手写读写锁
https://www.cnblogs.com/DarrenChan/p/8619476.html
首先我们对读写锁做一个概述:
假设你的程序中涉及到对一些共享资源的读和写操作,且写操作没有读操作那么频繁。在没有写操作的时候,两个线程同时读一个资源没有任何问题,所以应该允许多个线程能在同时读取共享资源。但是如果有一个线程想去写这些共享资源,就不应该再有其它线程对该资源进行读或写,也就是说:读-读能共存,读-写不能共存,写-写不能共存。这就需要一个读/写锁来解决这个问题。
package com.darrenchan.lock; /** * 用两个int变量实现读写锁 * @author Think * */ public class MyReadWriteLock { private int readcount = 0; private int writecount = 0; public void lockread() throws InterruptedException{ while(writecount > 0){ synchronized(this){ wait(); } } readcount++; //进行读取操作 System.out.println("读操作"); } public void unlockread(){ readcount--; synchronized(this){ notifyAll(); } } public void lockwrite() throws InterruptedException{ while(writecount > 0){ synchronized(this){ wait(); } } //之所以在这里先++,是先占一个坑,避免读操作太多,从而产生写的饥饿等待 writecount++; while(readcount > 0){ synchronized(this){ wait(); } } //进行写入操作 System.out.println("写操作"); } public void unlockwrite(){ writecount--; synchronized(this){ notifyAll(); } } public static void main(String[] args) throws InterruptedException { MyReadWriteLock readWriteLock = new MyReadWriteLock(); for(int i = 0; i < 2; i++){ Thread2 thread2 = new Thread2(i, readWriteLock); thread2.start(); } for (int i = 0; i < 10; i++) { Thread1 thread1 = new Thread1(i, readWriteLock); thread1.start(); } } } class Thread1 extends Thread{ public int i; public MyReadWriteLock readWriteLock; public Thread1(int i, MyReadWriteLock readWriteLock) { this.i = i; this.readWriteLock = readWriteLock; } @Override public void run() { try { readWriteLock.lockread(); Thread.sleep(1000);//模拟耗时 System.out.println("第"+i+"个读任务"); } catch (InterruptedException e) { e.printStackTrace(); } finally { readWriteLock.unlockread(); } } } class Thread2 extends Thread{ public int i; public MyReadWriteLock readWriteLock; public Thread2(int i, MyReadWriteLock readWriteLock) { this.i = i; this.readWriteLock = readWriteLock; } @Override public void run() { try { readWriteLock.lockwrite(); Thread.sleep(1000); System.out.println("第"+i+"个写任务"); } catch (InterruptedException e) { e.printStackTrace(); } finally { readWriteLock.unlockwrite(); } } }
当然这是简单的,还可以进一步加深,可以参考博客:http://ifeve.com/read-write-locks/#simple
注:以上代码在++和--的时候仍然会产生并发异常,建议用AtomicInteger类型,在硬件上保证++和--操作不会出现并发异常。