并发编程学习总结(六) ：java 显式锁ReentrantLock使用详解之测试锁与超时

我在上面的这篇博客中，曾经写到过线程在获取锁的过程中是无法中断的。

但是ReentrantLock提供了tryLock()、tryLock(long timeout, TimeUnit unit)、lock.lockInterruptibly()

tryLock() 方法试图申请一个锁，在成功获得锁后返回true，否则，立即返回false,而且线程可以立即离开去做其他的事情。

tryLock(long timeout, TimeUnit unit) 是一个具有超时参数的尝试申请锁的方法，阻塞时间不会超过给定的值；如果成功则返回true

lockInterruptibly() 获得锁，但是会不确定地发生阻塞。如果线程被中断，抛出一个InterruptedException异常。

如果当前获得锁的线程在做大量耗时的工作，使用lock.lock()方法申请锁的线程会一直阻塞，这样就降低了多线程的效率。而使用tryLock()方法申请锁，如果锁不可用则线程不会阻塞，转而可以去做其他工作。代码实例如下：

public class ReentrantLockTest4 {

private ReentrantLock lock = new ReentrantLock();

public void tryLockTest() throws InterruptedException {

long beginTime = System.currentTimeMillis();
while(System.currentTimeMillis() - beginTime <100) {}
// 当前线程尝试获得锁，如果获得锁返回true，否则返回false
if(lock.tryLock()) {
try{
System.out.println(Thread.currentThread().getName() + " tryLock get lock");
} finally {
lock.unlock();
System.out.println(Thread.currentThread().getName() + " tryLock release lock");
}
} else {
System.out.println(Thread.currentThread().getName() + " tryLock can not get lock");
}

}


public void lockTest() {
try{
// 当前线程在锁可用时直接获得锁，锁不可用时阻塞当前线程
lock.lock();
System.out.println(Thread.currentThread().getName() + " lock get lock");
long beginTime = System.currentTimeMillis();
while(System.currentTimeMillis() - beginTime <1000) {}
} finally {
lock.unlock();
System.out.println(Thread.currentThread().getName() + " lock release lock");
}

}

public static void main(String[] args) {
// TODO Auto-generated method stub
final ReentrantLockTest4 test = new ReentrantLockTest4();
Thread tryLock = new Thread(new Runnable() {
public void run() {
try {
test.tryLockTest();
} catch (InterruptedException e) {
// TODO Auto-generated catch block
e.printStackTrace();
}
}
},"tryLock_thread");

Thread lock = new Thread(new Runnable() {
public void run() {
test.lockTest();
}
},"lock_thread");

tryLock.start();
lock.start();
}

}

输出结果：
lock_thread lock get lock
tryLock_thread tryLock can not get lock
lock_thread lock release lock

lock方法不能被中断。如果一个线程在等待获得一个锁时被中断，中断线程在获得锁之前会一直处于阻塞状态。如果出现死锁，那么lock方法就无法被终止。但是tryLock和lockInterruptibly方法在申请锁的过程中是可以被中断的。代码如下

public class ReentrantLockTest5 {

private ReentrantLock lock = new ReentrantLock();

public void tryLockInterruptTest() {
long beginTime = System.currentTimeMillis();
while(System.currentTimeMillis() - beginTime <100) {}
try {
if (lock.tryLock(5000, TimeUnit.MILLISECONDS)) {
try{
System.out.println(Thread.currentThread().getName() + " tryLock get lock");
}finally {
lock.unlock();
}
}
} catch (InterruptedException e) {
// TODO Auto-generated catch block
System.out.println(Thread.currentThread().getName() + " was interrupted");
}
}


public void lockTest() {
try{
// 当前线程在锁可用时直接获得锁，锁不可用时阻塞当前线程
lock.lock();
System.out.println(Thread.currentThread().getName() + " lock get lock");
long beginTime = System.currentTimeMillis();
while(System.currentTimeMillis() - beginTime <1000) {}
} finally {
lock.unlock();
System.out.println(Thread.currentThread().getName() + " lock release lock");
}

}

public static void main(String[] args) {
// TODO Auto-generated method stub
final ReentrantLockTest5 test =  new ReentrantLockTest5();
Thread thread_tryLock = new Thread(new Runnable(){
@Override
public void run() {
// TODO Auto-generated method stub
test.tryLockInterruptTest();
}
},"tryLockInterruptTest");
Thread thread_lock = new Thread(new Runnable(){
@Override
public void run() {
// TODO Auto-generated method stub
test.lockTest();
}
},"lockTest");
thread_tryLock.start();
thread_lock.start();
try {
Thread.sleep(200);
} catch (InterruptedException e) {
// TODO Auto-generated catch block
System.out.println("main thread was interrupted");
}
thread_tryLock.interrupt();
}

}


输出结果：
lockTest lock get lock
tryLockInterruptTest was interrupted
lockTest lock release lock

秒客网

并发编程学习总结(六) ：java 显式锁ReentrantLock使用详解之测试锁与超时

相关文章