关于java:阅读-JDK-8-源码AQS-对-Condition-的实现

20次阅读

共计 16852 个字符,预计需要花费 43 分钟才能阅读完成。

前两篇文章别离介绍了 AQS 框架中的独占模式和共享模式,本篇将介绍 AQS 对 Condition 接口的实现。
在浏览本篇之前,倡议先理解 AQS 中的数据结构和独占模式的实现原理。

JUC 通过 Lock 和 Condition 两个接口实现管程(Monitor),其中 Lock 用于解决互斥问题,而 Condition 用于解决同步问题,而 AQS 对 Lock 和 Condition 接口的实现提供了一个根底的框架。

1. Condition 接口

Lock 代替了 synchronized 办法和语句的应用,Condition 代替了 Object 监视器办法的应用。

Condition 实现能够提供不同于 Object 监视器办法的行为和语义,比方受保障的告诉排序,或者在执行告诉时不须要放弃一个锁。
如果某个实现提供了这样非凡的语义,则该实现必须记录这些语义。

办法摘要:

// 造成以后线程在接到信号或被中断之前始终处于期待状态。void await() 

// 造成以后线程在接到信号、被中断或达到指定等待时间之前始终处于期待状态。boolean await(long time, TimeUnit unit) 

// 造成以后线程在接到信号、被中断或达到指定等待时间之前始终处于期待状态。long awaitNanos(long nanosTimeout) 

// 造成以后线程在接到信号之前始终处于期待状态。void awaitUninterruptibly() 

// 造成以后线程在接到信号、被中断或达到指定最初期限之前始终处于期待状态。boolean awaitUntil(Date deadline) 

// 唤醒一个期待线程。void signal() 

// 唤醒所有期待线程。void signalAll() 

Condition 实质上是一个队列(称为条件队列),线程期待某个条件成立时,在队列中阻塞,直到其余线程查看条件成立起初告诉它。
对于同一个锁,只会存在一个同步队列,然而可能会有多个条件队列,只有在应用了 Condition 才会存在条件队列。

AQS 中对条件队列的应用:

当线程获取锁之后,执行 Condition.await() 会开释锁并进入条件队列,阻塞期待直到被其余线程唤醒。
当其余线程执行 Condition.signal() 唤醒以后线程时,以后线程会从条件队列转移到同步队列来期待再次获取锁。
以后线程再一次获取锁之后,须要在 while 循环中判断条件是否成立,若不成立需从新执行 Condition.await() 去期待。

2. Condition 应用

Condition 实例本质上被绑定到一个锁上。要为特定 Lock 实例取得 Condition 实例,请应用其 newCondition() 办法。

Java 官网文档提供 Condition 接口的应用示例:

对于一个有界阻塞数组,当数组非满时才能够往数组中存放数据,否则阻塞;当数据非空时才能够往数组中取元素,否则阻塞。

class BoundedBuffer {final Lock lock = new ReentrantLock();
    final Condition notFull = lock.newCondition();
    final Condition notEmpty = lock.newCondition();

    final Object[] items = new Object[100];
    int putptr, takeptr, count;

    // 生产者办法,往数组外面写数据
    public void put(Object x) throws InterruptedException {lock.lock();
        try {while (count == items.length)
                notFull.await(); // 阻塞直到非满
            items[putptr] = x;
            if (++putptr == items.length) putptr = 0;
            ++count;
            notEmpty.signal(); // 告诉非空} finally {lock.unlock();
        }
    }

    // 消费者办法,从数组外面拿数据
    public Object take() throws InterruptedException {lock.lock();
        try {while (count == 0)
                notEmpty.await(); // 阻塞直到非空
            Object x = items[takeptr];
            if (++takeptr == items.length) takeptr = 0;
            --count;
            notFull.signal(); // 告诉非满
            return x;
        } finally {lock.unlock();
        }
    }
}

在 JDK 的实现中,独占模式才可应用 Condition,共享模式不反对 Condition。
因为 AQS 的外部类 ConditionObject 只反对独占模式。

java.util.concurrent.locks.ReentrantLock.Sync#newCondition

final ConditionObject newCondition() {return new ConditionObject();
}

java.util.concurrent.locks.ReentrantReadWriteLock.WriteLock#newCondition

public Condition newCondition() {return sync.newCondition();
}

java.util.concurrent.locks.ReentrantReadWriteLock.ReadLock#newCondition

/**
 * Throws {@code UnsupportedOperationException} because
 * {@code ReadLocks} do not support conditions.
 *
 * @throws UnsupportedOperationException always
 */
public Condition newCondition() {throw new UnsupportedOperationException();
}

3. Condition 期待办法

3.1 Condition#await

代码流程:

  1. 判断线程是否被中断,如果是,间接抛出 InterruptedException,否则进入下一步
  2. 将以后线程封装为节点,存入条件队列。
  3. 开释以后线程已取得的全副的锁,若无持有锁则抛异样。
  4. 在条件队列中,阻塞以后节点。
  5. 以后节点从阻塞中被唤醒(signalled or interrupted),则会从条件队列转移到同步队列(被动或被动地)。
  6. 在同步队列中,自旋、阻塞期待获取锁胜利。
  7. 判断整个过程中是否产生过中断,进行不同的解决(抛异样 或 从新中断)。

java.util.concurrent.locks.AbstractQueuedSynchronizer.ConditionObject#await()

/**
 * Implements interruptible condition wait.
 * <ol>
 * <li> If current thread is interrupted, throw InterruptedException.
 * <li> Save lock state returned by {@link #getState}.
 * <li> Invoke {@link #release} with saved state as argument,
 *      throwing IllegalMonitorStateException if it fails.
 * <li> Block until signalled or interrupted.
 * <li> Reacquire by invoking specialized version of
 *      {@link #acquire} with saved state as argument.
 * <li> If interrupted while blocked in step 4, throw InterruptedException.
 * </ol>
 */
public final void await() throws InterruptedException {if (Thread.interrupted())
        throw new InterruptedException();
    Node node = addConditionWaiter();    // 将以后线程封装成节点存入条件队列
    int savedState = fullyRelease(node); // 开释曾经持有的锁(就是在调用 Condition#await 之前持有的 Lock#lock 锁),并返回开释前的锁状态
    int interruptMode = 0;
    while (!isOnSyncQueue(node)) {       // 查看节点是否在同步队列上
        LockSupport.park(this);          // 节点还在条件队列中,则阻塞   
        // 节点从阻塞中被唤醒(condition#signal,Thread#interrupt),查看中断状态,设置中断解决模式
        // 补充:被 condition#signal 唤醒后的线程会从条件队列转移到同步队列(先出队再入队)// 补充:若在条件队列中就产生了中断,也会被转移到同步队列(不出队,只入队,见 checkInterruptWhileWaiting -> transferAfterCancelledWait)if ((interruptMode = checkInterruptWhileWaiting(node)) != 0)  
            break;                                                    
    }                                                                 
    if (acquireQueued(node, savedState) && interruptMode != THROW_IE) // 在同步队列期待获取资源直到胜利,判断设置中断解决模式
        interruptMode = REINTERRUPT;
    if (node.nextWaiter != null)  // clean up if cancelled 
        // nextWaiter 不为空,阐明以后节点是由 Thread#interrupt 唤醒的(condition#signal 唤醒阻塞节点会设置 nextWaiter 为空)// 此时以后节点同时存在于同步队列、条件队列上!然而 waitStatus 不是 CONDITION
        // 须要革除条件队列中已勾销的节点
        unlinkCancelledWaiters();
    if (interruptMode != 0)
        reportInterruptAfterWait(interruptMode); // 解决中断:抛异样,或者补上中断状态
}

留神:

  1. 线程被唤醒,可能是执行了 Condition#signal(其中执行 LockSupport#unpark 来唤醒条件队列的头节点),也可能是调用了 Thread#interrupt (会更新线程的中断标识)。
  2. 如果是由 Condition#signal 唤醒的,则以后节点被唤醒后,曾经位于同步队列。
  3. 如果是由 Thread#interrupt 唤醒的,则以后节点被唤醒后,须要判断是位于同步队列还是条件队列。
    3.1 如果是位于同步队列,阐明是先失去 Condition#signal 告诉,再被 Thread#interrupt 中断。
    3.2 如果是位于条件队列,阐明未失去 Condition#signal 告诉就被 Thread#interrupt 中断了,须要自行退出到同步队列中,再从条件队列中移除。
  4. 以后节点从条件队列转移到同步队列的过程中,产生了中断,该节点依旧会在同步队列中自旋、阻塞直到获取锁,再响应中断(抛异样或从新中断)。

3.1.1 addConditionWaiter

将以后线程封装为节点(waitStatus 为 CONDITION),增加到条件队列尾部。
若条件队列不存在则进行初始化,把以后节点作为头节点(不应用 dummy node)。

java.util.concurrent.locks.AbstractQueuedSynchronizer.ConditionObject#addConditionWaiter

/**
 * Adds a new waiter to wait queue.
 * @return its new wait node
 */
private Node addConditionWaiter() {
    Node t = lastWaiter;
    // If lastWaiter is cancelled, clean out. // 清理条件队列中已勾销的尾节点
    if (t != null && t.waitStatus != Node.CONDITION) {unlinkCancelledWaiters();
        t = lastWaiter;
    }
    Node node = new Node(Thread.currentThread(), Node.CONDITION); // 构建节点,尾插法
    if (t == null)
        firstWaiter = node;
    else
        t.nextWaiter = node;
    lastWaiter = node;
    return node;
}

3.1.2 fullyRelease

开释以后线程已持有的锁 / 资源,返回开释之前的锁 / 资源。
若未持有锁,报错。

这里存在 BUG:报错之前,以后线程曾经退出到条件队列之中了,会导致条件队列存储有效的节点数据。
应该将是否持有锁的校验提前到 addConditionWaiter 之前,JDK 11 中已修复该问题。

java.util.concurrent.locks.AbstractQueuedSynchronizer#fullyRelease

final int fullyRelease(Node node) { // 开释以后线程已持有的锁
    boolean failed = true;
    try {int savedState = getState(); // 获取 volatile 的 state,独占模式下示意以后线程锁持有的全副锁
        if (release(savedState)) {   // 开释全副的锁
            failed = false;
            return savedState;
        } else {throw new IllegalMonitorStateException(); // 未持有锁,报错
        }
    } finally {if (failed)
            node.waitStatus = Node.CANCELLED;
    }
}

3.1.3 isOnSyncQueue

判断节点是否在同步队列上。

java.util.concurrent.locks.AbstractQueuedSynchronizer#isOnSyncQueue

/**
 * Returns true if a node, always one that was initially placed on
 * a condition queue, is now waiting to reacquire on sync queue.
 * @param node the node
 * @return true if is reacquiring
 */
final boolean isOnSyncQueue(Node node) {if (node.waitStatus == Node.CONDITION || node.prev == null)
        return false;
    if (node.next != null) // If has successor, it must be on queue
        return true;
    return findNodeFromTail(node); // 从尾节点向前遍历查找
}

/**
 * Returns true if node is on sync queue by searching backwards from tail.
 * Called only when needed by isOnSyncQueue.
 * @return true if present
 */
private boolean findNodeFromTail(Node node) {
    Node t = tail;
    for (;;) {if (t == node)
            return true;
        if (t == null)
            return false;
        t = t.prev;
    }
}

1. 如果 waitStatus == CONDITION 阐明肯定是位于条件队列上。

从条件队列入队,结构节点的时候默认就为 CONDITION 状态。
将节点从条件队列转移到同步队列,首先会 CAS 设置 waitStatus 状态为 CONDITION,再执行入队操作。

2. node.prev == null 阐明肯定是位于条件队列上。

同步队列只有头节点合乎 node.prev == null,然而同步队列的头节点是 dummy node,其 thread 为空。
也就是说,来调用 isOnSyncQueue 办法且合乎 node.prev == null 条件的节点,只可能是位于条件队列上的节点。

3. 如果 node.next != null 阐明肯定是处于同步队列上。

节点退出同步队列是个复合操作,最初一步是设置 node.next,当 node.next != null 阐明入队操作已执行实现。

4. 如果以上都无奈判断节点是否位于同步队列,则遍历链表查找节点。

存在 node.prev != null 然而节点还没有齐全入队胜利的状况,因为入队操作设置 prev -> tail -> next 是非原子操作。
所以须要从 tail 向前遍历,能力精确判断 node 是否位于同步队列上。
调用 findNodeFromTail 办法前,节点个别位于尾节点左近,不会遍历过多节点。

3.1.4 checkInterruptWhileWaiting

阻塞在 Condition#await 的线程被唤醒之后,调用 checkInterruptWhileWaiting 来查看是否是由线程中断唤醒的。

如果是由线程中断唤醒的,须要进一步判断如何解决中断:

  1. THROW_IE:throw new InterruptedException();
  2. REINTERRUPT:Thread.currentThread().interrupt();

java.util.concurrent.locks.AbstractQueuedSynchronizer.ConditionObject#checkInterruptWhileWaiting

/**
 * Checks for interrupt, returning THROW_IE if interrupted
 * before signalled, REINTERRUPT if after signalled, or
 * 0 if not interrupted.
 */
private int checkInterruptWhileWaiting(Node node) {
    // 校验以后线程是否被中断,并革除线程的中断状态
    return Thread.interrupted() ? 
        (transferAfterCancelledWait(node) ? THROW_IE : REINTERRUPT) : 
        0; // 如果线程未被中断,返回 0
}

3.1.5 transferAfterCancelledWait

如果阻塞在 Condition#await 的线程是被中断唤醒的,执行 transferAfterCancelledWait 判断产生中断产生时节点所在的地位。
如果是位于条件队列,则将其增加到同步队列,返回 true;否则返回 false。

如何判断中断产生时节点所在的地位?

  1. 已知 Condition#signal 办法会批改状态(CONDITION -> 0),并操作节点从条件队列出队,从同步队列入队。
  2. 如果 Condition#await 执行 CAS 批改状态胜利(CONDITION -> 0),阐明线程中断产生时 Condition#signal 还没执行,此时节点是位于条件队列,须要将节点退出同步队列。
  3. 如果 Condition#await 执行 CAS 批改状态失败(CONDITION -> 0),阐明线程中断产生时 Condition#signal 曾经执行,以后线程须要自旋期待 Condition#signal 执行完。

java.util.concurrent.locks.AbstractQueuedSynchronizer#transferAfterCancelledWait

/**
 * Transfers node, if necessary, to sync queue after a cancelled wait.
 * Returns true if thread was cancelled before being signalled.
 *
 * @param node the node
 * @return true if cancelled before the node was signalled
 */
final boolean transferAfterCancelledWait(Node node) {if (compareAndSetWaitStatus(node, Node.CONDITION, 0)) { // 通过 CAS 胜利与否来判断节点地位
        enq(node); // 如果 CAS 胜利,阐明节点是位于条件队列,须要将它增加到同步队列
        return true;
    }
    /*
     * If we lost out to a signal(), then we can't proceed
     * until it finishes its enq().  Cancelling during an
     * incomplete transfer is both rare and transient, so just
     * spin.
     */                          // 条件队列上的节点失去告诉(Condition#signal)之后,会增加到同步队列中去。while (!isOnSyncQueue(node)) // 这里循环检测,直到确认节点曾经胜利增加到同步队列中。Thread.yield();
    return false;
}

3.1.6 unlinkCancelledWaiters

在 Condition#await 办法中,当线程从阻塞中被线程中断唤醒后,判断节点是位于条件队列中,除了将节点退出同步队列之外,还须要将节点从条件队列中移除。

官网的阐明:

  1. 持有锁时才可调用该办法。
  2. 以后线程在 Condition#wait 中阻塞,在被 Condition#signal 唤醒之前,由线程中断或期待超时唤醒。此时须要调用该办法革除条件队列中的有效节点。
  3. 只管该办法会遍历整个队列,然而只有在 Condition#signal 没有执行之前产生中断或勾销才会调用。
  4. 该办法会遍历整个条件队列,一次性把所有有效节点都革除掉,当队列中短时间呈现大量有效节点时(cancellation storms)可防止屡次遍历队列。

java.util.concurrent.locks.AbstractQueuedSynchronizer.ConditionObject#unlinkCancelledWaiters

/**
 * Unlinks cancelled waiter nodes from condition queue.
 * Called only while holding lock. This is called when
 * cancellation occurred during condition wait, and upon
 * insertion of a new waiter when lastWaiter is seen to have
 * been cancelled. This method is needed to avoid garbage
 * retention in the absence of signals. So even though it may
 * require a full traversal, it comes into play only when
 * timeouts or cancellations occur in the absence of
 * signals. It traverses all nodes rather than stopping at a
 * particular target to unlink all pointers to garbage nodes
 * without requiring many re-traversals during cancellation
 * storms.
 */ 
private void unlinkCancelledWaiters() { // 革除条件队列中状态不为 CONDITION 的节点
    Node t = firstWaiter; // 游标节点,记录以后遍历的节点
    Node trail = null;    // 游标节点,记录遍历过的最初一个无效节点(状态为 CONDITION)while (t != null) {   // 从条件队列的头节点开始遍历(上面正文用 next 代表下一个节点)Node next = t.nextWaiter;
        if (t.waitStatus != Node.CONDITION) { // 以后 t 为有效节点
            t.nextWaiter = null;
            if (trail == null)           // 首次遍历到 t 为无效节点时,才会初始化 trail
                firstWaiter = next;      // 设置 t.next 为新的头节点(下一次循环校验 t.next:若 t.next 有效,则把 t.next.next 设为新的头节点)else
                trail.nextWaiter = next; // 设置 trail.next 为 t.next(把 t 出队,下一次循环校验 t.next:若 t.next 有效,则把 t.next.next 设为 trail.next)if (next == null)
                lastWaiter = trail;      // 设置 trail 为新的尾节点
        }
        else      // 以后 t 为无效节点
            trail = t;
        t = next; // 持续遍历 t.next
    }
}

3.1.7 reportInterruptAfterWait

Condition#await 执行到最初,从阻塞中被唤醒且从新获得锁,判断 interruptMode != 0,即 Condition#await 整个过程中产生过中断,须要对中断进行对立解决。

具体见设置 interruptMode 的代码:checkInterruptWhileWaiting

  1. 如果在失去告诉之前被中断(在条件队列中),返回 THROW_IE
  2. 如果在失去告诉之后被中断(在同步队列中),返回 REINTERRUPT

java.util.concurrent.locks.AbstractQueuedSynchronizer.ConditionObject#reportInterruptAfterWait

/**
 * Throws InterruptedException, reinterrupts current thread, or
 * does nothing, depending on mode.
 */
private void reportInterruptAfterWait(int interruptMode)
    throws InterruptedException {if (interruptMode == THROW_IE)
        throw new InterruptedException();
    else if (interruptMode == REINTERRUPT)
        selfInterrupt();}

java.util.concurrent.locks.AbstractQueuedSynchronizer#selfInterrupt

/**
 * Convenience method to interrupt current thread.
 */
static void selfInterrupt() {Thread.currentThread().interrupt();}

3.2 Condition#awaitNanos

在 Condition 条件上阻塞,具备超时工夫。

  1. Condition#await:Block until signalled or interrupted.
  2. Condition#awaitNanos:Block until signalled, interrupted, or timed out.

java.util.concurrent.locks.AbstractQueuedSynchronizer.ConditionObject#awaitNanos

/**
 * Implements timed condition wait.
 * <ol>
 * <li> If current thread is interrupted, throw InterruptedException.
 * <li> Save lock state returned by {@link #getState}.
 * <li> Invoke {@link #release} with saved state as argument,
 *      throwing IllegalMonitorStateException if it fails.
 * <li> Block until signalled, interrupted, or timed out.
 * <li> Reacquire by invoking specialized version of
 *      {@link #acquire} with saved state as argument.
 * <li> If interrupted while blocked in step 4, throw InterruptedException.
 * </ol>
 */
public final long awaitNanos(long nanosTimeout)
        throws InterruptedException {if (Thread.interrupted())
        throw new InterruptedException();
    Node node = addConditionWaiter();
    int savedState = fullyRelease(node);
    final long deadline = System.nanoTime() + nanosTimeout; // 期待截止的工夫戳
    int interruptMode = 0;
    while (!isOnSyncQueue(node)) {if (nanosTimeout <= 0L) { // 已超时,查看节点所在位置,判断是否把节点退出同步队列
            transferAfterCancelledWait(node);
            break;
        }
        if (nanosTimeout >= spinForTimeoutThreshold) // 大于工夫阈值,进行阻塞;小于工夫阈值,进行自旋
            LockSupport.parkNanos(this, nanosTimeout);
        if ((interruptMode = checkInterruptWhileWaiting(node)) != 0) // 从阻塞中唤醒后,查看是否产生中断,如有中断则完结自旋
            break;
        nanosTimeout = deadline - System.nanoTime(); // 残余等待时间}
    if (acquireQueued(node, savedState) && interruptMode != THROW_IE)
        interruptMode = REINTERRUPT;
    if (node.nextWaiter != null)
        unlinkCancelledWaiters();
    if (interruptMode != 0)
        reportInterruptAfterWait(interruptMode);
    return deadline - System.nanoTime(); // 返回残余的阻塞工夫}

java.util.concurrent.locks.AbstractQueuedSynchronizer#spinForTimeoutThreshold

/**
 * The number of nanoseconds for which it is faster to spin
 * rather than to use timed park. A rough estimate suffices
 * to improve responsiveness with very short timeouts.
 */
static final long spinForTimeoutThreshold = 1000L;

3.3 Condition#awaitUninterruptibly

在 Condition 条件上阻塞,只能被 Condition#signal 唤醒。

  1. Condition#await:Block until signalled or interrupted.
  2. Condition#awaitNanos:Block until signalled, interrupted, or timed out.
  3. Condition#awaitUninterruptibly:Block until signalled.

java.util.concurrent.locks.AbstractQueuedSynchronizer.ConditionObject#awaitUninterruptibly

/**
 * Implements uninterruptible condition wait.
 * <ol>
 * <li> Save lock state returned by {@link #getState}.
 * <li> Invoke {@link #release} with saved state as argument,
 *      throwing IllegalMonitorStateException if it fails.
 * <li> Block until signalled.
 * <li> Reacquire by invoking specialized version of
 *      {@link #acquire} with saved state as argument.
 * </ol>
 */
public final void awaitUninterruptibly() {Node node = addConditionWaiter();
    int savedState = fullyRelease(node);
    boolean interrupted = false;
    while (!isOnSyncQueue(node)) {LockSupport.park(this);
        if (Thread.interrupted())
            interrupted = true;
    }
    if (acquireQueued(node, savedState) || interrupted)
        selfInterrupt();}

4. Condition 告诉办法

4.1 Condition#signal

唤醒在 Condition#await 上期待最久的线程。
把条件队列的头节点出队,把它退出同步队列,并唤醒节点中的线程。
被唤醒的线程从 Condition#await 中醒来后,执行 AbstractQueuedSynchronizer#acquireQueued 期待再次获取锁。

java.util.concurrent.locks.AbstractQueuedSynchronizer.ConditionObject#signal

/**
 * Moves the longest-waiting thread, if one exists, from the
 * wait queue for this condition to the wait queue for the
 * owning lock.
 *
 * @throws IllegalMonitorStateException if {@link #isHeldExclusively}
 *         returns {@code false}
 */
public final void signal() {if (!isHeldExclusively()) // 未持有独占锁,报错
        throw new IllegalMonitorStateException();
    Node first = firstWaiter;
    if (first != null)
        doSignal(first); // 唤醒队首节点(等待时间最长)}

4.1.1 doSignal

  1. 找到适合的可唤醒的节点,个别是条件队列的头节点,将它从条件队列出队。
  2. 如果头节点是有效节点,则出队直到找到无效节点。

java.util.concurrent.locks.AbstractQueuedSynchronizer.ConditionObject#doSignal

/**
 * Removes and transfers nodes until hit non-cancelled one or
 * null. Split out from signal in part to encourage compilers
 * to inline the case of no waiters.
 * @param first (non-null) the first node on condition queue
 */
private void doSignal(Node first) { // 把条件队列的头节点转移到同步队列
    do {if ( (firstWaiter = first.nextWaiter) == null) // 以后节点的后继节点作为新的头节点(出队),若为空,阐明队列为空
            lastWaiter = null;
        first.nextWaiter = null;
    } while (!transferForSignal(first) &&    // 把以后节点转移到同步队列(入队),并唤醒节点上的线程(阐明条件队列的头节点不是 dummy node)(first = firstWaiter) != null); // 转移失败,取最新的 firstWaiter,若不为空则重试,若为空,阐明队列为空
}

4.1.2 transferForSignal

  1. 批改节点状态:CONDITION -> 0
  2. 退出到同步队列
  3. 唤醒节点

java.util.concurrent.locks.AbstractQueuedSynchronizer#transferForSignal

/**
 * Transfers a node from a condition queue onto sync queue. // 将一个节点从条件队列转移到同步队列
 * Returns true if successful.
 * @param node the node
 * @return true if successfully transferred (else the node was
 * cancelled before signal)
 */
final boolean transferForSignal(Node node) {
    /*
     * If cannot change waitStatus, the node has been cancelled. // 条件队列上的节点状态不为 CONDITION,阐明是已勾销
     */
    if (!compareAndSetWaitStatus(node, Node.CONDITION, 0))
        return false;

    /*
     * Splice onto queue and try to set waitStatus of predecessor to
     * indicate that thread is (probably) waiting. If cancelled or
     * attempt to set waitStatus fails, wake up to resync (in which
     * case the waitStatus can be transiently and harmlessly wrong).
     */
    Node p = enq(node); // 增加到同步队列,返回上一个节点
    int ws = p.waitStatus;
    if (ws > 0 || !compareAndSetWaitStatus(p, ws, Node.SIGNAL))
        LockSupport.unpark(node.thread); // 查看上一个节点,发现不具备唤醒以后节点条件,则立刻唤醒以后节点
    return true;                         // 补充:node.thread 从 Condition#await 之中被唤醒,后续执行 acquireQueued 尝试获取锁
}

4.2 Condition#signalAll

遍历条件队列,顺次唤醒所有节点。
所有节点都会迁徙到同步队列期待获取锁。

java.util.concurrent.locks.AbstractQueuedSynchronizer.ConditionObject#signalAll

/**
 * Moves all threads from the wait queue for this condition to
 * the wait queue for the owning lock.
 *
 * @throws IllegalMonitorStateException if {@link #isHeldExclusively}
 *         returns {@code false}
 */
public final void signalAll() {if (!isHeldExclusively())
        throw new IllegalMonitorStateException();
    Node first = firstWaiter;
    if (first != null)
        doSignalAll(first);
}

java.util.concurrent.locks.AbstractQueuedSynchronizer.ConditionObject#doSignalAll

/**
 * Removes and transfers all nodes.
 * @param first (non-null) the first node on condition queue
 */
private void doSignalAll(Node first) {
    lastWaiter = firstWaiter = null;
    do {
        Node next = first.nextWaiter;
        first.nextWaiter = null;
        transferForSignal(first);
        first = next;
    } while (first != null);
}

相干浏览:
浏览 JDK 8 源码:AQS 中的独占模式
浏览 JDK 8 源码:AQS 中的共享模式
浏览 JDK 8 源码:AQS 对 Condition 的实现

作者:Sumkor

正文完
 0