Why use a ReentrantLock if one can use synchronized(this)?

Viewed 178923

I'm trying to understand what makes the lock in concurrency so important if one can use synchronized (this). In the dummy code below, I can do either:

  1. synchronized the entire method or synchronize the vulnerable area (synchronized(this){...})
  2. OR lock the vulnerable code area with a ReentrantLock.

Code:

    private final ReentrantLock lock = new ReentrantLock(); 
    private static List<Integer> ints;

    public Integer getResult(String name) { 
        .
        .
        .
        lock.lock();
        try {
            if (ints.size()==3) {
                ints=null;
                return -9;
            }   

            for (int x=0; x<ints.size(); x++) {
                System.out.println("["+name+"] "+x+"/"+ints.size()+". values >>>>"+ints.get(x));
            }

        } finally {
            lock.unlock();
        } 
        return random;
}
8 Answers

Synchronized locks does not offer any mechanism of waiting queue in which after the execution of one thread any thread running in parallel can acquire the lock. Due to which the thread which is there in the system and running for a longer period of time never gets chance to access the shared resource thus leading to starvation.

Reentrant locks are very much flexible and has a fairness policy in which if a thread is waiting for a longer time and after the completion of the currently executing thread we can make sure that the longer waiting thread gets the chance of accessing the shared resource hereby decreasing the throughput of the system and making it more time consuming.

One thing to keep in mind is :

The name 'ReentrantLock' gives out a wrong message about other locking mechanism that they are not re-entrant. This is not true. Lock acquired via 'synchronized' is also re-entrant in Java.

Key difference is that 'synchronized' uses intrinsic lock ( one that every Object has ) while Lock API doesn't.

I think the wait/notify/notifyAll methods don't belong on the Object class as it pollutes all objects with methods that are rarely used. They make much more sense on a dedicated Lock class. So from this point of view, perhaps it's better to use a tool that is explicitly designed for the job at hand - ie ReentrantLock.

Lets assume this code is running in a thread:

private static ReentrantLock lock = new ReentrantLock();

void accessResource() {
    lock.lock();
    if( checkSomeCondition() ) {
        accessResource();
    }
    lock.unlock();
}

Because the thread owns the lock it will allow multiple calls to lock(), so it re-enter the lock. This can be achieved with a reference count so it doesn't has to acquire lock again.

Related