Concurrent Semaphore release may cause some require thread not signaled

FULL PRODUCT VERSION :
java version "1.6.0_11"
Java(TM) SE Runtime Environment (build 1.6.0_11-b03)
Java HotSpot(TM) Client VM (build 11.0-b16, mixed mode, sharing)

ADDITIONAL OS VERSION INFORMATION :
Windows XP SP3

EXTRA RELEVANT SYSTEM CONFIGURATION :
P4 2.8HT

A DESCRIPTION OF THE PROBLEM :
Semaphore initial state 0. 4 threads run 4 tasks.
Two threads run acquire semaphore once, the other two threads run release semaphore once.
One possible result is the semaphore state value is 1 and one thread still waiting.

The possible reason:
When AbstractQueuedSynchronizer#release are called, head.waitStatus may be 0 because the previous acquire thread may run at AbstractQueuedSynchronizer#doAcquireShared before setHeadAndPropagate is called.

STEPS TO FOLLOW TO REPRODUCE THE PROBLEM :
The given executable test case will hung. Suggest using dual-core CPU or HT CPU to reproduce.


REPRODUCIBILITY :
This bug can be reproduced occasionally.

---------- BEGIN SOURCE ----------
import java.util.concurrent.Semaphore;

public class TestSemaphore {

    private static Semaphore sem = new Semaphore(0);

    private static class Thread1 extends Thread {
        @Override
        public void run() {
            sem.acquireUninterruptibly();
        }
    }

    private static class Thread2 extends Thread {
        @Override
        public void run() {
            sem.release();
        }
    }

    public static void main(String[] args) throws InterruptedException {
        for (int i = 0; i < 10000000; i++) {
            Thread t1 = new Thread1();
            Thread t2 = new Thread1();
            Thread t3 = new Thread2();
            Thread t4 = new Thread2();
            t1.start();
            t2.start();
            t3.start();
            t4.start();
            t1.join();
            t2.join();
            t3.join();
            t4.join();
            System.out.println(i);
        }
    }
}

---------- END SOURCE ----------