在java的锁机制中,公平和非公平的参考物是什么,个人而言觉得是相对产生的结果而立,简单的来说,如果一个线程组里,能保证每个线程都能拿到锁,那么这个锁就是公平锁。相反,如果保证不了每个线程都能拿到锁,也就是存在有线程饿死,那么这个锁就是非公平锁。本文围绕ReenTrantLock来讲。
实现原理
那如何能保证每个线程都能拿到锁呢,队列FIFO是一个完美的解决方案,也就是先进先出,java的ReenTrantLock也就是用队列实现的公平锁和非公平锁。
在公平的锁中,如果有另一个线程持有锁或者有其他线程在等待队列中等待这个所,那么新发出的请求的线程将被放入到队列中。而非公平锁上,只有当锁被某个线程持有时,新发出请求的线程才会被放入队列中(此时和公平锁是一样的)。所以,它们的差别在于非公平锁会有更多的机会去抢占锁。
公平锁:
/**
* Sync object for fair locks
*/
static final class FairSync extends Sync {
private static final long serialVersionUID = -3000897897090466540L;
final void lock() {
acquire(1);
}
/**
* Fair version of tryAcquire. Don't grant access unless
* recursive call or no waiters or is first.
*/
protected final boolean tryAcquire(int acquires) {
final Thread current = Thread.currentThread();
int c = getState();
if (c == 0) {
if (!hasQueuedPredecessors() &&
compareAndSetState(0, acquires)) {
setExclusiveOwnerThread(current);
return true;
}
}
else if (current == getExclusiveOwnerThread()) {
int nextc = c + acquires;
if (nextc < 0)
throw new Error("Maximum lock count exceeded");
setState(nextc);
return true;
}
return false;
}
}
其中 hasQueuedPredecessors() 方法实现如下:
/**
* 看看是否有其他想要获取锁的并且等待时间比当前线程长的线程
*/
public final boolean hasQueuedPredecessors() {
// The correctness of this depends on head being initialized
// before tail and on head.next being accurate if the current
// thread is first in queue.
Node t = tail; // Read fields in reverse initialization order
Node h = head;
Node s;
return h != t &&
((s = h.next) == null || s.thread != Thread.currentThread());
}
非公平锁:
/**
* Sync object for non-fair locks
*/
static final class NonfairSync extends Sync {
private static final long serialVersionUID = 7316153563782823691L;
/**
* Performs lock. Try immediate barge, backing up to normal
* acquire on failure.
*/
final void lock() {
if (compareAndSetState(0, 1))
setExclusiveOwnerThread(Thread.currentThread());
else
acquire(1);
}
protected final boolean tryAcquire(int acquires) {
return nonfairTryAcquire(acquires);
}
}
示例
公平锁:
package com.thread.fair;
import java.util.concurrent.locks.ReentrantLock;
/**
* Created by Fant.J.
*/
public class MyFairLock {
/**
* true 表示 ReentrantLock 的公平锁
*/
private ReentrantLock lock = new ReentrantLock(true);
public void testFail(){
try {
lock.lock();
System.out.println(Thread.currentThread().getName() +"获得了锁");
}finally {
lock.unlock();
}
}
public static void main(String[] args) {
MyFairLock fairLock = new MyFairLock();
Runnable runnable = () -> {
System.out.println(Thread.currentThread().getName()+"启动");
fairLock.testFail();
};
Thread[] threadArray = new Thread[10];
for (int i=0; i<10; i++) {
threadArray[i] = new Thread(runnable);
}
for (int i=0; i<10; i++) {
threadArray[i].start();
}
}
}
执行结果:
Thread-0启动
Thread-0获得了锁
Thread-1启动
Thread-1获得了锁
Thread-2启动
Thread-2获得了锁
Thread-3启动
Thread-3获得了锁
Thread-4启动
Thread-4获得了锁
Thread-5启动
Thread-5获得了锁
Thread-6启动
Thread-6获得了锁
Thread-8启动
Thread-8获得了锁
Thread-7启动
Thread-7获得了锁
Thread-9启动
Thread-9获得了锁
可以看到,获取锁的线程顺序正是线程启动的顺序。
非公平锁:
/**
* Created by Fant.J.
*/
public class MyNonfairLock {
/**
* false 表示 ReentrantLock 的非公平锁
*/
private ReentrantLock lock = new ReentrantLock(false);
public void testFail(){
try {
lock.lock();
System.out.println(Thread.currentThread().getName() +"获得了锁");
}finally {
lock.unlock();
}
}
public static void main(String[] args) {
MyNonfairLock nonfairLock = new MyNonfairLock();
Runnable runnable = () -> {
System.out.println(Thread.currentThread().getName()+"启动");
nonfairLock.testFail();
};
Thread[] threadArray = new Thread[10];
for (int i=0; i<10; i++) {
threadArray[i] = new Thread(runnable);
}
for (int i=0; i<10; i++) {
threadArray[i].start();
}
}
}
执行结果:
Thread-1启动
Thread-0启动
Thread-0获得了锁
Thread-1获得了锁
Thread-8启动
Thread-8获得了锁
Thread-3启动
Thread-3获得了锁
Thread-4启动
Thread-4获得了锁
Thread-5启动
Thread-2启动
Thread-9启动
Thread-5获得了锁
Thread-2获得了锁
Thread-9获得了锁
Thread-6启动
Thread-7启动
Thread-6获得了锁
Thread-7获得了锁
可以看出非公平锁对锁的获取是乱序的,即有一个抢占锁的过程。
最后
那非公平锁和公平锁适合什么场合使用呢,他们的优缺点又是什么呢?
优缺点:
非公平锁性能高于公平锁性能。首先,在恢复一个被挂起的线程与该线程真正运行之间存在着严重的延迟。而且,非公平锁能更充分的利用cpu的时间片,尽量的减少cpu空闲的状态时间。
使用场景:
使用场景的话呢,其实还是和他们的属性一一相关,举个栗子:如果业务中线程占用(处理)时间要远长于线程等待,那用非公平锁其实效率并不明显,但是用公平锁会给业务增强很多的可控制性。