上一篇我们讲解了synchronized
的使用,用它就可以满足数据的同步,但是为什么有时我还是会使用Lock
呢?因此在这里我们不得不说说synchronized
的不足之处,例如当获取锁的线程执行完要释放锁时,由于某些原因该线程被阻塞了,那么此时并没有将获取的锁释放掉,别的线程也就一直等着锁的释放,这样就很影响程序的执行效率。要解决这个问题,只要让等待的线程不无限期的等待下去就可以了,Lock
就完全可以解决这样的问题。
Lock
类的使用:
public interface Lock {
void lock();
void lockInterruptibly() throws InterruptedException;
boolean tryLock();
boolean tryLock(long var1, TimeUnit var3) throws InterruptedException;
void unlock();
Condition newCondition();
}
它是个接口类,有6个方法,前面4个就是用来获取锁的,第5个用来释放锁的,第6个用来设置等待和通知;那么下面我们具体操作下这些方法的使用。
lock()的使用
private Lock lock = new ReentrantLock();
public void testLock(String name) {
lock.lock();
System.out.println(Thread.currentThread().getName() + ",Request Lock...");
try {
for (int i = 0; i < 100; i++) {
y++;
System.out.println(name + "-->CurrThread:" + Thread.currentThread().getName() + " y=" + y);
}
} catch (Exception e) {
e.printStackTrace();
} finally {
lock.unlock();
System.out.println(Thread.currentThread().getName() + ",Release Lock ...");
}
}
ReentrantLock
类实现了接口Lock
,新增了一些方法;Lock
自己加锁,用完后自己在释放掉锁;如果加锁的程序在运行过程中发生异常,也不会释放锁,那么就需要我们自己进行处理了,所以我们在使用Lock
时会将其放入try{}catch{}finally{}
中,将释放锁的操作放入finally
中。
传入同一实例对象,在两个线程中调用此方法,运行,结果为:
pool-4-thread-1,Request Lock...
TempRunnable->testLock-->CurrThread:pool-4-thread-1 y=1
TempRunnable->testLock-->CurrThread:pool-4-thread-1 y=2
TempRunnable->testLock-->CurrThread:pool-4-thread-1 y=3
pool-4-thread-1,Release Lock...
pool-4-thread-1,Request Lock...
Count Instance 1:-->CurrThread:pool-4-thread-1 y=4
Count Instance 1:-->CurrThread:pool-4-thread-1 y=5
Count Instance 1:-->CurrThread:pool-4-thread-1 y=6
pool-4-thread-1,Release Lock...
从结果上就可以看出,当释放完锁后另一线程才开始申请锁,运行。
tryLock()
方法的使用
tryLock()
方法具有返回值,得到锁返回true
,否则返回false
, 看代码:
public void tryLock(String name) {
if (lockTry.tryLock()) {
System.out.println(Thread.currentThread().getName() + ",Request Lock success...");
try {
for (int i = 0; i < 3; i++) {
y++;
System.out.println(name + "-->CurrThread:" + Thread.currentThread().getName() + " y=" + y);
}
} catch (Exception e) {
e.printStackTrace();
} finally {
System.out.println(Thread.currentThread().getName() + ",Release Lock ...");
}
} else {
System.out.println(Thread.currentThread().getName() + ",Request Lock failed...");
}
}
传入同一实例对象,在两个线程中调用此方法,运行结果为:
Thread-1,Request Lock success...
Thread-0,Request Lock failed...
Count Instance 1:-->CurrThread:Thread-1 y=1
Count Instance 1:-->CurrThread:Thread-1 y=2
Count Instance 1:-->CurrThread:Thread-1 y=3
Thread-1,Release Lock ...
线程1申请锁成功,线程0申请锁失败了,此时他并没有继续在等待线程1的执行完成。
tryLock(long l, TimeUnit t)
方法的使用
tryLock(long l, TimeUnit t)
第一个参数表示要等待的时间,第二个参数表示时间单位。
更改上面的方法,如下:
public void tryLock(String name) {
try {
if (lockTry.tryLock(2000, TimeUnit.MILLISECONDS)) {
System.out.println(Thread.currentThread().getName() + ",Request Lock success...");
for (int i = 0; i < 3; i++) {
y++;
System.out.println(name + "-->CurrThread:" + Thread.currentThread().getName() + " y=" + y);
}
} else {
System.out.println(Thread.currentThread().getName() + ",Request Lock failed...");
}
} catch (Exception e) {
e.printStackTrace();
} finally {
lockTry.unlock();
System.out.println(Thread.currentThread().getName() + ",Release Lock ...");
}
}
传入同一实例对象,在两个线程中调用此方法,运行结果为:
Thread-1,Request Lock success...
Count Instance 1:-->CurrThread:Thread-1 y=1
Count Instance 1:-->CurrThread:Thread-1 y=2
Count Instance 1:-->CurrThread:Thread-1 y=3
Thread-1,Release Lock ...
Thread-0,Request Lock success...
TempRunnable->TryLock-->CurrThread:Thread-0 y=4
TempRunnable->TryLock-->CurrThread:Thread-0 y=5
TempRunnable->TryLock-->CurrThread:Thread-0 y=6
Thread-0,Release Lock ...
重新设置时间tryLock(1, TimeUnit.MICROSECONDS)
之后,传入同一实例对象,在两个线程中调用此方法,运行结果为:
Thread-1,Request Lock success...
Thread-0,Request Lock failed...
Count Instance 1:-->CurrThread:Thread-1 y=1
Count Instance 1:-->CurrThread:Thread-1 y=2
Count Instance 1:-->CurrThread:Thread-1 y=3
Thread-1,Release Lock ...
一个线程在获取到锁之后,另一线程在获取的话肯定就失败了,但是这里设置了时间之后,在获取不到线程的时候就进行等待设定的时间之后在进行获取,获取到了就返回true
,获取不到或者中途线程中断了就返回false
。
lockInterruptibly()
的使用
lockInterruptibly()
通过这个方法去获取锁时,如果线程 正在等待获取锁,则这个线程能够 响应中断,即中断线程的等待状态。看代码:
public void interrupt(String name) throws InterruptedException {
lockTry.lockInterruptibly();
try {
for (int i = 0; i < 1000; i++) {
y++;
System.out.println(name + "-->CurrThread:" + Thread.currentThread().getName() + " y=" + y);
}
} catch (Exception e) {
e.printStackTrace();
} finally {
lockTry.unlock();
System.out.println(Thread.currentThread().getName() + ",Release Lock ...");
}
}
生成一个线程类TestRunnable
,在里面调用interrupt
方法,并处理的异常,代码为:
public class TestRunnable implements Runnable {
private Count count;
public TestRunnable(Count count) {
this.count = count;
}
@Override
public void run() {
try {
count.interrupt(count.getName());
} catch (InterruptedException e) {
e.printStackTrace();
System.out.println(Thread.currentThread().getName() + ",Request Lock Interrupted...");
}
}
}
设置中断线程:
Count count1 = new Count("Count Instance 1:");
Thread thread1 = new Thread(new TestRunnable(count1));
Thread thread2 = new Thread(new TestRunnable(count1));
thread1.start();
thread2.start();
try {
Thread.sleep(10);
} catch (InterruptedException e) {
e.printStackTrace();
}
thread2.interrupt();
上面代码设置的是休眠10毫秒后,中断thread2
;运行代码,看看结果:
MainThread:main started...
TempRunnable->Interrupt-->CurrThread:Thread-0 y=1
......
java.lang.InterruptedException
at java.util.concurrent.locks.AbstractQueuedSynchronizer.doAcquireInterruptibly(AbstractQueuedSynchronizer.java:898)
......
at java.lang.Thread.run(Thread.java:745)
......
TempRunnable->Interrupt-->CurrThread:Thread-0 y=507
Thread-1,Request Lock Interrupted...
......
TempRunnable->Interrupt-->CurrThread:Thread-0 y=1000
Thread-0,Release Lock ...
从结果看,10毫秒后,我们中断了正在等待锁的线程thread2
,这也说明线程运行时间超出了10毫秒,如果小于10毫秒,那么线程不会被中断,线程thread2
会获得锁,并将代码执行完成。
我们在写interrupt
方法时,我们将lockTry.lockInterruptibly();
写在了try{}catch{}
之外,这样异常就从方法抛出,代码为:public void interrupt(String name) throws InterruptedException {}
;那么我们为什么不把lockTry.lockInterruptibly();
写在了try{}catch{}
之内呢?我们用代码验证下,将方法更改为:
public void interrupt(String name) {
try {
lockTry.lockInterruptibly();
for (int i = 0; i < 1000; i++) {
y++;
System.out.println(name + "-->CurrThread:" + Thread.currentThread().getName() + " y=" + y);
}
} catch (Exception e) {
e.printStackTrace();
System.out.println(Thread.currentThread().getName() + ",Request Lock Interrupted...");
} finally {
lockTry.unlock();
System.out.println(Thread.currentThread().getName() + ",Release Lock ...");
}
}
运行,结果为:
MainThread:main started...
TempRunnable->Interrupt-->CurrThread:Thread-0 y=1
......
Exception in thread "Thread-1" java.lang.IllegalMonitorStateException
at java.util.concurrent.locks.ReentrantLock$Sync.tryRelease(ReentrantLock.java:151)
......
at java.lang.Thread.run(Thread.java:745)
......
TempRunnable->Interrupt-->CurrThread:Thread-0 y=507
Thread-1,Request Lock Interrupted...
......
TempRunnable->Interrupt-->CurrThread:Thread-0 y=1000
Thread-0,Release Lock ...
结果中抛出了异常"Thread-1" IllegalMonitorStateException
,它是"Thread-1"
抛出来的,但不是中断异常。当线程1在等待锁的过程中,10毫秒后进行了线程中断,执行完成后就会去继续执行finally
中的解锁操作,而线程1并没有获取到锁,这样就造成了异常。
因此,对于方法void lockInterruptibly() throws InterruptedException;
和boolean tryLock(long var1, TimeUnit var3) throws InterruptedException;
都应该将调用语句放在try{}catch{}
之外。
公平锁
的使用
公平锁即尽量以请求锁的顺序来获取锁
看代码的实现:
private Lock lockFair = new ReentrantLock(true);//true 设置公平锁 false 不设置
public void fairLock() {
try {
lockFair.lock();
System.out.println(Thread.currentThread().getName() + ",Request Lock...");
} catch (Exception e) {
e.printStackTrace();
} finally {
lockFair.unlock();
System.out.println(Thread.currentThread().getName() + ",Release Lock...");
}
}
建立线程类,在生成10个线程,调用此方法:
public class MyRunnable implements Runnable {
private Count count;
public MyRunnable(Count count) {
this.count = count;
}
@Override
public void run() {
System.out.println(Thread.currentThread().getName() + " running...");
count.fairLock();
}
}
List<Thread> threadList = new ArrayList<>();
for (int i = 0; i < 5; i++) {
threadList.add(new Thread(new MyRunnable()));
}
for (Thread thread1 : threadList) {
thread1.start();
}
运行,结果为:
Thread-0 running...
Thread-1 running...
Thread-2 running...
Thread-0,Request Lock...
Thread-4 running...
Thread-3 running...
Thread-1,Request Lock...
Thread-0,Release Lock...
Thread-2,Request Lock...
Thread-1,Release Lock...
Thread-4,Request Lock...
Thread-2,Release Lock...
Thread-4,Release Lock...
Thread-3,Request Lock...
Thread-3,Release Lock...
从结果看,它们是按照请求的顺序,进行锁的申请的。
ReadWriteLock
类的使用
来看看的源代码:
public interface ReadWriteLock {
Lock readLock();
Lock writeLock();
}
ReadWriteLock
类是一个接口类,里面就实现了两个方法:读锁和写锁。它的具体实现类为ReentrantReadWriteLock
。我们用代码进行使用演示:
private ReadWriteLock lock = new ReentrantReadWriteLock();
private Lock writeLock = lock.writeLock();
private Lock readLock = lock.readLock();
public void write(String name) {
writeLock.lock();
try {
for (int i = 0; i < 3; i++) {
y++;
System.out.println(name + "-->CurrThread:" + Thread.currentThread().getName() + " y=" + y);
}
} catch (Exception e) {
e.printStackTrace();
System.out.println("Abnormal exit of write method...");
} finally {
System.out.println(Thread.currentThread().getName() + ",Release Write Lock ...");
writeLock.unlock();
}
}
public void read(String name) {
readLock.lock();
try {
for (int i = 0; i < 3; i++) {
System.out.println(name + "-->CurrThread:" + Thread.currentThread().getName() + " output=" + i);
}
} catch (Exception e) {
e.printStackTrace();
System.out.println("Abnormal exit of read method...");
} finally {
System.out.println(Thread.currentThread().getName() + ",Release Read Lock ...");
readLock.unlock();
}
}
上面的两个方法,一个加的是读锁,一个加的是写锁。
生成两个线程,在传入同一实例对象
- 两个线程都调用
read
方法,运行:
Thread-0,Request Read Lock ...
Count Instance 1:-->CurrThread:Thread-0 output=0
Count Instance 1:-->CurrThread:Thread-0 output=1
Thread-1,Request Read Lock ...
Count Instance 1:-->CurrThread:Thread-0 output=2
Count Instance 1:-->CurrThread:Thread-1 output=0
Thread-0,Release Read Lock ...
Count Instance 1:-->CurrThread:Thread-1 output=1
Count Instance 1:-->CurrThread:Thread-1 output=2
Thread-1,Release Read Lock ...
从结果看,可以多线程一起读共享数据。
- 一个线程调用方法
read
,一个线程调用方法write
,运行:
Thread-0,Request Write Lock ...
TempRunnable->WriteLock-->CurrThread:Thread-0 y=1
TempRunnable->WriteLock-->CurrThread:Thread-0 y=2
TempRunnable->WriteLock-->CurrThread:Thread-0 y=3
Thread-0,Release Write Lock ...
Thread-1,Request Read Lock ...
Count Instance 1:-->CurrThread:Thread-1 output=0
Count Instance 1:-->CurrThread:Thread-1 output=1
Count Instance 1:-->CurrThread:Thread-1 output=2
Thread-1,Release Read Lock ...
Thread-0,Request Read Lock ...
Count Instance 1:-->CurrThread:Thread-0 output=0
Count Instance 1:-->CurrThread:Thread-0 output=1
Count Instance 1:-->CurrThread:Thread-0 output=2
Thread-0,Release Read Lock ...
Thread-1,Request Write Lock ...
TempRunnable->WriteLock-->CurrThread:Thread-1 y=1
TempRunnable->WriteLock-->CurrThread:Thread-1 y=2
TempRunnable->WriteLock-->CurrThread:Thread-1 y=3
Thread-1,Release Write Lock ...
从结果看,不管是先调用读还是写,都是一个获取锁执行完成释放锁后,另一个线程才执行。它们之间是互斥的。
- 两个线程都调用
write
方法,运行:
Thread-1,Request Write Lock ...
TempRunnable->WriteLock-->CurrThread:Thread-1 y=1
TempRunnable->WriteLock-->CurrThread:Thread-1 y=2
TempRunnable->WriteLock-->CurrThread:Thread-1 y=3
Thread-1,Release Write Lock ...
Thread-0,Request Write Lock ...
TempRunnable->WriteLock-->CurrThread:Thread-0 y=4
TempRunnable->WriteLock-->CurrThread:Thread-0 y=5
TempRunnable->WriteLock-->CurrThread:Thread-0 y=6
Thread-0,Release Write Lock ...
从结果看,它们也是互斥的,只有一个线程执行完释放锁之后,另一个线程才会执行。
读锁和写锁,除了读读外,读写,写读,写写之间都是互斥的。
condition
类的使用
condition
类,是Java
提供的等待/通知类。看看其源码:
public interface Condition {
//使当前线程加入 await() 等待队列中,并释放当锁,当其他线程调用signal()会重新请求锁。与Object.wait()类似。
void await() throws InterruptedException;
//调用该方法的前提是,当前线程已经成功获得与该条件对象绑定的重入锁,否则调用该方法时会抛出IllegalMonitorStateException。
//调用该方法后,结束等待的唯一方法是其它线程调用该条件对象的signal()或signalALL()方法。等待过程中如果当前线程被中断,该方法仍然会继续等待,同时保留该线程的中断状态。
void awaitUninterruptibly();
// 调用该方法的前提是,当前线程已经成功获得与该条件对象绑定的重入锁,否则调用该方法时会抛出IllegalMonitorStateException。
//nanosTimeout指定该方法等待信号的的最大时间(单位为纳秒)。若指定时间内收到signal()或signalALL()则返回nanosTimeout减去已经等待的时间;
//若指定时间内有其它线程中断该线程,则抛出InterruptedException并清除当前线程的打断状态;若指定时间内未收到通知,则返回0或负数。
long awaitNanos(long var1) throws InterruptedException;
//与await()基本一致,唯一不同点在于,指定时间之内没有收到signal()或signalALL()信号或者线程中断时该方法会返回false;其它情况返回true。
boolean await(long var1, TimeUnit var3) throws InterruptedException;
//适用条件与行为与awaitNanos(long nanosTimeout)完全一样,唯一不同点在于它不是等待指定时间,而是等待由参数指定的某一时刻。
boolean awaitUntil(Date var1) throws InterruptedException;
//唤醒一个在 await()等待队列中的线程。与Object.notify()相似
void signal();
//唤醒 await()等待队列中所有的线程。与object.notifyAll()相似
void signalAll();
}
使用await()
和signal()
方法,来进行一个简单的实例:
private Lock lockTry = new ReentrantLock();
private Condition condition = lockTry.newCondition();
public void signal(String name) {
lockTry.lock();
System.out.println(name+ "signal:"+ Thread.currentThread().getName() + ",Request Lock...");
try {
condition.signal();
System.out.println(name + "-->CurrThread: Wait for the Thread to wake up...");
} catch (Exception e) {
e.printStackTrace();
} finally {
lockTry.unlock();
System.out.println(Thread.currentThread().getName() + ",Release Lock...");
}
}
public void wait(String name) {
lockTry.lock();
System.out.println(name+ "wait:"+Thread.currentThread().getName() + ",Request Lock...");
try {
for (int i = 0; i < 5; i++) {
y++;
if (i == 2) {
System.out.println(name + "-->CurrThread:" + Thread.currentThread().getName() + " is waiting...");
condition.await();
}
System.out.println(name + "-->CurrThread:" + Thread.currentThread().getName() + " y=" + y);
}
} catch (Exception e) {
e.printStackTrace();
} finally {
lockTry.unlock();
System.out.println(Thread.currentThread().getName() + ",Release Lock...");
}
}
建立两个线程,分别调用这两个方法;注意线程在调用signal
方法时,让线程睡眠几秒,这样在输出结果是就能看出很明显的结果。
运行,结果为:
TempRunnable->Waitwait:Thread-0,Request Lock...
TempRunnable->Wait-->CurrThread:Thread-0 y=1
TempRunnable->Wait-->CurrThread:Thread-0 y=2
TempRunnable->Wait-->CurrThread:Thread-0 is waiting...
Count Instance 1:signal:Thread-1,Request Lock...
Count Instance 1:-->CurrThread: Wait for the Thread to wake up...
Thread-1,Release Lock...
TempRunnable->Wait-->CurrThread:Thread-0 y=3
TempRunnable->Wait-->CurrThread:Thread-0 y=4
TempRunnable->Wait-->CurrThread:Thread-0 y=5
Thread-0,Release Lock...
网友评论