ReentrantReadWriteLock 深入源码解析
ReentrantReadWriteLock是JUC提供的读写锁，在某些应用场景下，读操作要比写操作频繁的多，此时应该尽可能利用读写之间协作，减少共享资源的竞争。
| 写操作 | 读操作 | |
|---|---|---|
| 写操作 | 禁止 | 禁止 |
| 读操作 | 禁止 | 允许 |
读写锁的三个重要特性：
· 重入性：无论读锁和写锁都支持线程重入。
· 公平性：支持公平锁和非公平锁，默认是非公平锁，非公平锁吞吐量较高。
· 锁降级：线程在持有写锁时，可以获取读锁，然后释放写锁，最后释放读锁，这就是锁的降级。
· 锁升级：不支持读锁到写锁的升级转换。
演示示例：
首先，新建两个线程类，用于模拟多线程环境下的读写线程：
package com.securitit.serialize.locks;
import java.util.List;
import java.util.concurrent.locks.Lock;
public class ReentrantReadLockThread extends Thread {
// 锁实例.
private Lock lock;
// 模拟数据存储.
private List<String> dataList;
// 构造方法.
public ReentrantReadLockThread(Lock lock, List<String> dataList) {
this.lock = lock;
this.dataList = dataList;
}
@Override
public void run() {
try {
lock.lock();
System.out.println(Thread.currentThread().getName() + "：读锁获得锁.");
for(String data : dataList) {
System.out.println(Thread.currentThread().getName() + "：读锁读数据.");
}
Thread.sleep(2000);
} catch (Exception ex) {
ex.printStackTrace();
} finally {
System.out.println(Thread.currentThread().getName() + "：读锁释放锁.");
lock.unlock();
}
}
}
package com.securitit.serialize.locks;
import java.util.List;
import java.util.concurrent.locks.Lock;
public class ReentrantWriteLockThread extends Thread {
// 锁实例.
private Lock lock;
// 模拟数据存储.
private List<String> dataList;
// 构造方法.
public ReentrantWriteLockThread(Lock lock, List<String> dataList) {
this.lock = lock;
this.dataList = dataList;
}
@Override
public void run() {
try {
lock.lock();
System.out.println(Thread.currentThread().getName() + "：写锁获得锁.");
dataList.add("写锁写入数据");
Thread.sleep(2000);
} catch (Exception ex) {
ex.printStackTrace();
} finally {
System.out.println(Thread.currentThread().getName() + "：写锁释放锁.");
lock.unlock();
}
}
}
然后，在测试类中，启动五个读线程和五个写线程，同时操作数据模拟存储List：
package com.securitit.serialize.locks;
import java.util.ArrayList;
import java.util.List;
import java.util.concurrent.locks.Lock;
import java.util.concurrent.locks.ReentrantReadWriteLock;
public class ReentrantReadWriteLockTester {
// 读写锁实例.
private static ReentrantReadWriteLock readWriteLock = new ReentrantReadWriteLock();
// 读锁.
private static Lock readLock = readWriteLock.readLock();
// 写锁.
private static Lock writeLock = readWriteLock.writeLock();
// 模拟数据存储.
private static List<String> dataList = new ArrayList<String>();
public static void main(String[] args) {
new ReentrantReadLockThread(readLock, dataList).start();
new ReentrantWriteLockThread(writeLock, dataList).start();
new ReentrantReadLockThread(readLock, dataList).start();
new ReentrantWriteLockThread(writeLock, dataList).start();
new ReentrantReadLockThread(readLock, dataList).start();
new ReentrantWriteLockThread(writeLock, dataList).start();
new ReentrantReadLockThread(readLock, dataList).start();
new ReentrantWriteLockThread(writeLock, dataList).start();
new ReentrantReadLockThread(readLock, dataList).start();
new ReentrantWriteLockThread(writeLock, dataList).start();
}
}
输出结果：
Thread-0：读锁获得锁.
Thread-0：读锁释放锁.
Thread-1：写锁获得锁.
Thread-1：写锁释放锁.
Thread-3：写锁获得锁.
Thread-3：写锁释放锁.
Thread-5：写锁获得锁.
Thread-5：写锁释放锁.
Thread-4：读锁获得锁.
Thread-4：读锁读数据.
Thread-4：读锁读数据.
Thread-6：读锁获得锁.
Thread-4：读锁读数据.
Thread-2：读锁获得锁.
Thread-6：读锁读数据.
Thread-2：读锁读数据.
Thread-6：读锁读数据.
Thread-2：读锁读数据.
Thread-6：读锁读数据.
Thread-2：读锁读数据.
Thread-4：读锁释放锁.
Thread-6：读锁释放锁.
Thread-2：读锁释放锁.
Thread-7：写锁获得锁.
Thread-7：写锁释放锁.
Thread-8：读锁获得锁.
Thread-8：读锁读数据.
Thread-8：读锁读数据.
Thread-8：读锁读数据.
Thread-8：读锁读数据.
Thread-8：读锁释放锁.
Thread-9：写锁获得锁.
Thread-9：写锁释放锁.
文中最初提到的锁的其他特性，应用起来也比较简单，可以对上面的示例稍作变换即可实现重入性、公平性、锁降级。
源码分析：
实现基础：
ReentrantReadWriteLock与ReentrantLock实现类似，通过内部类FairSync和NonfairSync来实现作为实现基础：
FairSync：
abstract static class Sync extends AbstractQueuedSynchronizer {
// 序列化版本号.
private static final long serialVersionUID = 6317671515068378041L;
// 高16位为读锁，低16位为写锁.
static final int SHARED_SHIFT = 16;
// 读锁单位.
static final int SHARED_UNIT = (1 << SHARED_SHIFT);
// 读锁最大数量.
static final int MAX_COUNT = (1 << SHARED_SHIFT) - 1;
// 写锁最大数量.
static final int EXCLUSIVE_MASK = (1 << SHARED_SHIFT) - 1;
// 读锁，即共享锁的持有者数量.
static int sharedCount(int c) { return c >>> SHARED_SHIFT; }
// 写锁，即独占锁的持有者数量.
static int exclusiveCount(int c) { return c & EXCLUSIVE_MASK; }
// 计数实现.
static final class HoldCounter {
int count = 0;
final long tid = getThreadId(Thread.currentThread());
}
// ThreadLocal默认实现.
static final class ThreadLocalHoldCounter
extends ThreadLocal<HoldCounter> {
public HoldCounter initialValue() {
return new HoldCounter();
}
}
// 读锁持有者数量.
private transient ThreadLocalHoldCounter readHolds;
// 缓存持有者数量.
private transient HoldCounter cachedHoldCounter;
// 第一个读锁持有者.
private transient Thread firstReader = null;
// 第一个读锁持有者数量.
private transient int firstReaderHoldCount;
Sync() {
readHolds = new ThreadLocalHoldCounter();
setState(getState()); // ensures visibility of readHolds
}
// 读锁持有者应该阻塞.
abstract boolean readerShouldBlock();
// 写锁持有者应该阻塞.
abstract boolean writerShouldBlock();
// 释放锁.
protected final boolean tryRelease(int releases) {
// 是否持有独占锁.
if (!isHeldExclusively())
throw new IllegalMonitorStateException();
// 计算重入次数.
int nextc = getState() - releases;
boolean free = exclusiveCount(nextc) == 0;
if (free)
setExclusiveOwnerThread(null);
setState(nextc);
return free;
}
// 尝试获取独占锁.
protected final boolean tryAcquire(int acquires) {
// 取得当前线程.
Thread current = Thread.currentThread();
int c = getState();
// 独占锁数量.
int w = exclusiveCount(c);
if (c != 0) {
// 条件翻译：独占锁数量为0 || 独占锁持有者非当前线程.
if (w == 0 || current != getExclusiveOwnerThread())
return false;
// 独占锁总数超过上限.
if (w + exclusiveCount(acquires) > MAX_COUNT)
throw new Error("Maximum lock count exceeded");
// Reentrant acquire
setState(c + acquires);
return true;
}
if (writerShouldBlock() ||
!compareAndSetState(c, c + acquires))
return false;
// 设置独占锁持有者.
setExclusiveOwnerThread(current);
return true;
}
// 释放共享锁.
protected final boolean tryReleaseShared(int unused) {
Thread current = Thread.currentThread();
// 当前线程是第一个读锁持有者.
if (firstReader == current) {
// assert firstReaderHoldCount > 0;
if (firstReaderHoldCount == 1)
firstReader = null;
else
firstReaderHoldCount--;
} else {
// 设置cachedHoldCounter.
HoldCounter rh = cachedHoldCounter;
if (rh == null || rh.tid != getThreadId(current))
rh = readHolds.get();
int count = rh.count;
if (count <= 1) {
readHolds.remove();
if (count <= 0)
throw unmatchedUnlockException();
}
--rh.count;
}
for (;;) {
int c = getState();
int nextc = c - SHARED_UNIT;
if (compareAndSetState(c, nextc))
return nextc == 0;
}
}
// 尝试解锁读锁时，当前线程没有持有读锁.
private IllegalMonitorStateException unmatchedUnlockException() {
return new IllegalMonitorStateException(
"attempt to unlock read lock, not locked by current thread");
}
// 获取共享锁.
protected final int tryAcquireShared(int unused) {
// 取得当前线程.
Thread current = Thread.currentThread();
int c = getState();
// 锁被当前线程以外的其他线程持有.
if (exclusiveCount(c) != 0 &&
getExclusiveOwnerThread() != current)
return -1;
// 计算共享数量.
int r = sharedCount(c);
// 读锁是否阻塞.
// 共享数量是否在上限之内.
// 尝试以CAS方式设置锁状态.
if (!readerShouldBlock() &&
r < MAX_COUNT &&
compareAndSetState(c, c + SHARED_UNIT)) {
// 设置firstReader、firstReaderHoldCount的值.
// 设置cachedHoldCounter的值.
// 将HoldCounter加入readHolds.
if (r == 0) {
firstReader = current;
firstReaderHoldCount = 1;
} else if (firstReader == current) {
firstReaderHoldCount++;
} else {
HoldCounter rh = cachedHoldCounter;
if (rh == null || rh.tid != getThreadId(current))
cachedHoldCounter = rh = readHolds.get();
else if (rh.count == 0)
readHolds.set(rh);
rh.count++;
}
return 1;
}
return fullTryAcquireShared(current);
}
// 获取读锁，即共享锁.
final int fullTryAcquireShared(Thread current) {
HoldCounter rh = null;
for (;;) {
int c = getState();
// 锁被当前线程以外的其他线程持有，则返回失败.
if (exclusiveCount(c) != 0) {
if (getExclusiveOwnerThread() != current)
return -1;
} else
// 读锁是否阻塞.
if (readerShouldBlock()) {
// firstReader是当前线程.
if (firstReader == current) {
// assert firstReaderHoldCount > 0;
} else {
// 初始化HoldCounter.
if (rh == null) {
rh = cachedHoldCounter;
if (rh == null || rh.tid != getThreadId(current)) {
rh = readHolds.get();
if (rh.count == 0)
readHolds.remove();
}
}
if (rh.count == 0)
return -1;
}
}
// 共享锁数量是否在上限之内.
if (sharedCount(c) == MAX_COUNT)
throw new Error("Maximum lock count exceeded");
// 尝试以CAS方式设置锁状态.
if (compareAndSetState(c, c + SHARED_UNIT)) {
// 设置firstReader、firstReaderHoldCount的值.
// 设置cachedHoldCounter的值.
// 将HoldCounter加入readHolds.
if (sharedCount(c) == 0) {
firstReader = current;
firstReaderHoldCount = 1;
} else if (firstReader == current) {
firstReaderHoldCount++;
} else {
if (rh == null)
rh = cachedHoldCounter;
if (rh == null || rh.tid != getThreadId(current))
rh = readHolds.get();
else if (rh.count == 0)
readHolds.set(rh);
rh.count++;
cachedHoldCounter = rh; // cache for release
}
return 1;
}
}
}
// 尝试获取写锁.
final boolean tryWriteLock() {
Thread current = Thread.currentThread();
int c = getState();
if (c != 0) {
int w = exclusiveCount(c);
// 若锁被当前线程外的其他线程持有，则获取失败.
if (w == 0 || current != getExclusiveOwnerThread())
return false;
if (w == MAX_COUNT)
throw new Error("Maximum lock count exceeded");
}
// 尝试以CAS方式设置锁状态.
if (!compareAndSetState(c, c + 1))
return false;
// 设置当前线程为锁持有者.
setExclusiveOwnerThread(current);
return true;
}
// 尝试获取读锁.
final boolean tryReadLock() {
Thread current = Thread.currentThread();
for (;;) {
int c = getState();
// 若锁被当前线程以外的线程持有，则获取失败.
if (exclusiveCount(c) != 0 &&
getExclusiveOwnerThread() != current)
return false;
// 计算共享数量.
int r = sharedCount(c);
if (r == MAX_COUNT)
throw new Error("Maximum lock count exceeded");
// 尝试以CAS方式设置锁状态.
if (compareAndSetState(c, c + SHARED_UNIT)) {
// 设置firstReader、firstReaderHoldCount的值.
// 设置cachedHoldCounter的值.
// 将HoldCounter加入readHolds.
if (r == 0) {
firstReader = current;
firstReaderHoldCount = 1;
} else if (firstReader == current) {
firstReaderHoldCount++;
} else {
HoldCounter rh = cachedHoldCounter;
if (rh == null || rh.tid != getThreadId(current))
cachedHoldCounter = rh = readHolds.get();
else if (rh.count == 0)
readHolds.set(rh);
rh.count++;
}
return true;
}
}
}
// 是否独占锁持有者.
protected final boolean isHeldExclusively() {
// 当前线程是否独占锁持有者.
return getExclusiveOwnerThread() == Thread.currentThread();
}
// 获取Condition.
final ConditionObject newCondition() {
return new ConditionObject();
}
// 获取持有者.
final Thread getOwner() {
// Must read state before owner to ensure memory consistency
return ((exclusiveCount(getState()) == 0) ?
null :
getExclusiveOwnerThread());
}
// 获取读锁持有者数量.
final int getReadLockCount() {
return sharedCount(getState());
}
// 当前线程是否写锁持有者.
final boolean isWriteLocked() {
return exclusiveCount(getState()) != 0;
}
// 获取写锁持有者数量.
final int getWriteHoldCount() {
// 若是独占锁持有者，则返回独占锁数量，否则返回0.
return isHeldExclusively() ? exclusiveCount(getState()) : 0;
}
// 获取读锁持有者数量.
final int getReadHoldCount() {
// 若为0，返回0.
if (getReadLockCount() == 0)
return 0;
// 如果firstReader是当前线程，则返回firstReaderHoldCount.
Thread current = Thread.currentThread();
if (firstReader == current)
return firstReaderHoldCount;
// 若缓存持有者为当前线程，则返回.
HoldCounter rh = cachedHoldCounter;
if (rh != null && rh.tid == getThreadId(current))
return rh.count;
// 到读锁集合获取数量，并返回.
int count = readHolds.get().count;
if (count == 0) readHolds.remove();
return count;
}
// 反序列化.
private void readObject(java.io.ObjectInputStream s)
throws java.io.IOException, ClassNotFoundException {
s.defaultReadObject();
readHolds = new ThreadLocalHoldCounter();
setState(0); // reset to unlocked state
}
// 获取锁状态.
final int getCount() { return getState(); }
}
NonFairSync：
static final class NonfairSync extends Sync {
// 序列化版本号.
private static final long serialVersionUID = -8159625535654395037L;
// 写锁是否阻塞.
final boolean writerShouldBlock() {
return false; // writers can always barge
}
// 读锁是否阻塞.
final boolean readerShouldBlock() {
//
return apparentlyFirstQueuedIsExclusive();
}
}
ReadLock：
public static class ReadLock implements Lock, java.io.Serializable {
private static final long serialVersionUID = -5992448646407690164L;
private final Sync sync;
// Constructor.
protected ReadLock(ReentrantReadWriteLock lock) {
sync = lock.sync;
}
// 获得锁,
public void lock() {
sync.acquireShared(1);
}
// 带中断的获取锁.
public void lockInterruptibly() throws InterruptedException {
sync.acquireSharedInterruptibly(1);
}
// 尝试获得锁.
public boolean tryLock() {
return sync.tryReadLock();
}
// 尝试指定等待时间的获取锁.
public boolean tryLock(long timeout, TimeUnit unit)
throws InterruptedException {
return sync.tryAcquireSharedNanos(1, unit.toNanos(timeout));
}
// 尝试释放当前共享锁.
public void unlock() {
sync.releaseShared(1);
}
// 读锁不支持Condition.
public Condition newCondition() {
throw new UnsupportedOperationException();
}
}
WriteLock：
public static class WriteLock implements Lock, java.io.Serializable {
// 序列化版本号.
private static final long serialVersionUID = -4992448646407690164L;
private final Sync sync;
// Constructor.
protected WriteLock(ReentrantReadWriteLock lock) {
sync = lock.sync;
}
// 获得锁.
public void lock() {
sync.acquire(1);
}
// 带中断的获取锁.
public void lockInterruptibly() throws InterruptedException {
sync.acquireInterruptibly(1);
}
// 尝试获取锁.
public boolean tryLock( ) {
return sync.tryWriteLock();
}
// 尝试指定等待时间的获取锁.
public boolean tryLock(long timeout, TimeUnit unit)
throws InterruptedException {
return sync.tryAcquireNanos(1, unit.toNanos(timeout));
}
// 释放锁.
public void unlock() {
sync.release(1);
}
// 获得锁的Condition.
public Condition newCondition() {
return sync.newCondition();
}
// toString().
public String toString() {
Thread o = sync.getOwner();
return super.toString() + ((o == null) ?
"[Unlocked]" :
"[Locked by thread " + o.getName() + "]");
}
// 是否被当前线程持有.
public boolean isHeldByCurrentThread() {
return sync.isHeldExclusively();
}
// 获得持有者数量.
public int getHoldCount() {
return sync.getWriteHoldCount();
}
}
上面的四个内部类是ReentrantReadWriteLock实现的主要基础，首先Sync继承自AQS，AQS本身对独占锁和共享锁进行了抽象和实现，其实独占锁和共享锁对应的就是写锁和读锁。ReadLock和WriteLock实现了Lock接口，具体实现使用内部类Sync已经封装好的关于独占锁和共享锁的实现。
基本方法：
// 是否公平锁.
public final boolean isFair() {
return sync instanceof FairSync;
}
// 获得锁的持有者.
protected Thread getOwner() {
return sync.getOwner();
}
// 获得读锁的数量.
public int getReadLockCount() {
return sync.getReadLockCount();
}
// 写锁是否已被持有.
public boolean isWriteLocked() {
return sync.isWriteLocked();
}
// 当前线程是否获得写锁.
public boolean isWriteLockedByCurrentThread() {
return sync.isHeldExclusively();
}
// 获得写锁持有者数量.
public int getWriteHoldCount() {
return sync.getWriteHoldCount();
}
// 获得读锁持有者数量.
public int getReadHoldCount() {
return sync.getReadHoldCount();
}
// 获得写锁持有者集合.
protected Collection<Thread> getQueuedWriterThreads() {
return sync.getExclusiveQueuedThreads();
}
// 获得读锁持有者集合.
protected Collection<Thread> getQueuedReaderThreads() {
return sync.getSharedQueuedThreads();
}
// 同步队列中是否线程.
public final boolean hasQueuedThreads() {
return sync.hasQueuedThreads();
}
// 同步队列中是否有指定线程.
public final boolean hasQueuedThread(Thread thread) {
return sync.isQueued(thread);
}
// 同步队列长度.
public final int getQueueLength() {
return sync.getQueueLength();
}
// 同步队列的所有线程集合.
protected Collection<Thread> getQueuedThreads() {
return sync.getQueuedThreads();
}
// 是否有指定Condition的等待者.
public boolean hasWaiters(Condition condition) {
if (condition == null)
throw new NullPointerException();
if (!(condition instanceof AbstractQueuedSynchronizer.ConditionObject))
throw new IllegalArgumentException("not owner");
return sync.hasWaiters((AbstractQueuedSynchronizer.ConditionObject)condition);
}
// 指定Condition的等待者数量.
public int getWaitQueueLength(Condition condition) {
if (condition == null)
throw new NullPointerException();
if (!(condition instanceof AbstractQueuedSynchronizer.ConditionObject))
throw new IllegalArgumentException("not owner");
return sync.getWaitQueueLength((AbstractQueuedSynchronizer.ConditionObject)condition);
}
上面是提供的针对同步队列的一些信息获取，可以通过这些方法进行同步队列监控，以便更好的处理业务逻辑。
总结：
在多线程高并发的情况下，如何合理减少资源的竞争是提高并发效率的根本，从Java提供的语言级的synchronized、到JUC的ReentrantLock重入锁、再到读写分离的ReentrantReadWriteLock，或者基于CAS的其他同步工具，都是在减少锁的使用或减少锁的作用空间或时间，以此来削减独占锁（也叫互斥锁、排它锁）带来的负面影响。
注：文中源码均来自于JDK1.8版本，不同版本间可能存在差异。
如果有哪里有不明白或不清楚的内容，欢迎留言哦！