心线程；在没有设置allowCoreThreadTimeOut为true情况下，核心线程不会销毁；

2、

    /**
     * Maximum pool size. Note that the actual maximum is internally
     * bounded by CAPACITY.
     */
    private volatile int maximumPoolSize;

线程池线程数最大值，达到最大值后线程池不会再增加线程执行任务，任务会进入等待队列或者由拒绝策略处理；

该值实际的可设置最大值不是Integer.MAX_VALUE，而是常量CAPACITY（后面再解析常量）

3、

    /**
     * Timeout in nanoseconds for idle threads waiting for work.
     * Threads use this timeout when there are more than corePoolSize
     * present or if allowCoreThreadTimeOut. Otherwise they wait
     * forever for new work.
     */
    private volatile long keepAliveTime;

空闲工作线程的空闲时间；超过corePoolSize的线程或者allowCoreThreadTimeOut为true的主线程使用这个作为超时时间；

否则线程一直等待任务或关闭；

4、

    /**
     * The queue used for holding tasks and handing off to worker
     * threads.  We do not require that workQueue.poll() returning
     * null necessarily means that workQueue.isEmpty(), so rely
     * solely on isEmpty to see if the queue is empty (which we must
     * do for example when deciding whether to transition from
     * SHUTDOWN to TIDYING).  This accommodates special-purpose
     * queues such as DelayQueues for which poll() is allowed to
     * return null even if it may later return non-null when delays
     * expire.
     */
    private final BlockingQueue<Runnable> workQueue;

  这个队列用于保存任