保證線程安全的方式有很多,比如CAS操作、synchronized、原子類、volatile保證可見性和ReentrantLock等,這篇文章我們主要探討ReentrantLock的相關(guān)內(nèi)容。本文基于JDK1.8講述ReentrantLock.
一、可重入鎖
所謂可重入鎖,即一個(gè)線程已經(jīng)獲得了某個(gè)鎖,當(dāng)這個(gè)線程要再次獲取這個(gè)鎖時(shí),依然可以獲取成功,不會(huì)發(fā)生死鎖的情況。synchronized就是一個(gè)可重入鎖,除此之外,JDK提供的ReentrantLock也是一種可重入鎖。
二、ReentrantLock
2.1 ReentrantLock的簡單使用
public class TestReentrantLock {
private static int i = 0;
public static void main(String[] args) {
ReentrantLock lock = new ReentrantLock();
try {
lock.lock();
i++;
} finally {
lock.unlock();
}
System.out.println(i);
}
}
上面是ReentrantLock的一個(gè)簡單使用案列,進(jìn)入同步代碼塊之前,需要調(diào)用lock()方法進(jìn)行加鎖,執(zhí)行完同步代碼塊之后,為了防止異常發(fā)生時(shí)造成死鎖,需要在finally塊中調(diào)用unlock()方法進(jìn)行解鎖。
2.2 ReentrantLock UML圖
2.3 lock()方法調(diào)用鏈
上圖描述了ReentrantLock.lock()加鎖的方法調(diào)用過程。在ReentrantLock中有一個(gè)成員變量private final Sync sync,Sync是AQS的一個(gè)子類。ReentrantLock的lock()方法中,調(diào)用了sync的lock()方法,這個(gè)方法為抽象方法,具體調(diào)用的是NonfairSync中實(shí)現(xiàn)的lock()方法:
/**
* Performs lock. Try immediate barge, backing up to normal
* acquire on failure.
*/
final void lock() {
if (compareAndSetState(0, 1))
setExclusiveOwnerThread(Thread.currentThread());
else
acquire(1);
}
在這個(gè)方法中,先嘗試通過CAS操作進(jìn)行加鎖。如果加鎖失敗,會(huì)調(diào)用AQS的acquire()方法:
/**
* Acquires in exclusive mode, ignoring interrupts. Implemented
* by invoking at least once {@link #tryAcquire},
* returning on success. Otherwise the thread is queued, possibly
* repeatedly blocking and unblocking, invoking {@link
* #tryAcquire} until success. This method can be used
* to implement method {@link Lock#lock}.
*
* @param arg the acquire argument. This value is conveyed to
* {@link #tryAcquire} but is otherwise uninterpreted and
* can represent anything you like.
*/
public final void acquire(int arg) {
if (!tryAcquire(arg) &&
acquireQueued(addWaiter(Node.EXCLUSIVE), arg))
selfInterrupt();
}
在AQS的acquire方法中,先嘗試調(diào)用tryAcquire方法進(jìn)行加鎖,如果失敗,會(huì)調(diào)用acquireQueued進(jìn)入等待隊(duì)列當(dāng)中。acquireQueued方法將會(huì)在第三章中講解,先來看tryAcquire方法的內(nèi)容。AQS的tryAcquire方法是一個(gè)模板方法,其具體實(shí)現(xiàn)在NonfairSync的tryAcquire方法中,里面僅僅是調(diào)用了nonfairTryAcquire方法:
/**
* Performs non-fair tryLock. tryAcquire is implemented in
* subclasses, but both need nonfair try for trylock method.
*/
final boolean nonfairTryAcquire(int acquires) {
final Thread current = Thread.currentThread();
int c = getState();
if (c == 0) {
if (compareAndSetState(0, acquires)) {
setExclusiveOwnerThread(current);
return true;
}
}
else if (current == getExclusiveOwnerThread()) {
int nextc = c + acquires;
if (nextc < 0) // overflow
throw new Error("Maximum lock count exceeded");
setState(nextc);
return true;
}
return false;
}
在這個(gè)方法中,先獲取state判斷其是否為0。如果為0表示沒有其他線程占用鎖,會(huì)嘗試通過CAS操作將state設(shè)為1進(jìn)行加鎖;如果state不為0,表示某個(gè)線程已經(jīng)占用了鎖,判斷占用鎖的線程是否為當(dāng)前線程,如果是,則將state進(jìn)行加1的操作,這就是ReentrantLock可重入的實(shí)現(xiàn)原理。
三、AQS
AQS即AbstractQueuedSynchronizer。AQS提供了一個(gè)基于FIFO隊(duì)列,可以用于構(gòu)建鎖或者其他相關(guān)同步裝置的基礎(chǔ)框架。AQS其實(shí)是CLH(Craig,Landin,Hagersten)鎖的一個(gè)變種,下面來講解AQS的核心思想及其具體實(shí)現(xiàn)。
3.1 state
/**
* The synchronization state.
*/
private volatile int state;
state是AQS中最核心的成員變量。這是一個(gè)volatile變量,當(dāng)其為0時(shí),表示沒有任何線程占用鎖。線程通過CAS將state從0置為1進(jìn)行加鎖,當(dāng)線程持有鎖的情況下,再次進(jìn)行加鎖,會(huì)將state加1,即重入。
3.2 exclusiveOwnerThread
/**
* The current owner of exclusive mode synchronization.
*/
private transient Thread exclusiveOwnerThread;
exclusiveOwnerThread是AQS的父類AbstractOwnableSynchronizer中的成員變量,其作用是實(shí)現(xiàn)可重入機(jī)制時(shí),用于判斷持有鎖的線程是否為當(dāng)前線程。
3.3 AQS等待隊(duì)列
除了以上state和exclusiveOwnerThread兩個(gè)重要的成員變量以外,AQS還維護(hù)了一個(gè)等待隊(duì)列。當(dāng)線程嘗試加鎖失敗時(shí),會(huì)進(jìn)入這個(gè)等待隊(duì)列中,這也是整個(gè)AQS中最核心的內(nèi)容。這個(gè)等待隊(duì)列是一個(gè)雙向鏈表,其節(jié)點(diǎn)Node對等待加鎖的線程進(jìn)行封裝。
/**
* Creates and enqueues node for current thread and given mode.
*
* @param mode Node.EXCLUSIVE for exclusive, Node.SHARED for shared
* @return the new node
*/
private Node addWaiter(Node mode) {
Node node = new Node(Thread.currentThread(), mode);
// Try the fast path of enq; backup to full enq on failure
Node pred = tail;
if (pred != null) {
node.prev = pred;
// 通過CAS操作將自身追加到鏈表尾部
if (compareAndSetTail(pred, node)) {
pred.next = node;
return node;
}
}
enq(node);
return node;
}
當(dāng)線程嘗試加鎖失敗時(shí),通過CAS操作將自身追加到鏈表尾部。入隊(duì)之后,會(huì)調(diào)用acquireQueued在隊(duì)列中嘗試加鎖:
/**
* Acquires in exclusive uninterruptible mode for thread already in
* queue. Used by condition wait methods as well as acquire.
*
* @param node the node
* @param arg the acquire argument
* @return {@code true} if interrupted while waiting
*/
final boolean acquireQueued(final Node node, int arg) {
boolean failed = true;
try {
boolean interrupted = false;
for (;;) {
final Node p = node.predecessor();
if (p == head && tryAcquire(arg)) {
setHead(node);
p.next = null; // help GC
failed = false;
return interrupted;
}
if (shouldParkAfterFailedAcquire(p, node) &&
parkAndCheckInterrupt())
interrupted = true;
}
} finally {
if (failed)
cancelAcquire(node);
}
}
在這個(gè)方法中,會(huì)判斷其前置節(jié)點(diǎn)是否為頭節(jié)點(diǎn),如果是,則嘗試進(jìn)行加鎖。如果加鎖失敗,則調(diào)用LockSupport.park方法進(jìn)入阻塞狀態(tài),等待其前置節(jié)點(diǎn)釋放鎖之后將其喚醒。
3.4 AQS中的模板方法設(shè)計(jì)模式
AQS完美地運(yùn)用了模板方法設(shè)計(jì)模式,其中定義了一系列的模板方法。比如以下方法:
// 互斥模式下使用:嘗試獲取鎖
protected boolean tryAcquire(int arg) {
throw new UnsupportedOperationException();
}
// 互斥模式下使用:嘗試釋放鎖
protected boolean tryRelease(int arg) {
throw new UnsupportedOperationException();
}
// 共享模式下使用:嘗試獲取鎖
protected int tryAcquireShared(int arg) {
throw new UnsupportedOperationException();
}
// 共享模式下使用:嘗試釋放鎖
protected boolean tryReleaseShared(int arg) {
throw new UnsupportedOperationException();
}
這些方法在AQS中只拋出了UnsupportedOperationException異常,所以需要子類去實(shí)現(xiàn)它們。之所以沒有將這些方法設(shè)計(jì)成為抽象方法,是因?yàn)锳QS的子類可能只需要實(shí)現(xiàn)其中的某些方法即可實(shí)現(xiàn)其功能。
總結(jié)
不同版本的JDK,AQS的實(shí)現(xiàn)可能會(huì)有細(xì)微的差異,但其核心思想是不會(huì)變的,即線程加鎖失敗后,通過CAS進(jìn)行入隊(duì)的操作,并通過CAS的方法設(shè)置state來獲得鎖。
到這里,本篇關(guān)于Java多線程中的可重入鎖ReentrantLock詳細(xì)講解的文章就到此結(jié)束了,如果您還想要了解更多關(guān)于Java多線程相關(guān)知識,推薦您可以看一看我們的Java多線程視頻,適合想要了解Java多線程的小伙伴們。