Java中的互斥锁，了解一下...

Posted 2021-04-24 Goach的开发日记

在 Java 的世界里面，我们平时使用最多的就是 synchronized 。但在 Java 里面还有一种互斥锁，它是通过 ReentrantLock 来实现的。那么什么是互斥锁呢，先来看一个例子。

开启三个线程，它们的 ID 分别为 A，B，C。然后依次轮训打印他们的 ID 。这样来回打印 10 次

这是一个面试题目，如果通过 synchronized 来实现的话，代码如下：

public class ThreadTest { public static void main(String[] args) {        Loop loop = new Loop(); new Thread(new Runnable() { @Override public void run() { for (int i = 0; i < 20; i++) { loop.thread1(); } System.out.println("线程1退出"); } }, "A").start(); new Thread(new Runnable() { @Override public void run() { for (int i = 0; i < 20; i++) { loop.thread2(); } System.out.println("线程2退出"); } }, "B").start(); new Thread(new Runnable() { @Override public void run() { for (int i = 0; i < 20; i++) { loop.thread3(); } System.out.println("线程3退出"); } }, "C").start();
 }
 private static final class Loop { private static int num = 0; private final Object lock = new Object();
 private void thread1() { synchronized (lock) { if (num == 1 || num == 2) { try { lock.wait(); } catch (InterruptedException e) { e.printStackTrace(); } } else { System.out.println("线程1" + Thread.currentThread().getName()); num = 1; lock.notifyAll(); } } }
 private void thread2() { synchronized (lock) { if (num == 0 || num == 2) { try { lock.wait(); } catch (InterruptedException e) { e.printStackTrace(); } } else { System.out.println("线程2" + Thread.currentThread().getName()); num = 2; lock.notifyAll(); } } }
 private void thread3() { synchronized (lock) { if (num == 0 || num == 1) { try { lock.wait(); } catch (InterruptedException e) { e.printStackTrace(); } } else { System.out.println("线程3" + Thread.currentThread().getName()); num = 0; lock.notifyAll(); } } } } }

虽然能基本上实现，但是当一个线程通知其他线程的时候，其他两个线程会同时去抢这个锁，没法指定一个线程获取当前锁。而互斥锁正好有这个特点，所以我们就可以通过互斥锁来实现。

public class ThreadTest { public static void main(String[] args) { Loop2 loop = new Loop2(); new Thread(new Runnable() { @Override public void run() { for (int i = 0; i < 20; i++) { loop.thread1(); } System.out.println("线程1退出"); } }, "A").start(); new Thread(new Runnable() { @Override public void run() { for (int i = 0; i < 20; i++) { loop.thread2(); } System.out.println("线程2退出"); } }, "B").start(); new Thread(new Runnable() { @Override public void run() { for (int i = 0; i < 20; i++) { loop.thread3(); } System.out.println("线程3退出"); } }, "C").start();
    } /** * 互斥锁 */ private static final class Loop2 { private static int num = 0; private final Lock lock = new ReentrantLock(); private final Condition tread1Condition = lock.newCondition(); private final Condition tread2Condition = lock.newCondition(); private final Condition tread3Condition = lock.newCondition();
 private void thread1() { lock.lock(); try { if (num != 0) { tread1Condition.await(); } System.out.println("线程1" + Thread.currentThread().getName()); num = 1; tread2Condition.signal(); } catch (Exception e) { e.printStackTrace(); } finally { lock.unlock(); } }
 private void thread2() { lock.lock(); try { if (num != 1) { tread2Condition.await(); } System.out.println("线程2" + Thread.currentThread().getName()); num = 2; tread3Condition.signal(); } catch (Exception e) { e.printStackTrace(); } finally { lock.unlock(); } }
 private void thread3() { lock.lock(); try { if(num != 2){ tread3Condition.await(); } System.out.println("线程3" + Thread.currentThread().getName()); num = 0; tread1Condition.signal(); } catch (Exception e) { e.printStackTrace(); }finally { lock.unlock(); } } }}

上面的 Condition 是 ReentrantLock 的一个约束条件，它包括 await ， sigal 等方法来指定某个线程获取锁。

总结

1. 互斥锁可以指定某个线程获取锁，并同步

2. 互斥锁是通过 ReentrantLock + Condition 来实现的

3. 互斥锁一般有 await 和 signal 等方法来通信