解析Java线程同步锁的选择方法
在需要线程同步的时候如何选择合适的线程锁?
例:选择可以存入到常量池当中的对象,String对象等
代码如下:
public class SyncTest
{
private String name = "name";
public void method(String flag)
{
synchronized (name)
{
System.out.println(flag + ", invoke method ....");
try
{
Thread.sleep(1000);
}
catch (InterruptedException e)
{
e.printStackTrace();
}
}
}
public static void main(String[] args)
{
SyncTest test1 = new SyncTest();
SyncTest test2 = new SyncTest();
MyThread1 myThread1 = new MyThread1();
MyThread1 myThread2 = new MyThread1();
myThread1.syncTest = test1;
myThread2.syncTest = test1;
MyThread1 myThread3 = new MyThread1();
MyThread1 myThread4 = new MyThread1();
myThread3.syncTest = test2;
myThread4.syncTest = test2;
myThread1.start();
myThread2.start();
myThread3.start();
myThread4.start();
}
}
public class MyThread1 extends Thread
{
SyncTest syncTest;
@Override
public void run()
{
syncTest.method(this.getName());
}
}
本来应该是要实现线程thread1和thread2同步,线程thread3和thread4同步的,但结果呢?
却是使得线程thread1、thread2、thread3、thread4同步了,很是郁闷。
我推荐选用的同步锁对象:
代码如下:
package com.rcx.thread;
public class SyncTest
{
// 特殊的instance变量,用于充当同步锁的对象
private byte[] lock = new byte[0];
public void method(String flag)
{
synchronized (lock)
{
System.out.println(flag + ", invoke method f....");
try
{
Thread.sleep(1000);
}
catch (InterruptedException e)
{
e.printStackTrace();
}
}
}
public static void main(String[] args)
{
SyncTest test1 = new SyncTest();
SyncTest test2 = new SyncTest();
MyThread1 myThread1 = new MyThread1();
MyThread1 myThread2 = new MyThread1();
myThread1.syncTest = test1;
myThread2.syncTest = test1;
MyThread1 myThread3 = new MyThread1();
MyThread1 myThread4 = new MyThread1();
myThread3.syncTest = test2;
myThread4.syncTest = test2;
myThread1.start();
myThread2.start();
myThread3.start();
myThread4.start();
}
}
推荐使用0长度的byte数组充当同步锁对象,不会产生很诧异的错误同时不会占用很大内存。