标签:
ReentrantLock是一个较为常用的锁对象。在上次分析的uil开源项目中也多次被用到,下面谈谈其概念和基本使用。
一个可重入的互斥锁定 Lock,它具有与使用 synchronized 相同的一些基本行为和语义,但功能更强大。
表示同一时刻,多个线程中,只能有一个线程能获得该锁。但是多个线程都可以调用lock方法,只有一个会成功,其他的线程会被阻塞,直到该锁被释放
模仿synchronized 的语义;如果线程进入由线程已经拥有的监控器保护的 synchronized 块,就允许线程继续进行,当线程退出第二个(或者后续)synchronized 块的时候,不释放锁,只有线程退出它进入的监控器保护的第一个 synchronized 块时,才释放锁。
对于ReentrantLock,每次获得锁,并将请求计数置为一,如果同一个线程再次lock,计数器将递增,每次unlock时计数器值递减,直到计数器为0,锁释放
如果该锁没有被另一个线程保持,则lock时获取该锁定并立即返回,将锁定的保持计数设置为 1。
如果当前线程已经保持该锁定,则将保持计数加 1,并且该方法立即返回。
如果该锁定被另一个线程保持,则出于线程调度的目的,禁用当前线程,并且在获得锁定之前,该线程将一直处于休眠状态,此时锁定保持计数被设置为 1。
每次unlock时计数器值递减,直到计数器为0,释放锁
该类与lock绑定,用newCondition()方法创建,提供了线程之间通信的方式(类似信号量)。其使用基本与object类的wait,notify,notifyAll相同。
用condition.await()替换Object,wait(),调用时该线程阻塞,释放该线程的锁。
用condition.signal()替换Object.notify(),用condition.signalAll()替换Object.notifyAll(),唤醒该condition await方法所阻塞的线程
锁投票(我也不是特别理解,可以通过投票获取锁?)
定时锁等候
中断锁等候
线程A和B都要获取对象O的锁定,假设A获取了对象O锁,B将等待A释放对O的锁定,
如果使用 synchronized ,如果A不释放,B将一直等下去,不能被中断
如果 使用ReentrantLock,如果A不释放,可以使B在等待了足够长的时间以后,中断等待,而干别的事情
以下以linkedBlokingQueue源码为例子,来学习其使用。
public class LinkedBlockingQueue<E> extends AbstractQueue<E> implements BlockingQueue<E>, java.io.Serializable { //链表节点node类结构 static class Node<E> { volatile E item;//volatile,保证了数据的可见性 Node<E> next; Node(E x) { item = x; } } //容量 private final int capacity; //用原子变量,当前元素个数 private final AtomicInteger count = new AtomicInteger(0); //头节点 private transient Node<E> head; //表尾节点 private transient Node<E> last; //获取元素或删除元素时,要加的takeLock锁 private final ReentrantLock takeLock = new ReentrantLock(); //获取元素时若队列为空,线程阻塞,直至notEmpty条件满足(被通知) private final Condition notEmpty = takeLock.newCondition(); //插入元素时 要加putLock锁 private final ReentrantLock putLock = new ReentrantLock(); //插入时,若队列已满,线程阻塞,直至notFull条件满足(被通知) private final Condition notFull = putLock.newCondition(); // 唤醒等待的take操作,插入数据时若插入前链表中无数据,则调用,表示链表不再为空 private void signalNotEmpty() { final ReentrantLock takeLock = this.takeLock; takeLock.lock(); try { notEmpty.signal(); } finally { takeLock.unlock(); } } //唤醒等待插入操作,移除数据时若链表原先已满则调用,表示链表不再满 private void signalNotFull() { final ReentrantLock putLock = this.putLock; putLock.lock(); try { notFull.signal(); } finally { putLock.unlock(); } } // 插入到链表尾部 private void insert(E x) { last = last.next = new Node<E>(x); } //获取并移除头元素 private E extract() { Node<E> first = head.next; head = first; E x = first.item; first.item = null; return x; } //锁住两把锁,在remove,clear等方法中调用 private void fullyLock() { putLock.lock(); takeLock.lock(); } //和fullyLock成对使用 private void fullyUnlock() { takeLock.unlock(); putLock.unlock(); } //默认构造,容量为 Integer.MAX_VALUE public LinkedBlockingQueue() { this(Integer.MAX_VALUE); } //指定容量的构造 public LinkedBlockingQueue(int capacity) { if (capacity <= 0) throw new IllegalArgumentException(); this.capacity = capacity; last = head = new Node<E>(null); } //指定初始化集合的构造 public LinkedBlockingQueue(Collection<? extends E> c) { this(Integer.MAX_VALUE); for (E e : c) add(e); } //获得大小 public int size() { return count.get(); } //剩余容量 public int remainingCapacity() { return capacity - count.get(); } // 将指定元素插入到此队列的尾部,如已满,阻塞至队列中有元素被移除 public void put(E e) throws InterruptedException { if (e == null) throw new NullPointerException(); int c = -1; final ReentrantLock putLock = this.putLock; final AtomicInteger count = this.count; //加put锁,多个线程不能同时进入 putLock.lockInterruptibly(); try { try { //容量已满,则一直阻塞 while (count.get() == capacity) notFull.await(); } catch (InterruptedException ie) { notFull.signal(); // propagate to a non-interrupted thread throw ie; } //插入 insert(e); c = count.getAndIncrement(); //通知链表未满 if (c + 1 < capacity) notFull.signal(); } finally { //解锁,注意必须在finally里调用,反正各种异常导致没有unlock使线程死锁 putLock.unlock(); } //通知链表非空 if (c == 0) signalNotEmpty(); } // 将指定元素插入到此队列的尾部,如有必要,则等待一定时间以使空间变得可用。 public boolean offer(E e, long timeout, TimeUnit unit) throws InterruptedException { if (e == null) throw new NullPointerException(); long nanos = unit.toNanos(timeout); int c = -1; final ReentrantLock putLock = this.putLock; final AtomicInteger count = this.count; //加锁 putLock.lockInterruptibly(); try { for (;;) { //未满可插入 if (count.get() < capacity) { insert(e); c = count.getAndIncrement(); //通知未满 if (c + 1 < capacity) notFull.signal(); //跳出循环 break; } //队列已满,未能插入,等待时间是负的,直接返回 if (nanos <= 0) return false; try { //等待一定时间后再次尝试 nanos = notFull.awaitNanos(nanos); } catch (InterruptedException ie) { notFull.signal(); // propagate to a non-interrupted thread throw ie; } } } finally { //解锁 putLock.unlock(); } //通知已插入数据,链表非空 if (c == 0) signalNotEmpty(); return true; } //将指定元素插入到此队列的尾部(如果立即可行且不会超出此队列的容量), 在成功时返回 true,如果此队列已满,则返回 false。 public boolean offer(E e) { if (e == null) throw new NullPointerException(); final AtomicInteger count = this.count; if (count.get() == capacity) return false; int c = -1; final ReentrantLock putLock = this.putLock; putLock.lock(); try { //由于可能在lock被阻塞时其他线程进行了插入操作,需再次判断count if (count.get() < capacity) { insert(e); c = count.getAndIncrement(); //通知未满 if (c + 1 < capacity) notFull.signal(); } } finally { putLock.unlock(); } //通知非空 if (c == 0) signalNotEmpty(); // >0表示已成功插入 return c >= 0; } //获取并移除此队列的头部,若队列为空,则阻塞。 public E take() throws InterruptedException { E x; int c = -1; final AtomicInteger count = this.count; final ReentrantLock takeLock = this.takeLock; //加锁 takeLock.lockInterruptibly(); try { try { //队列为空时阻塞 while (count.get() == 0) notEmpty.await(); } catch (InterruptedException ie) { notEmpty.signal(); // propagate to a non-interrupted thread throw ie; } //获取数据 x = extract(); c = count.getAndDecrement(); //通知非空 if (c > 1) notEmpty.signal(); } finally { takeLock.unlock(); } //通知未满 if (c == capacity) signalNotFull(); return x; } //与offer方法结构基本一致,若队列为空,则阻塞一段时间,一段时间后仍为空,则返回null public E poll(long timeout, TimeUnit unit) throws InterruptedException { E x = null; int c = -1; long nanos = unit.toNanos(timeout); final AtomicInteger count = this.count; final ReentrantLock takeLock = this.takeLock; takeLock.lockInterruptibly(); try { for (;;) { if (count.get() > 0) { x = extract(); c = count.getAndDecrement(); if (c > 1) notEmpty.signal(); break; } if (nanos <= 0) return null; try { nanos = notEmpty.awaitNanos(nanos); } catch (InterruptedException ie) { notEmpty.signal(); // propagate to a non-interrupted thread throw ie; } } } finally { takeLock.unlock(); } if (c == capacity) signalNotFull(); return x; } ////与offer方法结构基本一致 队列为空,不阻塞,直接返回null public E poll() { final AtomicInteger count = this.count; if (count.get() == 0) return null; E x = null; int c = -1; final ReentrantLock takeLock = this.takeLock; takeLock.lock(); try { if (count.get() > 0) { x = extract(); c = count.getAndDecrement(); if (c > 1) notEmpty.signal(); } } finally { takeLock.unlock(); } if (c == capacity) signalNotFull(); return x; } //获取但不移除此队列的头;如果此队列为空,则返回 null。 public E peek() { if (count.get() == 0) return null; final ReentrantLock takeLock = this.takeLock; takeLock.lock(); try { Node<E> first = head.next; if (first == null) return null; else return first.item; } finally { takeLock.unlock(); } } /** * 从此队列移除指定元素的单个实例(如果存在)。 */ public boolean remove(Object o) { if (o == null) return false; boolean removed = false; //同时加锁,此时其他线程不能插入,不能移除 fullyLock(); try { Node<E> trail = head; Node<E> p = head.next; //遍历,获取到该元素 while (p != null) { if (o.equals(p.item)) { removed = true; break; } trail = p; p = p.next; } //删除该元素 if (removed) { p.item = null; trail.next = p.next; if (last == p) last = trail; if (count.getAndDecrement() == capacity) notFull.signalAll(); } } finally { fullyUnlock(); } return removed; } …… }
http://coderrobin.com/2015/02/12/java%E5%B9%B6%E5%8F%91%E7%BC%96%E7%A8%8B%E5%9F%BA%E7%A1%80-ReentrantLock/
java并发编程基础-ReentrantLock及LinkedBlockingQueue源码分析
标签:
原文地址:http://www.cnblogs.com/chenying99/p/4343976.html