目錄1. 介紹2. 源碼剖析2.1 上鎖(獲取資源)2.2 釋放資源2.3 公平鎖與非公平鎖的區(qū)別1. 介紹

結(jié)合上面的ReentrantLock類(lèi)圖，ReentrantLock實(shí)現(xiàn)了Lock接口，它的內(nèi)部類(lèi)Sync繼承自AQS，絕大部分使用AQS的子類(lèi)需要自定義的方法存在Sync中。而ReentrantLock有公平與非公平的區(qū)別，即’是否先阻塞就先獲取資源’，它的主要實(shí)現(xiàn)就是FairSync與NonfairSync，后面會(huì)從源碼角度看看它們的區(qū)別。

2. 源碼剖析

Sync是ReentrantLock控制同步的基礎(chǔ)。它的子類(lèi)分為了公平與非公平。使用AQS的state代表獲取鎖的數(shù)量

abstract static class Sync extends AbstractQueuedSynchronizer { private static final long serialVersionUID = -5179523762034025860L; /*** Performs {@link Lock#lock}. The main reason for subclassing* is to allow fast path for nonfair version.*/ abstract void lock(); ...}

我們可以看出內(nèi)部類(lèi)Sync是一個(gè)抽象類(lèi)，繼承它的子類(lèi)(FairSync與NonfairSync)需要實(shí)現(xiàn)抽象方法lock。

下面我們先從非公平鎖的角度來(lái)看看獲取資源與釋放資源的原理

故事就從就兩個(gè)變量開(kāi)始:

// 獲取一個(gè)非公平的獨(dú)占鎖/*** public ReentrantLock() {* sync = new ReentrantLock.NonfairSync();* }*/private Lock lock = new ReentrantLock();// 獲取條件變量private Condition condition = lock.newCondition();2.1 上鎖(獲取資源)

lock.lock()

public void lock() { sync.lock();}

static final class NonfairSync extends Sync { private static final long serialVersionUID = 7316153563782823691L; // 獲取資源 final void lock() {// 若此時(shí)沒(méi)有線(xiàn)程獲取到資源，直接設(shè)置當(dāng)前線(xiàn)程獨(dú)占訪(fǎng)問(wèn)資源。if (compareAndSetState(0, 1)) setExclusiveOwnerThread(Thread.currentThread());else // AQS的方法 acquire(1); } protected final boolean tryAcquire(int acquires) {// 實(shí)現(xiàn)在父類(lèi)Sync中return nonfairTryAcquire(acquires); }}

AQS的acquire

public final void acquire(int arg) { if (!tryAcquire(arg) &&acquireQueued(addWaiter(Node.EXCLUSIVE), arg))selfInterrupt();}

// Sync實(shí)現(xiàn)的非公平的tryAcquirefinal boolean nonfairTryAcquire(int acquires) { final Thread current = Thread.currentThread(); int c = getState(); // 此時(shí)若沒(méi)有線(xiàn)程獲取到資源，當(dāng)前線(xiàn)程就直接占用該資源 if (c == 0) {if (compareAndSetState(0, acquires)) { setExclusiveOwnerThread(current); return true;} } // 若當(dāng)前線(xiàn)程已經(jīng)占用了該資源，可以再次獲取該資源 ->這個(gè)行為就是可重入鎖的支撐 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;}

嘗試獲取資源的過(guò)程是非常簡(jiǎn)單的，這里再貼一下acquire的流程

2.2 釋放資源

lock.unlock();

public void unlock() { // AQS的方法 sync.release(1);}

AQS的release

public final boolean release(int arg) { if (tryRelease(arg)) {Node h = head;if (h != null && h.waitStatus != 0) unparkSuccessor(h);return true; } return false;}

release的流程已經(jīng)剖析過(guò)了，接下來(lái)看看tryRelease的實(shí)現(xiàn)

protected final boolean tryRelease(int releases) { int c = getState() - releases; // 這里可以看出若沒(méi)有持有鎖，就釋放資源，就會(huì)報(bào)錯(cuò) if (Thread.currentThread() != getExclusiveOwnerThread())throw new IllegalMonitorStateException(); boolean free = false; if (c == 0) {free = true;setExclusiveOwnerThread(null); } setState(c); return free;}

tryRelease的實(shí)現(xiàn)也很簡(jiǎn)單，這里再貼一下release的流程圖

2.3 公平鎖與非公平鎖的區(qū)別

公平鎖與非公平鎖，即’是否先阻塞就先獲取資源’, ReentrantLock中公平與否的控制就在tryAcquire中。下面我們看看，公平鎖的tryAcquire

static final class FairSync extends Sync {private static final long serialVersionUID = -3000897897090466540L;final void lock() { acquire(1);}protected final boolean tryAcquire(int acquires) { final Thread current = Thread.currentThread(); int c = getState(); if (c == 0) {// (2.3.1)// sync queue中是否存在前驅(qū)結(jié)點(diǎn)if (!hasQueuedPredecessors() && compareAndSetState(0, acquires)) { setExclusiveOwnerThread(current); return true;} } else if (current == getExclusiveOwnerThread()) {int nextc = c + acquires;if (nextc < 0) throw new Error('Maximum lock count exceeded');setState(nextc);return true; } return false;} }

區(qū)別在代碼(2.3.1)

hasQueuedPredecessors

判斷當(dāng)前線(xiàn)程的前面有無(wú)其他線(xiàn)程排隊(duì)；若當(dāng)前線(xiàn)程在隊(duì)列頭部或者隊(duì)列為空返回false

public final boolean hasQueuedPredecessors() { // The correctness of this depends on head being initialized // before tail and on head.next being accurate if the current // thread is first in queue. Node t = tail; // Read fields in reverse initialization order Node h = head; Node s; return h != t &&((s = h.next) == null || s.thread != Thread.currentThread());}

結(jié)合下面的入隊(duì)代碼(enq), 我們分析hasQueuedPredecessors為true的情況:

1.h != t ,表示此時(shí)queue不為空; (s = h.next) == null, 表示另一個(gè)結(jié)點(diǎn)已經(jīng)運(yùn)行了下面的步驟(2)，還沒(méi)來(lái)得及運(yùn)行步驟(3)。簡(jiǎn)言之，就是B線(xiàn)程想要獲取鎖的同時(shí)，A線(xiàn)程獲取鎖失敗剛好在入隊(duì)(B入隊(duì)的同時(shí)，之前占有的資源的線(xiàn)程，剛好釋放資源)

2.h != t 且 (s = h.next) ！= null，表示此時(shí)至少有一個(gè)結(jié)點(diǎn)在sync queue中；s.thread != Thread.currentThread()，這個(gè)情況比較復(fù)雜，設(shè)想一下有這三個(gè)結(jié)點(diǎn) A -> B C， A此時(shí)獲取到資源，而B(niǎo)此時(shí)因?yàn)楂@取資源失敗正在sync queue阻塞，C還沒(méi)有獲取資源(還沒(méi)有執(zhí)行tryAcquire)。

時(shí)刻一：A釋放資源成功后(執(zhí)行tryRelease成功)，B此時(shí)還沒(méi)有成功獲取資源(C執(zhí)行s = h.next時(shí)，B還在sync queue中且是老二)

時(shí)刻二: C此時(shí)執(zhí)行hasQueuedPredecessors，s.thread != Thread.currentThread()成立，此時(shí)s.thread表示的是B

private Node enq(final Node node) { for (;;) {Node t = tail;if (t == null) { // Must initialize if (compareAndSetHead(new Node())) // (1) 第一次初始化tail = head;} else { node.prev = t; if (compareAndSetTail(t, node)) { // (2) 設(shè)置queue的tailt.next = node; // (3)return t; }} }}

Note that 1. because cancellations due to interrupts and timeouts may occur at any time, a true return does not guarantee that some other thread will acquire before the current thread（虛假true）. 2. Likewise, it is possible for another thread to win a race to enqueue after this method has returned false, due to the queue being empty（虛假false）.

這位大佬對(duì)hasQueuedPredecessors進(jìn)行詳細(xì)的分析，他文中解釋了虛假true以及虛假false。我這里簡(jiǎn)單解釋一下:

1.虛假true, 當(dāng)兩個(gè)線(xiàn)程都執(zhí)行tryAcquire，都執(zhí)行到hasQueuedPredecessors，都返回true，但是只有一個(gè)線(xiàn)程執(zhí)行compareAndSetState(0, acquires)成功

2.虛假false，當(dāng)一個(gè)線(xiàn)程A執(zhí)行doAcquireInterruptibly，發(fā)生了中斷，還沒(méi)有清除掉該結(jié)點(diǎn)時(shí)；此時(shí)，線(xiàn)程B執(zhí)行hasQueuedPredecessors時(shí)，返回true

以上就是淺談JAVA并發(fā)之ReentrantLock的詳細(xì)內(nèi)容，更多關(guān)于JAVA并發(fā)之ReentrantLock的資料請(qǐng)關(guān)注好吧啦網(wǎng)其它相關(guān)文章！