java线程锁

出处;https://www.liaoxuefeng.com/wiki/1252599548343744/1306581033549858

 

java线程锁是重入锁：一个线程可以重复获取锁

 

if (lock.tryLock(1, TimeUnit.SECONDS)) {
    try {
        ...
    } finally {
        lock.unlock();
    }
}

使用private final Lock lock = new ReentrantLock();
先去获取锁，然后再去铺获代码块

、最多等待1秒。如果1秒后仍未获取到锁，tryLock()返回false，不会无限等待下去。

所以，使用ReentrantLock比直接使用synchronized更安全，线程在tryLock()失败的时候不会导致死锁

 

package com.example.demo.config;

public class Thread1 extends Thread{
    public void run() {
        System.out.println("Thread-1: try get lock A...");
        synchronized (Demo.lockA) { //给A加锁
            System.out.println("Thread-1: lock A got.");
            Demo.sleep1s();
            System.out.println("Thread-1: try get lock B...");
            synchronized (Demo.lockB) {//获取B锁
                System.out.println("Thread-1: lock B got.");
                Demo.sleep1s();
            }
            System.out.println("Thread-1: lock B released.");
        }
        System.out.println("Thread-1: lock A released.");
    }
}

 

package com.example.demo.config;

public class Thread2 extends Thread{
    public void run() {
        System.out.println("Thread-2: try get lock B...");
        synchronized (Demo.lockB) {//给B加锁
            System.out.println("Thread-2: lock B got.");
            Demo.sleep1s();
            System.out.println("Thread-2: try get lock A...");
            synchronized (Demo.lockA) {//获取A锁
                System.out.println("Thread-2: lock A got.");
                Demo.sleep1s();
            }
            System.out.println("Thread-2: lock A released.");
        }
        System.out.println("Thread-2: lock B released.");
    }
}

 

package com.example.demo.config;


public class Demo {
    static final Object lockA = new Object();
    static final Object lockB = new Object();
    private Integer value=0;
    private Integer another=0;

    public Integer getValue() {
        return value;
    }

    public void setValue(Integer value) {
        this.value = value;
    }

    public Integer getAnother() {
        return another;
    }

    public void setAnother(Integer another) {
        this.another = another;
    }

    public void add(int m) {
        synchronized (lockA) { // 获得lockA的锁
            System.out.println("获得lockA的锁0");
            this.value += m;
            synchronized (lockB) { // 获得lockB的锁
                System.out.println("获得lockB的锁0");
                this.another += m;
            } // 释放lockB的锁
        } // 释放lockA的锁
    }

    public void dec(int m) {
        synchronized (lockB) { // 获得lockB的锁
            System.out.println("获得lockB的锁1");
            this.another -= m;
            synchronized (lockA) { // 获得lockA的锁
                System.out.println("获得lockA的锁1");

                this.value -= m;
            } // 释放lockA的锁
        } // 释放lockB的锁
    }

    static void sleep1s() {
        try {
            Thread.sleep(1000);
        } catch (InterruptedException e) {
            e.printStackTrace();
        }
    }
    
    public static void main(String[] args) {
                 new  Thread1().start();
                 new  Thread2().start();

        Demo demo=new Demo();
        try {
            demo.add(2);
            demo.dec(2);
        } catch (Exception e) {
            e.printStackTrace();
        }

    }


}

 

结果：
Connected to the target VM, address: '127.0.0.1:51315', transport: 'socket'
Thread-1: try get lock A...
Thread-2: try get lock B...
Thread-2: lock B got.
Thread-1: lock A got.
Thread-1: try get lock B...
Thread-2: try get lock A...
造成死锁，
线程1 A加锁后获取B
线程2 B加锁后获取A
造成线程1获取B锁进入等待
线程2获取A锁进入等待，死循环下取
思维没有好的解决办法，只有停止虚拟机进行解决

 

 

 

使用ReentrantLock比直接使用synchronized更安全，可以替代synchronized进行线程同步。

但是，synchronized可以配合wait和notify实现线程在条件不满足时等待，条件满足时唤醒，用ReentrantLock我们怎么编写wait和notify的功能呢？

答案是使用Condition对象来实现wait和notify的功能。

我们仍然以TaskQueue为例，把前面用synchronized实现的功能通过ReentrantLock和Condition来实现

 

 

package com.example.demo.config;

import java.util.LinkedList;
import java.util.Queue;
import java.util.concurrent.Executors;
import java.util.concurrent.LinkedBlockingDeque;
import java.util.concurrent.ThreadPoolExecutor;
import java.util.concurrent.TimeUnit;
import java.util.concurrent.locks.Condition;
import java.util.concurrent.locks.Lock;
import java.util.concurrent.locks.ReentrantLock;

public class Demo {
    private final Lock lock = new ReentrantLock();
    private final Condition condition = lock.newCondition();
    private Queue<String> queue = new LinkedList<>();

    public void addTask(String s) throws InterruptedException {
        lock.lock();
        try {
            queue.add(s);
            sleep();
            condition.signalAll();
        } finally {
            System.out.println("锁取消，全部唤醒");
            lock.unlock();
        }
    }

    public String getTask() throws InterruptedException {
        lock.lock();
        try {
            while (queue.isEmpty()) {
                condition.await();
                sleep();
                System.out.println("等待了");
            }
            return queue.remove();
        } finally {
            System.out.println("锁取消，进行等待");
            lock.unlock();
        }
    }

    static void sleep() throws InterruptedException {
        Thread.sleep(1000);
    }

    public static void main(String[] args) throws InterruptedException {

        ThreadPoolExecutor threadPoolExecutor = new ThreadPoolExecutor(30,
                100,//当前线程最大并发数
                3,
                TimeUnit.SECONDS,
                new LinkedBlockingDeque<>(30),//候客区
                Executors.defaultThreadFactory(),
                //new ThreadPoolExecutor.AbortPolicy() //满了，再进来的线程，不处理，直接抛异常
                //new ThreadPoolExecutor.CallerRunsPolicy()//满了，返回回去，这里会让main线程处理
                //new ThreadPoolExecutor.DiscardPolicy()//满了，再进来的线程，不处理，不会抛异常
                new ThreadPoolExecutor.DiscardOldestPolicy()//满了，再进来的线程，尝试去获取cpu时间片，会与最早的线程竞争，不会抛异常
        );


        Demo demo = new Demo();
        for (int i = 0; i < 9; i++) {
            threadPoolExecutor.execute(() -> {
                demo.addTask("abcd");
                demo.getTask();
            });
        }
        boolean shutdown = threadPoolExecutor.isShutdown();
//        threadPoolExecutor.wait(30000);
        if (shutdown) {
            threadPoolExecutor.shutdown();
        }
    }
}

 

可见，使用Condition时，引用的Condition对象必须从Lock实例的newCondition()返回，这样才能获得一个绑定了Lock实例的Condition实例。

Condition提供的await()、signal()、signalAll()原理和synchronized锁对象的wait()、notify()、notifyAll()是一致的，并且其行为也是一样的：

    await()会释放当前锁，进入等待状态；

    signal()会唤醒某个等待线程；

    signalAll()会唤醒所有等待线程；

    唤醒线程从await()返回后需要重新获得锁。

此外，和tryLock()类似，await()可以在等待指定时间后，如果还没有被其他线程通过signal()或signalAll()唤醒，可以自己醒来：

if (condition.await(1, TimeUnit.SECOND)) {
    // 被其他线程唤醒
} else {
    // 指定时间内没有被其他线程唤醒
}

可见，使用Condition配合Lock，我们可以实现更灵活的线程同步。