若有不正之处请多多谅解，并欢迎批评指正。

请尊重作者劳动成果，转载请标明原文链接：

http://www.cnblogs.com/go2sea/p/5625536.html

Semaphore是JUC包提供的一个共享锁，一般称之为信号量。

Semaphore通过自定义的同步器维护了一个或多个共享资源，线程通过调用acquire获取共享资源，通过调用release释放。

源代码：

/*
 * ORACLE PROPRIETARY/CONFIDENTIAL. Use is subject to license terms.
 *
 *
 *
 *
 *
 *
 *
 *
 *
 *
 *
 *
 *
 *
 *
 *
 *
 *
 *
 *
 */

/*
 *
 *
 *
 *
 *
 * Written by Doug Lea with assistance from members of JCP JSR-166
 * Expert Group and released to the public domain, as explained at
 * http://creativecommons.org/publicdomain/zero/1.0/
 */

package java.util.concurrent;
import java.util.*;
import java.util.concurrent.locks.*;
import java.util.concurrent.atomic.*;

/**
 * A counting semaphore.  Conceptually, a semaphore maintains a set of
 * permits.  Each {@link #acquire} blocks if necessary until a permit is
 * available, and then takes it.  Each {@link #release} adds a permit,
 * potentially releasing a blocking acquirer.
 * However, no actual permit objects are used; the {@code Semaphore} just
 * keeps a count of the number available and acts accordingly.
 *
 * <p>Semaphores are often used to restrict the number of threads than can
 * access some (physical or logical) resource. For example, here is
 * a class that uses a semaphore to control access to a pool of items:
 * <pre>
 * class Pool {
 *   private static final int MAX_AVAILABLE = 100;
 *   private final Semaphore available = new Semaphore(MAX_AVAILABLE, true);
 *
 *   public Object getItem() throws InterruptedException {
 *     available.acquire();
 *     return getNextAvailableItem();
 *   }
 *
 *   public void putItem(Object x) {
 *     if (markAsUnused(x))
 *       available.release();
 *   }
 *
 *   // Not a particularly efficient data structure; just for demo
 *
 *   protected Object[] items = ... whatever kinds of items being managed
 *   protected boolean[] used = new boolean[MAX_AVAILABLE];
 *
 *   protected synchronized Object getNextAvailableItem() {
 *     for (int i = 0; i < MAX_AVAILABLE; ++i) {
 *       if (!used[i]) {
 *          used[i] = true;
 *          return items[i];
 *       }
 *     }
 *     return null; // not reached
 *   }
 *
 *   protected synchronized boolean markAsUnused(Object item) {
 *     for (int i = 0; i < MAX_AVAILABLE; ++i) {
 *       if (item == items[i]) {
 *          if (used[i]) {
 *            used[i] = false;
 *            return true;
 *          } else
 *            return false;
 *       }
 *     }
 *     return false;
 *   }
 *
 * }
 * </pre>
 *
 * <p>Before obtaining an item each thread must acquire a permit from
 * the semaphore, guaranteeing that an item is available for use. When
 * the thread has finished with the item it is returned back to the
 * pool and a permit is returned to the semaphore, allowing another
 * thread to acquire that item.  Note that no synchronization lock is
 * held when {@link #acquire} is called as that would prevent an item
 * from being returned to the pool.  The semaphore encapsulates the
 * synchronization needed to restrict access to the pool, separately
 * from any synchronization needed to maintain the consistency of the
 * pool itself.
 *
 * <p>A semaphore initialized to one, and which is used such that it
 * only has at most one permit available, can serve as a mutual
 * exclusion lock.  This is more commonly known as a <em>binary
 * semaphore</em>, because it only has two states: one permit
 * available, or zero permits available.  When used in this way, the
 * binary semaphore has the property (unlike many {@link Lock}
 * implementations), that the &quot;lock&quot; can be released by a
 * thread other than the owner (as semaphores have no notion of
 * ownership).  This can be useful in some specialized contexts, such
 * as deadlock recovery.
 *
 * <p> The constructor for this class optionally accepts a
 * <em>fairness</em> parameter. When set false, this class makes no
 * guarantees about the order in which threads acquire permits. In
 * particular, <em>barging</em> is permitted, that is, a thread
 * invoking {@link #acquire} can be allocated a permit ahead of a
 * thread that has been waiting - logically the new thread places itself at
 * the head of the queue of waiting threads. When fairness is set true, the
 * semaphore guarantees that threads invoking any of the {@link
 * #acquire() acquire} methods are selected to obtain permits in the order in
 * which their invocation of those methods was processed
 * (first-in-first-out; FIFO). Note that FIFO ordering necessarily
 * applies to specific internal points of execution within these
 * methods.  So, it is possible for one thread to invoke
 * {@code acquire} before another, but reach the ordering point after
 * the other, and similarly upon return from the method.
 * Also note that the untimed {@link #tryAcquire() tryAcquire} methods do not
 * honor the fairness setting, but will take any permits that are
 * available.
 *
 * <p>Generally, semaphores used to control resource access should be
 * initialized as fair, to ensure that no thread is starved out from
 * accessing a resource. When using semaphores for other kinds of
 * synchronization control, the throughput advantages of non-fair
 * ordering often outweigh fairness considerations.
 *
 * <p>This class also provides convenience methods to {@link
 * #acquire(int) acquire} and {@link #release(int) release} multiple
 * permits at a time.  Beware of the increased risk of indefinite
 * postponement when these methods are used without fairness set true.
 *
 * <p>Memory consistency effects: Actions in a thread prior to calling
 * a "release" method such as {@code release()}
 * <a href="package-summary.html#MemoryVisibility"><i>happen-before</i></a>
 * actions following a successful "acquire" method such as {@code acquire()}
 * in another thread.
 *
 * @since 1.5
 * @author Doug Lea
 *
 */

public class Semaphore implements java.io.Serializable {
    private static final long serialVersionUID = -3222578661600680210L;
    /** All mechanics via AbstractQueuedSynchronizer subclass */
    private final Sync sync;

    /**
     * Synchronization implementation for semaphore.  Uses AQS state
     * to represent permits. Subclassed into fair and nonfair
     * versions.
     */
    abstract static class Sync extends AbstractQueuedSynchronizer {
        private static final long serialVersionUID = 1192457210091910933L;

        Sync(int permits) {
            setState(permits);
        }

        final int getPermits() {
            return getState();
        }

        final int nonfairTryAcquireShared(int acquires) {
            for (;;) {
                int available = getState();
                int remaining = available - acquires;
                if (remaining < 0 ||
                    compareAndSetState(available, remaining))
                    return remaining;
            }
        }

        protected final boolean tryReleaseShared(int releases) {
            for (;;) {
                int current = getState();
                int next = current + releases;
                if (next < current) // overflow
                    throw new Error("Maximum permit count exceeded");
                if (compareAndSetState(current, next))
                    return true;
            }
        }

        final void reducePermits(int reductions) {
            for (;;) {
                int current = getState();
                int next = current - reductions;
                if (next > current) // underflow
                    throw new Error("Permit count underflow");
                if (compareAndSetState(current, next))
                    return;
            }
        }

        final int drainPermits() {
            for (;;) {
                int current = getState();
                if (current == 0 || compareAndSetState(current, 0))
                    return current;
            }
        }
    }

    /**
     * NonFair version
     */
    static final class NonfairSync extends Sync {
        private static final long serialVersionUID = -2694183684443567898L;

        NonfairSync(int permits) {
            super(permits);
        }

        protected int tryAcquireShared(int acquires) {
            return nonfairTryAcquireShared(acquires);
        }
    }

    /**
     * Fair version
     */
    static final class FairSync extends Sync {
        private static final long serialVersionUID = 2014338818796000944L;

        FairSync(int permits) {
            super(permits);
        }

        protected int tryAcquireShared(int acquires) {
            for (;;) {
                if (hasQueuedPredecessors())
                    return -1;
                int available = getState();
                int remaining = available - acquires;
                if (remaining < 0 ||
                    compareAndSetState(available, remaining))
                    return remaining;
            }
        }
    }

    /**
     * Creates a {@code Semaphore} with the given number of
     * permits and nonfair fairness setting.
     *
     * @param permits the initial number of permits available.
     *        This value may be negative, in which case releases
     *        must occur before any acquires will be granted.
     */
    public Semaphore(int permits) {
        sync = new NonfairSync(permits);
    }

    /**
     * Creates a {@code Semaphore} with the given number of
     * permits and the given fairness setting.
     *
     * @param permits the initial number of permits available.
     *        This value may be negative, in which case releases
     *        must occur before any acquires will be granted.
     * @param fair {@code true} if this semaphore will guarantee
     *        first-in first-out granting of permits under contention,
     *        else {@code false}
     */
    public Semaphore(int permits, boolean fair) {
        sync = fair ? new FairSync(permits) : new NonfairSync(permits);
    }

    /**
     * Acquires a permit from this semaphore, blocking until one is
     * available, or the thread is {@linkplain Thread#interrupt interrupted}.
     *
     * <p>Acquires a permit, if one is available and returns immediately,
     * reducing the number of available permits by one.
     *
     * <p>If no permit is available then the current thread becomes
     * disabled for thread scheduling purposes and lies dormant until
     * one of two things happens:
     * <ul>
     * <li>Some other thread invokes the {@link #release} method for this
     * semaphore and the current thread is next to be assigned a permit; or
     * <li>Some other thread {@linkplain Thread#interrupt interrupts}
     * the current thread.
     * </ul>
     *
     * <p>If the current thread:
     * <ul>
     * <li>has its interrupted status set on entry to this method; or
     * <li>is {@linkplain Thread#interrupt interrupted} while waiting
     * for a permit,
     * </ul>
     * then {@link InterruptedException} is thrown and the current thread's
     * interrupted status is cleared.
     *
     * @throws InterruptedException if the current thread is interrupted
     */
    public void acquire() throws InterruptedException {
        sync.acquireSharedInterruptibly(1);
    }

    /**
     * Acquires a permit from this semaphore, blocking until one is
     * available.
     *
     * <p>Acquires a permit, if one is available and returns immediately,
     * reducing the number of available permits by one.
     *
     * <p>If no permit is available then the current thread becomes
     * disabled for thread scheduling purposes and lies dormant until
     * some other thread invokes the {@link #release} method for this
     * semaphore and the current thread is next to be assigned a permit.
     *
     * <p>If the current thread is {@linkplain Thread#interrupt interrupted}
     * while waiting for a permit then it will continue to wait, but the
     * time at which the thread is assigned a permit may change compared to
     * the time it would have received the permit had no interruption
     * occurred.  When the thread does return from this method its interrupt
     * status will be set.
     */
    public void acquireUninterruptibly() {
        sync.acquireShared(1);
    }

    /**
     * Acquires a permit from this semaphore, only if one is available at the
     * time of invocation.
     *
     * <p>Acquires a permit, if one is available and returns immediately,
     * with the value {@code true},
     * reducing the number of available permits by one.
     *
     * <p>If no permit is available then this method will return
     * immediately with the value {@code false}.
     *
     * <p>Even when this semaphore has been set to use a
     * fair ordering policy, a call to {@code tryAcquire()} <em>will</em>
     * immediately acquire a permit if one is available, whether or not
     * other threads are currently waiting.
     * This &quot;barging&quot; behavior can be useful in certain
     * circumstances, even though it breaks fairness. If you want to honor
     * the fairness setting, then use
     * {@link #tryAcquire(long, TimeUnit) tryAcquire(0, TimeUnit.SECONDS) }
     * which is almost equivalent (it also detects interruption).
     *
     * @return {@code true} if a permit was acquired and {@code false}
     *         otherwise
     */
    public boolean tryAcquire() {
        return sync.nonfairTryAcquireShared(1) >= 0;
    }

    /**
     * Acquires a permit from this semaphore, if one becomes available
     * within the given waiting time and the current thread has not
     * been {@linkplain Thread#interrupt interrupted}.
     *
     * <p>Acquires a permit, if one is available and returns immediately,
     * with the value {@code true},
     * reducing the number of available permits by one.
     *
     * <p>If no permit is available then the current thread becomes
     * disabled for thread scheduling purposes and lies dormant until
     * one of three things happens:
     * <ul>
     * <li>Some other thread invokes the {@link #release} method for this
     * semaphore and the current thread is next to be assigned a permit; or
     * <li>Some other thread {@linkplain Thread#interrupt interrupts}
     * the current thread; or
     * <li>The specified waiting time elapses.
     * </ul>
     *
     * <p>If a permit is acquired then the value {@code true} is returned.
     *
     * <p>If the current thread:
     * <ul>
     * <li>has its interrupted status set on entry to this method; or
     * <li>is {@linkplain Thread#interrupt interrupted} while waiting
     * to acquire a permit,
     * </ul>
     * then {@link InterruptedException} is thrown and the current thread's
     * interrupted status is cleared.
     *
     * <p>If the specified waiting time elapses then the value {@code false}
     * is returned.  If the time is less than or equal to zero, the method
     * will not wait at all.
     *
     * @param timeout the maximum time to wait for a permit
     * @param unit the time unit of the {@code timeout} argument
     * @return {@code true} if a permit was acquired and {@code false}
     *         if the waiting time elapsed before a permit was acquired
     * @throws InterruptedException if the current thread is interrupted
     */
    public boolean tryAcquire(long timeout, TimeUnit unit)
        throws InterruptedException {
        return sync.tryAcquireSharedNanos(1, unit.toNanos(timeout));
    }

    /**
     * Releases a permit, returning it to the semaphore.
     *
     * <p>Releases a permit, increasing the number of available permits by
     * one.  If any threads are trying to acquire a permit, then one is
     * selected and given the permit that was just released.  That thread
     * is (re)enabled for thread scheduling purposes.
     *
     * <p>There is no requirement that a thread that releases a permit must
     * have acquired that permit by calling {@link #acquire}.
     * Correct usage of a semaphore is established by programming convention
     * in the application.
     */
    public void release() {
        sync.releaseShared(1);
    }

    /**
     * Acquires the given number of permits from this semaphore,
     * blocking until all are available,
     * or the thread is {@linkplain Thread#interrupt interrupted}.
     *
     * <p>Acquires the given number of permits, if they are available,
     * and returns immediately, reducing the number of available permits
     * by the given amount.
     *
     * <p>If insufficient permits are available then the current thread becomes
     * disabled for thread scheduling purposes and lies dormant until
     * one of two things happens:
     * <ul>
     * <li>Some other thread invokes one of the {@link #release() release}
     * methods for this semaphore, the current thread is next to be assigned
     * permits and the number of available permits satisfies this request; or
     * <li>Some other thread {@linkplain Thread#interrupt interrupts}
     * the current thread.
     * </ul>
     *
     * <p>If the current thread:
     * <ul>
     * <li>has its interrupted status set on entry to this method; or
     * <li>is {@linkplain Thread#interrupt interrupted} while waiting
     * for a permit,
     * </ul>
     * then {@link InterruptedException} is thrown and the current thread's
     * interrupted status is cleared.
     * Any permits that were to be assigned to this thread are instead
     * assigned to other threads trying to acquire permits, as if
     * permits had been made available by a call to {@link #release()}.
     *
     * @param permits the number of permits to acquire
     * @throws InterruptedException if the current thread is interrupted
     * @throws IllegalArgumentException if {@code permits} is negative
     */
    public void acquire(int permits) throws InterruptedException {
        if (permits < 0) throw new IllegalArgumentException();
        sync.acquireSharedInterruptibly(permits);
    }

    /**
     * Acquires the given number of permits from this semaphore,
     * blocking until all are available.
     *
     * <p>Acquires the given number of permits, if they are available,
     * and returns immediately, reducing the number of available permits
     * by the given amount.
     *
     * <p>If insufficient permits are available then the current thread becomes
     * disabled for thread scheduling purposes and lies dormant until
     * some other thread invokes one of the {@link #release() release}
     * methods for this semaphore, the current thread is next to be assigned
     * permits and the number of available permits satisfies this request.
     *
     * <p>If the current thread is {@linkplain Thread#interrupt interrupted}
     * while waiting for permits then it will continue to wait and its
     * position in the queue is not affected.  When the thread does return
     * from this method its interrupt status will be set.
     *
     * @param permits the number of permits to acquire
     * @throws IllegalArgumentException if {@code permits} is negative
     *
     */
    public void acquireUninterruptibly(int permits) {
        if (permits < 0) throw new IllegalArgumentException();
        sync.acquireShared(permits);
    }

    /**
     * Acquires the given number of permits from this semaphore, only
     * if all are available at the time of invocation.
     *
     * <p>Acquires the given number of permits, if they are available, and
     * returns immediately, with the value {@code true},
     * reducing the number of available permits by the given amount.
     *
     * <p>If insufficient permits are available then this method will return
     * immediately with the value {@code false} and the number of available
     * permits is unchanged.
     *
     * <p>Even when this semaphore has been set to use a fair ordering
     * policy, a call to {@code tryAcquire} <em>will</em>
     * immediately acquire a permit if one is available, whether or
     * not other threads are currently waiting.  This
     * &quot;barging&quot; behavior can be useful in certain
     * circumstances, even though it breaks fairness. If you want to
     * honor the fairness setting, then use {@link #tryAcquire(int,
     * long, TimeUnit) tryAcquire(permits, 0, TimeUnit.SECONDS) }
     * which is almost equivalent (it also detects interruption).
     *
     * @param permits the number of permits to acquire
     * @return {@code true} if the permits were acquired and
     *         {@code false} otherwise
     * @throws IllegalArgumentException if {@code permits} is negative
     */
    public boolean tryAcquire(int permits) {
        if (permits < 0) throw new IllegalArgumentException();
        return sync.nonfairTryAcquireShared(permits) >= 0;
    }

    /**
     * Acquires the given number of permits from this semaphore, if all
     * become available within the given waiting time and the current
     * thread has not been {@linkplain Thread#interrupt interrupted}.
     *
     * <p>Acquires the given number of permits, if they are available and
     * returns immediately, with the value {@code true},
     * reducing the number of available permits by the given amount.
     *
     * <p>If insufficient permits are available then
     * the current thread becomes disabled for thread scheduling
     * purposes and lies dormant until one of three things happens:
     * <ul>
     * <li>Some other thread invokes one of the {@link #release() release}
     * methods for this semaphore, the current thread is next to be assigned
     * permits and the number of available permits satisfies this request; or
     * <li>Some other thread {@linkplain Thread#interrupt interrupts}
     * the current thread; or
     * <li>The specified waiting time elapses.
     * </ul>
     *
     * <p>If the permits are acquired then the value {@code true} is returned.
     *
     * <p>If the current thread:
     * <ul>
     * <li>has its interrupted status set on entry to this method; or
     * <li>is {@linkplain Thread#interrupt interrupted} while waiting
     * to acquire the permits,
     * </ul>
     * then {@link InterruptedException} is thrown and the current thread's
     * interrupted status is cleared.
     * Any permits that were to be assigned to this thread, are instead
     * assigned to other threads trying to acquire permits, as if
     * the permits had been made available by a call to {@link #release()}.
     *
     * <p>If the specified waiting time elapses then the value {@code false}
     * is returned.  If the time is less than or equal to zero, the method
     * will not wait at all.  Any permits that were to be assigned to this
     * thread, are instead assigned to other threads trying to acquire
     * permits, as if the permits had been made available by a call to
     * {@link #release()}.
     *
     * @param permits the number of permits to acquire
     * @param timeout the maximum time to wait for the permits
     * @param unit the time unit of the {@code timeout} argument
     * @return {@code true} if all permits were acquired and {@code false}
     *         if the waiting time elapsed before all permits were acquired
     * @throws InterruptedException if the current thread is interrupted
     * @throws IllegalArgumentException if {@code permits} is negative
     */
    public boolean tryAcquire(int permits, long timeout, TimeUnit unit)
        throws InterruptedException {
        if (permits < 0) throw new IllegalArgumentException();
        return sync.tryAcquireSharedNanos(permits, unit.toNanos(timeout));
    }

    /**
     * Releases the given number of permits, returning them to the semaphore.
     *
     * <p>Releases the given number of permits, increasing the number of
     * available permits by that amount.
     * If any threads are trying to acquire permits, then one
     * is selected and given the permits that were just released.
     * If the number of available permits satisfies that thread's request
     * then that thread is (re)enabled for thread scheduling purposes;
     * otherwise the thread will wait until sufficient permits are available.
     * If there are still permits available
     * after this thread's request has been satisfied, then those permits
     * are assigned in turn to other threads trying to acquire permits.
     *
     * <p>There is no requirement that a thread that releases a permit must
     * have acquired that permit by calling {@link Semaphore#acquire acquire}.
     * Correct usage of a semaphore is established by programming convention
     * in the application.
     *
     * @param permits the number of permits to release
     * @throws IllegalArgumentException if {@code permits} is negative
     */
    public void release(int permits) {
        if (permits < 0) throw new IllegalArgumentException();
        sync.releaseShared(permits);
    }

    /**
     * Returns the current number of permits available in this semaphore.
     *
     * <p>This method is typically used for debugging and testing purposes.
     *
     * @return the number of permits available in this semaphore
     */
    public int availablePermits() {
        return sync.getPermits();
    }

    /**
     * Acquires and returns all permits that are immediately available.
     *
     * @return the number of permits acquired
     */
    public int drainPermits() {
        return sync.drainPermits();
    }

    /**
     * Shrinks the number of available permits by the indicated
     * reduction. This method can be useful in subclasses that use
     * semaphores to track resources that become unavailable. This
     * method differs from {@code acquire} in that it does not block
     * waiting for permits to become available.
     *
     * @param reduction the number of permits to remove
     * @throws IllegalArgumentException if {@code reduction} is negative
     */
    protected void reducePermits(int reduction) {
        if (reduction < 0) throw new IllegalArgumentException();
        sync.reducePermits(reduction);
    }

    /**
     * Returns {@code true} if this semaphore has fairness set true.
     *
     * @return {@code true} if this semaphore has fairness set true
     */
    public boolean isFair() {
        return sync instanceof FairSync;
    }

    /**
     * Queries whether any threads are waiting to acquire. Note that
     * because cancellations may occur at any time, a {@code true}
     * return does not guarantee that any other thread will ever
     * acquire.  This method is designed primarily for use in
     * monitoring of the system state.
     *
     * @return {@code true} if there may be other threads waiting to
     *         acquire the lock
     */
    public final boolean hasQueuedThreads() {
        return sync.hasQueuedThreads();
    }

    /**
     * Returns an estimate of the number of threads waiting to acquire.
     * The value is only an estimate because the number of threads may
     * change dynamically while this method traverses internal data
     * structures.  This method is designed for use in monitoring of the
     * system state, not for synchronization control.
     *
     * @return the estimated number of threads waiting for this lock
     */
    public final int getQueueLength() {
        return sync.getQueueLength();
    }

    /**
     * Returns a collection containing threads that may be waiting to acquire.
     * Because the actual set of threads may change dynamically while
     * constructing this result, the returned collection is only a best-effort
     * estimate.  The elements of the returned collection are in no particular
     * order.  This method is designed to facilitate construction of
     * subclasses that provide more extensive monitoring facilities.
     *
     * @return the collection of threads
     */
    protected Collection<Thread> getQueuedThreads() {
        return sync.getQueuedThreads();
    }

    /**
     * Returns a string identifying this semaphore, as well as its state.
     * The state, in brackets, includes the String {@code "Permits ="}
     * followed by the number of permits.
     *
     * @return a string identifying this semaphore, as well as its state
     */
    public String toString() {
        return super.toString() + "[Permits = " + sync.getPermits() + "]";
    }
}

View Code

下面我们来详细分下下Semaphore的工作原理。

一、构造函数

    public Semaphore(int permits) {
        sync = new NonfairSync(permits);
    }

    public Semaphore(int permits, boolean fair) {
        sync = fair ? new FairSync(permits) : new NonfairSync(permits);
    }

初始化Semaphore时需要指定共享资源的个数。Semaphore提供了两种模式：公平模式&非公平模式。如果不指定工作模式的话，默认工作在非公平模式下。后面我们将看到，两种模式的区别在于获取共享资源时的排序策略。Semaphore有三个内部类：Sync&NonfairSync&FairSync。后两个继承自Sync，Sync继承自AQS。除了序列化版本号之外，Semaphore只有一个成员变量sync，公平模式下sync初始化为FairSync，非公平模式下sync初始化为NonfairSync。

二、acquire 响应中断获取资源

Semaphore提供了两种获取资源的方式：响应中断&不响应中断。我们先来看一下响应中断的获取。

    public void acquire() throws InterruptedException {
        sync.acquireSharedInterruptibly(1);
    }

acquire方法由同步器sync调用上层AQS提供的acquireSharedInterruptibly方法获取：

    public final void acquireSharedInterruptibly(int arg)
            throws InterruptedException {
        if (Thread.interrupted())
            throw new InterruptedException();
        if (tryAcquireShared(arg) < 0)
            doAcquireSharedInterruptibly(arg);
    }

acquireSharedInterruptibly方法先检测中断。然后调用tryAcquireShared方法试图获取共享资源。这时公平模式和非公平模式的代码执行路径发生分叉，FairSync和NonfairSync各自重写了tryAcquireShared方法。

我们先来看下非公平模式下的tryAcquireShared方法：

        protected int tryAcquireShared(int acquires) {
            return nonfairTryAcquireShared(acquires);
        }

它直接代用了父类Sync提供的nonfairTryAcquireShared方法：

        final int nonfairTryAcquireShared(int acquires) {
            for (;;) {
                int available = getState();
                int remaining = available - acquires;
                if (remaining < 0 ||
                    compareAndSetState(available, remaining))
                    return remaining;
            }
        }

注意，这里是一个CAS自旋。因为Semaphore是一个共享锁，可能有多个线程同时申请共享资源，因此CAS操作可能失败。直到成功获取返回剩余资源数目，或者发现没有剩余资源返回负值代表申请失败。有一个问题，为什么我们不在CAS操作失败后就直接返回失败呢？因为这样做虽然不会导致错误，但会降低效率：在还有剩余资源的情况下，一个线程因为竞争导致CAS失败后被放入等待序列尾，一定在队列头部有一个线程被唤醒去试图获取资源，这比直接自旋继续获取多了操作等待队列的开销。

这里“非公平”的语义体现在：如果一个线程通过nonfairTryAcquireShared成功获取了共享资源，对于此时正在等待队列中的线程来说，可能是不公平的：队列中线程先到，却没能先获取资源。

如果tryAcquireShared没能成功获取，acquireSharedInterruptibly方法调用doAcquireSharedInterruptibly方法将当前线程放入等待队列并开始自旋检测获取资源：

    private void doAcquireSharedInterruptibly(int arg)
        throws InterruptedException {
        final Node node = addWaiter(Node.SHARED);
        boolean failed = true;
        try {
            for (;;) {
                final Node p = node.predecessor();
                if (p == head) {
                    int r = tryAcquireShared(arg);
                    if (r >= 0) {
                        setHeadAndPropagate(node, r);
                        p.next = null; // help GC
                        failed = false;
                        return;
                    }
                }
                if (shouldParkAfterFailedAcquire(p, node) &&
                    parkAndCheckInterrupt())
                    throw new InterruptedException();
            }
        } finally {
            if (failed)
                cancelAcquire(node);
        }
    }

我们注意到，doAcquireSharedInterruptibly中，当一个线程从parkAndCheckInterrupt方法中被中断唤醒之后，直接抛出了中断异常。还记得我们分析AQS时的doAcquireShared方法吗，它在这里的处理方式是用一个局部变量interrupted记录下这个异常但不立即处理，而是等到成功获取资源之后返回这个中断标志，并在上层调用selfInterrupt方法补上中断。这正是两个方法的关键区别：是否及时响应中断。

我们再来看公平模式下的tryAcquireShared方法：

        protected int tryAcquireShared(int acquires) {
            for (;;) {
                if (hasQueuedPredecessors())
                    return -1;
                int available = getState();
                int remaining = available - acquires;
                if (remaining < 0 ||
                    compareAndSetState(available, remaining))
                    return remaining;
            }
        }

相比较非公平模式的nonfairTryAcquireShared方法，公平模式下的tryAcquireShared方法在试图获取之前做了一个判断，如果发现等对队列中有线程在等待获取资源，就直接返回-1表示获取失败。当前线程会被上层的acquireSharedInterruptibly方法调用doAcquireShared方法放入等待队列中。这正是“公平”模式的语义：如果有线程先于我进入等待队列且正在等待，就直接进入等待队列，效果便是各个线程按照申请的顺序获得共享资源，具有公平性。

三、acquireUnInterruptibly 不响应中断获取资源

    public void acquireUninterruptibly() {
        sync.acquireShared(1);
    }

acquireUnInterruptibly方法调用AQS提供的acquireShared方法：

    public final void acquireShared(int arg) {
        if (tryAcquireShared(arg) < 0)
            doAcquireShared(arg);
    }

acquireShared方法首先试图获取资源，这与acquireSharedInterruptibly方法相比，没有先检测中断的这一步。紧接着调用doAcquireShared方法，由于这个方法我在另一篇博文AQS源码学习笔记中已经详细分析过，这里我们只关注它与doAcquireSharedInterruptibly方法的区别：

    private void doAcquireShared(int arg) {
        final Node node = addWaiter(Node.SHARED);
        boolean failed = true;
        try {
            boolean interrupted = false;
            for (;;) {
                final Node p = node.predecessor();
                if (p == head) {
                    int r = tryAcquireShared(arg);
                    if (r >= 0) {
                        setHeadAndPropagate(node, r);
                        p.next = null; // help GC
                        if (interrupted)
                            selfInterrupt();
                        failed = false;
                        return;
                    }
                }
                if (shouldParkAfterFailedAcquire(p, node) &&
                    parkAndCheckInterrupt())
                    interrupted = true;
            }
        } finally {
            if (failed)
                cancelAcquire(node);
        }
    }

正如刚刚说过的，区别只在线程从parkAndCheckInterrupt方法中因中断而返回时的处理：在这里它没有抛出异常，而是用一个局部变量interrupted记录下这个异常但不立即处理，而是等到成功获取资源之后返回这个中断标志，并在上层调用selfInterrupt方法补上中断。

四、acquire(int) & acquireUninterruptibly(int) 指定申请的资源数目的获取

    public void acquire(int permits) throws InterruptedException {
        if (permits < 0) throw new IllegalArgumentException();
        sync.acquireSharedInterruptibly(permits);
    }

    public void acquireUninterruptibly(int permits) {
        if (permits < 0) throw new IllegalArgumentException();
        sync.acquireShared(permits);
    }

可以看到，与不指定数目时的获取的区别仅在参数值，不再赘述。

五、release 释放资源

公平模式和非公平模式的释放资源操作是一样的：

    public void release() {
        sync.releaseShared(1);
    }
    
    public void release(int permits) {
        if (permits < 0) throw new IllegalArgumentException();
        sync.releaseShared(permits);
    }

调用AQS提供的releaseShared方法：

    public final boolean releaseShared(int arg) {
        if (tryReleaseShared(arg)) {
            doReleaseShared();
            return true;
        }
        return false;
    }

releaseShared方法首先调用我们重写的tryReleaseShared方法试图释放资源。然后调用doReleaseShared方法唤醒队列之后的等待线程。由于在我的另一篇博文AQS源码学习笔记中已经详细分析了doReleaseShared方法，因此不再赘述。我们主要关注tryReleaseShared方法：

        protected final boolean tryReleaseShared(int releases) {
            for (;;) {
                int current = getState();
                int next = current + releases;
                if (next < current) // overflow
                    throw new Error("Maximum permit count exceeded");
                if (compareAndSetState(current, next))
                    return true;
            }
        }

这个方法也是一个CAS自旋，原因是应为Semaphore是一个共享锁，可能有多个线程同时释放资源，因此CAS操作可能失败。最后方法总会成功释放并返回true（如果不出错的话）。

六、tryAcquire & tryAcquire(timeout) 方法

    public boolean tryAcquire() {
        return sync.nonfairTryAcquireShared(1) >= 0;
    }

    public boolean tryAcquire(long timeout, TimeUnit unit)
        throws InterruptedException {
        return sync.tryAcquireSharedNanos(1, unit.toNanos(timeout));
    }

    public boolean tryAcquire(int permits) {
        if (permits < 0) throw new IllegalArgumentException();
        return sync.nonfairTryAcquireShared(permits) >= 0;
    }

    public boolean tryAcquire(int permits, long timeout, TimeUnit unit)
        throws InterruptedException {
        if (permits < 0) throw new IllegalArgumentException();
        return sync.tryAcquireSharedNanos(permits, unit.toNanos(timeout));
    }

没有指定等待时间的tryAcquire调用的是nonfairTryAcquireShared方法，我们已经分析过，不再赘述。我们重点关注指定等待时长的方法。限时等待是通过调用AQS提供的tryAcquireSharedNanos方法实现的：

    public final boolean tryAcquireSharedNanos(int arg, long nanosTimeout)
            throws InterruptedException {
        if (Thread.interrupted())
            throw new InterruptedException();
        return tryAcquireShared(arg) >= 0 ||
            doAcquireSharedNanos(arg, nanosTimeout);
    }

注意：限时等待默认都是及时响应中断的。方法开始先检测中断，然后调用tryAcquireShared方法试图获取资源，如果成功的话直接返回true，不成功则调用doAcquireSharedNanos方法：

    private boolean doAcquireSharedNanos(int arg, long nanosTimeout)
            throws InterruptedException {
        if (nanosTimeout <= 0L)
            return false;
        final long deadline = System.nanoTime() + nanosTimeout;
        final Node node = addWaiter(Node.SHARED);
        boolean failed = true;
        try {
            for (;;) {
                final Node p = node.predecessor();
                if (p == head) {
                    int r = tryAcquireShared(arg);
                    if (r >= 0) {
                        setHeadAndPropagate(node, r);
                        p.next = null; // help GC
                        failed = false;
                        return true;
                    }
                }
                nanosTimeout = deadline - System.nanoTime();
                if (nanosTimeout <= 0L)
                    return false;
                if (shouldParkAfterFailedAcquire(p, node) &&
                    nanosTimeout > spinForTimeoutThreshold)
                    LockSupport.parkNanos(this, nanosTimeout);
                if (Thread.interrupted())
                    throw new InterruptedException();
            }
        } finally {
            if (failed)
                cancelAcquire(node);
        }
    }

方法在自旋之前先计算了一个结束等待的时间节点deadline，然后便开始自旋，每次自旋都要计算一下剩余等待时间nanosTimeout，如果nanosTimeout小于等于0，说明已经到达deadline，直接返回false表示超时。

有一点值得注意，spinForTimeoutThreshold这个值规定了一个阈值，当剩余等待时间小于这个值的时候，线程将不再被park，而是一直在自旋试图获取资源。关于这个值的作用Doug Lea是这样注释的：

    /**
     * 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.
     */

我的理解是，park和unpark操作需要一定的开销，当nanosTimeout很小的时候，这个开销就相对很大了。这个阈值的设置可以让短时等待的线程一直保持自旋，可以提高短时等待的反应效率，而由于nanosTimeout很小，自旋又不会有过多的开销。

除此之外，doAcquireSharedNanos方法与不限时等待的doAcquireShared方法还有两点重要区别：①由于有等待时限，所以线程从park方法返回时我们不能确定返回的原因是中断还是超时，因此需要调用interrupted方法检测一下中断标志；②doAcquireSharedNanos方法是及时响应中断的，而doAcquireShared方法延迟处理中断。

七、drainPermits & reducePermits 修改剩余共享资源数量

Semaphore提供了“耗尽”所有剩余共享资源的操作：

    public int drainPermits() {
        return sync.drainPermits();
    }

drainPermits调用了自定义同步器Sync的同名方法：

        final int drainPermits() {
            for (;;) {
                int current = getState();
                if (current == 0 || compareAndSetState(current, 0))
                    return current;
            }
        }

用CAS自旋将剩余资源清空。

我们再来看看“缩减”剩余共享资源的操作：

    protected void reducePermits(int reduction) {
        if (reduction < 0) throw new IllegalArgumentException();
        sync.reducePermits(reduction);
    }

首先，缩减必须是单向的，即只能减少不能增加，然后调用Sync的同名方法：

        final void reducePermits(int reductions) {
            for (;;) {
                int current = getState();
                int next = current - reductions;
                if (next > current) // underflow
                    throw new Error("Permit count underflow");
                if (compareAndSetState(current, next))
                    return;
            }
        }

用CAS自旋在剩余共享资源上做缩减。

上述两个对共享资源数量的修改操作有两点需要注意：①是不可逆的②是对剩余资源的操作而不是全部资源，当剩余资源数目不足或已经为0时，方法就返回，正咋被占用的资源不参与。

八、其他

    public int availablePermits() {
        return sync.getPermits();
    }public boolean isFair() {
        return sync instanceof FairSync;
    }

    public final boolean hasQueuedThreads() {
        return sync.hasQueuedThreads();
    }

    public final int getQueueLength() {
        return sync.getQueueLength();
    }

    protected Collection<Thread> getQueuedThreads() {
        return sync.getQueuedThreads();
    }

    public String toString() {
        return super.toString() + "[Permits = " + sync.getPermits() + "]";
    }

这些方法比较简单，不再赘述。

总结：

Semaphore是JUC包提供的一个典型的共享锁，它通过自定义两种不同的同步器（FairSync&NonfairSync）提供了公平&非公平两种工作模式，两种模式下分别提供了限时/不限时、响应中断/不响应中断的获取资源的方法（限时获取总是及时响应中断的），而所有的释放资源的release操作是统一的。