设为首页收藏本站

IT技术擎 - 最棒的IT web技术交流社区

 找回密码
 注册为IT技术擎人

QQ登录

只需一步,快速开始

搜索
热搜: php h5 jquery
查看: 12|回复: 0

[框架开发] AQS 框架之 Lock 接口

[复制链接]

1万

主题

1万

帖子

3万

积分

版主

Rank: 7Rank: 7Rank: 7

积分
37750
发表于 2018-9-10 21:20:51 | 显示全部楼层 |阅读模式

■ 前言 - 内存锁的不足

  • 不可中断:使用内部锁(指的是 synchronized) 时,不能中断正在等待获取锁的线程
  • 不可超时:使用内部锁时,在请求锁失败情况下,必须无限等待,没有超时效果
  • 自动释放:使用内部锁时,内部锁必须在获取它们的代码块中被自动释放(虽然对代码来说是种简化且对异常友好)
  • 不可伸缩:使用内部锁时,无法细粒度控制锁(伸缩性不足),即无法实现锁分离和锁联结,比如为每个链表节点(或部分)加锁从而允许不同的线程能够独立操作链表的不同节点(部分),遍历或修改链表时,需先获取该节点锁并直到获取下一个节点锁时才释放当前节点锁
  • 性能问题:使用内部锁时,在有竞争情况下仍会出现性能问题,尽管JDK6对内部锁进行了优化,但无论是偏向锁或是轻量级锁都是针对无竞争情况的优化,无竞争情况下与 ReentractLock性能一致,但有竞争时Lock明显更高效

■ Lock 接口综述


  • 定义: JDK1.5 引入Lock接口,其定义了一些抽象的锁操作,相比synchronized,Lock 提供了无条件、可轮询、可定时、可中断的锁获取操作,所有加锁和解锁的方法都是显式的
  • 实现: Lock 的实现必须提供具有与 synchronized 相同的内存语义,但加锁的语义、调度算法、顺序保证、性能特性可以有所不同
  • 使用: Lock接口的实现基本是通过聚合一个同步器 AbstractQueuedSynchronized 的子类来完成线程的访问控制
  • 对比内部锁: Lock缺少隐式获取/释放锁的便捷,但却拥有了锁获取与释放的可操作性、可中断的获取锁以及超时获取锁等多种内部锁不具备的同步性,甚至还支持读写锁分离,同时允许获取和释放可以不在同一个块中
  • 补充:此番为 AQS 框架之综述 (排期中) 的子番,同时也是 AQS 框架之 ReentractLock (排期中) 的预备番

  并发包中的Lock接口实现类,其中ReadLock和WriteLock是ReentrantReadWriteLock的静态内部类,如图:

■ Lock 接口方法标准使用
  1. //标准用法
  2. Lock lock = newReentrantLock();
  3. //获取锁应在try之前,因为若获取锁时发生异常,异常抛出同时会导致锁无故释放
  4. lock.lock();
  5. try{ doSometing();
  6. }
  7. finally{//注意:必须在finally块中释放锁,目的是保证在获取到锁之后,最终能被释放 lock.unlock();
  8. }
复制代码

■ Lock 接口方重点方法 1. lock()

  • lock方法应具有与内部锁加锁相同的内存语义,即无锁阻塞和支持可重入
  • lock方法必须搭配unlock方法使用,同时必须在finally中显式调用unlock方法释放锁
  1. /** * Acquires the lock. * 获取锁,调用该方法的当前线程将会获取锁,当锁获得后,从该方法返回 * <p >If the lock is not available then the current thread becomes * disabled for thread scheduling purposes and lies dormant until the * lock has been acquired. * 若当前锁不可用(已被占有),当前线程会一直休眠直到锁为可被获取状态 * <p > <b >Implementation Considerations </b > * 实现该方法的注意事项 * <p >A {
  2. @codeLock} implementation may be able to detect erroneous use * of the lock, such as an invocation that would cause deadlock, and * may throw an (unchecked) exception in such circumstances. The * circumstances and the exception type must be documented by that * {
  3. @codeLock} implementation. * 该方法的实现需要能发现lock被错误使用,如死锁或抛出不可查异常(即可运行期异常和Error) * 此时该实现必须用文档注明其可能出现的异常或需要的使用环境 */
  4. voidlock();
复制代码
2. lockInterruptibly()

  • lockInterruptibly 方法提供可中断的锁获取操作并允许在可取消的活动中使用
  1. //样例代码<br>publicbooleandoTask throwsInterruptedException(){ lock.lockInterruptibly();try{returncancelTask(); }
  2. finally{ lock.unlock(); }
  3. }
  4. //取消任务
  5. privatebooleancancelTask() throwsInterruptedException {...}
复制代码
  1. /** * Acquires the lock unless the current thread is * {
  2. @linkplainThread#interrupt interrupted}. * 可中断地获取锁,即在锁的获取中可以中断当前线程 * <p >Acquires the lock if it is available and returns immediately. * 当获取锁时锁可用就立即返回 * <p >If the lock is not available then the current thread becomes * disabled for thread scheduling purposes and lies dormant until * one of two things happens: * <ul > * <li >The lock is acquired by the current thread; or * <li >Some other thread {
  3. @linkplainThread#interrupt interrupts} the * current thread, and interruption of lock acquisition is supported. * </ul > * 若当前锁不可用(已被占有),当前线程会一直休眠直到以下两种情况发生: * 1.锁被当前线程获取 * 2.其他线程中断当前线程,同时锁的获取允许被中断 * <p > <b >Implementation Considerations </b > * 实现该方法的注意事项 * <p >The ability to interrupt a lock acquisition in some * implementations may not be possible, and if possible may be an * expensive operation. The programmer should be aware that this * may be the case. An implementation should document when this is * the case. * 该方法属于拓展方法,只有需要中断服务的时候才需要实现它 * <p >An implementation can favor responding to an interrupt over * normal method return. * 相对于返回,该方法更适合抛出一个中断响应,比如中断异常 * @throwsInterruptedException if the current thread is * interrupted while acquiring the lock (and interruption * of lock acquisition is supported) */
  4. voidlockInterruptibly() throwsInterruptedException;
复制代码
3. tryLock()

  • tryLock 方法提供可定时与可轮询的锁获取方式,与无条件的锁获取相比,具有更完善的错误恢复机制
  • tryLock 方法能够有效的防止死锁的发生,比如使用轮询锁优雅失败规避死锁
  • tryLock 方法同时提供定时锁的功能,其允许在限时活动内部使用独占锁,当线程获取锁、被中断或超时后返回
  • tryLock 方法支持轮询获取锁:通过一个循环配合tryLock()来不断尝试获取锁,由于tryLock()非阻塞因此会立即返回是否成功获取锁的结果;当不能获取所有的锁时,应释放已获得的所有锁并重新尝试获取
  • tryLock 方法同时支持响应中断
  1. /** * Acquires the lock only if it is free at the time of invocation. * 尝试非阻塞的获取锁,调用该方法后立即返回是否成功获取锁true/false * <p >Acquires the lock if it is available and returns immediately * with the value {
  2. @codetrue}. * If the lock is not available then this method will return * immediately with the value {
  3. @codefalse}. * 当锁不可用时立即返回false * This usage ensures that the lock is unlocked if it was acquired, and * doesn't try to unlock if the lock was not acquired. * 该实现应确保当锁被获取时是未锁状态,当未被获取时不会尝试解锁 * @return{
  4. @codetrue} if the lock was acquired and * {
  5. @codefalse} otherwise */
  6. booleantryLock();
  7. /** * Acquires the lock if it is free within the given waiting time and the * current thread has not been {
  8. @linkplainThread#interrupt interrupted}. * 没有被中断当前线程在指定超时时间内获取锁 * If the lock is not available then the current thread becomes disabled for * thread scheduling purposes and lies dormant until one of three things happens: * <ul > * <li >The lock is acquired by the current thread; or * <li >Some other thread {
  9. @linkplainThread#interrupt interrupts} the * current thread, and interruption of lock acquisition is supported; or * <li >The specified waiting time elapses * </ul > * <p >If the specified waiting time elapses then the value {
  10. @codefalse} is returned. * If the time is less than or equal to zero, the method will not wait at all. * 当前线程在以下三种情况下会返回: * 1.当前线程在超时时间内获得锁 * 2.当前线程在超时时间内被中断 * 3.超时时间结束,返回false,线程不再被阻塞 * @paramtime the maximum time to wait for the lock * @paramunit the time unit of the {
  11. @codetime} argument * @return{
  12. @codetrue} if the lock was acquired and {
  13. @codefalse} * if the waiting time elapsed before the lock was acquired * @throwsInterruptedException if the current thread is interrupted * while acquiring the lock (and interruption of lock * acquisition is supported) */
  14. booleantryLock(longtime, TimeUnit unit) throwsInterruptedException;
复制代码
4. unLock()

  • 使用lock方法、lockInterruptibly方法、tryLock方法都必须显式调用unlock方法释放锁
  • unlock方法必须在finally块中执行,这也是Lock使用的一个代码隐患(容易忘记执行)
  • unlock方法允许与lock方法不在同一个块(即{})中执行,但业务代码必须保证在try-finally块中执行
  1. /** * Releases the lock. * 释放锁 */
  2. voidunlock();
复制代码
5. newCondition 方法

  • Lock搭配Condition可以实现更加灵活的锁获取与释放的条件控制
  1. /** * Returns a new {
  2. @linkCondition} instance that is bound to this * {
  3. @codeLock} instance. * 返回一个等待通知组(条)件 * <p >Before waiting on the condition the lock must be held by the current thread. * A call to {
  4. @linkCondition#await()} will atomically release the lock * before waiting and re-acquire the lock before the wait returns. * 该组件与当前锁绑定,当先线程只有获得锁才能调用该组件的await方法并释放锁 * @returnA new {
  5. @linkCondition} instance for this {
  6. @codeLock} instance * @throwsUnsupportedOperationException if this {
  7. @codeLock} * implementation does not support conditions */Condition newCondition();
复制代码
该用户未在地球留下任何的痕迹

本版积分规则

QQ|小黑屋|帮助|IT技术擎 ( 沪ICP备15054863号  

GMT+8, 2018-9-19 01:50

Powered by Discuz! X3.2 Licensed

© 2001-2013 Comsenz Inc.

快速回复 返回顶部 返回列表