Java高并发编程:线程锁技术

时间:2022-06-15 15:10:13

笔记摘要

这里介绍了java5中的线程锁技术:Lock和Condition,实现线程间的通信,其中的读锁和写锁的使用通过一个缓存系统进行了演示,对于Condition的应用通过一个阻塞队列进行演示。

线程锁技术:Lock & Condition 实现线程同步通信所属包:java.util.concurrent.locks

线程锁 说明
Synchronized 同步方法,锁对象是this;同步静态方法,锁对象是字节码.class;同步代码块,锁对象是任意对象,但必须是同一个对象
Lock 同步锁接口
ReentrantLock lock(),unlock(),newCondition()
ReadWriteLock 读写锁接口
ReentrantReadWriteLock readLock()获取读锁,writeLock()获取写锁
Condition 线程间通信 await()等待 signal()唤醒

1. Lock

Lock比传统线程模型中的synchronized方式更加面向对象,相对于synchronized 方法和语句它具有更广泛的锁定操作,此实现允许更灵活的结构,可以具有差别很大的属性,可以支持多个相关的 Condition 对象。

于现实生活中类似,锁本身也是一个对象。两个线程执行的代码片段要实现同步互斥的结果,它们必须用同一个Lock对象,锁是上在代表要操作的资源的类的内部方法中,而不是线程代码中。

ReentrantLock

方法声明 功能描述
lock() 获取锁
tryLock() 尝试获取锁
unock() 释放锁
newCondition() 获取锁的Condition

常用形式如下

Lock lock = new ReentrantLock();

public void doSth(){
lock.lock();
try {
// 执行某些操作
}finally {
lock.unlock();
}
}

读写锁

分为读锁和写锁,多个读锁不互斥,读锁与写锁互斥,写锁与写锁互斥,这是由JVM自己控制的。你只要上好相应的锁即可。如果你的代码只读数据,可以很多人同时读,但不能同时写,那就上读锁;如果你的代码修改数据,只能有一个人在写,且不能同时读取,那就上写锁。总之,读的时候上读锁,写的时候上写锁!

读写锁的使用情景:

  • 如果代码只读数据,就可以很多人共同读取,但不能同时写。
  • 如果代码修改数据,只能有一个人在写,且不能同时读数据。

API中ReentrantReadWriteLock类提供的一个读写锁缓存示例:

class CachedData {  

  Object data;
volatile boolean cacheValid;

  ReentrantReadWriteLock rwl = new ReentrantReadWriteLock();

  void processCachedData() {

   rwl.readLock().lock();

   if (!cacheValid) {

   // Must release read lock before acquiring write lock

   rwl.readLock().unlock();
   rwl.writeLock().lock();

   // Recheck state because another thread might have acquired
    // write lock and changed state before we did.
   if (!cacheValid) {

   data = ...
   cacheValid = true;

   }
   // Downgrade by acquiring read lock before releasing write lock
rwl.readLock().lock();

   rwl.writeLock().unlock(); // Unlock write, still hold read

   }
   use(data);
   rwl.readLock().unlock();

  }
}

读写锁的应用:编写一个缓存系统

注解:为了避免线程的安全问题,synchronized和ReadWriteLock都可以,synchronized也防止了并发读取,性能较低有一个线程先进去,开始读取数据,进行判断,发现没有数据,其他线程就没有必要进去了,就释放读锁,加上写锁,去查找数据写入,为了避免写入的其他对象等待,再做一次判断,数据写入完成后,释放写锁,上读锁,防止写入,还原原来的状态。

两次判断:第一次为了写入数据,所以释放读锁,上写锁。第二次为了防止阻塞的线程重复写入

import java.util.HashMap;  
import java.util.Map;
import java.util.concurrent.locks.ReadWriteLock;
import java.util.concurrent.locks.ReentrantReadWriteLock;

public class CacheDemo {

//定义一个map用于缓存对象
private Map<String, Object> cache = new HashMap<String, Object>();

//获取一个读写锁对象
private ReadWriteLock rwl = new ReentrantReadWriteLock();

//带有缓存的获取指定值的方法
public Object getData(String key){
rwl.readLock().lock(); //上读锁
Object value = null;
try{
value = cache.get(key); //获取要查询的值
if(value == null){ //线程出现安全问题的地方

rwl.readLock().unlock(); //没有数据,释放读锁,上写锁
// 多个线程去上写锁,第一个上成功后,其他线程阻塞,第一个线程开始执行下面的代码,最后
// 释放写锁后,后面的线程继续上写锁,为了避免后面的线程重复写入,进行二次判断
rwl.writeLock().lock();
try{
if(value==null){ //二次判断,防止其他线程重复写数据
value = "aaaa"; //实际是去查询数据库
}
}finally{
rwl.writeLock().unlock(); //写完数据,释放写锁
}
rwl.readLock().lock(); //恢复读锁
}
}finally{
rwl.readLock().unlock(); //最终释放读锁
}
return value; //返回获取到的值
}
}

虚假唤醒:用while代替if

Lock lock = new ReentrantLock();
try {
lock.lock();
//需要加锁的代码
}finally {
lock.unlock();
}

读写锁测试

public class ReadWriteLockTest {
public static void main(String[] args) {
final Queue3 q3 = new Queue3();
for(int i=0;i<3;i++)
{
new Thread(){
public void run(){
while(true){
q3.get();
}
}
}.start();
new Thread(){
public void run(){
while(true){
q3.put(new Random().nextInt(10000));
}
}
}.start();
}
}
}

class Queue3{
private Object data = null;
ReadWriteLock rwl = new ReentrantReadWriteLock
();
public void get(){
rwl.readLock().lock();
try {
System.out.println(Thread.currentThread().getName() + " be ready to read data!");
Thread.sleep((long)(Math.random()*1000));
System.out.println(Thread.currentThread().getName() + "have read data :" + data);
} catch (InterruptedException e) {
e.printStackTrace();
}finally{
rwl.readLock().unlock();
}
}

public void put(Object data){
rwl.writeLock().lock();
try {
System.out.println(Thread.currentThread().getName() + " be ready to write data!");
Thread.sleep((long)(Math.random()*1000));
this.data = data;
System.out.println(Thread.currentThread().getName() + " have write data: " + data);
} catch (InterruptedException e) {
e.printStackTrace();
}finally{
rwl.writeLock().unlock();
}
}
}
Thread-0 be ready to read data!
Thread-2 be ready to read data!
Thread-4 be ready to read data!
Thread-0have read data :null
Thread-2have read data :null
Thread-4have read data :null
Thread-5 be ready to write data!
Thread-5 have write data: 7975
Thread-5 be ready to write data!
Thread-5 have write data: 9832
Thread-3 be ready to write data!
Thread-3 have write data: 2813
Thread-3 be ready to write data!
Thread-3 have write data: 7998
Thread-1 be ready to write data!
Thread-1 have write data: 6737
Thread-1 be ready to write data!
...

2. Condition

用于实现线程间的通信,是为了解决Object.wait()、nitify()、notifyAll()难以使用的问题

Condition 将 Object 监视器方法(wait、notify 和 notifyAll)分解成截然不同的对象,以便通过将这些对象与任意 Lock 实现组合使用,为每个对象提供多个等待 set(wait-set)。其中,Lock 替代了 synchronized 方法和语句的使用,Condition 替代了 Object 监视器方法wait和notify的使用

一个锁内部可以有多个Condition,即有多路等待通知,传统的线程机制中一个监视器对象上只能有一路等待和通知,要想实现多路等待和通知,必须嵌套使用多个同步监视器对象。使用一个监视器往往会产生顾此失彼的情况。

在等待 Condition 时,允许发生“虚假唤醒”,这通常作为对基础平台语义的让步。对于大多数应用程序,这带来的实际影响很小,因为 Condition 应该总是在一个循环中被等待,并测试正被等待的状态声明。某个实现可以随意移除可能的虚假唤醒,但建议应用程序程序员总是假定这些虚假唤醒可能发生,因此总是在一个循环中等待。

方法声明 功能描述
await() 线程等待
await(long time, TimeUnit unit) 线程等待特定的时间,超过等待时间则为超时
signal() 随机唤醒某个等待线程
signalAll() 唤醒所有等待中的线程

Condition的应用:阻塞队列(使用了两个监视器)

说明:该应用是 java.util.concurrent.locks包中Condition接口中的示例代码。使用了两个Condition分别用于管理取数据的线程,和存数据的线程,这样就可以明确的唤醒需要的一类线程,如果使用一个Condition,当队列满了之后,唤醒的并不一定就是取数据的线程

class BoundedBuffer {  

final Lock lock = new ReentrantLock();
final Condition notFull = lock.newCondition();
final Condition notEmpty = lock.newCondition();

final Object[] items = new Object[100];
int putptr, takeptr, count;

public void put(Object x) throws InterruptedException {
lock.lock();
try {
while (count == items.length) //循环判断队列是否已存满
notFull.await(); //如果队列存满了,则要存入数据的线程等待
items[putptr] = x;
if (++putptr == items.length) putptr = 0;//当队列放满,指针回到0
++count; //添加了一个数据
notEmpty.signal(); //队列中有数据了,所以就唤醒取数据的线程
} finally {
lock.unlock();
}
}

public Object take() throws InterruptedException {
lock.lock();
try {
while (count == 0) //循环判断,队列是否有空位
notEmpty.await(); //要取的线程等待
Object x = items[takeptr];
if (++takeptr == items.length) takeptr = 0;
--count; //取走一个,说明队列有空闲的位置,
notFull.signal(); //所以通知存入的线程
return x;
} finally {
lock.unlock();
}
}
}

Condition测试

public class ConditionCommunication {
public static void main(String[] args) {
final Business business = new Business();
new Thread(
new Runnable() {
@Override
public void run() {
for(int i=1;i<=5;i++){
business.sub(i);
}
}
}
).start();

for(int i=1;i<=5;i++){
business.main(i);
}
}

class Business {
Lock lock = new ReentrantLock();
Condition condition = lock.newCondition();
private boolean bShouldSub = true;
public void sub(int i){
lock.lock();
try{
while(!bShouldSub){
try {
condition.await();
} catch (Exception e) {
e.printStackTrace();
}
}
for(int j=1;j<=2;j++){
System.out.println("sub thread sequence of " + j + ",loop of " + i);
}
bShouldSub = false;
condition.signal();
}finally{
lock.unlock();
}
}

public void main(int i){
lock.lock();
try{
while(bShouldSub){
try {
condition.await();
} catch (Exception e) {
e.printStackTrace();
}
}
for(int j=1;j<=4;j++){
System.out.println("main thread sequence of " + j + ",loop of " + i);
}
bShouldSub = true;
condition.signal();
}finally{
lock.unlock();
}
}
}
}

输出结果

sub thread sequence of 1,loop of 1
sub thread sequence of 2,loop of 1
main thread sequence of 1,loop of 1
main thread sequence of 2,loop of 1
main thread sequence of 3,loop of 1
main thread sequence of 4,loop of 1
sub thread sequence of 1,loop of 2
sub thread sequence of 2,loop of 2
main thread sequence of 1,loop of 2
main thread sequence of 2,loop of 2
main thread sequence of 3,loop of 2
main thread sequence of 4,loop of 2
sub thread sequence of 1,loop of 3
sub thread sequence of 2,loop of 3
main thread sequence of 1,loop of 3
main thread sequence of 2,loop of 3
main thread sequence of 3,loop of 3
main thread sequence of 4,loop of 3
sub thread sequence of 1,loop of 4
sub thread sequence of 2,loop of 4
main thread sequence of 1,loop of 4
main thread sequence of 2,loop of 4
main thread sequence of 3,loop of 4
main thread sequence of 4,loop of 4
sub thread sequence of 1,loop of 5
sub thread sequence of 2,loop of 5
main thread sequence of 1,loop of 5
main thread sequence of 2,loop of 5
main thread sequence of 3,loop of 5
main thread sequence of 4,loop of 5

使用ReentrantLock和Condition实现一个简单的阻塞队列MyArrayBlockingQueue,如果调用take方法时集合中没有数据,那么调用线程就阻塞;如果调用put方法时,集合数据已满,那么也会引起调用线程阻塞。但是,这两个阻塞的条件时不同的,分别为为notFull和notEmpty

import java.util.concurrent.locks.Condition;
import java.util.concurrent.locks.Lock;
import java.util.concurrent.locks.ReentrantLock;

public class MyArrayBlockingQueue<T> {
// 数据数组
private final T[] items;
// 锁
private final Lock lock = new ReentrantLock();
// 队满的条件
private Condition notFull = lock.newCondition();
// 队空条件
private Condition notEmpty = lock.newCondition();
// 头部索引
private int head;
// 尾部索引
private int tail;
// 数据的个数
private int count;

public MyArrayBlockingQueue(int maxSize) {
items = (T[]) new Object[maxSize];
}

public MyArrayBlockingQueue() {
this(10);
}

public void put(T t) {
lock.lock();
try {
while (count == getCapacity()) {
System.out.println("数据已满,等待");
notFull.await();
}
items[tail] = t;
if (++tail == getCapacity()) {
tail = 0;
}
++count;
notEmpty.signalAll(); // 唤醒等待数据的线程
} catch (InterruptedException e) {
e.printStackTrace();
} finally {
lock.unlock();
}
}

public T take() {
lock.lock();
try {
while (count == 0) {
System.out.println("还没有数据,请等待");
notEmpty.await();
}
T ret = items[head];
items[head] = null;
if (++head == getCapacity()) {
head = 0;
}
--count;
notFull.signalAll(); // 唤醒添加数据的线程
return ret;
} catch (InterruptedException e) {
e.printStackTrace();
} finally {
lock.unlock();
}
return null;
}

public int getCapacity() {
return items.length;
}

public int size() {
lock.lock();
try {
return count;
} finally {
lock.unlock();
}
}

public static void main(String[] args) {
MyArrayBlockingQueue<Integer> aQueue = new MyArrayBlockingQueue<Integer>();
aQueue.put(3);
aQueue.put(24);
for (int i = 0; i < 5; i++) {
System.out.println(aQueue.take());
}
}
}

输出结果

3
24
还没有数据,请等待

3. Condition练习

一共有3个线程,两个子线程先后循环2次,接着主线程循环3次,接着又回到两 个子线程先后循环2次,再回到主线程又循环3次,如此循环5次。

思路:老二先执行,执行完唤醒老三,老三执行完唤醒老大,老大执行完唤醒老二,以此循环,所以定义3个Condition对象和一个执行标识即可

示例出现的问题:两个文件中有同名类的情况

解决方案:可以将一个文件中的那个同名外部类放进类中,但是静态不能创建内部类的实例对象,所以需要加上static,这样两个类的名称就不一样了。 一个是原来的类名,一个是在自己类名前面加上外部类的类名。

import java.util.concurrent.locks.Condition;  
import java.util.concurrent.locks.Lock;
import java.util.concurrent.locks.ReentrantLock;

public class ThreeConditionCommunication {
public static void main(String[] args){

final Business business = new Business();

   //创建并启动子线程老二
new Thread(new Runnable(){
@Override
public void run() {
for(int i=1;i<=5;i++){
business.sub2(i);
}
}
}).start();

   //创建并启动子线程老三
new Thread(new Runnable(){
@Override
public void run() {
for(int i=1;i<=5;i++){
business.sub3(i);
}
}
}).start();

//主线程
for(int i=1;i<=5;i++){
business.main(i);
}
}

static class Business{

Lock lock = new ReentrantLock();
Condition condition1 = lock.newCondition();
Condition condition2 = lock.newCondition();
Condition condition3 = lock.newCondition();

//定义一个变量来决定线程的执行权
private int ShouldSub = 1;

public void sub2(int i){
//上锁,不让其他线程执行
lock.lock();
try{
if(ShouldSub != 2){ //如果不该老二执行,就等待
try {
condition2.await();
} catch (InterruptedException e) {
e.printStackTrace();
}
}

for(int j=1;j<=2;j++){
System.out.println("sub thread sequence of"+i+",loop of "+j);
}
ShouldSub = 3; //准备让老三执行
condition3.signal(); //唤醒老三
}finally{
lock.unlock();
}
}

public void sub3(int i){

lock.lock();
try{
if(ShouldSub != 3){
try {
condition3.await();
} catch (InterruptedException e) {
e.printStackTrace();
}
}

for(int j=1;j<=2;j++){
System.out.println("sub2 thread sequence of"+i+",loop of "+j);
}
ShouldSub = 1; //准备让老大执行
condition1.signal(); //唤醒老大
}finally{
lock.unlock();
}
}

   //主线程
public void main(int i){

lock.lock();
try{
if(ShouldSub!=1){
try {
condition1.await();
} catch (InterruptedException e) {
e.printStackTrace();
}
}
for(int j=1;j<=3;j++){
System.out.println("main thread sequence of"+i+", loop of "+j);
}
ShouldSub = 2; //准备让老二执行
condition2.signal(); //唤醒老二
}finally{
lock.unlock();
}
}
}
}

输出结果

main thread sequence of1, loop of 1
main thread sequence of1, loop of 2
main thread sequence of1, loop of 3
sub thread sequence of1,loop of 1
sub thread sequence of1,loop of 2
sub2 thread sequence of1,loop of 1
sub2 thread sequence of1,loop of 2
main thread sequence of2, loop of 1
main thread sequence of2, loop of 2
main thread sequence of2, loop of 3
sub thread sequence of2,loop of 1
sub thread sequence of2,loop of 2
sub2 thread sequence of2,loop of 1
sub2 thread sequence of2,loop of 2
main thread sequence of3, loop of 1
main thread sequence of3, loop of 2
main thread sequence of3, loop of 3
sub thread sequence of3,loop of 1
sub thread sequence of3,loop of 2
sub2 thread sequence of3,loop of 1
sub2 thread sequence of3,loop of 2
main thread sequence of4, loop of 1
main thread sequence of4, loop of 2
main thread sequence of4, loop of 3
sub thread sequence of4,loop of 1
sub thread sequence of4,loop of 2
sub2 thread sequence of4,loop of 1
sub2 thread sequence of4,loop of 2
main thread sequence of5, loop of 1
main thread sequence of5, loop of 2
main thread sequence of5, loop of 3
sub thread sequence of5,loop of 1
sub thread sequence of5,loop of 2
sub2 thread sequence of5,loop of 1
sub2 thread sequence of5,loop of 2

4. 多路等待和通知

class BoundedBuffer {
final Lock lock = new ReentrantLock();
final Condition notFull = lock.newCondition();
final Condition notEmpty = lock.newCondition();

final Object[] items = new Object[100];
int putptr, takeptr, count;

public void put(Object x) throws InterruptedException {
lock.lock();
try {
while (count == items.length)
notFull.await();
items[putptr] = x;
if (++putptr == items.length) putptr = 0;
++count;
notEmpty.signal();
} finally {
lock.unlock();
}
}

public Object take() throws InterruptedException {
lock.lock();
try {
while (count == 0)
notEmpty.await();
Object x = items[takeptr];
if (++takeptr == items.length) takeptr = 0;
--count;
notFull.signal();
return x;
} finally {
lock.unlock();
}
}
}