Sync java.util.concurrent library up to 12.06.2013.

CouncurrentHashMap was skipped from this sync.

Change-Id: I29c67698a2e706b22e3cb5920c5fe91f1f15461c
diff --git a/luni/src/main/java/java/util/concurrent/AbstractExecutorService.java b/luni/src/main/java/java/util/concurrent/AbstractExecutorService.java
index a7f7745..23e68bb 100644
--- a/luni/src/main/java/java/util/concurrent/AbstractExecutorService.java
+++ b/luni/src/main/java/java/util/concurrent/AbstractExecutorService.java
@@ -9,19 +9,19 @@
 
 /**
  * Provides default implementations of {@link ExecutorService}
- * execution methods. This class implements the <tt>submit</tt>,
- * <tt>invokeAny</tt> and <tt>invokeAll</tt> methods using a
- * {@link RunnableFuture} returned by <tt>newTaskFor</tt>, which defaults
+ * execution methods. This class implements the {@code submit},
+ * {@code invokeAny} and {@code invokeAll} methods using a
+ * {@link RunnableFuture} returned by {@code newTaskFor}, which defaults
  * to the {@link FutureTask} class provided in this package.  For example,
- * the implementation of <tt>submit(Runnable)</tt> creates an
- * associated <tt>RunnableFuture</tt> that is executed and
- * returned. Subclasses may override the <tt>newTaskFor</tt> methods
- * to return <tt>RunnableFuture</tt> implementations other than
- * <tt>FutureTask</tt>.
+ * the implementation of {@code submit(Runnable)} creates an
+ * associated {@code RunnableFuture} that is executed and
+ * returned. Subclasses may override the {@code newTaskFor} methods
+ * to return {@code RunnableFuture} implementations other than
+ * {@code FutureTask}.
  *
- * <p> <b>Extension example</b>. Here is a sketch of a class
+ * <p><b>Extension example</b>. Here is a sketch of a class
  * that customizes {@link ThreadPoolExecutor} to use
- * a <tt>CustomTask</tt> class instead of the default <tt>FutureTask</tt>:
+ * a {@code CustomTask} class instead of the default {@code FutureTask}:
  *  <pre> {@code
  * public class CustomThreadPoolExecutor extends ThreadPoolExecutor {
  *
@@ -42,15 +42,15 @@
 public abstract class AbstractExecutorService implements ExecutorService {
 
     /**
-     * Returns a <tt>RunnableFuture</tt> for the given runnable and default
+     * Returns a {@code RunnableFuture} for the given runnable and default
      * value.
      *
      * @param runnable the runnable task being wrapped
      * @param value the default value for the returned future
-     * @return a <tt>RunnableFuture</tt> which when run will run the
-     * underlying runnable and which, as a <tt>Future</tt>, will yield
+     * @return a {@code RunnableFuture} which, when run, will run the
+     * underlying runnable and which, as a {@code Future}, will yield
      * the given value as its result and provide for cancellation of
-     * the underlying task.
+     * the underlying task
      * @since 1.6
      */
     protected <T> RunnableFuture<T> newTaskFor(Runnable runnable, T value) {
@@ -58,13 +58,13 @@
     }
 
     /**
-     * Returns a <tt>RunnableFuture</tt> for the given callable task.
+     * Returns a {@code RunnableFuture} for the given callable task.
      *
      * @param callable the callable task being wrapped
-     * @return a <tt>RunnableFuture</tt> which when run will call the
-     * underlying callable and which, as a <tt>Future</tt>, will yield
+     * @return a {@code RunnableFuture} which, when run, will call the
+     * underlying callable and which, as a {@code Future}, will yield
      * the callable's result as its result and provide for
-     * cancellation of the underlying task.
+     * cancellation of the underlying task
      * @since 1.6
      */
     protected <T> RunnableFuture<T> newTaskFor(Callable<T> callable) {
@@ -108,14 +108,14 @@
      * the main mechanics of invokeAny.
      */
     private <T> T doInvokeAny(Collection<? extends Callable<T>> tasks,
-                            boolean timed, long nanos)
+                              boolean timed, long nanos)
         throws InterruptedException, ExecutionException, TimeoutException {
         if (tasks == null)
             throw new NullPointerException();
         int ntasks = tasks.size();
         if (ntasks == 0)
             throw new IllegalArgumentException();
-        List<Future<T>> futures= new ArrayList<Future<T>>(ntasks);
+        ArrayList<Future<T>> futures = new ArrayList<Future<T>>(ntasks);
         ExecutorCompletionService<T> ecs =
             new ExecutorCompletionService<T>(this);
 
@@ -129,7 +129,7 @@
             // Record exceptions so that if we fail to obtain any
             // result, we can throw the last exception we got.
             ExecutionException ee = null;
-            long lastTime = timed ? System.nanoTime() : 0;
+            final long deadline = timed ? System.nanoTime() + nanos : 0L;
             Iterator<? extends Callable<T>> it = tasks.iterator();
 
             // Start one task for sure; the rest incrementally
@@ -151,9 +151,7 @@
                         f = ecs.poll(nanos, TimeUnit.NANOSECONDS);
                         if (f == null)
                             throw new TimeoutException();
-                        long now = System.nanoTime();
-                        nanos -= now - lastTime;
-                        lastTime = now;
+                        nanos = deadline - System.nanoTime();
                     }
                     else
                         f = ecs.take();
@@ -175,8 +173,8 @@
             throw ee;
 
         } finally {
-            for (Future<T> f : futures)
-                f.cancel(true);
+            for (int i = 0, size = futures.size(); i < size; i++)
+                futures.get(i).cancel(true);
         }
     }
 
@@ -200,7 +198,7 @@
         throws InterruptedException {
         if (tasks == null)
             throw new NullPointerException();
-        List<Future<T>> futures = new ArrayList<Future<T>>(tasks.size());
+        ArrayList<Future<T>> futures = new ArrayList<Future<T>>(tasks.size());
         boolean done = false;
         try {
             for (Callable<T> t : tasks) {
@@ -208,7 +206,8 @@
                 futures.add(f);
                 execute(f);
             }
-            for (Future<T> f : futures) {
+            for (int i = 0, size = futures.size(); i < size; i++) {
+                Future<T> f = futures.get(i);
                 if (!f.isDone()) {
                     try {
                         f.get();
@@ -221,40 +220,39 @@
             return futures;
         } finally {
             if (!done)
-                for (Future<T> f : futures)
-                    f.cancel(true);
+                for (int i = 0, size = futures.size(); i < size; i++)
+                    futures.get(i).cancel(true);
         }
     }
 
     public <T> List<Future<T>> invokeAll(Collection<? extends Callable<T>> tasks,
                                          long timeout, TimeUnit unit)
         throws InterruptedException {
-        if (tasks == null || unit == null)
+        if (tasks == null)
             throw new NullPointerException();
         long nanos = unit.toNanos(timeout);
-        List<Future<T>> futures = new ArrayList<Future<T>>(tasks.size());
+        ArrayList<Future<T>> futures = new ArrayList<Future<T>>(tasks.size());
         boolean done = false;
         try {
             for (Callable<T> t : tasks)
                 futures.add(newTaskFor(t));
 
-            long lastTime = System.nanoTime();
+            final long deadline = System.nanoTime() + nanos;
+            final int size = futures.size();
 
             // Interleave time checks and calls to execute in case
             // executor doesn't have any/much parallelism.
-            Iterator<Future<T>> it = futures.iterator();
-            while (it.hasNext()) {
-                execute((Runnable)(it.next()));
-                long now = System.nanoTime();
-                nanos -= now - lastTime;
-                lastTime = now;
-                if (nanos <= 0)
+            for (int i = 0; i < size; i++) {
+                execute((Runnable)futures.get(i));
+                nanos = deadline - System.nanoTime();
+                if (nanos <= 0L)
                     return futures;
             }
 
-            for (Future<T> f : futures) {
+            for (int i = 0; i < size; i++) {
+                Future<T> f = futures.get(i);
                 if (!f.isDone()) {
-                    if (nanos <= 0)
+                    if (nanos <= 0L)
                         return futures;
                     try {
                         f.get(nanos, TimeUnit.NANOSECONDS);
@@ -263,17 +261,15 @@
                     } catch (TimeoutException toe) {
                         return futures;
                     }
-                    long now = System.nanoTime();
-                    nanos -= now - lastTime;
-                    lastTime = now;
+                    nanos = deadline - System.nanoTime();
                 }
             }
             done = true;
             return futures;
         } finally {
             if (!done)
-                for (Future<T> f : futures)
-                    f.cancel(true);
+                for (int i = 0, size = futures.size(); i < size; i++)
+                    futures.get(i).cancel(true);
         }
     }
 
diff --git a/luni/src/main/java/java/util/concurrent/ArrayBlockingQueue.java b/luni/src/main/java/java/util/concurrent/ArrayBlockingQueue.java
index e30ab67..3cfe6d5 100644
--- a/luni/src/main/java/java/util/concurrent/ArrayBlockingQueue.java
+++ b/luni/src/main/java/java/util/concurrent/ArrayBlockingQueue.java
@@ -111,9 +111,9 @@
     /**
      * Returns item at index i.
      */
+    @SuppressWarnings("unchecked")
     final E itemAt(int i) {
-        @SuppressWarnings("unchecked") E x = (E) items[i];
-        return x;
+        return (E) items[i];
     }
 
     /**
@@ -147,7 +147,8 @@
         // assert lock.getHoldCount() == 1;
         // assert items[takeIndex] != null;
         final Object[] items = this.items;
-        @SuppressWarnings("unchecked") E x = (E) items[takeIndex];
+        @SuppressWarnings("unchecked")
+        E x = (E) items[takeIndex];
         items[takeIndex] = null;
         takeIndex = inc(takeIndex);
         count--;
@@ -396,7 +397,7 @@
         final ReentrantLock lock = this.lock;
         lock.lock();
         try {
-            return (count == 0) ? null : itemAt(takeIndex);
+            return itemAt(takeIndex); // null when queue is empty
         } finally {
             lock.unlock();
         }
@@ -529,8 +530,13 @@
         try {
             final int count = this.count;
             Object[] a = new Object[count];
-            for (int i = takeIndex, k = 0; k < count; i = inc(i), k++)
-                a[k] = items[i];
+            int n = items.length - takeIndex;
+            if (count <= n) {
+                System.arraycopy(items, takeIndex, a, 0, count);
+            } else {
+                System.arraycopy(items, takeIndex, a, 0, n);
+                System.arraycopy(items, 0, a, n, count - n);
+            }
             return a;
         } finally {
             lock.unlock();
@@ -583,8 +589,13 @@
             if (len < count)
                 a = (T[])java.lang.reflect.Array.newInstance(
                     a.getClass().getComponentType(), count);
-            for (int i = takeIndex, k = 0; k < count; i = inc(i), k++)
-                a[k] = (T) items[i];
+            int n = items.length - takeIndex;
+            if (count <= n)
+                System.arraycopy(items, takeIndex, a, 0, count);
+            else {
+                System.arraycopy(items, takeIndex, a, 0, n);
+                System.arraycopy(items, 0, a, n, count - n);
+            }
             if (len > count)
                 a[count] = null;
             return a;
@@ -674,7 +685,8 @@
             int i = 0;
             try {
                 while (i < n) {
-                    @SuppressWarnings("unchecked") E x = (E) items[take];
+                    @SuppressWarnings("unchecked")
+                    E x = (E) items[take];
                     c.add(x);
                     items[take] = null;
                     take = inc(take);
diff --git a/luni/src/main/java/java/util/concurrent/BlockingDeque.java b/luni/src/main/java/java/util/concurrent/BlockingDeque.java
index 34f103d..8a393ba 100644
--- a/luni/src/main/java/java/util/concurrent/BlockingDeque.java
+++ b/luni/src/main/java/java/util/concurrent/BlockingDeque.java
@@ -12,11 +12,11 @@
  * for the deque to become non-empty when retrieving an element, and wait for
  * space to become available in the deque when storing an element.
  *
- * <p><tt>BlockingDeque</tt> methods come in four forms, with different ways
+ * <p>{@code BlockingDeque} methods come in four forms, with different ways
  * of handling operations that cannot be satisfied immediately, but may be
  * satisfied at some point in the future:
  * one throws an exception, the second returns a special value (either
- * <tt>null</tt> or <tt>false</tt>, depending on the operation), the third
+ * {@code null} or {@code false}, depending on the operation), the third
  * blocks the current thread indefinitely until the operation can succeed,
  * and the fourth blocks for only a given maximum time limit before giving
  * up.  These methods are summarized in the following table:
@@ -87,20 +87,20 @@
  *  </tr>
  * </table>
  *
- * <p>Like any {@link BlockingQueue}, a <tt>BlockingDeque</tt> is thread safe,
+ * <p>Like any {@link BlockingQueue}, a {@code BlockingDeque} is thread safe,
  * does not permit null elements, and may (or may not) be
  * capacity-constrained.
  *
- * <p>A <tt>BlockingDeque</tt> implementation may be used directly as a FIFO
- * <tt>BlockingQueue</tt>. The methods inherited from the
- * <tt>BlockingQueue</tt> interface are precisely equivalent to
- * <tt>BlockingDeque</tt> methods as indicated in the following table:
+ * <p>A {@code BlockingDeque} implementation may be used directly as a FIFO
+ * {@code BlockingQueue}. The methods inherited from the
+ * {@code BlockingQueue} interface are precisely equivalent to
+ * {@code BlockingDeque} methods as indicated in the following table:
  *
  * <p>
  * <table BORDER CELLPADDING=3 CELLSPACING=1>
  *  <tr>
- *    <td ALIGN=CENTER> <b><tt>BlockingQueue</tt> Method</b></td>
- *    <td ALIGN=CENTER> <b>Equivalent <tt>BlockingDeque</tt> Method</b></td>
+ *    <td ALIGN=CENTER> <b>{@code BlockingQueue} Method</b></td>
+ *    <td ALIGN=CENTER> <b>Equivalent {@code BlockingDeque} Method</b></td>
  *  </tr>
  *  <tr>
  *    <td ALIGN=CENTER COLSPAN = 2> <b>Insert</b></td>
@@ -179,7 +179,7 @@
     /**
      * Inserts the specified element at the front of this deque if it is
      * possible to do so immediately without violating capacity restrictions,
-     * throwing an <tt>IllegalStateException</tt> if no space is currently
+     * throwing an {@code IllegalStateException} if no space is currently
      * available.  When using a capacity-restricted deque, it is generally
      * preferable to use {@link #offerFirst(Object) offerFirst}.
      *
@@ -194,7 +194,7 @@
     /**
      * Inserts the specified element at the end of this deque if it is
      * possible to do so immediately without violating capacity restrictions,
-     * throwing an <tt>IllegalStateException</tt> if no space is currently
+     * throwing an {@code IllegalStateException} if no space is currently
      * available.  When using a capacity-restricted deque, it is generally
      * preferable to use {@link #offerLast(Object) offerLast}.
      *
@@ -209,7 +209,7 @@
     /**
      * Inserts the specified element at the front of this deque if it is
      * possible to do so immediately without violating capacity restrictions,
-     * returning <tt>true</tt> upon success and <tt>false</tt> if no space is
+     * returning {@code true} upon success and {@code false} if no space is
      * currently available.
      * When using a capacity-restricted deque, this method is generally
      * preferable to the {@link #addFirst(Object) addFirst} method, which can
@@ -225,7 +225,7 @@
     /**
      * Inserts the specified element at the end of this deque if it is
      * possible to do so immediately without violating capacity restrictions,
-     * returning <tt>true</tt> upon success and <tt>false</tt> if no space is
+     * returning {@code true} upon success and {@code false} if no space is
      * currently available.
      * When using a capacity-restricted deque, this method is generally
      * preferable to the {@link #addLast(Object) addLast} method, which can
@@ -273,10 +273,10 @@
      *
      * @param e the element to add
      * @param timeout how long to wait before giving up, in units of
-     *        <tt>unit</tt>
-     * @param unit a <tt>TimeUnit</tt> determining how to interpret the
-     *        <tt>timeout</tt> parameter
-     * @return <tt>true</tt> if successful, or <tt>false</tt> if
+     *        {@code unit}
+     * @param unit a {@code TimeUnit} determining how to interpret the
+     *        {@code timeout} parameter
+     * @return {@code true} if successful, or {@code false} if
      *         the specified waiting time elapses before space is available
      * @throws InterruptedException if interrupted while waiting
      * @throws ClassCastException if the class of the specified element
@@ -295,10 +295,10 @@
      *
      * @param e the element to add
      * @param timeout how long to wait before giving up, in units of
-     *        <tt>unit</tt>
-     * @param unit a <tt>TimeUnit</tt> determining how to interpret the
-     *        <tt>timeout</tt> parameter
-     * @return <tt>true</tt> if successful, or <tt>false</tt> if
+     *        {@code unit}
+     * @param unit a {@code TimeUnit} determining how to interpret the
+     *        {@code timeout} parameter
+     * @return {@code true} if successful, or {@code false} if
      *         the specified waiting time elapses before space is available
      * @throws InterruptedException if interrupted while waiting
      * @throws ClassCastException if the class of the specified element
@@ -334,10 +334,10 @@
      * become available.
      *
      * @param timeout how long to wait before giving up, in units of
-     *        <tt>unit</tt>
-     * @param unit a <tt>TimeUnit</tt> determining how to interpret the
-     *        <tt>timeout</tt> parameter
-     * @return the head of this deque, or <tt>null</tt> if the specified
+     *        {@code unit}
+     * @param unit a {@code TimeUnit} determining how to interpret the
+     *        {@code timeout} parameter
+     * @return the head of this deque, or {@code null} if the specified
      *         waiting time elapses before an element is available
      * @throws InterruptedException if interrupted while waiting
      */
@@ -350,10 +350,10 @@
      * become available.
      *
      * @param timeout how long to wait before giving up, in units of
-     *        <tt>unit</tt>
-     * @param unit a <tt>TimeUnit</tt> determining how to interpret the
-     *        <tt>timeout</tt> parameter
-     * @return the tail of this deque, or <tt>null</tt> if the specified
+     *        {@code unit}
+     * @param unit a {@code TimeUnit} determining how to interpret the
+     *        {@code timeout} parameter
+     * @return the tail of this deque, or {@code null} if the specified
      *         waiting time elapses before an element is available
      * @throws InterruptedException if interrupted while waiting
      */
@@ -363,13 +363,13 @@
     /**
      * Removes the first occurrence of the specified element from this deque.
      * If the deque does not contain the element, it is unchanged.
-     * More formally, removes the first element <tt>e</tt> such that
-     * <tt>o.equals(e)</tt> (if such an element exists).
-     * Returns <tt>true</tt> if this deque contained the specified element
+     * More formally, removes the first element {@code e} such that
+     * {@code o.equals(e)} (if such an element exists).
+     * Returns {@code true} if this deque contained the specified element
      * (or equivalently, if this deque changed as a result of the call).
      *
      * @param o element to be removed from this deque, if present
-     * @return <tt>true</tt> if an element was removed as a result of this call
+     * @return {@code true} if an element was removed as a result of this call
      * @throws ClassCastException if the class of the specified element
      *         is incompatible with this deque
      *         (<a href="../Collection.html#optional-restrictions">optional</a>)
@@ -381,13 +381,13 @@
     /**
      * Removes the last occurrence of the specified element from this deque.
      * If the deque does not contain the element, it is unchanged.
-     * More formally, removes the last element <tt>e</tt> such that
-     * <tt>o.equals(e)</tt> (if such an element exists).
-     * Returns <tt>true</tt> if this deque contained the specified element
+     * More formally, removes the last element {@code e} such that
+     * {@code o.equals(e)} (if such an element exists).
+     * Returns {@code true} if this deque contained the specified element
      * (or equivalently, if this deque changed as a result of the call).
      *
      * @param o element to be removed from this deque, if present
-     * @return <tt>true</tt> if an element was removed as a result of this call
+     * @return {@code true} if an element was removed as a result of this call
      * @throws ClassCastException if the class of the specified element
      *         is incompatible with this deque
      *         (<a href="../Collection.html#optional-restrictions">optional</a>)
@@ -402,8 +402,8 @@
      * Inserts the specified element into the queue represented by this deque
      * (in other words, at the tail of this deque) if it is possible to do so
      * immediately without violating capacity restrictions, returning
-     * <tt>true</tt> upon success and throwing an
-     * <tt>IllegalStateException</tt> if no space is currently available.
+     * {@code true} upon success and throwing an
+     * {@code IllegalStateException} if no space is currently available.
      * When using a capacity-restricted deque, it is generally preferable to
      * use {@link #offer(Object) offer}.
      *
@@ -423,7 +423,7 @@
      * Inserts the specified element into the queue represented by this deque
      * (in other words, at the tail of this deque) if it is possible to do so
      * immediately without violating capacity restrictions, returning
-     * <tt>true</tt> upon success and <tt>false</tt> if no space is currently
+     * {@code true} upon success and {@code false} if no space is currently
      * available.  When using a capacity-restricted deque, this method is
      * generally preferable to the {@link #add} method, which can fail to
      * insert an element only by throwing an exception.
@@ -465,8 +465,8 @@
      * {@link #offerLast(Object,long,TimeUnit) offerLast}.
      *
      * @param e the element to add
-     * @return <tt>true</tt> if the element was added to this deque, else
-     *         <tt>false</tt>
+     * @return {@code true} if the element was added to this deque, else
+     *         {@code false}
      * @throws InterruptedException {@inheritDoc}
      * @throws ClassCastException if the class of the specified element
      *         prevents it from being added to this deque
@@ -493,11 +493,11 @@
     /**
      * Retrieves and removes the head of the queue represented by this deque
      * (in other words, the first element of this deque), or returns
-     * <tt>null</tt> if this deque is empty.
+     * {@code null} if this deque is empty.
      *
      * <p>This method is equivalent to {@link #pollFirst()}.
      *
-     * @return the head of this deque, or <tt>null</tt> if this deque is empty
+     * @return the head of this deque, or {@code null} if this deque is empty
      */
     E poll();
 
@@ -521,7 +521,7 @@
      * <p>This method is equivalent to
      * {@link #pollFirst(long,TimeUnit) pollFirst}.
      *
-     * @return the head of this deque, or <tt>null</tt> if the
+     * @return the head of this deque, or {@code null} if the
      *         specified waiting time elapses before an element is available
      * @throws InterruptedException if interrupted while waiting
      */
@@ -544,27 +544,27 @@
     /**
      * Retrieves, but does not remove, the head of the queue represented by
      * this deque (in other words, the first element of this deque), or
-     * returns <tt>null</tt> if this deque is empty.
+     * returns {@code null} if this deque is empty.
      *
      * <p>This method is equivalent to {@link #peekFirst() peekFirst}.
      *
-     * @return the head of this deque, or <tt>null</tt> if this deque is empty
+     * @return the head of this deque, or {@code null} if this deque is empty
      */
     E peek();
 
     /**
      * Removes the first occurrence of the specified element from this deque.
      * If the deque does not contain the element, it is unchanged.
-     * More formally, removes the first element <tt>e</tt> such that
-     * <tt>o.equals(e)</tt> (if such an element exists).
-     * Returns <tt>true</tt> if this deque contained the specified element
+     * More formally, removes the first element {@code e} such that
+     * {@code o.equals(e)} (if such an element exists).
+     * Returns {@code true} if this deque contained the specified element
      * (or equivalently, if this deque changed as a result of the call).
      *
      * <p>This method is equivalent to
      * {@link #removeFirstOccurrence(Object) removeFirstOccurrence}.
      *
      * @param o element to be removed from this deque, if present
-     * @return <tt>true</tt> if this deque changed as a result of the call
+     * @return {@code true} if this deque changed as a result of the call
      * @throws ClassCastException if the class of the specified element
      *         is incompatible with this deque
      *         (<a href="../Collection.html#optional-restrictions">optional</a>)
@@ -574,12 +574,12 @@
     boolean remove(Object o);
 
     /**
-     * Returns <tt>true</tt> if this deque contains the specified element.
-     * More formally, returns <tt>true</tt> if and only if this deque contains
-     * at least one element <tt>e</tt> such that <tt>o.equals(e)</tt>.
+     * Returns {@code true} if this deque contains the specified element.
+     * More formally, returns {@code true} if and only if this deque contains
+     * at least one element {@code e} such that {@code o.equals(e)}.
      *
      * @param o object to be checked for containment in this deque
-     * @return <tt>true</tt> if this deque contains the specified element
+     * @return {@code true} if this deque contains the specified element
      * @throws ClassCastException if the class of the specified element
      *         is incompatible with this deque
      *         (<a href="../Collection.html#optional-restrictions">optional</a>)
diff --git a/luni/src/main/java/java/util/concurrent/BlockingQueue.java b/luni/src/main/java/java/util/concurrent/BlockingQueue.java
index 6cfe52b..cc6d541 100644
--- a/luni/src/main/java/java/util/concurrent/BlockingQueue.java
+++ b/luni/src/main/java/java/util/concurrent/BlockingQueue.java
@@ -19,11 +19,11 @@
  * element, and wait for space to become available in the queue when
  * storing an element.
  *
- * <p><tt>BlockingQueue</tt> methods come in four forms, with different ways
+ * <p>{@code BlockingQueue} methods come in four forms, with different ways
  * of handling operations that cannot be satisfied immediately, but may be
  * satisfied at some point in the future:
  * one throws an exception, the second returns a special value (either
- * <tt>null</tt> or <tt>false</tt>, depending on the operation), the third
+ * {@code null} or {@code false}, depending on the operation), the third
  * blocks the current thread indefinitely until the operation can succeed,
  * and the fourth blocks for only a given maximum time limit before giving
  * up.  These methods are summarized in the following table:
@@ -60,37 +60,37 @@
  *  </tr>
  * </table>
  *
- * <p>A <tt>BlockingQueue</tt> does not accept <tt>null</tt> elements.
- * Implementations throw <tt>NullPointerException</tt> on attempts
- * to <tt>add</tt>, <tt>put</tt> or <tt>offer</tt> a <tt>null</tt>.  A
- * <tt>null</tt> is used as a sentinel value to indicate failure of
- * <tt>poll</tt> operations.
+ * <p>A {@code BlockingQueue} does not accept {@code null} elements.
+ * Implementations throw {@code NullPointerException} on attempts
+ * to {@code add}, {@code put} or {@code offer} a {@code null}.  A
+ * {@code null} is used as a sentinel value to indicate failure of
+ * {@code poll} operations.
  *
- * <p>A <tt>BlockingQueue</tt> may be capacity bounded. At any given
- * time it may have a <tt>remainingCapacity</tt> beyond which no
- * additional elements can be <tt>put</tt> without blocking.
- * A <tt>BlockingQueue</tt> without any intrinsic capacity constraints always
- * reports a remaining capacity of <tt>Integer.MAX_VALUE</tt>.
+ * <p>A {@code BlockingQueue} may be capacity bounded. At any given
+ * time it may have a {@code remainingCapacity} beyond which no
+ * additional elements can be {@code put} without blocking.
+ * A {@code BlockingQueue} without any intrinsic capacity constraints always
+ * reports a remaining capacity of {@code Integer.MAX_VALUE}.
  *
- * <p> <tt>BlockingQueue</tt> implementations are designed to be used
+ * <p>{@code BlockingQueue} implementations are designed to be used
  * primarily for producer-consumer queues, but additionally support
  * the {@link java.util.Collection} interface.  So, for example, it is
  * possible to remove an arbitrary element from a queue using
- * <tt>remove(x)</tt>. However, such operations are in general
+ * {@code remove(x)}. However, such operations are in general
  * <em>not</em> performed very efficiently, and are intended for only
  * occasional use, such as when a queued message is cancelled.
  *
- * <p> <tt>BlockingQueue</tt> implementations are thread-safe.  All
+ * <p>{@code BlockingQueue} implementations are thread-safe.  All
  * queuing methods achieve their effects atomically using internal
  * locks or other forms of concurrency control. However, the
- * <em>bulk</em> Collection operations <tt>addAll</tt>,
- * <tt>containsAll</tt>, <tt>retainAll</tt> and <tt>removeAll</tt> are
+ * <em>bulk</em> Collection operations {@code addAll},
+ * {@code containsAll}, {@code retainAll} and {@code removeAll} are
  * <em>not</em> necessarily performed atomically unless specified
  * otherwise in an implementation. So it is possible, for example, for
- * <tt>addAll(c)</tt> to fail (throwing an exception) after adding
- * only some of the elements in <tt>c</tt>.
+ * {@code addAll(c)} to fail (throwing an exception) after adding
+ * only some of the elements in {@code c}.
  *
- * <p>A <tt>BlockingQueue</tt> does <em>not</em> intrinsically support
+ * <p>A {@code BlockingQueue} does <em>not</em> intrinsically support
  * any kind of &quot;close&quot; or &quot;shutdown&quot; operation to
  * indicate that no more items will be added.  The needs and usage of
  * such features tend to be implementation-dependent. For example, a
@@ -100,7 +100,7 @@
  *
  * <p>
  * Usage example, based on a typical producer-consumer scenario.
- * Note that a <tt>BlockingQueue</tt> can safely be used with multiple
+ * Note that a {@code BlockingQueue} can safely be used with multiple
  * producers and multiple consumers.
  *  <pre> {@code
  * class Producer implements Runnable {
@@ -152,13 +152,13 @@
     /**
      * Inserts the specified element into this queue if it is possible to do
      * so immediately without violating capacity restrictions, returning
-     * <tt>true</tt> upon success and throwing an
-     * <tt>IllegalStateException</tt> if no space is currently available.
+     * {@code true} upon success and throwing an
+     * {@code IllegalStateException} if no space is currently available.
      * When using a capacity-restricted queue, it is generally preferable to
      * use {@link #offer(Object) offer}.
      *
      * @param e the element to add
-     * @return <tt>true</tt> (as specified by {@link Collection#add})
+     * @return {@code true} (as specified by {@link Collection#add})
      * @throws IllegalStateException if the element cannot be added at this
      *         time due to capacity restrictions
      * @throws ClassCastException if the class of the specified element
@@ -172,14 +172,14 @@
     /**
      * Inserts the specified element into this queue if it is possible to do
      * so immediately without violating capacity restrictions, returning
-     * <tt>true</tt> upon success and <tt>false</tt> if no space is currently
+     * {@code true} upon success and {@code false} if no space is currently
      * available.  When using a capacity-restricted queue, this method is
      * generally preferable to {@link #add}, which can fail to insert an
      * element only by throwing an exception.
      *
      * @param e the element to add
-     * @return <tt>true</tt> if the element was added to this queue, else
-     *         <tt>false</tt>
+     * @return {@code true} if the element was added to this queue, else
+     *         {@code false}
      * @throws ClassCastException if the class of the specified element
      *         prevents it from being added to this queue
      * @throws NullPointerException if the specified element is null
@@ -208,10 +208,10 @@
      *
      * @param e the element to add
      * @param timeout how long to wait before giving up, in units of
-     *        <tt>unit</tt>
-     * @param unit a <tt>TimeUnit</tt> determining how to interpret the
-     *        <tt>timeout</tt> parameter
-     * @return <tt>true</tt> if successful, or <tt>false</tt> if
+     *        {@code unit}
+     * @param unit a {@code TimeUnit} determining how to interpret the
+     *        {@code timeout} parameter
+     * @return {@code true} if successful, or {@code false} if
      *         the specified waiting time elapses before space is available
      * @throws InterruptedException if interrupted while waiting
      * @throws ClassCastException if the class of the specified element
@@ -237,10 +237,10 @@
      * specified wait time if necessary for an element to become available.
      *
      * @param timeout how long to wait before giving up, in units of
-     *        <tt>unit</tt>
-     * @param unit a <tt>TimeUnit</tt> determining how to interpret the
-     *        <tt>timeout</tt> parameter
-     * @return the head of this queue, or <tt>null</tt> if the
+     *        {@code unit}
+     * @param unit a {@code TimeUnit} determining how to interpret the
+     *        {@code timeout} parameter
+     * @return the head of this queue, or {@code null} if the
      *         specified waiting time elapses before an element is available
      * @throws InterruptedException if interrupted while waiting
      */
@@ -250,11 +250,11 @@
     /**
      * Returns the number of additional elements that this queue can ideally
      * (in the absence of memory or resource constraints) accept without
-     * blocking, or <tt>Integer.MAX_VALUE</tt> if there is no intrinsic
+     * blocking, or {@code Integer.MAX_VALUE} if there is no intrinsic
      * limit.
      *
      * <p>Note that you <em>cannot</em> always tell if an attempt to insert
-     * an element will succeed by inspecting <tt>remainingCapacity</tt>
+     * an element will succeed by inspecting {@code remainingCapacity}
      * because it may be the case that another thread is about to
      * insert or remove an element.
      *
@@ -264,14 +264,14 @@
 
     /**
      * Removes a single instance of the specified element from this queue,
-     * if it is present.  More formally, removes an element <tt>e</tt> such
-     * that <tt>o.equals(e)</tt>, if this queue contains one or more such
+     * if it is present.  More formally, removes an element {@code e} such
+     * that {@code o.equals(e)}, if this queue contains one or more such
      * elements.
-     * Returns <tt>true</tt> if this queue contained the specified element
+     * Returns {@code true} if this queue contained the specified element
      * (or equivalently, if this queue changed as a result of the call).
      *
      * @param o element to be removed from this queue, if present
-     * @return <tt>true</tt> if this queue changed as a result of the call
+     * @return {@code true} if this queue changed as a result of the call
      * @throws ClassCastException if the class of the specified element
      *         is incompatible with this queue
      *         (<a href="../Collection.html#optional-restrictions">optional</a>)
@@ -281,12 +281,12 @@
     boolean remove(Object o);
 
     /**
-     * Returns <tt>true</tt> if this queue contains the specified element.
-     * More formally, returns <tt>true</tt> if and only if this queue contains
-     * at least one element <tt>e</tt> such that <tt>o.equals(e)</tt>.
+     * Returns {@code true} if this queue contains the specified element.
+     * More formally, returns {@code true} if and only if this queue contains
+     * at least one element {@code e} such that {@code o.equals(e)}.
      *
      * @param o object to be checked for containment in this queue
-     * @return <tt>true</tt> if this queue contains the specified element
+     * @return {@code true} if this queue contains the specified element
      * @throws ClassCastException if the class of the specified element
      *         is incompatible with this queue
      *         (<a href="../Collection.html#optional-restrictions">optional</a>)
@@ -300,10 +300,10 @@
      * to the given collection.  This operation may be more
      * efficient than repeatedly polling this queue.  A failure
      * encountered while attempting to add elements to
-     * collection <tt>c</tt> may result in elements being in neither,
+     * collection {@code c} may result in elements being in neither,
      * either or both collections when the associated exception is
      * thrown.  Attempts to drain a queue to itself result in
-     * <tt>IllegalArgumentException</tt>. Further, the behavior of
+     * {@code IllegalArgumentException}. Further, the behavior of
      * this operation is undefined if the specified collection is
      * modified while the operation is in progress.
      *
@@ -324,10 +324,10 @@
      * Removes at most the given number of available elements from
      * this queue and adds them to the given collection.  A failure
      * encountered while attempting to add elements to
-     * collection <tt>c</tt> may result in elements being in neither,
+     * collection {@code c} may result in elements being in neither,
      * either or both collections when the associated exception is
      * thrown.  Attempts to drain a queue to itself result in
-     * <tt>IllegalArgumentException</tt>. Further, the behavior of
+     * {@code IllegalArgumentException}. Further, the behavior of
      * this operation is undefined if the specified collection is
      * modified while the operation is in progress.
      *
diff --git a/luni/src/main/java/java/util/concurrent/BrokenBarrierException.java b/luni/src/main/java/java/util/concurrent/BrokenBarrierException.java
index 76fae13..9fe707d 100644
--- a/luni/src/main/java/java/util/concurrent/BrokenBarrierException.java
+++ b/luni/src/main/java/java/util/concurrent/BrokenBarrierException.java
@@ -15,19 +15,18 @@
  *
  * @since 1.5
  * @author Doug Lea
- *
  */
 public class BrokenBarrierException extends Exception {
     private static final long serialVersionUID = 7117394618823254244L;
 
     /**
-     * Constructs a <tt>BrokenBarrierException</tt> with no specified detail
+     * Constructs a {@code BrokenBarrierException} with no specified detail
      * message.
      */
     public BrokenBarrierException() {}
 
     /**
-     * Constructs a <tt>BrokenBarrierException</tt> with the specified
+     * Constructs a {@code BrokenBarrierException} with the specified
      * detail message.
      *
      * @param message the detail message
diff --git a/luni/src/main/java/java/util/concurrent/Callable.java b/luni/src/main/java/java/util/concurrent/Callable.java
index 293544b..a3b3883 100644
--- a/luni/src/main/java/java/util/concurrent/Callable.java
+++ b/luni/src/main/java/java/util/concurrent/Callable.java
@@ -9,21 +9,21 @@
 /**
  * A task that returns a result and may throw an exception.
  * Implementors define a single method with no arguments called
- * <tt>call</tt>.
+ * {@code call}.
  *
- * <p>The <tt>Callable</tt> interface is similar to {@link
+ * <p>The {@code Callable} interface is similar to {@link
  * java.lang.Runnable}, in that both are designed for classes whose
  * instances are potentially executed by another thread.  A
- * <tt>Runnable</tt>, however, does not return a result and cannot
+ * {@code Runnable}, however, does not return a result and cannot
  * throw a checked exception.
  *
- * <p> The {@link Executors} class contains utility methods to
- * convert from other common forms to <tt>Callable</tt> classes.
+ * <p>The {@link Executors} class contains utility methods to
+ * convert from other common forms to {@code Callable} classes.
  *
  * @see Executor
  * @since 1.5
  * @author Doug Lea
- * @param <V> the result type of method <tt>call</tt>
+ * @param <V> the result type of method {@code call}
  */
 public interface Callable<V> {
     /**
diff --git a/luni/src/main/java/java/util/concurrent/CancellationException.java b/luni/src/main/java/java/util/concurrent/CancellationException.java
index dc452e4..25ab271 100644
--- a/luni/src/main/java/java/util/concurrent/CancellationException.java
+++ b/luni/src/main/java/java/util/concurrent/CancellationException.java
@@ -18,12 +18,12 @@
     private static final long serialVersionUID = -9202173006928992231L;
 
     /**
-     * Constructs a <tt>CancellationException</tt> with no detail message.
+     * Constructs a {@code CancellationException} with no detail message.
      */
     public CancellationException() {}
 
     /**
-     * Constructs a <tt>CancellationException</tt> with the specified detail
+     * Constructs a {@code CancellationException} with the specified detail
      * message.
      *
      * @param message the detail message
diff --git a/luni/src/main/java/java/util/concurrent/CompletionService.java b/luni/src/main/java/java/util/concurrent/CompletionService.java
index 7b0931c..0607596 100644
--- a/luni/src/main/java/java/util/concurrent/CompletionService.java
+++ b/luni/src/main/java/java/util/concurrent/CompletionService.java
@@ -9,17 +9,17 @@
 /**
  * A service that decouples the production of new asynchronous tasks
  * from the consumption of the results of completed tasks.  Producers
- * <tt>submit</tt> tasks for execution. Consumers <tt>take</tt>
+ * {@code submit} tasks for execution. Consumers {@code take}
  * completed tasks and process their results in the order they
- * complete.  A <tt>CompletionService</tt> can for example be used to
- * manage asynchronous IO, in which tasks that perform reads are
+ * complete.  A {@code CompletionService} can for example be used to
+ * manage asynchronous I/O, in which tasks that perform reads are
  * submitted in one part of a program or system, and then acted upon
  * in a different part of the program when the reads complete,
  * possibly in a different order than they were requested.
  *
- * <p>Typically, a <tt>CompletionService</tt> relies on a separate
+ * <p>Typically, a {@code CompletionService} relies on a separate
  * {@link Executor} to actually execute the tasks, in which case the
- * <tt>CompletionService</tt> only manages an internal completion
+ * {@code CompletionService} only manages an internal completion
  * queue. The {@link ExecutorCompletionService} class provides an
  * implementation of this approach.
  *
@@ -28,7 +28,6 @@
  * <a href="package-summary.html#MemoryVisibility"><i>happen-before</i></a>
  * actions taken by that task, which in turn <i>happen-before</i>
  * actions following a successful return from the corresponding {@code take()}.
- *
  */
 public interface CompletionService<V> {
     /**
@@ -52,7 +51,7 @@
      * @param task the task to submit
      * @param result the result to return upon successful completion
      * @return a Future representing pending completion of the task,
-     *         and whose <tt>get()</tt> method will return the given
+     *         and whose {@code get()} method will return the given
      *         result value upon completion
      * @throws RejectedExecutionException if the task cannot be
      *         scheduled for execution
@@ -69,13 +68,12 @@
      */
     Future<V> take() throws InterruptedException;
 
-
     /**
      * Retrieves and removes the Future representing the next
-     * completed task or <tt>null</tt> if none are present.
+     * completed task, or {@code null} if none are present.
      *
      * @return the Future representing the next completed task, or
-     *         <tt>null</tt> if none are present
+     *         {@code null} if none are present
      */
     Future<V> poll();
 
@@ -85,11 +83,11 @@
      * time if none are yet present.
      *
      * @param timeout how long to wait before giving up, in units of
-     *        <tt>unit</tt>
-     * @param unit a <tt>TimeUnit</tt> determining how to interpret the
-     *        <tt>timeout</tt> parameter
+     *        {@code unit}
+     * @param unit a {@code TimeUnit} determining how to interpret the
+     *        {@code timeout} parameter
      * @return the Future representing the next completed task or
-     *         <tt>null</tt> if the specified waiting time elapses
+     *         {@code null} if the specified waiting time elapses
      *         before one is present
      * @throws InterruptedException if interrupted while waiting
      */
diff --git a/luni/src/main/java/java/util/concurrent/ConcurrentLinkedDeque.java b/luni/src/main/java/java/util/concurrent/ConcurrentLinkedDeque.java
index 30adb36..54b53ae 100644
--- a/luni/src/main/java/java/util/concurrent/ConcurrentLinkedDeque.java
+++ b/luni/src/main/java/java/util/concurrent/ConcurrentLinkedDeque.java
@@ -63,7 +63,6 @@
  * @author Martin Buchholz
  * @param <E> the type of elements held in this collection
  */
-
 public class ConcurrentLinkedDeque<E>
     extends AbstractCollection<E>
     implements Deque<E>, java.io.Serializable {
@@ -790,7 +789,7 @@
      * Creates an array list and fills it with elements of this list.
      * Used by toArray.
      *
-     * @return the arrayList
+     * @return the array list
      */
     private ArrayList<E> toArrayList() {
         ArrayList<E> list = new ArrayList<E>();
@@ -1360,11 +1359,10 @@
     }
 
     /**
-     * Saves the state to a stream (that is, serializes it).
+     * Saves this deque to a stream (that is, serializes it).
      *
      * @serialData All of the elements (each an {@code E}) in
      * the proper order, followed by a null
-     * @param s the stream
      */
     private void writeObject(java.io.ObjectOutputStream s)
         throws java.io.IOException {
@@ -1384,8 +1382,7 @@
     }
 
     /**
-     * Reconstitutes the instance from a stream (that is, deserializes it).
-     * @param s the stream
+     * Reconstitutes this deque from a stream (that is, deserializes it).
      */
     private void readObject(java.io.ObjectInputStream s)
         throws java.io.IOException, ClassNotFoundException {
@@ -1408,7 +1405,6 @@
         initHeadTail(h, t);
     }
 
-
     private boolean casHead(Node<E> cmp, Node<E> val) {
         return UNSAFE.compareAndSwapObject(this, headOffset, cmp, val);
     }
diff --git a/luni/src/main/java/java/util/concurrent/ConcurrentLinkedQueue.java b/luni/src/main/java/java/util/concurrent/ConcurrentLinkedQueue.java
index a0a26fd..873f825 100644
--- a/luni/src/main/java/java/util/concurrent/ConcurrentLinkedQueue.java
+++ b/luni/src/main/java/java/util/concurrent/ConcurrentLinkedQueue.java
@@ -69,7 +69,6 @@
  * @since 1.5
  * @author Doug Lea
  * @param <E> the type of elements held in this collection
- *
  */
 public class ConcurrentLinkedQueue<E> extends AbstractQueue<E>
         implements Queue<E>, java.io.Serializable {
@@ -218,7 +217,6 @@
      */
     private transient volatile Node<E> tail;
 
-
     /**
      * Creates a {@code ConcurrentLinkedQueue} that is initially empty.
      */
@@ -268,7 +266,7 @@
     }
 
     /**
-     * Try to CAS head to p. If successful, repoint old head to itself
+     * Tries to CAS head to p. If successful, repoint old head to itself
      * as sentinel for succ(), below.
      */
     final void updateHead(Node<E> h, Node<E> p) {
@@ -717,11 +715,10 @@
     }
 
     /**
-     * Saves the state to a stream (that is, serializes it).
+     * Saves this queue to a stream (that is, serializes it).
      *
      * @serialData All of the elements (each an {@code E}) in
      * the proper order, followed by a null
-     * @param s the stream
      */
     private void writeObject(java.io.ObjectOutputStream s)
         throws java.io.IOException {
@@ -741,8 +738,7 @@
     }
 
     /**
-     * Reconstitutes the instance from a stream (that is, deserializes it).
-     * @param s the stream
+     * Reconstitutes this queue from a stream (that is, deserializes it).
      */
     private void readObject(java.io.ObjectInputStream s)
         throws java.io.IOException, ClassNotFoundException {
diff --git a/luni/src/main/java/java/util/concurrent/ConcurrentMap.java b/luni/src/main/java/java/util/concurrent/ConcurrentMap.java
index 3405acf..27feeb2 100644
--- a/luni/src/main/java/java/util/concurrent/ConcurrentMap.java
+++ b/luni/src/main/java/java/util/concurrent/ConcurrentMap.java
@@ -13,7 +13,7 @@
 
 /**
  * A {@link java.util.Map} providing additional atomic
- * <tt>putIfAbsent</tt>, <tt>remove</tt>, and <tt>replace</tt> methods.
+ * {@code putIfAbsent}, {@code remove}, and {@code replace} methods.
  *
  * <p>Memory consistency effects: As with other concurrent
  * collections, actions in a thread prior to placing an object into a
@@ -27,7 +27,7 @@
  * @param <K> the type of keys maintained by this map
  * @param <V> the type of mapped values
  */
-public interface ConcurrentMap<K, V> extends Map<K, V> {
+public interface ConcurrentMap<K,V> extends Map<K,V> {
     /**
      * If the specified key is not already associated
      * with a value, associate it with the given value.
@@ -43,11 +43,11 @@
      * @param key key with which the specified value is to be associated
      * @param value value to be associated with the specified key
      * @return the previous value associated with the specified key, or
-     *         <tt>null</tt> if there was no mapping for the key.
-     *         (A <tt>null</tt> return can also indicate that the map
-     *         previously associated <tt>null</tt> with the key,
+     *         {@code null} if there was no mapping for the key.
+     *         (A {@code null} return can also indicate that the map
+     *         previously associated {@code null} with the key,
      *         if the implementation supports null values.)
-     * @throws UnsupportedOperationException if the <tt>put</tt> operation
+     * @throws UnsupportedOperationException if the {@code put} operation
      *         is not supported by this map
      * @throws ClassCastException if the class of the specified key or value
      *         prevents it from being stored in this map
@@ -55,7 +55,6 @@
      *         and this map does not permit null keys or values
      * @throws IllegalArgumentException if some property of the specified key
      *         or value prevents it from being stored in this map
-     *
      */
     V putIfAbsent(K key, V value);
 
@@ -73,8 +72,8 @@
      *
      * @param key key with which the specified value is associated
      * @param value value expected to be associated with the specified key
-     * @return <tt>true</tt> if the value was removed
-     * @throws UnsupportedOperationException if the <tt>remove</tt> operation
+     * @return {@code true} if the value was removed
+     * @throws UnsupportedOperationException if the {@code remove} operation
      *         is not supported by this map
      * @throws ClassCastException if the key or value is of an inappropriate
      *         type for this map
@@ -100,8 +99,8 @@
      * @param key key with which the specified value is associated
      * @param oldValue value expected to be associated with the specified key
      * @param newValue value to be associated with the specified key
-     * @return <tt>true</tt> if the value was replaced
-     * @throws UnsupportedOperationException if the <tt>put</tt> operation
+     * @return {@code true} if the value was replaced
+     * @throws UnsupportedOperationException if the {@code put} operation
      *         is not supported by this map
      * @throws ClassCastException if the class of a specified key or value
      *         prevents it from being stored in this map
@@ -126,11 +125,11 @@
      * @param key key with which the specified value is associated
      * @param value value to be associated with the specified key
      * @return the previous value associated with the specified key, or
-     *         <tt>null</tt> if there was no mapping for the key.
-     *         (A <tt>null</tt> return can also indicate that the map
-     *         previously associated <tt>null</tt> with the key,
+     *         {@code null} if there was no mapping for the key.
+     *         (A {@code null} return can also indicate that the map
+     *         previously associated {@code null} with the key,
      *         if the implementation supports null values.)
-     * @throws UnsupportedOperationException if the <tt>put</tt> operation
+     * @throws UnsupportedOperationException if the {@code put} operation
      *         is not supported by this map
      * @throws ClassCastException if the class of the specified key or value
      *         prevents it from being stored in this map
diff --git a/luni/src/main/java/java/util/concurrent/ConcurrentNavigableMap.java b/luni/src/main/java/java/util/concurrent/ConcurrentNavigableMap.java
index ea99886..e87fbee 100644
--- a/luni/src/main/java/java/util/concurrent/ConcurrentNavigableMap.java
+++ b/luni/src/main/java/java/util/concurrent/ConcurrentNavigableMap.java
@@ -38,7 +38,6 @@
      */
     ConcurrentNavigableMap<K,V> headMap(K toKey, boolean inclusive);
 
-
     /**
      * @throws ClassCastException       {@inheritDoc}
      * @throws NullPointerException     {@inheritDoc}
@@ -73,7 +72,7 @@
      * reflected in the descending map, and vice-versa.
      *
      * <p>The returned map has an ordering equivalent to
-     * <tt>{@link Collections#reverseOrder(Comparator) Collections.reverseOrder}(comparator())</tt>.
+     * {@link Collections#reverseOrder(Comparator) Collections.reverseOrder}{@code (comparator())}.
      * The expression {@code m.descendingMap().descendingMap()} returns a
      * view of {@code m} essentially equivalent to {@code m}.
      *
diff --git a/luni/src/main/java/java/util/concurrent/ConcurrentSkipListMap.java b/luni/src/main/java/java/util/concurrent/ConcurrentSkipListMap.java
index d0d6f14..803cd49 100644
--- a/luni/src/main/java/java/util/concurrent/ConcurrentSkipListMap.java
+++ b/luni/src/main/java/java/util/concurrent/ConcurrentSkipListMap.java
@@ -20,8 +20,8 @@
  * <p>This class implements a concurrent variant of <a
  * href="http://en.wikipedia.org/wiki/Skip_list" target="_top">SkipLists</a>
  * providing expected average <i>log(n)</i> time cost for the
- * <tt>containsKey</tt>, <tt>get</tt>, <tt>put</tt> and
- * <tt>remove</tt> operations and their variants.  Insertion, removal,
+ * {@code containsKey}, {@code get}, {@code put} and
+ * {@code remove} operations and their variants.  Insertion, removal,
  * update, and access operations safely execute concurrently by
  * multiple threads.  Iterators are <i>weakly consistent</i>, returning
  * elements reflecting the state of the map at some point at or since
@@ -30,28 +30,28 @@
  * other operations. Ascending key ordered views and their iterators
  * are faster than descending ones.
  *
- * <p>All <tt>Map.Entry</tt> pairs returned by methods in this class
+ * <p>All {@code Map.Entry} pairs returned by methods in this class
  * and its views represent snapshots of mappings at the time they were
- * produced. They do <em>not</em> support the <tt>Entry.setValue</tt>
+ * produced. They do <em>not</em> support the {@code Entry.setValue}
  * method. (Note however that it is possible to change mappings in the
- * associated map using <tt>put</tt>, <tt>putIfAbsent</tt>, or
- * <tt>replace</tt>, depending on exactly which effect you need.)
+ * associated map using {@code put}, {@code putIfAbsent}, or
+ * {@code replace}, depending on exactly which effect you need.)
  *
- * <p>Beware that, unlike in most collections, the <tt>size</tt>
+ * <p>Beware that, unlike in most collections, the {@code size}
  * method is <em>not</em> a constant-time operation. Because of the
  * asynchronous nature of these maps, determining the current number
  * of elements requires a traversal of the elements, and so may report
  * inaccurate results if this collection is modified during traversal.
- * Additionally, the bulk operations <tt>putAll</tt>, <tt>equals</tt>,
- * <tt>toArray</tt>, <tt>containsValue</tt>, and <tt>clear</tt> are
+ * Additionally, the bulk operations {@code putAll}, {@code equals},
+ * {@code toArray}, {@code containsValue}, and {@code clear} are
  * <em>not</em> guaranteed to be performed atomically. For example, an
- * iterator operating concurrently with a <tt>putAll</tt> operation
+ * iterator operating concurrently with a {@code putAll} operation
  * might view only some of the added elements.
  *
  * <p>This class and its views and iterators implement all of the
  * <em>optional</em> methods of the {@link Map} and {@link Iterator}
  * interfaces. Like most other concurrent collections, this class does
- * <em>not</em> permit the use of <tt>null</tt> keys or values because some
+ * <em>not</em> permit the use of {@code null} keys or values because some
  * null return values cannot be reliably distinguished from the absence of
  * elements.
  *
@@ -60,6 +60,7 @@
  * @param <V> the type of mapped values
  * @since 1.6
  */
+@SuppressWarnings("unchecked")
 public class ConcurrentSkipListMap<K,V> extends AbstractMap<K,V>
     implements ConcurrentNavigableMap<K,V>,
                Cloneable,
@@ -408,7 +409,7 @@
          * because callers will have already read value field and need
          * to use that read (not another done here) and so directly
          * test if value points to node.
-         * @param n a possibly null reference to a node
+         *
          * @return true if this node is a marker node
          */
         boolean isMarker() {
@@ -457,7 +458,7 @@
          * Returns value if this node contains a valid key-value pair,
          * else null.
          * @return this node's value if it isn't a marker or header or
-         * is deleted, else null.
+         * is deleted, else null
          */
         V getValidValue() {
             Object v = value;
@@ -810,7 +811,7 @@
      * Main insertion method.  Adds element if not present, or
      * replaces value if present and onlyIfAbsent is false.
      * @param kkey the key
-     * @param value  the value that must be associated with key
+     * @param value the value that must be associated with key
      * @param onlyIfAbsent if should not insert if already present
      * @return the old value, or null if newly inserted
      */
@@ -934,7 +935,7 @@
      * Adds given index nodes from given level down to 1.
      * @param idx the topmost index node being inserted
      * @param h the value of head to use to insert. This must be
-     * snapshotted by callers to provide correct insertion level
+     * snapshotted by callers to provide correct insertion level.
      * @param indexLevel the level of the index
      */
     private void addIndex(Index<K,V> idx, HeadIndex<K,V> h, int indexLevel) {
@@ -1359,7 +1360,7 @@
      * comparator.
      *
      * @param comparator the comparator that will be used to order this map.
-     *        If <tt>null</tt>, the {@linkplain Comparable natural
+     *        If {@code null}, the {@linkplain Comparable natural
      *        ordering} of the keys will be used.
      */
     public ConcurrentSkipListMap(Comparator<? super K> comparator) {
@@ -1373,7 +1374,7 @@
      * the keys.
      *
      * @param  m the map whose mappings are to be placed in this map
-     * @throws ClassCastException if the keys in <tt>m</tt> are not
+     * @throws ClassCastException if the keys in {@code m} are not
      *         {@link Comparable}, or are not mutually comparable
      * @throws NullPointerException if the specified map or any of its keys
      *         or values are null
@@ -1400,7 +1401,7 @@
     }
 
     /**
-     * Returns a shallow copy of this <tt>ConcurrentSkipListMap</tt>
+     * Returns a shallow copy of this {@code ConcurrentSkipListMap}
      * instance. (The keys and values themselves are not cloned.)
      *
      * @return a shallow copy of this map
@@ -1477,11 +1478,11 @@
     /* ---------------- Serialization -------------- */
 
     /**
-     * Saves the state of this map to a stream (that is, serializes it).
+     * Saves this map to a stream (that is, serializes it).
      *
      * @serialData The key (Object) and value (Object) for each
      * key-value mapping represented by the map, followed by
-     * <tt>null</tt>. The key-value mappings are emitted in key-order
+     * {@code null}. The key-value mappings are emitted in key-order
      * (as determined by the Comparator, or by the keys' natural
      * ordering if no Comparator).
      */
@@ -1502,9 +1503,7 @@
     }
 
     /**
-     * Reconstitutes the map from a stream (that is, deserializes it).
-     *
-     * @param s the stream
+     * Reconstitutes this map from a stream (that is, deserializes it).
      */
     private void readObject(final java.io.ObjectInputStream s)
         throws java.io.IOException, ClassNotFoundException {
@@ -1567,11 +1566,11 @@
     /* ------ Map API methods ------ */
 
     /**
-     * Returns <tt>true</tt> if this map contains a mapping for the specified
+     * Returns {@code true} if this map contains a mapping for the specified
      * key.
      *
      * @param key key whose presence in this map is to be tested
-     * @return <tt>true</tt> if this map contains a mapping for the specified key
+     * @return {@code true} if this map contains a mapping for the specified key
      * @throws ClassCastException if the specified key cannot be compared
      *         with the keys currently in the map
      * @throws NullPointerException if the specified key is null
@@ -1606,7 +1605,7 @@
      * @param key key with which the specified value is to be associated
      * @param value value to be associated with the specified key
      * @return the previous value associated with the specified key, or
-     *         <tt>null</tt> if there was no mapping for the key
+     *         {@code null} if there was no mapping for the key
      * @throws ClassCastException if the specified key cannot be compared
      *         with the keys currently in the map
      * @throws NullPointerException if the specified key or value is null
@@ -1622,7 +1621,7 @@
      *
      * @param  key key for which mapping should be removed
      * @return the previous value associated with the specified key, or
-     *         <tt>null</tt> if there was no mapping for the key
+     *         {@code null} if there was no mapping for the key
      * @throws ClassCastException if the specified key cannot be compared
      *         with the keys currently in the map
      * @throws NullPointerException if the specified key is null
@@ -1632,15 +1631,15 @@
     }
 
     /**
-     * Returns <tt>true</tt> if this map maps one or more keys to the
+     * Returns {@code true} if this map maps one or more keys to the
      * specified value.  This operation requires time linear in the
      * map size. Additionally, it is possible for the map to change
      * during execution of this method, in which case the returned
      * result may be inaccurate.
      *
      * @param value value whose presence in this map is to be tested
-     * @return <tt>true</tt> if a mapping to <tt>value</tt> exists;
-     *         <tt>false</tt> otherwise
+     * @return {@code true} if a mapping to {@code value} exists;
+     *         {@code false} otherwise
      * @throws NullPointerException if the specified value is null
      */
     public boolean containsValue(Object value) {
@@ -1656,8 +1655,8 @@
 
     /**
      * Returns the number of key-value mappings in this map.  If this map
-     * contains more than <tt>Integer.MAX_VALUE</tt> elements, it
-     * returns <tt>Integer.MAX_VALUE</tt>.
+     * contains more than {@code Integer.MAX_VALUE} elements, it
+     * returns {@code Integer.MAX_VALUE}.
      *
      * <p>Beware that, unlike in most collections, this method is
      * <em>NOT</em> a constant-time operation. Because of the
@@ -1680,8 +1679,8 @@
     }
 
     /**
-     * Returns <tt>true</tt> if this map contains no key-value mappings.
-     * @return <tt>true</tt> if this map contains no key-value mappings
+     * Returns {@code true} if this map contains no key-value mappings.
+     * @return {@code true} if this map contains no key-value mappings
      */
     public boolean isEmpty() {
         return findFirst() == null;
@@ -1743,12 +1742,12 @@
      * The collection is backed by the map, so changes to the map are
      * reflected in the collection, and vice-versa.  The collection
      * supports element removal, which removes the corresponding
-     * mapping from the map, via the <tt>Iterator.remove</tt>,
-     * <tt>Collection.remove</tt>, <tt>removeAll</tt>,
-     * <tt>retainAll</tt> and <tt>clear</tt> operations.  It does not
-     * support the <tt>add</tt> or <tt>addAll</tt> operations.
+     * mapping from the map, via the {@code Iterator.remove},
+     * {@code Collection.remove}, {@code removeAll},
+     * {@code retainAll} and {@code clear} operations.  It does not
+     * support the {@code add} or {@code addAll} operations.
      *
-     * <p>The view's <tt>iterator</tt> is a "weakly consistent" iterator
+     * <p>The view's {@code iterator} is a "weakly consistent" iterator
      * that will never throw {@link ConcurrentModificationException},
      * and guarantees to traverse elements as they existed upon
      * construction of the iterator, and may (but is not guaranteed to)
@@ -1765,20 +1764,20 @@
      * The set is backed by the map, so changes to the map are
      * reflected in the set, and vice-versa.  The set supports element
      * removal, which removes the corresponding mapping from the map,
-     * via the <tt>Iterator.remove</tt>, <tt>Set.remove</tt>,
-     * <tt>removeAll</tt>, <tt>retainAll</tt> and <tt>clear</tt>
-     * operations.  It does not support the <tt>add</tt> or
-     * <tt>addAll</tt> operations.
+     * via the {@code Iterator.remove}, {@code Set.remove},
+     * {@code removeAll}, {@code retainAll} and {@code clear}
+     * operations.  It does not support the {@code add} or
+     * {@code addAll} operations.
      *
-     * <p>The view's <tt>iterator</tt> is a "weakly consistent" iterator
+     * <p>The view's {@code iterator} is a "weakly consistent" iterator
      * that will never throw {@link ConcurrentModificationException},
      * and guarantees to traverse elements as they existed upon
      * construction of the iterator, and may (but is not guaranteed to)
      * reflect any modifications subsequent to construction.
      *
-     * <p>The <tt>Map.Entry</tt> elements returned by
-     * <tt>iterator.next()</tt> do <em>not</em> support the
-     * <tt>setValue</tt> operation.
+     * <p>The {@code Map.Entry} elements returned by
+     * {@code iterator.next()} do <em>not</em> support the
+     * {@code setValue} operation.
      *
      * @return a set view of the mappings contained in this map,
      *         sorted in ascending key order
@@ -1802,15 +1801,15 @@
 
     /**
      * Compares the specified object with this map for equality.
-     * Returns <tt>true</tt> if the given object is also a map and the
+     * Returns {@code true} if the given object is also a map and the
      * two maps represent the same mappings.  More formally, two maps
-     * <tt>m1</tt> and <tt>m2</tt> represent the same mappings if
-     * <tt>m1.entrySet().equals(m2.entrySet())</tt>.  This
+     * {@code m1} and {@code m2} represent the same mappings if
+     * {@code m1.entrySet().equals(m2.entrySet())}.  This
      * operation may return misleading results if either map is
      * concurrently modified during execution of this method.
      *
      * @param o object to be compared for equality with this map
-     * @return <tt>true</tt> if the specified object is equal to this map
+     * @return {@code true} if the specified object is equal to this map
      */
     public boolean equals(Object o) {
         if (o == this)
@@ -1842,7 +1841,7 @@
      * {@inheritDoc}
      *
      * @return the previous value associated with the specified key,
-     *         or <tt>null</tt> if there was no mapping for the key
+     *         or {@code null} if there was no mapping for the key
      * @throws ClassCastException if the specified key cannot be compared
      *         with the keys currently in the map
      * @throws NullPointerException if the specified key or value is null
@@ -1897,7 +1896,7 @@
      * {@inheritDoc}
      *
      * @return the previous value associated with the specified key,
-     *         or <tt>null</tt> if there was no mapping for the key
+     *         or {@code null} if there was no mapping for the key
      * @throws ClassCastException if the specified key cannot be compared
      *         with the keys currently in the map
      * @throws NullPointerException if the specified key or value is null
@@ -2014,9 +2013,9 @@
 
     /**
      * Returns a key-value mapping associated with the greatest key
-     * strictly less than the given key, or <tt>null</tt> if there is
+     * strictly less than the given key, or {@code null} if there is
      * no such key. The returned entry does <em>not</em> support the
-     * <tt>Entry.setValue</tt> method.
+     * {@code Entry.setValue} method.
      *
      * @throws ClassCastException {@inheritDoc}
      * @throws NullPointerException if the specified key is null
@@ -2036,9 +2035,9 @@
 
     /**
      * Returns a key-value mapping associated with the greatest key
-     * less than or equal to the given key, or <tt>null</tt> if there
+     * less than or equal to the given key, or {@code null} if there
      * is no such key. The returned entry does <em>not</em> support
-     * the <tt>Entry.setValue</tt> method.
+     * the {@code Entry.setValue} method.
      *
      * @param key the key
      * @throws ClassCastException {@inheritDoc}
@@ -2060,9 +2059,9 @@
 
     /**
      * Returns a key-value mapping associated with the least key
-     * greater than or equal to the given key, or <tt>null</tt> if
+     * greater than or equal to the given key, or {@code null} if
      * there is no such entry. The returned entry does <em>not</em>
-     * support the <tt>Entry.setValue</tt> method.
+     * support the {@code Entry.setValue} method.
      *
      * @throws ClassCastException {@inheritDoc}
      * @throws NullPointerException if the specified key is null
@@ -2082,9 +2081,9 @@
 
     /**
      * Returns a key-value mapping associated with the least key
-     * strictly greater than the given key, or <tt>null</tt> if there
+     * strictly greater than the given key, or {@code null} if there
      * is no such key. The returned entry does <em>not</em> support
-     * the <tt>Entry.setValue</tt> method.
+     * the {@code Entry.setValue} method.
      *
      * @param key the key
      * @throws ClassCastException {@inheritDoc}
@@ -2106,9 +2105,9 @@
 
     /**
      * Returns a key-value mapping associated with the least
-     * key in this map, or <tt>null</tt> if the map is empty.
+     * key in this map, or {@code null} if the map is empty.
      * The returned entry does <em>not</em> support
-     * the <tt>Entry.setValue</tt> method.
+     * the {@code Entry.setValue} method.
      */
     public Map.Entry<K,V> firstEntry() {
         for (;;) {
@@ -2123,9 +2122,9 @@
 
     /**
      * Returns a key-value mapping associated with the greatest
-     * key in this map, or <tt>null</tt> if the map is empty.
+     * key in this map, or {@code null} if the map is empty.
      * The returned entry does <em>not</em> support
-     * the <tt>Entry.setValue</tt> method.
+     * the {@code Entry.setValue} method.
      */
     public Map.Entry<K,V> lastEntry() {
         for (;;) {
@@ -2140,9 +2139,9 @@
 
     /**
      * Removes and returns a key-value mapping associated with
-     * the least key in this map, or <tt>null</tt> if the map is empty.
+     * the least key in this map, or {@code null} if the map is empty.
      * The returned entry does <em>not</em> support
-     * the <tt>Entry.setValue</tt> method.
+     * the {@code Entry.setValue} method.
      */
     public Map.Entry<K,V> pollFirstEntry() {
         return doRemoveFirstEntry();
@@ -2150,9 +2149,9 @@
 
     /**
      * Removes and returns a key-value mapping associated with
-     * the greatest key in this map, or <tt>null</tt> if the map is empty.
+     * the greatest key in this map, or {@code null} if the map is empty.
      * The returned entry does <em>not</em> support
-     * the <tt>Entry.setValue</tt> method.
+     * the {@code Entry.setValue} method.
      */
     public Map.Entry<K,V> pollLastEntry() {
         return doRemoveLastEntry();
@@ -2268,7 +2267,7 @@
 
     static final <E> List<E> toList(Collection<E> c) {
         // Using size() here would be a pessimization.
-        List<E> list = new ArrayList<E>();
+        ArrayList<E> list = new ArrayList<E>();
         for (E e : c)
             list.add(e);
         return list;
@@ -2312,7 +2311,7 @@
             Collection<?> c = (Collection<?>) o;
             try {
                 return containsAll(c) && c.containsAll(this);
-            } catch (ClassCastException unused)   {
+            } catch (ClassCastException unused) {
                 return false;
             } catch (NullPointerException unused) {
                 return false;
@@ -2421,7 +2420,7 @@
             Collection<?> c = (Collection<?>) o;
             try {
                 return containsAll(c) && c.containsAll(this);
-            } catch (ClassCastException unused)   {
+            } catch (ClassCastException unused) {
                 return false;
             } catch (NullPointerException unused) {
                 return false;
@@ -2438,8 +2437,8 @@
      * underlying maps, differing in that mappings outside their range are
      * ignored, and attempts to add mappings outside their ranges result
      * in {@link IllegalArgumentException}.  Instances of this class are
-     * constructed only using the <tt>subMap</tt>, <tt>headMap</tt>, and
-     * <tt>tailMap</tt> methods of their underlying maps.
+     * constructed only using the {@code subMap}, {@code headMap}, and
+     * {@code tailMap} methods of their underlying maps.
      *
      * @serial include
      */
@@ -2467,7 +2466,7 @@
         private transient Collection<V> valuesView;
 
         /**
-         * Creates a new submap, initializing all fields
+         * Creates a new submap, initializing all fields.
          */
         SubMap(ConcurrentSkipListMap<K,V> map,
                K fromKey, boolean fromInclusive,
@@ -2516,7 +2515,7 @@
         }
 
         /**
-         * Returns true if node key is less than upper bound of range
+         * Returns true if node key is less than upper bound of range.
          */
         private boolean isBeforeEnd(ConcurrentSkipListMap.Node<K,V> n) {
             if (n == null)
@@ -2534,7 +2533,7 @@
 
         /**
          * Returns lowest node. This node might not be in range, so
-         * most usages need to check bounds
+         * most usages need to check bounds.
          */
         private ConcurrentSkipListMap.Node<K,V> loNode() {
             if (lo == null)
@@ -2547,7 +2546,7 @@
 
         /**
          * Returns highest node. This node might not be in range, so
-         * most usages need to check bounds
+         * most usages need to check bounds.
          */
         private ConcurrentSkipListMap.Node<K,V> hiNode() {
             if (hi == null)
@@ -2559,7 +2558,7 @@
         }
 
         /**
-         * Returns lowest absolute key (ignoring directonality)
+         * Returns lowest absolute key (ignoring directonality).
          */
         private K lowestKey() {
             ConcurrentSkipListMap.Node<K,V> n = loNode();
@@ -2570,7 +2569,7 @@
         }
 
         /**
-         * Returns highest absolute key (ignoring directonality)
+         * Returns highest absolute key (ignoring directonality).
          */
         private K highestKey() {
             ConcurrentSkipListMap.Node<K,V> n = hiNode();
diff --git a/luni/src/main/java/java/util/concurrent/ConcurrentSkipListSet.java b/luni/src/main/java/java/util/concurrent/ConcurrentSkipListSet.java
index 71431a9..f1402b6 100644
--- a/luni/src/main/java/java/util/concurrent/ConcurrentSkipListSet.java
+++ b/luni/src/main/java/java/util/concurrent/ConcurrentSkipListSet.java
@@ -19,7 +19,7 @@
  * on which constructor is used.
  *
  * <p>This implementation provides expected average <i>log(n)</i> time
- * cost for the <tt>contains</tt>, <tt>add</tt>, and <tt>remove</tt>
+ * cost for the {@code contains}, {@code add}, and {@code remove}
  * operations and their variants.  Insertion, removal, and access
  * operations safely execute concurrently by multiple threads.
  * Iterators are <i>weakly consistent</i>, returning elements
@@ -29,23 +29,23 @@
  * other operations.  Ascending ordered views and their iterators are
  * faster than descending ones.
  *
- * <p>Beware that, unlike in most collections, the <tt>size</tt>
+ * <p>Beware that, unlike in most collections, the {@code size}
  * method is <em>not</em> a constant-time operation. Because of the
  * asynchronous nature of these sets, determining the current number
  * of elements requires a traversal of the elements, and so may report
  * inaccurate results if this collection is modified during traversal.
- * Additionally, the bulk operations <tt>addAll</tt>,
- * <tt>removeAll</tt>, <tt>retainAll</tt>, <tt>containsAll</tt>,
- * <tt>equals</tt>, and <tt>toArray</tt> are <em>not</em> guaranteed
+ * Additionally, the bulk operations {@code addAll},
+ * {@code removeAll}, {@code retainAll}, {@code containsAll},
+ * {@code equals}, and {@code toArray} are <em>not</em> guaranteed
  * to be performed atomically. For example, an iterator operating
- * concurrently with an <tt>addAll</tt> operation might view only some
+ * concurrently with an {@code addAll} operation might view only some
  * of the added elements.
  *
  * <p>This class and its iterators implement all of the
  * <em>optional</em> methods of the {@link Set} and {@link Iterator}
  * interfaces. Like most other concurrent collection implementations,
- * this class does not permit the use of <tt>null</tt> elements,
- * because <tt>null</tt> arguments and return values cannot be reliably
+ * this class does not permit the use of {@code null} elements,
+ * because {@code null} arguments and return values cannot be reliably
  * distinguished from the absence of elements.
  *
  * @author Doug Lea
@@ -61,7 +61,7 @@
     /**
      * The underlying map. Uses Boolean.TRUE as value for each
      * element.  This field is declared final for the sake of thread
-     * safety, which entails some ugliness in clone()
+     * safety, which entails some ugliness in clone().
      */
     private final ConcurrentNavigableMap<E,Object> m;
 
@@ -78,7 +78,7 @@
      * the specified comparator.
      *
      * @param comparator the comparator that will be used to order this set.
-     *        If <tt>null</tt>, the {@linkplain Comparable natural
+     *        If {@code null}, the {@linkplain Comparable natural
      *        ordering} of the elements will be used.
      */
     public ConcurrentSkipListSet(Comparator<? super E> comparator) {
@@ -91,7 +91,7 @@
      * {@linkplain Comparable natural ordering}.
      *
      * @param c The elements that will comprise the new set
-     * @throws ClassCastException if the elements in <tt>c</tt> are
+     * @throws ClassCastException if the elements in {@code c} are
      *         not {@link Comparable}, or are not mutually comparable
      * @throws NullPointerException if the specified collection or any
      *         of its elements are null
@@ -122,7 +122,7 @@
     }
 
     /**
-     * Returns a shallow copy of this <tt>ConcurrentSkipListSet</tt>
+     * Returns a shallow copy of this {@code ConcurrentSkipListSet}
      * instance. (The elements themselves are not cloned.)
      *
      * @return a shallow copy of this set
@@ -143,8 +143,8 @@
 
     /**
      * Returns the number of elements in this set.  If this set
-     * contains more than <tt>Integer.MAX_VALUE</tt> elements, it
-     * returns <tt>Integer.MAX_VALUE</tt>.
+     * contains more than {@code Integer.MAX_VALUE} elements, it
+     * returns {@code Integer.MAX_VALUE}.
      *
      * <p>Beware that, unlike in most collections, this method is
      * <em>NOT</em> a constant-time operation. Because of the
@@ -162,20 +162,20 @@
     }
 
     /**
-     * Returns <tt>true</tt> if this set contains no elements.
-     * @return <tt>true</tt> if this set contains no elements
+     * Returns {@code true} if this set contains no elements.
+     * @return {@code true} if this set contains no elements
      */
     public boolean isEmpty() {
         return m.isEmpty();
     }
 
     /**
-     * Returns <tt>true</tt> if this set contains the specified element.
-     * More formally, returns <tt>true</tt> if and only if this set
-     * contains an element <tt>e</tt> such that <tt>o.equals(e)</tt>.
+     * Returns {@code true} if this set contains the specified element.
+     * More formally, returns {@code true} if and only if this set
+     * contains an element {@code e} such that {@code o.equals(e)}.
      *
      * @param o object to be checked for containment in this set
-     * @return <tt>true</tt> if this set contains the specified element
+     * @return {@code true} if this set contains the specified element
      * @throws ClassCastException if the specified element cannot be
      *         compared with the elements currently in this set
      * @throws NullPointerException if the specified element is null
@@ -186,15 +186,15 @@
 
     /**
      * Adds the specified element to this set if it is not already present.
-     * More formally, adds the specified element <tt>e</tt> to this set if
-     * the set contains no element <tt>e2</tt> such that <tt>e.equals(e2)</tt>.
+     * More formally, adds the specified element {@code e} to this set if
+     * the set contains no element {@code e2} such that {@code e.equals(e2)}.
      * If this set already contains the element, the call leaves the set
-     * unchanged and returns <tt>false</tt>.
+     * unchanged and returns {@code false}.
      *
      * @param e element to be added to this set
-     * @return <tt>true</tt> if this set did not already contain the
+     * @return {@code true} if this set did not already contain the
      *         specified element
-     * @throws ClassCastException if <tt>e</tt> cannot be compared
+     * @throws ClassCastException if {@code e} cannot be compared
      *         with the elements currently in this set
      * @throws NullPointerException if the specified element is null
      */
@@ -204,15 +204,15 @@
 
     /**
      * Removes the specified element from this set if it is present.
-     * More formally, removes an element <tt>e</tt> such that
-     * <tt>o.equals(e)</tt>, if this set contains such an element.
-     * Returns <tt>true</tt> if this set contained the element (or
+     * More formally, removes an element {@code e} such that
+     * {@code o.equals(e)}, if this set contains such an element.
+     * Returns {@code true} if this set contained the element (or
      * equivalently, if this set changed as a result of the call).
      * (This set will not contain the element once the call returns.)
      *
      * @param o object to be removed from this set, if present
-     * @return <tt>true</tt> if this set contained the specified element
-     * @throws ClassCastException if <tt>o</tt> cannot be compared
+     * @return {@code true} if this set contained the specified element
+     * @throws ClassCastException if {@code o} cannot be compared
      *         with the elements currently in this set
      * @throws NullPointerException if the specified element is null
      */
@@ -250,7 +250,7 @@
 
     /**
      * Compares the specified object with this set for equality.  Returns
-     * <tt>true</tt> if the specified object is also a set, the two sets
+     * {@code true} if the specified object is also a set, the two sets
      * have the same size, and every member of the specified set is
      * contained in this set (or equivalently, every member of this set is
      * contained in the specified set).  This definition ensures that the
@@ -258,7 +258,7 @@
      * set interface.
      *
      * @param o the object to be compared for equality with this set
-     * @return <tt>true</tt> if the specified object is equal to this set
+     * @return {@code true} if the specified object is equal to this set
      */
     public boolean equals(Object o) {
         // Override AbstractSet version to avoid calling size()
@@ -269,7 +269,7 @@
         Collection<?> c = (Collection<?>) o;
         try {
             return containsAll(c) && c.containsAll(this);
-        } catch (ClassCastException unused)   {
+        } catch (ClassCastException unused) {
             return false;
         } catch (NullPointerException unused) {
             return false;
@@ -283,7 +283,7 @@
      * value is the <i>asymmetric set difference</i> of the two sets.
      *
      * @param  c collection containing elements to be removed from this set
-     * @return <tt>true</tt> if this set changed as a result of the call
+     * @return {@code true} if this set changed as a result of the call
      * @throws ClassCastException if the types of one or more elements in this
      *         set are incompatible with the specified collection
      * @throws NullPointerException if the specified collection or any
@@ -431,7 +431,7 @@
      * reflected in the descending set, and vice-versa.
      *
      * <p>The returned set has an ordering equivalent to
-     * <tt>{@link Collections#reverseOrder(Comparator) Collections.reverseOrder}(comparator())</tt>.
+     * {@link Collections#reverseOrder(Comparator) Collections.reverseOrder}{@code (comparator())}.
      * The expression {@code s.descendingSet().descendingSet()} returns a
      * view of {@code s} essentially equivalent to {@code s}.
      *
diff --git a/luni/src/main/java/java/util/concurrent/CopyOnWriteArraySet.java b/luni/src/main/java/java/util/concurrent/CopyOnWriteArraySet.java
index 1f37bc9..6fa8feb 100644
--- a/luni/src/main/java/java/util/concurrent/CopyOnWriteArraySet.java
+++ b/luni/src/main/java/java/util/concurrent/CopyOnWriteArraySet.java
@@ -20,17 +20,17 @@
  *       vastly outnumber mutative operations, and you need
  *       to prevent interference among threads during traversal.
  *  <li>It is thread-safe.
- *  <li>Mutative operations (<tt>add</tt>, <tt>set</tt>, <tt>remove</tt>, etc.)
+ *  <li>Mutative operations ({@code add}, {@code set}, {@code remove}, etc.)
  *      are expensive since they usually entail copying the entire underlying
  *      array.
- *  <li>Iterators do not support the mutative <tt>remove</tt> operation.
+ *  <li>Iterators do not support the mutative {@code remove} operation.
  *  <li>Traversal via iterators is fast and cannot encounter
  *      interference from other threads. Iterators rely on
  *      unchanging snapshots of the array at the time the iterators were
  *      constructed.
  * </ul>
  *
- * <p> <b>Sample Usage.</b> The following code sketch uses a
+ * <p><b>Sample Usage.</b> The following code sketch uses a
  * copy-on-write set to maintain a set of Handler objects that
  * perform some action upon state updates.
  *
@@ -48,7 +48,7 @@
  *   public void update() {
  *     changeState();
  *     for (Handler handler : handlers)
- *        handler.handle();
+ *       handler.handle();
  *   }
  * }}</pre>
  *
@@ -92,22 +92,22 @@
     }
 
     /**
-     * Returns <tt>true</tt> if this set contains no elements.
+     * Returns {@code true} if this set contains no elements.
      *
-     * @return <tt>true</tt> if this set contains no elements
+     * @return {@code true} if this set contains no elements
      */
     public boolean isEmpty() {
         return al.isEmpty();
     }
 
     /**
-     * Returns <tt>true</tt> if this set contains the specified element.
-     * More formally, returns <tt>true</tt> if and only if this set
-     * contains an element <tt>e</tt> such that
+     * Returns {@code true} if this set contains the specified element.
+     * More formally, returns {@code true} if and only if this set
+     * contains an element {@code e} such that
      * <tt>(o==null&nbsp;?&nbsp;e==null&nbsp;:&nbsp;o.equals(e))</tt>.
      *
      * @param o element whose presence in this set is to be tested
-     * @return <tt>true</tt> if this set contains the specified element
+     * @return {@code true} if this set contains the specified element
      */
     public boolean contains(Object o) {
         return al.contains(o);
@@ -143,7 +143,7 @@
      * <p>If this set fits in the specified array with room to spare
      * (i.e., the array has more elements than this set), the element in
      * the array immediately following the end of the set is set to
-     * <tt>null</tt>.  (This is useful in determining the length of this
+     * {@code null}.  (This is useful in determining the length of this
      * set <i>only</i> if the caller knows that this set does not contain
      * any null elements.)
      *
@@ -156,14 +156,14 @@
      * precise control over the runtime type of the output array, and may,
      * under certain circumstances, be used to save allocation costs.
      *
-     * <p>Suppose <tt>x</tt> is a set known to contain only strings.
+     * <p>Suppose {@code x} is a set known to contain only strings.
      * The following code can be used to dump the set into a newly allocated
-     * array of <tt>String</tt>:
+     * array of {@code String}:
      *
      *  <pre> {@code String[] y = x.toArray(new String[0]);}</pre>
      *
-     * Note that <tt>toArray(new Object[0])</tt> is identical in function to
-     * <tt>toArray()</tt>.
+     * Note that {@code toArray(new Object[0])} is identical in function to
+     * {@code toArray()}.
      *
      * @param a the array into which the elements of this set are to be
      *        stored, if it is big enough; otherwise, a new array of the same
@@ -188,15 +188,15 @@
 
     /**
      * Removes the specified element from this set if it is present.
-     * More formally, removes an element <tt>e</tt> such that
+     * More formally, removes an element {@code e} such that
      * <tt>(o==null&nbsp;?&nbsp;e==null&nbsp;:&nbsp;o.equals(e))</tt>,
-     * if this set contains such an element.  Returns <tt>true</tt> if
+     * if this set contains such an element.  Returns {@code true} if
      * this set contained the element (or equivalently, if this set
      * changed as a result of the call).  (This set will not contain the
      * element once the call returns.)
      *
      * @param o object to be removed from this set, if present
-     * @return <tt>true</tt> if this set contained the specified element
+     * @return {@code true} if this set contained the specified element
      */
     public boolean remove(Object o) {
         return al.remove(o);
@@ -204,14 +204,14 @@
 
     /**
      * Adds the specified element to this set if it is not already present.
-     * More formally, adds the specified element <tt>e</tt> to this set if
-     * the set contains no element <tt>e2</tt> such that
+     * More formally, adds the specified element {@code e} to this set if
+     * the set contains no element {@code e2} such that
      * <tt>(e==null&nbsp;?&nbsp;e2==null&nbsp;:&nbsp;e.equals(e2))</tt>.
      * If this set already contains the element, the call leaves the set
-     * unchanged and returns <tt>false</tt>.
+     * unchanged and returns {@code false}.
      *
      * @param e element to be added to this set
-     * @return <tt>true</tt> if this set did not already contain the specified
+     * @return {@code true} if this set did not already contain the specified
      *         element
      */
     public boolean add(E e) {
@@ -219,12 +219,12 @@
     }
 
     /**
-     * Returns <tt>true</tt> if this set contains all of the elements of the
+     * Returns {@code true} if this set contains all of the elements of the
      * specified collection.  If the specified collection is also a set, this
-     * method returns <tt>true</tt> if it is a <i>subset</i> of this set.
+     * method returns {@code true} if it is a <i>subset</i> of this set.
      *
      * @param  c collection to be checked for containment in this set
-     * @return <tt>true</tt> if this set contains all of the elements of the
+     * @return {@code true} if this set contains all of the elements of the
      *         specified collection
      * @throws NullPointerException if the specified collection is null
      * @see #contains(Object)
@@ -236,13 +236,13 @@
     /**
      * Adds all of the elements in the specified collection to this set if
      * they're not already present.  If the specified collection is also a
-     * set, the <tt>addAll</tt> operation effectively modifies this set so
+     * set, the {@code addAll} operation effectively modifies this set so
      * that its value is the <i>union</i> of the two sets.  The behavior of
      * this operation is undefined if the specified collection is modified
      * while the operation is in progress.
      *
      * @param  c collection containing elements to be added to this set
-     * @return <tt>true</tt> if this set changed as a result of the call
+     * @return {@code true} if this set changed as a result of the call
      * @throws NullPointerException if the specified collection is null
      * @see #add(Object)
      */
@@ -257,7 +257,7 @@
      * <i>asymmetric set difference</i> of the two sets.
      *
      * @param  c collection containing elements to be removed from this set
-     * @return <tt>true</tt> if this set changed as a result of the call
+     * @return {@code true} if this set changed as a result of the call
      * @throws ClassCastException if the class of an element of this set
      *         is incompatible with the specified collection (optional)
      * @throws NullPointerException if this set contains a null element and the
@@ -278,7 +278,7 @@
      * two sets.
      *
      * @param  c collection containing elements to be retained in this set
-     * @return <tt>true</tt> if this set changed as a result of the call
+     * @return {@code true} if this set changed as a result of the call
      * @throws ClassCastException if the class of an element of this set
      *         is incompatible with the specified collection (optional)
      * @throws NullPointerException if this set contains a null element and the
@@ -297,7 +297,7 @@
      * <p>The returned iterator provides a snapshot of the state of the set
      * when the iterator was constructed. No synchronization is needed while
      * traversing the iterator. The iterator does <em>NOT</em> support the
-     * <tt>remove</tt> method.
+     * {@code remove} method.
      *
      * @return an iterator over the elements in this set
      */
@@ -354,7 +354,7 @@
     }
 
     /**
-     * Test for equality, coping with nulls.
+     * Tests for equality, coping with nulls.
      */
     private static boolean eq(Object o1, Object o2) {
         return (o1 == null) ? o2 == null : o1.equals(o2);
diff --git a/luni/src/main/java/java/util/concurrent/CountDownLatch.java b/luni/src/main/java/java/util/concurrent/CountDownLatch.java
index e90badf..fe0fa65 100644
--- a/luni/src/main/java/java/util/concurrent/CountDownLatch.java
+++ b/luni/src/main/java/java/util/concurrent/CountDownLatch.java
@@ -5,7 +5,7 @@
  */
 
 package java.util.concurrent;
-import java.util.concurrent.locks.*;
+import java.util.concurrent.locks.AbstractQueuedSynchronizer;
 
 /**
  * A synchronization aid that allows one or more threads to wait until
@@ -63,15 +63,15 @@
  *   private final CountDownLatch startSignal;
  *   private final CountDownLatch doneSignal;
  *   Worker(CountDownLatch startSignal, CountDownLatch doneSignal) {
- *      this.startSignal = startSignal;
- *      this.doneSignal = doneSignal;
+ *     this.startSignal = startSignal;
+ *     this.doneSignal = doneSignal;
  *   }
  *   public void run() {
- *      try {
- *        startSignal.await();
- *        doWork();
- *        doneSignal.countDown();
- *      } catch (InterruptedException ex) {} // return;
+ *     try {
+ *       startSignal.await();
+ *       doWork();
+ *       doneSignal.countDown();
+ *     } catch (InterruptedException ex) {} // return;
  *   }
  *
  *   void doWork() { ... }
@@ -101,14 +101,14 @@
  *   private final CountDownLatch doneSignal;
  *   private final int i;
  *   WorkerRunnable(CountDownLatch doneSignal, int i) {
- *      this.doneSignal = doneSignal;
- *      this.i = i;
+ *     this.doneSignal = doneSignal;
+ *     this.i = i;
  *   }
  *   public void run() {
- *      try {
- *        doWork(i);
- *        doneSignal.countDown();
- *      } catch (InterruptedException ex) {} // return;
+ *     try {
+ *       doWork(i);
+ *       doneSignal.countDown();
+ *     } catch (InterruptedException ex) {} // return;
  *   }
  *
  *   void doWork() { ... }
diff --git a/luni/src/main/java/java/util/concurrent/CountedCompleter.java b/luni/src/main/java/java/util/concurrent/CountedCompleter.java
new file mode 100644
index 0000000..ffe7582
--- /dev/null
+++ b/luni/src/main/java/java/util/concurrent/CountedCompleter.java
@@ -0,0 +1,715 @@
+/*
+ * Written by Doug Lea with assistance from members of JCP JSR-166
+ * Expert Group and released to the public domain, as explained at
+ * http://creativecommons.org/publicdomain/zero/1.0/
+ */
+
+package java.util.concurrent;
+
+/**
+ * A {@link ForkJoinTask} with a completion action performed when
+ * triggered and there are no remaining pending
+ * actions. CountedCompleters are in general more robust in the
+ * presence of subtask stalls and blockage than are other forms of
+ * ForkJoinTasks, but are less intuitive to program.  Uses of
+ * CountedCompleter are similar to those of other completion based
+ * components (such as {@link java.nio.channels.CompletionHandler})
+ * except that multiple <em>pending</em> completions may be necessary
+ * to trigger the completion action {@link #onCompletion}, not just one.
+ * Unless initialized otherwise, the {@linkplain #getPendingCount pending
+ * count} starts at zero, but may be (atomically) changed using
+ * methods {@link #setPendingCount}, {@link #addToPendingCount}, and
+ * {@link #compareAndSetPendingCount}. Upon invocation of {@link
+ * #tryComplete}, if the pending action count is nonzero, it is
+ * decremented; otherwise, the completion action is performed, and if
+ * this completer itself has a completer, the process is continued
+ * with its completer.  As is the case with related synchronization
+ * components such as {@link java.util.concurrent.Phaser Phaser} and
+ * {@link java.util.concurrent.Semaphore Semaphore}, these methods
+ * affect only internal counts; they do not establish any further
+ * internal bookkeeping. In particular, the identities of pending
+ * tasks are not maintained. As illustrated below, you can create
+ * subclasses that do record some or all pending tasks or their
+ * results when needed.  As illustrated below, utility methods
+ * supporting customization of completion traversals are also
+ * provided. However, because CountedCompleters provide only basic
+ * synchronization mechanisms, it may be useful to create further
+ * abstract subclasses that maintain linkages, fields, and additional
+ * support methods appropriate for a set of related usages.
+ *
+ * <p>A concrete CountedCompleter class must define method {@link
+ * #compute}, that should in most cases (as illustrated below), invoke
+ * {@code tryComplete()} once before returning. The class may also
+ * optionally override method {@link #onCompletion} to perform an
+ * action upon normal completion, and method {@link
+ * #onExceptionalCompletion} to perform an action upon any exception.
+ *
+ * <p>CountedCompleters most often do not bear results, in which case
+ * they are normally declared as {@code CountedCompleter<Void>}, and
+ * will always return {@code null} as a result value.  In other cases,
+ * you should override method {@link #getRawResult} to provide a
+ * result from {@code join(), invoke()}, and related methods.  In
+ * general, this method should return the value of a field (or a
+ * function of one or more fields) of the CountedCompleter object that
+ * holds the result upon completion. Method {@link #setRawResult} by
+ * default plays no role in CountedCompleters.  It is possible, but
+ * rarely applicable, to override this method to maintain other
+ * objects or fields holding result data.
+ *
+ * <p>A CountedCompleter that does not itself have a completer (i.e.,
+ * one for which {@link #getCompleter} returns {@code null}) can be
+ * used as a regular ForkJoinTask with this added functionality.
+ * However, any completer that in turn has another completer serves
+ * only as an internal helper for other computations, so its own task
+ * status (as reported in methods such as {@link ForkJoinTask#isDone})
+ * is arbitrary; this status changes only upon explicit invocations of
+ * {@link #complete}, {@link ForkJoinTask#cancel}, {@link
+ * ForkJoinTask#completeExceptionally} or upon exceptional completion
+ * of method {@code compute}. Upon any exceptional completion, the
+ * exception may be relayed to a task's completer (and its completer,
+ * and so on), if one exists and it has not otherwise already
+ * completed. Similarly, cancelling an internal CountedCompleter has
+ * only a local effect on that completer, so is not often useful.
+ *
+ * <p><b>Sample Usages.</b>
+ *
+ * <p><b>Parallel recursive decomposition.</b> CountedCompleters may
+ * be arranged in trees similar to those often used with {@link
+ * RecursiveAction}s, although the constructions involved in setting
+ * them up typically vary. Here, the completer of each task is its
+ * parent in the computation tree. Even though they entail a bit more
+ * bookkeeping, CountedCompleters may be better choices when applying
+ * a possibly time-consuming operation (that cannot be further
+ * subdivided) to each element of an array or collection; especially
+ * when the operation takes a significantly different amount of time
+ * to complete for some elements than others, either because of
+ * intrinsic variation (for example I/O) or auxiliary effects such as
+ * garbage collection.  Because CountedCompleters provide their own
+ * continuations, other threads need not block waiting to perform
+ * them.
+ *
+ * <p>For example, here is an initial version of a class that uses
+ * divide-by-two recursive decomposition to divide work into single
+ * pieces (leaf tasks). Even when work is split into individual calls,
+ * tree-based techniques are usually preferable to directly forking
+ * leaf tasks, because they reduce inter-thread communication and
+ * improve load balancing. In the recursive case, the second of each
+ * pair of subtasks to finish triggers completion of its parent
+ * (because no result combination is performed, the default no-op
+ * implementation of method {@code onCompletion} is not overridden). A
+ * static utility method sets up the base task and invokes it
+ * (here, implicitly using the {@link ForkJoinPool#commonPool()}).
+ *
+ * <pre> {@code
+ * class MyOperation<E> { void apply(E e) { ... }  }
+ *
+ * class ForEach<E> extends CountedCompleter<Void> {
+ *
+ *   public static <E> void forEach(E[] array, MyOperation<E> op) {
+ *     new ForEach<E>(null, array, op, 0, array.length).invoke();
+ *   }
+ *
+ *   final E[] array; final MyOperation<E> op; final int lo, hi;
+ *   ForEach(CountedCompleter<?> p, E[] array, MyOperation<E> op, int lo, int hi) {
+ *     super(p);
+ *     this.array = array; this.op = op; this.lo = lo; this.hi = hi;
+ *   }
+ *
+ *   public void compute() { // version 1
+ *     if (hi - lo >= 2) {
+ *       int mid = (lo + hi) >>> 1;
+ *       setPendingCount(2); // must set pending count before fork
+ *       new ForEach(this, array, op, mid, hi).fork(); // right child
+ *       new ForEach(this, array, op, lo, mid).fork(); // left child
+ *     }
+ *     else if (hi > lo)
+ *       op.apply(array[lo]);
+ *     tryComplete();
+ *   }
+ * }}</pre>
+ *
+ * This design can be improved by noticing that in the recursive case,
+ * the task has nothing to do after forking its right task, so can
+ * directly invoke its left task before returning. (This is an analog
+ * of tail recursion removal.)  Also, because the task returns upon
+ * executing its left task (rather than falling through to invoke
+ * {@code tryComplete}) the pending count is set to one:
+ *
+ * <pre> {@code
+ * class ForEach<E> ...
+ *   public void compute() { // version 2
+ *     if (hi - lo >= 2) {
+ *       int mid = (lo + hi) >>> 1;
+ *       setPendingCount(1); // only one pending
+ *       new ForEach(this, array, op, mid, hi).fork(); // right child
+ *       new ForEach(this, array, op, lo, mid).compute(); // direct invoke
+ *     }
+ *     else {
+ *       if (hi > lo)
+ *         op.apply(array[lo]);
+ *       tryComplete();
+ *     }
+ *   }
+ * }</pre>
+ *
+ * As a further improvement, notice that the left task need not even
+ * exist.  Instead of creating a new one, we can iterate using the
+ * original task, and add a pending count for each fork. Additionally,
+ * because no task in this tree implements an {@link #onCompletion}
+ * method, {@code tryComplete()} can be replaced with {@link
+ * #propagateCompletion}.
+ *
+ * <pre> {@code
+ * class ForEach<E> ...
+ *   public void compute() { // version 3
+ *     int l = lo,  h = hi;
+ *     while (h - l >= 2) {
+ *       int mid = (l + h) >>> 1;
+ *       addToPendingCount(1);
+ *       new ForEach(this, array, op, mid, h).fork(); // right child
+ *       h = mid;
+ *     }
+ *     if (h > l)
+ *       op.apply(array[l]);
+ *     propagateCompletion();
+ *   }
+ * }</pre>
+ *
+ * Additional improvements of such classes might entail precomputing
+ * pending counts so that they can be established in constructors,
+ * specializing classes for leaf steps, subdividing by say, four,
+ * instead of two per iteration, and using an adaptive threshold
+ * instead of always subdividing down to single elements.
+ *
+ * <p><b>Searching.</b> A tree of CountedCompleters can search for a
+ * value or property in different parts of a data structure, and
+ * report a result in an {@link
+ * java.util.concurrent.atomic.AtomicReference AtomicReference} as
+ * soon as one is found. The others can poll the result to avoid
+ * unnecessary work. (You could additionally {@linkplain #cancel
+ * cancel} other tasks, but it is usually simpler and more efficient
+ * to just let them notice that the result is set and if so skip
+ * further processing.)  Illustrating again with an array using full
+ * partitioning (again, in practice, leaf tasks will almost always
+ * process more than one element):
+ *
+ * <pre> {@code
+ * class Searcher<E> extends CountedCompleter<E> {
+ *   final E[] array; final AtomicReference<E> result; final int lo, hi;
+ *   Searcher(CountedCompleter<?> p, E[] array, AtomicReference<E> result, int lo, int hi) {
+ *     super(p);
+ *     this.array = array; this.result = result; this.lo = lo; this.hi = hi;
+ *   }
+ *   public E getRawResult() { return result.get(); }
+ *   public void compute() { // similar to ForEach version 3
+ *     int l = lo,  h = hi;
+ *     while (result.get() == null && h >= l) {
+ *       if (h - l >= 2) {
+ *         int mid = (l + h) >>> 1;
+ *         addToPendingCount(1);
+ *         new Searcher(this, array, result, mid, h).fork();
+ *         h = mid;
+ *       }
+ *       else {
+ *         E x = array[l];
+ *         if (matches(x) && result.compareAndSet(null, x))
+ *           quietlyCompleteRoot(); // root task is now joinable
+ *         break;
+ *       }
+ *     }
+ *     tryComplete(); // normally complete whether or not found
+ *   }
+ *   boolean matches(E e) { ... } // return true if found
+ *
+ *   public static <E> E search(E[] array) {
+ *       return new Searcher<E>(null, array, new AtomicReference<E>(), 0, array.length).invoke();
+ *   }
+ * }}</pre>
+ *
+ * In this example, as well as others in which tasks have no other
+ * effects except to compareAndSet a common result, the trailing
+ * unconditional invocation of {@code tryComplete} could be made
+ * conditional ({@code if (result.get() == null) tryComplete();})
+ * because no further bookkeeping is required to manage completions
+ * once the root task completes.
+ *
+ * <p><b>Recording subtasks.</b> CountedCompleter tasks that combine
+ * results of multiple subtasks usually need to access these results
+ * in method {@link #onCompletion}. As illustrated in the following
+ * class (that performs a simplified form of map-reduce where mappings
+ * and reductions are all of type {@code E}), one way to do this in
+ * divide and conquer designs is to have each subtask record its
+ * sibling, so that it can be accessed in method {@code onCompletion}.
+ * This technique applies to reductions in which the order of
+ * combining left and right results does not matter; ordered
+ * reductions require explicit left/right designations.  Variants of
+ * other streamlinings seen in the above examples may also apply.
+ *
+ * <pre> {@code
+ * class MyMapper<E> { E apply(E v) {  ...  } }
+ * class MyReducer<E> { E apply(E x, E y) {  ...  } }
+ * class MapReducer<E> extends CountedCompleter<E> {
+ *   final E[] array; final MyMapper<E> mapper;
+ *   final MyReducer<E> reducer; final int lo, hi;
+ *   MapReducer<E> sibling;
+ *   E result;
+ *   MapReducer(CountedCompleter<?> p, E[] array, MyMapper<E> mapper,
+ *              MyReducer<E> reducer, int lo, int hi) {
+ *     super(p);
+ *     this.array = array; this.mapper = mapper;
+ *     this.reducer = reducer; this.lo = lo; this.hi = hi;
+ *   }
+ *   public void compute() {
+ *     if (hi - lo >= 2) {
+ *       int mid = (lo + hi) >>> 1;
+ *       MapReducer<E> left = new MapReducer(this, array, mapper, reducer, lo, mid);
+ *       MapReducer<E> right = new MapReducer(this, array, mapper, reducer, mid, hi);
+ *       left.sibling = right;
+ *       right.sibling = left;
+ *       setPendingCount(1); // only right is pending
+ *       right.fork();
+ *       left.compute();     // directly execute left
+ *     }
+ *     else {
+ *       if (hi > lo)
+ *           result = mapper.apply(array[lo]);
+ *       tryComplete();
+ *     }
+ *   }
+ *   public void onCompletion(CountedCompleter<?> caller) {
+ *     if (caller != this) {
+ *       MapReducer<E> child = (MapReducer<E>)caller;
+ *       MapReducer<E> sib = child.sibling;
+ *       if (sib == null || sib.result == null)
+ *         result = child.result;
+ *       else
+ *         result = reducer.apply(child.result, sib.result);
+ *     }
+ *   }
+ *   public E getRawResult() { return result; }
+ *
+ *   public static <E> E mapReduce(E[] array, MyMapper<E> mapper, MyReducer<E> reducer) {
+ *     return new MapReducer<E>(null, array, mapper, reducer,
+ *                              0, array.length).invoke();
+ *   }
+ * }}</pre>
+ *
+ * Here, method {@code onCompletion} takes a form common to many
+ * completion designs that combine results. This callback-style method
+ * is triggered once per task, in either of the two different contexts
+ * in which the pending count is, or becomes, zero: (1) by a task
+ * itself, if its pending count is zero upon invocation of {@code
+ * tryComplete}, or (2) by any of its subtasks when they complete and
+ * decrement the pending count to zero. The {@code caller} argument
+ * distinguishes cases.  Most often, when the caller is {@code this},
+ * no action is necessary. Otherwise the caller argument can be used
+ * (usually via a cast) to supply a value (and/or links to other
+ * values) to be combined.  Assuming proper use of pending counts, the
+ * actions inside {@code onCompletion} occur (once) upon completion of
+ * a task and its subtasks. No additional synchronization is required
+ * within this method to ensure thread safety of accesses to fields of
+ * this task or other completed tasks.
+ *
+ * <p><b>Completion Traversals</b>. If using {@code onCompletion} to
+ * process completions is inapplicable or inconvenient, you can use
+ * methods {@link #firstComplete} and {@link #nextComplete} to create
+ * custom traversals.  For example, to define a MapReducer that only
+ * splits out right-hand tasks in the form of the third ForEach
+ * example, the completions must cooperatively reduce along
+ * unexhausted subtask links, which can be done as follows:
+ *
+ * <pre> {@code
+ * class MapReducer<E> extends CountedCompleter<E> { // version 2
+ *   final E[] array; final MyMapper<E> mapper;
+ *   final MyReducer<E> reducer; final int lo, hi;
+ *   MapReducer<E> forks, next; // record subtask forks in list
+ *   E result;
+ *   MapReducer(CountedCompleter<?> p, E[] array, MyMapper<E> mapper,
+ *              MyReducer<E> reducer, int lo, int hi, MapReducer<E> next) {
+ *     super(p);
+ *     this.array = array; this.mapper = mapper;
+ *     this.reducer = reducer; this.lo = lo; this.hi = hi;
+ *     this.next = next;
+ *   }
+ *   public void compute() {
+ *     int l = lo,  h = hi;
+ *     while (h - l >= 2) {
+ *       int mid = (l + h) >>> 1;
+ *       addToPendingCount(1);
+ *       (forks = new MapReducer(this, array, mapper, reducer, mid, h, forks)).fork();
+ *       h = mid;
+ *     }
+ *     if (h > l)
+ *       result = mapper.apply(array[l]);
+ *     // process completions by reducing along and advancing subtask links
+ *     for (CountedCompleter<?> c = firstComplete(); c != null; c = c.nextComplete()) {
+ *       for (MapReducer t = (MapReducer)c, s = t.forks;  s != null; s = t.forks = s.next)
+ *         t.result = reducer.apply(t.result, s.result);
+ *     }
+ *   }
+ *   public E getRawResult() { return result; }
+ *
+ *   public static <E> E mapReduce(E[] array, MyMapper<E> mapper, MyReducer<E> reducer) {
+ *     return new MapReducer<E>(null, array, mapper, reducer,
+ *                              0, array.length, null).invoke();
+ *   }
+ * }}</pre>
+ *
+ * <p><b>Triggers.</b> Some CountedCompleters are themselves never
+ * forked, but instead serve as bits of plumbing in other designs;
+ * including those in which the completion of one of more async tasks
+ * triggers another async task. For example:
+ *
+ * <pre> {@code
+ * class HeaderBuilder extends CountedCompleter<...> { ... }
+ * class BodyBuilder extends CountedCompleter<...> { ... }
+ * class PacketSender extends CountedCompleter<...> {
+ *   PacketSender(...) { super(null, 1); ... } // trigger on second completion
+ *   public void compute() { } // never called
+ *   public void onCompletion(CountedCompleter<?> caller) { sendPacket(); }
+ * }
+ * // sample use:
+ * PacketSender p = new PacketSender();
+ * new HeaderBuilder(p, ...).fork();
+ * new BodyBuilder(p, ...).fork();
+ * }</pre>
+ *
+ * @since 1.8
+ * @hide
+ * @author Doug Lea
+ */
+public abstract class CountedCompleter<T> extends ForkJoinTask<T> {
+    private static final long serialVersionUID = 5232453752276485070L;
+
+    /** This task's completer, or null if none */
+    final CountedCompleter<?> completer;
+    /** The number of pending tasks until completion */
+    volatile int pending;
+
+    /**
+     * Creates a new CountedCompleter with the given completer
+     * and initial pending count.
+     *
+     * @param completer this task's completer, or {@code null} if none
+     * @param initialPendingCount the initial pending count
+     */
+    protected CountedCompleter(CountedCompleter<?> completer,
+                               int initialPendingCount) {
+        this.completer = completer;
+        this.pending = initialPendingCount;
+    }
+
+    /**
+     * Creates a new CountedCompleter with the given completer
+     * and an initial pending count of zero.
+     *
+     * @param completer this task's completer, or {@code null} if none
+     */
+    protected CountedCompleter(CountedCompleter<?> completer) {
+        this.completer = completer;
+    }
+
+    /**
+     * Creates a new CountedCompleter with no completer
+     * and an initial pending count of zero.
+     */
+    protected CountedCompleter() {
+        this.completer = null;
+    }
+
+    /**
+     * The main computation performed by this task.
+     */
+    public abstract void compute();
+
+    /**
+     * Performs an action when method {@link #tryComplete} is invoked
+     * and the pending count is zero, or when the unconditional
+     * method {@link #complete} is invoked.  By default, this method
+     * does nothing. You can distinguish cases by checking the
+     * identity of the given caller argument. If not equal to {@code
+     * this}, then it is typically a subtask that may contain results
+     * (and/or links to other results) to combine.
+     *
+     * @param caller the task invoking this method (which may
+     * be this task itself)
+     */
+    public void onCompletion(CountedCompleter<?> caller) {
+    }
+
+    /**
+     * Performs an action when method {@link #completeExceptionally}
+     * is invoked or method {@link #compute} throws an exception, and
+     * this task has not otherwise already completed normally. On
+     * entry to this method, this task {@link
+     * ForkJoinTask#isCompletedAbnormally}.  The return value of this
+     * method controls further propagation: If {@code true} and this
+     * task has a completer, then this completer is also completed
+     * exceptionally.  The default implementation of this method does
+     * nothing except return {@code true}.
+     *
+     * @param ex the exception
+     * @param caller the task invoking this method (which may
+     * be this task itself)
+     * @return true if this exception should be propagated to this
+     * task's completer, if one exists
+     */
+    public boolean onExceptionalCompletion(Throwable ex, CountedCompleter<?> caller) {
+        return true;
+    }
+
+    /**
+     * Returns the completer established in this task's constructor,
+     * or {@code null} if none.
+     *
+     * @return the completer
+     */
+    public final CountedCompleter<?> getCompleter() {
+        return completer;
+    }
+
+    /**
+     * Returns the current pending count.
+     *
+     * @return the current pending count
+     */
+    public final int getPendingCount() {
+        return pending;
+    }
+
+    /**
+     * Sets the pending count to the given value.
+     *
+     * @param count the count
+     */
+    public final void setPendingCount(int count) {
+        pending = count;
+    }
+
+    /**
+     * Adds (atomically) the given value to the pending count.
+     *
+     * @param delta the value to add
+     */
+    public final void addToPendingCount(int delta) {
+        int c; // note: can replace with intrinsic in jdk8
+        do {} while (!U.compareAndSwapInt(this, PENDING, c = pending, c+delta));
+    }
+
+    /**
+     * Sets (atomically) the pending count to the given count only if
+     * it currently holds the given expected value.
+     *
+     * @param expected the expected value
+     * @param count the new value
+     * @return true if successful
+     */
+    public final boolean compareAndSetPendingCount(int expected, int count) {
+        return U.compareAndSwapInt(this, PENDING, expected, count);
+    }
+
+    /**
+     * If the pending count is nonzero, (atomically) decrements it.
+     *
+     * @return the initial (undecremented) pending count holding on entry
+     * to this method
+     */
+    public final int decrementPendingCountUnlessZero() {
+        int c;
+        do {} while ((c = pending) != 0 &&
+                     !U.compareAndSwapInt(this, PENDING, c, c - 1));
+        return c;
+    }
+
+    /**
+     * Returns the root of the current computation; i.e., this
+     * task if it has no completer, else its completer's root.
+     *
+     * @return the root of the current computation
+     */
+    public final CountedCompleter<?> getRoot() {
+        CountedCompleter<?> a = this, p;
+        while ((p = a.completer) != null)
+            a = p;
+        return a;
+    }
+
+    /**
+     * If the pending count is nonzero, decrements the count;
+     * otherwise invokes {@link #onCompletion} and then similarly
+     * tries to complete this task's completer, if one exists,
+     * else marks this task as complete.
+     */
+    public final void tryComplete() {
+        CountedCompleter<?> a = this, s = a;
+        for (int c;;) {
+            if ((c = a.pending) == 0) {
+                a.onCompletion(s);
+                if ((a = (s = a).completer) == null) {
+                    s.quietlyComplete();
+                    return;
+                }
+            }
+            else if (U.compareAndSwapInt(a, PENDING, c, c - 1))
+                return;
+        }
+    }
+
+    /**
+     * Equivalent to {@link #tryComplete} but does not invoke {@link
+     * #onCompletion} along the completion path: If the pending count
+     * is nonzero, decrements the count; otherwise, similarly tries to
+     * complete this task's completer, if one exists, else marks this
+     * task as complete. This method may be useful in cases where
+     * {@code onCompletion} should not, or need not, be invoked for
+     * each completer in a computation.
+     */
+    public final void propagateCompletion() {
+        CountedCompleter<?> a = this, s = a;
+        for (int c;;) {
+            if ((c = a.pending) == 0) {
+                if ((a = (s = a).completer) == null) {
+                    s.quietlyComplete();
+                    return;
+                }
+            }
+            else if (U.compareAndSwapInt(a, PENDING, c, c - 1))
+                return;
+        }
+    }
+
+    /**
+     * Regardless of pending count, invokes {@link #onCompletion},
+     * marks this task as complete and further triggers {@link
+     * #tryComplete} on this task's completer, if one exists.  The
+     * given rawResult is used as an argument to {@link #setRawResult}
+     * before invoking {@link #onCompletion} or marking this task as
+     * complete; its value is meaningful only for classes overriding
+     * {@code setRawResult}.
+     *
+     * <p>This method may be useful when forcing completion as soon as
+     * any one (versus all) of several subtask results are obtained.
+     * However, in the common (and recommended) case in which {@code
+     * setRawResult} is not overridden, this effect can be obtained
+     * more simply using {@code quietlyCompleteRoot();}.
+     *
+     * @param rawResult the raw result
+     */
+    public void complete(T rawResult) {
+        CountedCompleter<?> p;
+        setRawResult(rawResult);
+        onCompletion(this);
+        quietlyComplete();
+        if ((p = completer) != null)
+            p.tryComplete();
+    }
+
+
+    /**
+     * If this task's pending count is zero, returns this task;
+     * otherwise decrements its pending count and returns {@code
+     * null}. This method is designed to be used with {@link
+     * #nextComplete} in completion traversal loops.
+     *
+     * @return this task, if pending count was zero, else {@code null}
+     */
+    public final CountedCompleter<?> firstComplete() {
+        for (int c;;) {
+            if ((c = pending) == 0)
+                return this;
+            else if (U.compareAndSwapInt(this, PENDING, c, c - 1))
+                return null;
+        }
+    }
+
+    /**
+     * If this task does not have a completer, invokes {@link
+     * ForkJoinTask#quietlyComplete} and returns {@code null}.  Or, if
+     * the completer's pending count is non-zero, decrements that
+     * pending count and returns {@code null}.  Otherwise, returns the
+     * completer.  This method can be used as part of a completion
+     * traversal loop for homogeneous task hierarchies:
+     *
+     * <pre> {@code
+     * for (CountedCompleter<?> c = firstComplete();
+     *      c != null;
+     *      c = c.nextComplete()) {
+     *   // ... process c ...
+     * }}</pre>
+     *
+     * @return the completer, or {@code null} if none
+     */
+    public final CountedCompleter<?> nextComplete() {
+        CountedCompleter<?> p;
+        if ((p = completer) != null)
+            return p.firstComplete();
+        else {
+            quietlyComplete();
+            return null;
+        }
+    }
+
+    /**
+     * Equivalent to {@code getRoot().quietlyComplete()}.
+     */
+    public final void quietlyCompleteRoot() {
+        for (CountedCompleter<?> a = this, p;;) {
+            if ((p = a.completer) == null) {
+                a.quietlyComplete();
+                return;
+            }
+            a = p;
+        }
+    }
+
+    /**
+     * Supports ForkJoinTask exception propagation.
+     */
+    void internalPropagateException(Throwable ex) {
+        CountedCompleter<?> a = this, s = a;
+        while (a.onExceptionalCompletion(ex, s) &&
+               (a = (s = a).completer) != null && a.status >= 0)
+            a.recordExceptionalCompletion(ex);
+    }
+
+    /**
+     * Implements execution conventions for CountedCompleters.
+     */
+    protected final boolean exec() {
+        compute();
+        return false;
+    }
+
+    /**
+     * Returns the result of the computation. By default
+     * returns {@code null}, which is appropriate for {@code Void}
+     * actions, but in other cases should be overridden, almost
+     * always to return a field or function of a field that
+     * holds the result upon completion.
+     *
+     * @return the result of the computation
+     */
+    public T getRawResult() { return null; }
+
+    /**
+     * A method that result-bearing CountedCompleters may optionally
+     * use to help maintain result data.  By default, does nothing.
+     * Overrides are not recommended. However, if this method is
+     * overridden to update existing objects or fields, then it must
+     * in general be defined to be thread-safe.
+     */
+    protected void setRawResult(T t) { }
+
+    // Unsafe mechanics
+    private static final sun.misc.Unsafe U;
+    private static final long PENDING;
+    static {
+        try {
+            U = sun.misc.Unsafe.getUnsafe();
+            PENDING = U.objectFieldOffset
+                (CountedCompleter.class.getDeclaredField("pending"));
+        } catch (Exception e) {
+            throw new Error(e);
+        }
+    }
+}
diff --git a/luni/src/main/java/java/util/concurrent/CyclicBarrier.java b/luni/src/main/java/java/util/concurrent/CyclicBarrier.java
index cf0b46e..e1a7bee 100644
--- a/luni/src/main/java/java/util/concurrent/CyclicBarrier.java
+++ b/luni/src/main/java/java/util/concurrent/CyclicBarrier.java
@@ -5,7 +5,8 @@
  */
 
 package java.util.concurrent;
-import java.util.concurrent.locks.*;
+import java.util.concurrent.locks.Condition;
+import java.util.concurrent.locks.ReentrantLock;
 
 /**
  * A synchronization aid that allows a set of threads to all wait for
@@ -15,7 +16,7 @@
  * <em>cyclic</em> because it can be re-used after the waiting threads
  * are released.
  *
- * <p>A <tt>CyclicBarrier</tt> supports an optional {@link Runnable} command
+ * <p>A {@code CyclicBarrier} supports an optional {@link Runnable} command
  * that is run once per barrier point, after the last thread in the party
  * arrives, but before any threads are released.
  * This <em>barrier action</em> is useful
@@ -68,8 +69,8 @@
  * barrier until all rows have been processed. When all rows are processed
  * the supplied {@link Runnable} barrier action is executed and merges the
  * rows. If the merger
- * determines that a solution has been found then <tt>done()</tt> will return
- * <tt>true</tt> and each worker will terminate.
+ * determines that a solution has been found then {@code done()} will return
+ * {@code true} and each worker will terminate.
  *
  * <p>If the barrier action does not rely on the parties being suspended when
  * it is executed, then any of the threads in the party could execute that
@@ -82,7 +83,7 @@
  *   // log the completion of this iteration
  * }}</pre>
  *
- * <p>The <tt>CyclicBarrier</tt> uses an all-or-none breakage model
+ * <p>The {@code CyclicBarrier} uses an all-or-none breakage model
  * for failed synchronization attempts: If a thread leaves a barrier
  * point prematurely because of interruption, failure, or timeout, all
  * other threads waiting at that barrier point will also leave
@@ -109,7 +110,7 @@
      * is reset. There can be many generations associated with threads
      * using the barrier - due to the non-deterministic way the lock
      * may be allocated to waiting threads - but only one of these
-     * can be active at a time (the one to which <tt>count</tt> applies)
+     * can be active at a time (the one to which {@code count} applies)
      * and all the rest are either broken or tripped.
      * There need not be an active generation if there has been a break
      * but no subsequent reset.
@@ -229,7 +230,7 @@
     }
 
     /**
-     * Creates a new <tt>CyclicBarrier</tt> that will trip when the
+     * Creates a new {@code CyclicBarrier} that will trip when the
      * given number of parties (threads) are waiting upon it, and which
      * will execute the given barrier action when the barrier is tripped,
      * performed by the last thread entering the barrier.
@@ -248,7 +249,7 @@
     }
 
     /**
-     * Creates a new <tt>CyclicBarrier</tt> that will trip when the
+     * Creates a new {@code CyclicBarrier} that will trip when the
      * given number of parties (threads) are waiting upon it, and
      * does not perform a predefined action when the barrier is tripped.
      *
@@ -271,7 +272,7 @@
 
     /**
      * Waits until all {@linkplain #getParties parties} have invoked
-     * <tt>await</tt> on this barrier.
+     * {@code await} on this barrier.
      *
      * <p>If the current thread is not the last to arrive then it is
      * disabled for thread scheduling purposes and lies dormant until
@@ -296,7 +297,7 @@
      *
      * <p>If the barrier is {@link #reset} while any thread is waiting,
      * or if the barrier {@linkplain #isBroken is broken} when
-     * <tt>await</tt> is invoked, or while any thread is waiting, then
+     * {@code await} is invoked, or while any thread is waiting, then
      * {@link BrokenBarrierException} is thrown.
      *
      * <p>If any thread is {@linkplain Thread#interrupt interrupted} while waiting,
@@ -313,7 +314,7 @@
      * the broken state.
      *
      * @return the arrival index of the current thread, where index
-     *         <tt>{@link #getParties()} - 1</tt> indicates the first
+     *         {@code getParties() - 1} indicates the first
      *         to arrive and zero indicates the last to arrive
      * @throws InterruptedException if the current thread was interrupted
      *         while waiting
@@ -321,7 +322,7 @@
      *         interrupted or timed out while the current thread was
      *         waiting, or the barrier was reset, or the barrier was
      *         broken when {@code await} was called, or the barrier
-     *         action (if present) failed due an exception.
+     *         action (if present) failed due to an exception
      */
     public int await() throws InterruptedException, BrokenBarrierException {
         try {
@@ -333,7 +334,7 @@
 
     /**
      * Waits until all {@linkplain #getParties parties} have invoked
-     * <tt>await</tt> on this barrier, or the specified waiting time elapses.
+     * {@code await} on this barrier, or the specified waiting time elapses.
      *
      * <p>If the current thread is not the last to arrive then it is
      * disabled for thread scheduling purposes and lies dormant until
@@ -363,7 +364,7 @@
      *
      * <p>If the barrier is {@link #reset} while any thread is waiting,
      * or if the barrier {@linkplain #isBroken is broken} when
-     * <tt>await</tt> is invoked, or while any thread is waiting, then
+     * {@code await} is invoked, or while any thread is waiting, then
      * {@link BrokenBarrierException} is thrown.
      *
      * <p>If any thread is {@linkplain Thread#interrupt interrupted} while
@@ -382,7 +383,7 @@
      * @param timeout the time to wait for the barrier
      * @param unit the time unit of the timeout parameter
      * @return the arrival index of the current thread, where index
-     *         <tt>{@link #getParties()} - 1</tt> indicates the first
+     *         {@code getParties() - 1} indicates the first
      *         to arrive and zero indicates the last to arrive
      * @throws InterruptedException if the current thread was interrupted
      *         while waiting
@@ -391,7 +392,7 @@
      *         interrupted or timed out while the current thread was
      *         waiting, or the barrier was reset, or the barrier was broken
      *         when {@code await} was called, or the barrier action (if
-     *         present) failed due an exception
+     *         present) failed due to an exception
      */
     public int await(long timeout, TimeUnit unit)
         throws InterruptedException,
diff --git a/luni/src/main/java/java/util/concurrent/DelayQueue.java b/luni/src/main/java/java/util/concurrent/DelayQueue.java
index 52028cb..945249e 100644
--- a/luni/src/main/java/java/util/concurrent/DelayQueue.java
+++ b/luni/src/main/java/java/util/concurrent/DelayQueue.java
@@ -4,7 +4,6 @@
  * http://creativecommons.org/publicdomain/zero/1.0/
  */
 
-
 package java.util.concurrent;
 import static java.util.concurrent.TimeUnit.NANOSECONDS;
 import java.util.concurrent.locks.Condition;
@@ -17,15 +16,15 @@
 
 /**
  * An unbounded {@linkplain BlockingQueue blocking queue} of
- * <tt>Delayed</tt> elements, in which an element can only be taken
+ * {@code Delayed} elements, in which an element can only be taken
  * when its delay has expired.  The <em>head</em> of the queue is that
- * <tt>Delayed</tt> element whose delay expired furthest in the
- * past.  If no delay has expired there is no head and <tt>poll</tt>
- * will return <tt>null</tt>. Expiration occurs when an element's
- * <tt>getDelay(TimeUnit.NANOSECONDS)</tt> method returns a value less
+ * {@code Delayed} element whose delay expired furthest in the
+ * past.  If no delay has expired there is no head and {@code poll}
+ * will return {@code null}. Expiration occurs when an element's
+ * {@code getDelay(TimeUnit.NANOSECONDS)} method returns a value less
  * than or equal to zero.  Even though unexpired elements cannot be
- * removed using <tt>take</tt> or <tt>poll</tt>, they are otherwise
- * treated as normal elements. For example, the <tt>size</tt> method
+ * removed using {@code take} or {@code poll}, they are otherwise
+ * treated as normal elements. For example, the {@code size} method
  * returns the count of both expired and unexpired elements.
  * This queue does not permit null elements.
  *
@@ -39,11 +38,10 @@
  * @author Doug Lea
  * @param <E> the type of elements held in this collection
  */
-
 public class DelayQueue<E extends Delayed> extends AbstractQueue<E>
     implements BlockingQueue<E> {
 
-    private transient final ReentrantLock lock = new ReentrantLock();
+    private final transient ReentrantLock lock = new ReentrantLock();
     private final PriorityQueue<E> q = new PriorityQueue<E>();
 
     /**
@@ -72,12 +70,12 @@
     private final Condition available = lock.newCondition();
 
     /**
-     * Creates a new <tt>DelayQueue</tt> that is initially empty.
+     * Creates a new {@code DelayQueue} that is initially empty.
      */
     public DelayQueue() {}
 
     /**
-     * Creates a <tt>DelayQueue</tt> initially containing the elements of the
+     * Creates a {@code DelayQueue} initially containing the elements of the
      * given collection of {@link Delayed} instances.
      *
      * @param c the collection of elements to initially contain
@@ -92,7 +90,7 @@
      * Inserts the specified element into this delay queue.
      *
      * @param e the element to add
-     * @return <tt>true</tt> (as specified by {@link Collection#add})
+     * @return {@code true} (as specified by {@link Collection#add})
      * @throws NullPointerException if the specified element is null
      */
     public boolean add(E e) {
@@ -103,7 +101,7 @@
      * Inserts the specified element into this delay queue.
      *
      * @param e the element to add
-     * @return <tt>true</tt>
+     * @return {@code true}
      * @throws NullPointerException if the specified element is null
      */
     public boolean offer(E e) {
@@ -139,7 +137,7 @@
      * @param e the element to add
      * @param timeout This parameter is ignored as the method never blocks
      * @param unit This parameter is ignored as the method never blocks
-     * @return <tt>true</tt>
+     * @return {@code true}
      * @throws NullPointerException {@inheritDoc}
      */
     public boolean offer(E e, long timeout, TimeUnit unit) {
@@ -147,10 +145,10 @@
     }
 
     /**
-     * Retrieves and removes the head of this queue, or returns <tt>null</tt>
+     * Retrieves and removes the head of this queue, or returns {@code null}
      * if this queue has no elements with an expired delay.
      *
-     * @return the head of this queue, or <tt>null</tt> if this
+     * @return the head of this queue, or {@code null} if this
      *         queue has no elements with an expired delay
      */
     public E poll() {
@@ -186,7 +184,8 @@
                     long delay = first.getDelay(NANOSECONDS);
                     if (delay <= 0)
                         return q.poll();
-                    else if (leader != null)
+                    first = null; // don't retain ref while waiting
+                    if (leader != null)
                         available.await();
                     else {
                         Thread thisThread = Thread.currentThread();
@@ -212,7 +211,7 @@
      * until an element with an expired delay is available on this queue,
      * or the specified wait time expires.
      *
-     * @return the head of this queue, or <tt>null</tt> if the
+     * @return the head of this queue, or {@code null} if the
      *         specified waiting time elapses before an element with
      *         an expired delay becomes available
      * @throws InterruptedException {@inheritDoc}
@@ -235,6 +234,7 @@
                         return q.poll();
                     if (nanos <= 0)
                         return null;
+                    first = null; // don't retain ref while waiting
                     if (nanos < delay || leader != null)
                         nanos = available.awaitNanos(nanos);
                     else {
@@ -259,13 +259,13 @@
 
     /**
      * Retrieves, but does not remove, the head of this queue, or
-     * returns <tt>null</tt> if this queue is empty.  Unlike
-     * <tt>poll</tt>, if no expired elements are available in the queue,
+     * returns {@code null} if this queue is empty.  Unlike
+     * {@code poll}, if no expired elements are available in the queue,
      * this method returns the element that will expire next,
      * if one exists.
      *
-     * @return the head of this queue, or <tt>null</tt> if this
-     *         queue is empty.
+     * @return the head of this queue, or {@code null} if this
+     *         queue is empty
      */
     public E peek() {
         final ReentrantLock lock = this.lock;
@@ -288,7 +288,7 @@
     }
 
     /**
-     * Return first element only if it is expired.
+     * Returns first element only if it is expired.
      * Used only by drainTo.  Call only when holding lock.
      */
     private E peekExpired() {
@@ -369,10 +369,10 @@
     }
 
     /**
-     * Always returns <tt>Integer.MAX_VALUE</tt> because
-     * a <tt>DelayQueue</tt> is not capacity constrained.
+     * Always returns {@code Integer.MAX_VALUE} because
+     * a {@code DelayQueue} is not capacity constrained.
      *
-     * @return <tt>Integer.MAX_VALUE</tt>
+     * @return {@code Integer.MAX_VALUE}
      */
     public int remainingCapacity() {
         return Integer.MAX_VALUE;
@@ -412,7 +412,7 @@
      * <p>If this queue fits in the specified array with room to spare
      * (i.e., the array has more elements than this queue), the element in
      * the array immediately following the end of the queue is set to
-     * <tt>null</tt>.
+     * {@code null}.
      *
      * <p>Like the {@link #toArray()} method, this method acts as bridge between
      * array-based and collection-based APIs.  Further, this method allows
@@ -420,12 +420,12 @@
      * under certain circumstances, be used to save allocation costs.
      *
      * <p>The following code can be used to dump a delay queue into a newly
-     * allocated array of <tt>Delayed</tt>:
+     * allocated array of {@code Delayed}:
      *
      * <pre> {@code Delayed[] a = q.toArray(new Delayed[0]);}</pre>
      *
-     * Note that <tt>toArray(new Object[0])</tt> is identical in function to
-     * <tt>toArray()</tt>.
+     * Note that {@code toArray(new Object[0])} is identical in function to
+     * {@code toArray()}.
      *
      * @param a the array into which the elements of the queue are to
      *          be stored, if it is big enough; otherwise, a new array of the
diff --git a/luni/src/main/java/java/util/concurrent/Delayed.java b/luni/src/main/java/java/util/concurrent/Delayed.java
index 39d927c..6a9527d 100644
--- a/luni/src/main/java/java/util/concurrent/Delayed.java
+++ b/luni/src/main/java/java/util/concurrent/Delayed.java
@@ -11,8 +11,8 @@
  * acted upon after a given delay.
  *
  * <p>An implementation of this interface must define a
- * <tt>compareTo</tt> method that provides an ordering consistent with
- * its <tt>getDelay</tt> method.
+ * {@code compareTo} method that provides an ordering consistent with
+ * its {@code getDelay} method.
  *
  * @since 1.5
  * @author Doug Lea
diff --git a/luni/src/main/java/java/util/concurrent/Exchanger.java b/luni/src/main/java/java/util/concurrent/Exchanger.java
index 6069dce..01d5960f 100644
--- a/luni/src/main/java/java/util/concurrent/Exchanger.java
+++ b/luni/src/main/java/java/util/concurrent/Exchanger.java
@@ -6,7 +6,8 @@
  */
 
 package java.util.concurrent;
-import java.util.concurrent.atomic.*;
+import java.util.concurrent.atomic.AtomicInteger;
+import java.util.concurrent.atomic.AtomicReference;
 import java.util.concurrent.locks.LockSupport;
 
 /**
@@ -73,486 +74,427 @@
  * @param <V> The type of objects that may be exchanged
  */
 public class Exchanger<V> {
+
     /*
-     * Algorithm Description:
+     * Overview: The core algorithm is, for an exchange "slot",
+     * and a participant (caller) with an item:
      *
-     * The basic idea is to maintain a "slot", which is a reference to
-     * a Node containing both an Item to offer and a "hole" waiting to
-     * get filled in.  If an incoming "occupying" thread sees that the
-     * slot is null, it CAS'es (compareAndSets) a Node there and waits
-     * for another to invoke exchange.  That second "fulfilling" thread
-     * sees that the slot is non-null, and so CASes it back to null,
-     * also exchanging items by CASing the hole, plus waking up the
-     * occupying thread if it is blocked.  In each case CAS'es may
-     * fail because a slot at first appears non-null but is null upon
-     * CAS, or vice-versa.  So threads may need to retry these
-     * actions.
+     * for (;;) {
+     *   if (slot is empty) {                       // offer
+     *     place item in a Node;
+     *     if (can CAS slot from empty to node) {
+     *       wait for release;
+     *       return matching item in node;
+     *     }
+     *   }
+     *   else if (can CAS slot from node to empty) { // release
+     *     get the item in node;
+     *     set matching item in node;
+     *     release waiting thread;
+     *   }
+     *   // else retry on CAS failure
+     * }
      *
-     * This simple approach works great when there are only a few
-     * threads using an Exchanger, but performance rapidly
-     * deteriorates due to CAS contention on the single slot when
-     * there are lots of threads using an exchanger.  So instead we use
-     * an "arena"; basically a kind of hash table with a dynamically
-     * varying number of slots, any one of which can be used by
-     * threads performing an exchange.  Incoming threads pick slots
-     * based on a hash of their Thread ids.  If an incoming thread
-     * fails to CAS in its chosen slot, it picks an alternative slot
-     * instead.  And similarly from there.  If a thread successfully
-     * CASes into a slot but no other thread arrives, it tries
-     * another, heading toward the zero slot, which always exists even
-     * if the table shrinks.  The particular mechanics controlling this
-     * are as follows:
+     * This is among the simplest forms of a "dual data structure" --
+     * see Scott and Scherer's DISC 04 paper and
+     * http://www.cs.rochester.edu/research/synchronization/pseudocode/duals.html
      *
-     * Waiting: Slot zero is special in that it is the only slot that
-     * exists when there is no contention.  A thread occupying slot
-     * zero will block if no thread fulfills it after a short spin.
-     * In other cases, occupying threads eventually give up and try
-     * another slot.  Waiting threads spin for a while (a period that
-     * should be a little less than a typical context-switch time)
-     * before either blocking (if slot zero) or giving up (if other
-     * slots) and restarting.  There is no reason for threads to block
-     * unless there are unlikely to be any other threads present.
-     * Occupants are mainly avoiding memory contention so sit there
-     * quietly polling for a shorter period than it would take to
-     * block and then unblock them.  Non-slot-zero waits that elapse
-     * because of lack of other threads waste around one extra
-     * context-switch time per try, which is still on average much
-     * faster than alternative approaches.
+     * This works great in principle. But in practice, like many
+     * algorithms centered on atomic updates to a single location, it
+     * scales horribly when there are more than a few participants
+     * using the same Exchanger. So the implementation instead uses a
+     * form of elimination arena, that spreads out this contention by
+     * arranging that some threads typically use different slots,
+     * while still ensuring that eventually, any two parties will be
+     * able to exchange items. That is, we cannot completely partition
+     * across threads, but instead give threads arena indices that
+     * will on average grow under contention and shrink under lack of
+     * contention. We approach this by defining the Nodes that we need
+     * anyway as ThreadLocals, and include in them per-thread index
+     * and related bookkeeping state. (We can safely reuse per-thread
+     * nodes rather than creating them fresh each time because slots
+     * alternate between pointing to a node vs null, so cannot
+     * encounter ABA problems. However, we do need some care in
+     * resetting them between uses.)
      *
-     * Sizing: Usually, using only a few slots suffices to reduce
-     * contention.  Especially with small numbers of threads, using
-     * too many slots can lead to just as poor performance as using
-     * too few of them, and there's not much room for error.  The
-     * variable "max" maintains the number of slots actually in
-     * use.  It is increased when a thread sees too many CAS
-     * failures.  (This is analogous to resizing a regular hash table
-     * based on a target load factor, except here, growth steps are
-     * just one-by-one rather than proportional.)  Growth requires
-     * contention failures in each of three tried slots.  Requiring
-     * multiple failures for expansion copes with the fact that some
-     * failed CASes are not due to contention but instead to simple
-     * races between two threads or thread pre-emptions occurring
-     * between reading and CASing.  Also, very transient peak
-     * contention can be much higher than the average sustainable
-     * levels.  An attempt to decrease the max limit is usually made
-     * when a non-slot-zero wait elapses without being fulfilled.
-     * Threads experiencing elapsed waits move closer to zero, so
-     * eventually find existing (or future) threads even if the table
-     * has been shrunk due to inactivity.  The chosen mechanics and
-     * thresholds for growing and shrinking are intrinsically
-     * entangled with indexing and hashing inside the exchange code,
-     * and can't be nicely abstracted out.
+     * Implementing an effective arena requires allocating a bunch of
+     * space, so we only do so upon detecting contention (except on
+     * uniprocessors, where they wouldn't help, so aren't used).
+     * Otherwise, exchanges use the single-slot slotExchange method.
+     * On contention, not only must the slots be in different
+     * locations, but the locations must not encounter memory
+     * contention due to being on the same cache line (or more
+     * generally, the same coherence unit).  Because, as of this
+     * writing, there is no way to determine cacheline size, we define
+     * a value that is enough for common platforms.  Additionally,
+     * extra care elsewhere is taken to avoid other false/unintended
+     * sharing and to enhance locality, including adding padding to
+     * Nodes, embedding "bound" as an Exchanger field, and reworking
+     * some park/unpark mechanics compared to LockSupport versions.
      *
-     * Hashing: Each thread picks its initial slot to use in accord
-     * with a simple hashcode.  The sequence is the same on each
-     * encounter by any given thread, but effectively random across
-     * threads.  Using arenas encounters the classic cost vs quality
-     * tradeoffs of all hash tables.  Here, we use a one-step FNV-1a
-     * hash code based on the current thread's Thread.getId(), along
-     * with a cheap approximation to a mod operation to select an
-     * index.  The downside of optimizing index selection in this way
-     * is that the code is hardwired to use a maximum table size of
-     * 32.  But this value more than suffices for known platforms and
-     * applications.
+     * The arena starts out with only one used slot. We expand the
+     * effective arena size by tracking collisions; i.e., failed CASes
+     * while trying to exchange. By nature of the above algorithm, the
+     * only kinds of collision that reliably indicate contention are
+     * when two attempted releases collide -- one of two attempted
+     * offers can legitimately fail to CAS without indicating
+     * contention by more than one other thread. (Note: it is possible
+     * but not worthwhile to more precisely detect contention by
+     * reading slot values after CAS failures.)  When a thread has
+     * collided at each slot within the current arena bound, it tries
+     * to expand the arena size by one. We track collisions within
+     * bounds by using a version (sequence) number on the "bound"
+     * field, and conservatively reset collision counts when a
+     * participant notices that bound has been updated (in either
+     * direction).
      *
-     * Probing: On sensed contention of a selected slot, we probe
-     * sequentially through the table, analogously to linear probing
-     * after collision in a hash table.  (We move circularly, in
-     * reverse order, to mesh best with table growth and shrinkage
-     * rules.)  Except that to minimize the effects of false-alarms
-     * and cache thrashing, we try the first selected slot twice
-     * before moving.
+     * The effective arena size is reduced (when there is more than
+     * one slot) by giving up on waiting after a while and trying to
+     * decrement the arena size on expiration. The value of "a while"
+     * is an empirical matter.  We implement by piggybacking on the
+     * use of spin->yield->block that is essential for reasonable
+     * waiting performance anyway -- in a busy exchanger, offers are
+     * usually almost immediately released, in which case context
+     * switching on multiprocessors is extremely slow/wasteful.  Arena
+     * waits just omit the blocking part, and instead cancel. The spin
+     * count is empirically chosen to be a value that avoids blocking
+     * 99% of the time under maximum sustained exchange rates on a
+     * range of test machines. Spins and yields entail some limited
+     * randomness (using a cheap xorshift) to avoid regular patterns
+     * that can induce unproductive grow/shrink cycles. (Using a
+     * pseudorandom also helps regularize spin cycle duration by
+     * making branches unpredictable.)  Also, during an offer, a
+     * waiter can "know" that it will be released when its slot has
+     * changed, but cannot yet proceed until match is set.  In the
+     * mean time it cannot cancel the offer, so instead spins/yields.
+     * Note: It is possible to avoid this secondary check by changing
+     * the linearization point to be a CAS of the match field (as done
+     * in one case in the Scott & Scherer DISC paper), which also
+     * increases asynchrony a bit, at the expense of poorer collision
+     * detection and inability to always reuse per-thread nodes. So
+     * the current scheme is typically a better tradeoff.
      *
-     * Padding: Even with contention management, slots are heavily
-     * contended, so use cache-padding to avoid poor memory
-     * performance.  Because of this, slots are lazily constructed
-     * only when used, to avoid wasting this space unnecessarily.
-     * While isolation of locations is not much of an issue at first
-     * in an application, as time goes on and garbage-collectors
-     * perform compaction, slots are very likely to be moved adjacent
-     * to each other, which can cause much thrashing of cache lines on
-     * MPs unless padding is employed.
+     * On collisions, indices traverse the arena cyclically in reverse
+     * order, restarting at the maximum index (which will tend to be
+     * sparsest) when bounds change. (On expirations, indices instead
+     * are halved until reaching 0.) It is possible (and has been
+     * tried) to use randomized, prime-value-stepped, or double-hash
+     * style traversal instead of simple cyclic traversal to reduce
+     * bunching.  But empirically, whatever benefits these may have
+     * don't overcome their added overhead: We are managing operations
+     * that occur very quickly unless there is sustained contention,
+     * so simpler/faster control policies work better than more
+     * accurate but slower ones.
      *
-     * This is an improvement of the algorithm described in the paper
-     * "A Scalable Elimination-based Exchange Channel" by William
-     * Scherer, Doug Lea, and Michael Scott in Proceedings of SCOOL05
-     * workshop.  Available at: http://hdl.handle.net/1802/2104
+     * Because we use expiration for arena size control, we cannot
+     * throw TimeoutExceptions in the timed version of the public
+     * exchange method until the arena size has shrunken to zero (or
+     * the arena isn't enabled). This may delay response to timeout
+     * but is still within spec.
+     *
+     * Essentially all of the implementation is in methods
+     * slotExchange and arenaExchange. These have similar overall
+     * structure, but differ in too many details to combine. The
+     * slotExchange method uses the single Exchanger field "slot"
+     * rather than arena array elements. However, it still needs
+     * minimal collision detection to trigger arena construction.
+     * (The messiest part is making sure interrupt status and
+     * InterruptedExceptions come out right during transitions when
+     * both methods may be called. This is done by using null return
+     * as a sentinel to recheck interrupt status.)
+     *
+     * As is too common in this sort of code, methods are monolithic
+     * because most of the logic relies on reads of fields that are
+     * maintained as local variables so can't be nicely factored --
+     * mainly, here, bulky spin->yield->block/cancel code), and
+     * heavily dependent on intrinsics (Unsafe) to use inlined
+     * embedded CAS and related memory access operations (that tend
+     * not to be as readily inlined by dynamic compilers when they are
+     * hidden behind other methods that would more nicely name and
+     * encapsulate the intended effects). This includes the use of
+     * putOrderedX to clear fields of the per-thread Nodes between
+     * uses. Note that field Node.item is not declared as volatile
+     * even though it is read by releasing threads, because they only
+     * do so after CAS operations that must precede access, and all
+     * uses by the owning thread are otherwise acceptably ordered by
+     * other operations. (Because the actual points of atomicity are
+     * slot CASes, it would also be legal for the write to Node.match
+     * in a release to be weaker than a full volatile write. However,
+     * this is not done because it could allow further postponement of
+     * the write, delaying progress.)
      */
 
+    /**
+     * The byte distance (as a shift value) between any two used slots
+     * in the arena.  1 << ASHIFT should be at least cacheline size.
+     */
+    private static final int ASHIFT = 7;
+
+    /**
+     * The maximum supported arena index. The maximum allocatable
+     * arena size is MMASK + 1. Must be a power of two minus one, less
+     * than (1<<(31-ASHIFT)). The cap of 255 (0xff) more than suffices
+     * for the expected scaling limits of the main algorithms.
+     */
+    private static final int MMASK = 0xff;
+
+    /**
+     * Unit for sequence/version bits of bound field. Each successful
+     * change to the bound also adds SEQ.
+     */
+    private static final int SEQ = MMASK + 1;
+
     /** The number of CPUs, for sizing and spin control */
     private static final int NCPU = Runtime.getRuntime().availableProcessors();
 
     /**
-     * The capacity of the arena.  Set to a value that provides more
-     * than enough space to handle contention.  On small machines
-     * most slots won't be used, but it is still not wasted because
-     * the extra space provides some machine-level address padding
-     * to minimize interference with heavily CAS'ed Slot locations.
-     * And on very large machines, performance eventually becomes
-     * bounded by memory bandwidth, not numbers of threads/CPUs.
-     * This constant cannot be changed without also modifying
-     * indexing and hashing algorithms.
+     * The maximum slot index of the arena: The number of slots that
+     * can in principle hold all threads without contention, or at
+     * most the maximum indexable value.
      */
-    private static final int CAPACITY = 32;
+    static final int FULL = (NCPU >= (MMASK << 1)) ? MMASK : NCPU >>> 1;
 
     /**
-     * The value of "max" that will hold all threads without
-     * contention.  When this value is less than CAPACITY, some
-     * otherwise wasted expansion can be avoided.
+     * The bound for spins while waiting for a match. The actual
+     * number of iterations will on average be about twice this value
+     * due to randomization. Note: Spinning is disabled when NCPU==1.
      */
-    private static final int FULL =
-        Math.max(0, Math.min(CAPACITY, NCPU / 2) - 1);
-
-    /**
-     * The number of times to spin (doing nothing except polling a
-     * memory location) before blocking or giving up while waiting to
-     * be fulfilled.  Should be zero on uniprocessors.  On
-     * multiprocessors, this value should be large enough so that two
-     * threads exchanging items as fast as possible block only when
-     * one of them is stalled (due to GC or preemption), but not much
-     * longer, to avoid wasting CPU resources.  Seen differently, this
-     * value is a little over half the number of cycles of an average
-     * context switch time on most systems.  The value here is
-     * approximately the average of those across a range of tested
-     * systems.
-     */
-    private static final int SPINS = (NCPU == 1) ? 0 : 2000;
-
-    /**
-     * The number of times to spin before blocking in timed waits.
-     * Timed waits spin more slowly because checking the time takes
-     * time.  The best value relies mainly on the relative rate of
-     * System.nanoTime vs memory accesses.  The value is empirically
-     * derived to work well across a variety of systems.
-     */
-    private static final int TIMED_SPINS = SPINS / 20;
-
-    /**
-     * Sentinel item representing cancellation of a wait due to
-     * interruption, timeout, or elapsed spin-waits.  This value is
-     * placed in holes on cancellation, and used as a return value
-     * from waiting methods to indicate failure to set or get hole.
-     */
-    private static final Object CANCEL = new Object();
+    private static final int SPINS = 1 << 10;
 
     /**
      * Value representing null arguments/returns from public
-     * methods.  This disambiguates from internal requirement that
-     * holes start out as null to mean they are not yet set.
+     * methods. Needed because the API originally didn't disallow null
+     * arguments, which it should have.
      */
     private static final Object NULL_ITEM = new Object();
 
     /**
-     * Nodes hold partially exchanged data.  This class
-     * opportunistically subclasses AtomicReference to represent the
-     * hole.  So get() returns hole, and compareAndSet CAS'es value
-     * into hole.  This class cannot be parameterized as "V" because
-     * of the use of non-V CANCEL sentinels.
+     * Sentinel value returned by internal exchange methods upon
+     * timeout, to avoid need for separate timed versions of these
+     * methods.
      */
-    private static final class Node extends AtomicReference<Object> {
-        /** The element offered by the Thread creating this node. */
-        public final Object item;
+    private static final Object TIMED_OUT = new Object();
 
-        /** The Thread waiting to be signalled; null until waiting. */
-        public volatile Thread waiter;
+    /**
+     * Nodes hold partially exchanged data, plus other per-thread
+     * bookkeeping.
+     */
+    static final class Node {
+        int index;              // Arena index
+        int bound;              // Last recorded value of Exchanger.bound
+        int collides;           // Number of CAS failures at current bound
+        int hash;               // Pseudo-random for spins
+        Object item;            // This thread's current item
+        volatile Object match;  // Item provided by releasing thread
+        volatile Thread parked; // Set to this thread when parked, else null
 
-        /**
-         * Creates node with given item and empty hole.
-         * @param item the item
-         */
-        public Node(Object item) {
-            this.item = item;
-        }
+        // Padding to ameliorate unfortunate memory placements
+        Object p0, p1, p2, p3, p4, p5, p6, p7, p8, p9, pa, pb, pc, pd, pe, pf;
+        Object q0, q1, q2, q3, q4, q5, q6, q7, q8, q9, qa, qb, qc, qd, qe, qf;
+    }
+
+    /** The corresponding thread local class */
+    static final class Participant extends ThreadLocal<Node> {
+        public Node initialValue() { return new Node(); }
     }
 
     /**
-     * A Slot is an AtomicReference with heuristic padding to lessen
-     * cache effects of this heavily CAS'ed location.  While the
-     * padding adds noticeable space, all slots are created only on
-     * demand, and there will be more than one of them only when it
-     * would improve throughput more than enough to outweigh using
-     * extra space.
+     * Per-thread state
      */
-    private static final class Slot extends AtomicReference<Object> {
-        // Improve likelihood of isolation on <= 128 byte cache lines.
-        // We used to target 64 byte cache lines, but some x86s (including
-        // i7 under some BIOSes) actually use 128 byte cache lines.
-        long q0, q1, q2, q3, q4, q5, q6, q7, q8, q9, qa, qb, qc, qd, qe;
-    }
+    private final Participant participant;
 
     /**
-     * Slot array.  Elements are lazily initialized when needed.
-     * Declared volatile to enable double-checked lazy construction.
+     * Elimination array; null until enabled (within slotExchange).
+     * Element accesses use emulation of volatile gets and CAS.
      */
-    private volatile Slot[] arena = new Slot[CAPACITY];
+    private volatile Node[] arena;
 
     /**
-     * The maximum slot index being used.  The value sometimes
-     * increases when a thread experiences too many CAS contentions,
-     * and sometimes decreases when a spin-wait elapses.  Changes
-     * are performed only via compareAndSet, to avoid stale values
-     * when a thread happens to stall right before setting.
+     * Slot used until contention detected.
      */
-    private final AtomicInteger max = new AtomicInteger();
+    private volatile Node slot;
 
     /**
-     * Main exchange function, handling the different policy variants.
-     * Uses Object, not "V" as argument and return value to simplify
-     * handling of sentinel values.  Callers from public methods decode
-     * and cast accordingly.
+     * The index of the largest valid arena position, OR'ed with SEQ
+     * number in high bits, incremented on each update.  The initial
+     * update from 0 to SEQ is used to ensure that the arena array is
+     * constructed only once.
+     */
+    private volatile int bound;
+
+    /**
+     * Exchange function when arenas enabled. See above for explanation.
      *
      * @param item the (non-null) item to exchange
      * @param timed true if the wait is timed
-     * @param nanos if timed, the maximum wait time
-     * @return the other thread's item, or CANCEL if interrupted or timed out
+     * @param ns if timed, the maximum wait time, else 0L
+     * @return the other thread's item; or null if interrupted; or
+     * TIMED_OUT if timed and timed out
      */
-    private Object doExchange(Object item, boolean timed, long nanos) {
-        Node me = new Node(item);                 // Create in case occupying
-        int index = hashIndex();                  // Index of current slot
-        int fails = 0;                            // Number of CAS failures
-
-        for (;;) {
-            Object y;                             // Contents of current slot
-            Slot slot = arena[index];
-            if (slot == null)                     // Lazily initialize slots
-                createSlot(index);                // Continue loop to reread
-            else if ((y = slot.get()) != null &&  // Try to fulfill
-                     slot.compareAndSet(y, null)) {
-                Node you = (Node)y;               // Transfer item
-                if (you.compareAndSet(null, item)) {
-                    LockSupport.unpark(you.waiter);
-                    return you.item;
-                }                                 // Else cancelled; continue
-            }
-            else if (y == null &&                 // Try to occupy
-                     slot.compareAndSet(null, me)) {
-                if (index == 0)                   // Blocking wait for slot 0
-                    return timed ?
-                        awaitNanos(me, slot, nanos) :
-                        await(me, slot);
-                Object v = spinWait(me, slot);    // Spin wait for non-0
-                if (v != CANCEL)
-                    return v;
-                me = new Node(item);              // Throw away cancelled node
-                int m = max.get();
-                if (m > (index >>>= 1))           // Decrease index
-                    max.compareAndSet(m, m - 1);  // Maybe shrink table
-            }
-            else if (++fails > 1) {               // Allow 2 fails on 1st slot
-                int m = max.get();
-                if (fails > 3 && m < FULL && max.compareAndSet(m, m + 1))
-                    index = m + 1;                // Grow on 3rd failed slot
-                else if (--index < 0)
-                    index = m;                    // Circularly traverse
-            }
-        }
-    }
-
-    /**
-     * Returns a hash index for the current thread.  Uses a one-step
-     * FNV-1a hash code (http://www.isthe.com/chongo/tech/comp/fnv/)
-     * based on the current thread's Thread.getId().  These hash codes
-     * have more uniform distribution properties with respect to small
-     * moduli (here 1-31) than do other simple hashing functions.
-     *
-     * <p>To return an index between 0 and max, we use a cheap
-     * approximation to a mod operation, that also corrects for bias
-     * due to non-power-of-2 remaindering (see {@link
-     * java.util.Random#nextInt}).  Bits of the hashcode are masked
-     * with "nbits", the ceiling power of two of table size (looked up
-     * in a table packed into three ints).  If too large, this is
-     * retried after rotating the hash by nbits bits, while forcing new
-     * top bit to 0, which guarantees eventual termination (although
-     * with a non-random-bias).  This requires an average of less than
-     * 2 tries for all table sizes, and has a maximum 2% difference
-     * from perfectly uniform slot probabilities when applied to all
-     * possible hash codes for sizes less than 32.
-     *
-     * @return a per-thread-random index, 0 <= index < max
-     */
-    private final int hashIndex() {
-        long id = Thread.currentThread().getId();
-        int hash = (((int)(id ^ (id >>> 32))) ^ 0x811c9dc5) * 0x01000193;
-
-        int m = max.get();
-        int nbits = (((0xfffffc00  >> m) & 4) | // Compute ceil(log2(m+1))
-                     ((0x000001f8 >>> m) & 2) | // The constants hold
-                     ((0xffff00f2 >>> m) & 1)); // a lookup table
-        int index;
-        while ((index = hash & ((1 << nbits) - 1)) > m)       // May retry on
-            hash = (hash >>> nbits) | (hash << (33 - nbits)); // non-power-2 m
-        return index;
-    }
-
-    /**
-     * Creates a new slot at given index.  Called only when the slot
-     * appears to be null.  Relies on double-check using builtin
-     * locks, since they rarely contend.  This in turn relies on the
-     * arena array being declared volatile.
-     *
-     * @param index the index to add slot at
-     */
-    private void createSlot(int index) {
-        // Create slot outside of lock to narrow sync region
-        Slot newSlot = new Slot();
-        Slot[] a = arena;
-        synchronized (a) {
-            if (a[index] == null)
-                a[index] = newSlot;
-        }
-    }
-
-    /**
-     * Tries to cancel a wait for the given node waiting in the given
-     * slot, if so, helping clear the node from its slot to avoid
-     * garbage retention.
-     *
-     * @param node the waiting node
-     * @param slot the slot it is waiting in
-     * @return true if successfully cancelled
-     */
-    private static boolean tryCancel(Node node, Slot slot) {
-        if (!node.compareAndSet(null, CANCEL))
-            return false;
-        if (slot.get() == node) // pre-check to minimize contention
-            slot.compareAndSet(node, null);
-        return true;
-    }
-
-    // Three forms of waiting. Each just different enough not to merge
-    // code with others.
-
-    /**
-     * Spin-waits for hole for a non-0 slot.  Fails if spin elapses
-     * before hole filled.  Does not check interrupt, relying on check
-     * in public exchange method to abort if interrupted on entry.
-     *
-     * @param node the waiting node
-     * @return on success, the hole; on failure, CANCEL
-     */
-    private static Object spinWait(Node node, Slot slot) {
-        int spins = SPINS;
-        for (;;) {
-            Object v = node.get();
-            if (v != null)
+    private final Object arenaExchange(Object item, boolean timed, long ns) {
+        Node[] a = arena;
+        Node p = participant.get();
+        for (int i = p.index;;) {                      // access slot at i
+            int b, m, c; long j;                       // j is raw array offset
+            Node q = (Node)U.getObjectVolatile(a, j = (i << ASHIFT) + ABASE);
+            if (q != null && U.compareAndSwapObject(a, j, q, null)) {
+                Object v = q.item;                     // release
+                q.match = item;
+                Thread w = q.parked;
+                if (w != null)
+                    U.unpark(w);
                 return v;
-            else if (spins > 0)
-                --spins;
-            else
-                tryCancel(node, slot);
-        }
-    }
-
-    /**
-     * Waits for (by spinning and/or blocking) and gets the hole
-     * filled in by another thread.  Fails if interrupted before
-     * hole filled.
-     *
-     * When a node/thread is about to block, it sets its waiter field
-     * and then rechecks state at least one more time before actually
-     * parking, thus covering race vs fulfiller noticing that waiter
-     * is non-null so should be woken.
-     *
-     * Thread interruption status is checked only surrounding calls to
-     * park.  The caller is assumed to have checked interrupt status
-     * on entry.
-     *
-     * @param node the waiting node
-     * @return on success, the hole; on failure, CANCEL
-     */
-    private static Object await(Node node, Slot slot) {
-        Thread w = Thread.currentThread();
-        int spins = SPINS;
-        for (;;) {
-            Object v = node.get();
-            if (v != null)
-                return v;
-            else if (spins > 0)                 // Spin-wait phase
-                --spins;
-            else if (node.waiter == null)       // Set up to block next
-                node.waiter = w;
-            else if (w.isInterrupted())         // Abort on interrupt
-                tryCancel(node, slot);
-            else                                // Block
-                LockSupport.park(node);
-        }
-    }
-
-    /**
-     * Waits for (at index 0) and gets the hole filled in by another
-     * thread.  Fails if timed out or interrupted before hole filled.
-     * Same basic logic as untimed version, but a bit messier.
-     *
-     * @param node the waiting node
-     * @param nanos the wait time
-     * @return on success, the hole; on failure, CANCEL
-     */
-    private Object awaitNanos(Node node, Slot slot, long nanos) {
-        int spins = TIMED_SPINS;
-        long lastTime = 0;
-        Thread w = null;
-        for (;;) {
-            Object v = node.get();
-            if (v != null)
-                return v;
-            long now = System.nanoTime();
-            if (w == null)
-                w = Thread.currentThread();
-            else
-                nanos -= now - lastTime;
-            lastTime = now;
-            if (nanos > 0) {
-                if (spins > 0)
-                    --spins;
-                else if (node.waiter == null)
-                    node.waiter = w;
-                else if (w.isInterrupted())
-                    tryCancel(node, slot);
-                else
-                    LockSupport.parkNanos(node, nanos);
             }
-            else if (tryCancel(node, slot) && !w.isInterrupted())
-                return scanOnTimeout(node);
-        }
-    }
-
-    /**
-     * Sweeps through arena checking for any waiting threads.  Called
-     * only upon return from timeout while waiting in slot 0.  When a
-     * thread gives up on a timed wait, it is possible that a
-     * previously-entered thread is still waiting in some other
-     * slot.  So we scan to check for any.  This is almost always
-     * overkill, but decreases the likelihood of timeouts when there
-     * are other threads present to far less than that in lock-based
-     * exchangers in which earlier-arriving threads may still be
-     * waiting on entry locks.
-     *
-     * @param node the waiting node
-     * @return another thread's item, or CANCEL
-     */
-    private Object scanOnTimeout(Node node) {
-        Object y;
-        for (int j = arena.length - 1; j >= 0; --j) {
-            Slot slot = arena[j];
-            if (slot != null) {
-                while ((y = slot.get()) != null) {
-                    if (slot.compareAndSet(y, null)) {
-                        Node you = (Node)y;
-                        if (you.compareAndSet(null, node.item)) {
-                            LockSupport.unpark(you.waiter);
-                            return you.item;
+            else if (i <= (m = (b = bound) & MMASK) && q == null) {
+                p.item = item;                         // offer
+                if (U.compareAndSwapObject(a, j, null, p)) {
+                    long end = (timed && m == 0) ? System.nanoTime() + ns : 0L;
+                    Thread t = Thread.currentThread(); // wait
+                    for (int h = p.hash, spins = SPINS;;) {
+                        Object v = p.match;
+                        if (v != null) {
+                            U.putOrderedObject(p, MATCH, null);
+                            p.item = null;             // clear for next use
+                            p.hash = h;
+                            return v;
+                        }
+                        else if (spins > 0) {
+                            h ^= h << 1; h ^= h >>> 3; h ^= h << 10; // xorshift
+                            if (h == 0)                // initialize hash
+                                h = SPINS | (int)t.getId();
+                            else if (h < 0 &&          // approx 50% true
+                                     (--spins & ((SPINS >>> 1) - 1)) == 0)
+                                Thread.yield();        // two yields per wait
+                        }
+                        else if (U.getObjectVolatile(a, j) != p)
+                            spins = SPINS;       // releaser hasn't set match yet
+                        else if (!t.isInterrupted() && m == 0 &&
+                                 (!timed ||
+                                  (ns = end - System.nanoTime()) > 0L)) {
+                            U.putObject(t, BLOCKER, this); // emulate LockSupport
+                            p.parked = t;              // minimize window
+                            if (U.getObjectVolatile(a, j) == p)
+                                U.park(false, ns);
+                            p.parked = null;
+                            U.putObject(t, BLOCKER, null);
+                        }
+                        else if (U.getObjectVolatile(a, j) == p &&
+                                 U.compareAndSwapObject(a, j, p, null)) {
+                            if (m != 0)                // try to shrink
+                                U.compareAndSwapInt(this, BOUND, b, b + SEQ - 1);
+                            p.item = null;
+                            p.hash = h;
+                            i = p.index >>>= 1;        // descend
+                            if (Thread.interrupted())
+                                return null;
+                            if (timed && m == 0 && ns <= 0L)
+                                return TIMED_OUT;
+                            break;                     // expired; restart
                         }
                     }
                 }
+                else
+                    p.item = null;                     // clear offer
+            }
+            else {
+                if (p.bound != b) {                    // stale; reset
+                    p.bound = b;
+                    p.collides = 0;
+                    i = (i != m || m == 0) ? m : m - 1;
+                }
+                else if ((c = p.collides) < m || m == FULL ||
+                         !U.compareAndSwapInt(this, BOUND, b, b + SEQ + 1)) {
+                    p.collides = c + 1;
+                    i = (i == 0) ? m : i - 1;          // cyclically traverse
+                }
+                else
+                    i = m + 1;                         // grow
+                p.index = i;
             }
         }
-        return CANCEL;
+    }
+
+    /**
+     * Exchange function used until arenas enabled. See above for explanation.
+     *
+     * @param item the item to exchange
+     * @param timed true if the wait is timed
+     * @param ns if timed, the maximum wait time, else 0L
+     * @return the other thread's item; or null if either the arena
+     * was enabled or the thread was interrupted before completion; or
+     * TIMED_OUT if timed and timed out
+     */
+    private final Object slotExchange(Object item, boolean timed, long ns) {
+        Node p = participant.get();
+        Thread t = Thread.currentThread();
+        if (t.isInterrupted()) // preserve interrupt status so caller can recheck
+            return null;
+
+        for (Node q;;) {
+            if ((q = slot) != null) {
+                if (U.compareAndSwapObject(this, SLOT, q, null)) {
+                    Object v = q.item;
+                    q.match = item;
+                    Thread w = q.parked;
+                    if (w != null)
+                        U.unpark(w);
+                    return v;
+                }
+                // create arena on contention, but continue until slot null
+                if (NCPU > 1 && bound == 0 &&
+                    U.compareAndSwapInt(this, BOUND, 0, SEQ))
+                    arena = new Node[(FULL + 2) << ASHIFT];
+            }
+            else if (arena != null)
+                return null; // caller must reroute to arenaExchange
+            else {
+                p.item = item;
+                if (U.compareAndSwapObject(this, SLOT, null, p))
+                    break;
+                p.item = null;
+            }
+        }
+
+        // await release
+        int h = p.hash;
+        long end = timed ? System.nanoTime() + ns : 0L;
+        int spins = (NCPU > 1) ? SPINS : 1;
+        Object v;
+        while ((v = p.match) == null) {
+            if (spins > 0) {
+                h ^= h << 1; h ^= h >>> 3; h ^= h << 10;
+                if (h == 0)
+                    h = SPINS | (int)t.getId();
+                else if (h < 0 && (--spins & ((SPINS >>> 1) - 1)) == 0)
+                    Thread.yield();
+            }
+            else if (slot != p)
+                spins = SPINS;
+            else if (!t.isInterrupted() && arena == null &&
+                     (!timed || (ns = end - System.nanoTime()) > 0L)) {
+                U.putObject(t, BLOCKER, this);
+                p.parked = t;
+                if (slot == p)
+                    U.park(false, ns);
+                p.parked = null;
+                U.putObject(t, BLOCKER, null);
+            }
+            else if (U.compareAndSwapObject(this, SLOT, p, null)) {
+                v = timed && ns <= 0L && !t.isInterrupted() ? TIMED_OUT : null;
+                break;
+            }
+        }
+        U.putOrderedObject(p, MATCH, null);
+        p.item = null;
+        p.hash = h;
+        return v;
     }
 
     /**
      * Creates a new Exchanger.
      */
     public Exchanger() {
+        participant = new Participant();
     }
 
     /**
@@ -588,16 +530,16 @@
      * @throws InterruptedException if the current thread was
      *         interrupted while waiting
      */
+    @SuppressWarnings("unchecked")
     public V exchange(V x) throws InterruptedException {
-        if (!Thread.interrupted()) {
-            Object v = doExchange((x == null) ? NULL_ITEM : x, false, 0);
-            if (v == NULL_ITEM)
-                return null;
-            if (v != CANCEL)
-                return (V)v;
-            Thread.interrupted(); // Clear interrupt status on IE throw
-        }
-        throw new InterruptedException();
+        Object v;
+        Object item = (x == null) ? NULL_ITEM : x; // translate null args
+        if ((arena != null ||
+             (v = slotExchange(item, false, 0L)) == null) &&
+            ((Thread.interrupted() || // disambiguates null return
+              (v = arenaExchange(item, false, 0L)) == null)))
+            throw new InterruptedException();
+        return (v == NULL_ITEM) ? null : (V)v;
     }
 
     /**
@@ -635,25 +577,61 @@
      *
      * @param x the object to exchange
      * @param timeout the maximum time to wait
-     * @param unit the time unit of the <tt>timeout</tt> argument
+     * @param unit the time unit of the {@code timeout} argument
      * @return the object provided by the other thread
      * @throws InterruptedException if the current thread was
      *         interrupted while waiting
      * @throws TimeoutException if the specified waiting time elapses
      *         before another thread enters the exchange
      */
+    @SuppressWarnings("unchecked")
     public V exchange(V x, long timeout, TimeUnit unit)
         throws InterruptedException, TimeoutException {
-        if (!Thread.interrupted()) {
-            Object v = doExchange((x == null) ? NULL_ITEM : x,
-                                  true, unit.toNanos(timeout));
-            if (v == NULL_ITEM)
-                return null;
-            if (v != CANCEL)
-                return (V)v;
-            if (!Thread.interrupted())
-                throw new TimeoutException();
-        }
-        throw new InterruptedException();
+        Object v;
+        Object item = (x == null) ? NULL_ITEM : x;
+        long ns = unit.toNanos(timeout);
+        if ((arena != null ||
+             (v = slotExchange(item, true, ns)) == null) &&
+            ((Thread.interrupted() ||
+              (v = arenaExchange(item, true, ns)) == null)))
+            throw new InterruptedException();
+        if (v == TIMED_OUT)
+            throw new TimeoutException();
+        return (v == NULL_ITEM) ? null : (V)v;
     }
+
+    // Unsafe mechanics
+    private static final sun.misc.Unsafe U;
+    private static final long BOUND;
+    private static final long SLOT;
+    private static final long MATCH;
+    private static final long BLOCKER;
+    private static final int ABASE;
+    static {
+        int s;
+        try {
+            U = sun.misc.Unsafe.getUnsafe();
+            Class<?> ek = Exchanger.class;
+            Class<?> nk = Node.class;
+            Class<?> ak = Node[].class;
+            Class<?> tk = Thread.class;
+            BOUND = U.objectFieldOffset
+                (ek.getDeclaredField("bound"));
+            SLOT = U.objectFieldOffset
+                (ek.getDeclaredField("slot"));
+            MATCH = U.objectFieldOffset
+                (nk.getDeclaredField("match"));
+            BLOCKER = U.objectFieldOffset
+                (tk.getDeclaredField("parkBlocker"));
+            s = U.arrayIndexScale(ak);
+            // ABASE absorbs padding in front of element 0
+            ABASE = U.arrayBaseOffset(ak) + (1 << ASHIFT);
+
+        } catch (Exception e) {
+            throw new Error(e);
+        }
+        if ((s & (s-1)) != 0 || s > (1 << ASHIFT))
+            throw new Error("Unsupported array scale");
+    }
+
 }
diff --git a/luni/src/main/java/java/util/concurrent/ExecutionException.java b/luni/src/main/java/java/util/concurrent/ExecutionException.java
index 9bb8dee..dbfbe65 100644
--- a/luni/src/main/java/java/util/concurrent/ExecutionException.java
+++ b/luni/src/main/java/java/util/concurrent/ExecutionException.java
@@ -19,14 +19,14 @@
     private static final long serialVersionUID = 7830266012832686185L;
 
     /**
-     * Constructs an <tt>ExecutionException</tt> with no detail message.
+     * Constructs an {@code ExecutionException} with no detail message.
      * The cause is not initialized, and may subsequently be
      * initialized by a call to {@link #initCause(Throwable) initCause}.
      */
     protected ExecutionException() { }
 
     /**
-     * Constructs an <tt>ExecutionException</tt> with the specified detail
+     * Constructs an {@code ExecutionException} with the specified detail
      * message. The cause is not initialized, and may subsequently be
      * initialized by a call to {@link #initCause(Throwable) initCause}.
      *
@@ -37,7 +37,7 @@
     }
 
     /**
-     * Constructs an <tt>ExecutionException</tt> with the specified detail
+     * Constructs an {@code ExecutionException} with the specified detail
      * message and cause.
      *
      * @param  message the detail message
@@ -49,10 +49,10 @@
     }
 
     /**
-     * Constructs an <tt>ExecutionException</tt> with the specified cause.
+     * Constructs an {@code ExecutionException} with the specified cause.
      * The detail message is set to {@code (cause == null ? null :
      * cause.toString())} (which typically contains the class and
-     * detail message of <tt>cause</tt>).
+     * detail message of {@code cause}).
      *
      * @param  cause the cause (which is saved for later retrieval by the
      *         {@link #getCause()} method)
diff --git a/luni/src/main/java/java/util/concurrent/Executor.java b/luni/src/main/java/java/util/concurrent/Executor.java
index 831bf46..f55209a 100644
--- a/luni/src/main/java/java/util/concurrent/Executor.java
+++ b/luni/src/main/java/java/util/concurrent/Executor.java
@@ -10,9 +10,9 @@
  * An object that executes submitted {@link Runnable} tasks. This
  * interface provides a way of decoupling task submission from the
  * mechanics of how each task will be run, including details of thread
- * use, scheduling, etc.  An <tt>Executor</tt> is normally used
+ * use, scheduling, etc.  An {@code Executor} is normally used
  * instead of explicitly creating threads. For example, rather than
- * invoking <tt>new Thread(new(RunnableTask())).start()</tt> for each
+ * invoking {@code new Thread(new(RunnableTask())).start()} for each
  * of a set of tasks, you might use:
  *
  * <pre>
@@ -22,7 +22,7 @@
  * ...
  * </pre>
  *
- * However, the <tt>Executor</tt> interface does not strictly
+ * However, the {@code Executor} interface does not strictly
  * require that execution be asynchronous. In the simplest case, an
  * executor can run the submitted task immediately in the caller's
  * thread:
@@ -45,7 +45,7 @@
  *   }
  * }}</pre>
  *
- * Many <tt>Executor</tt> implementations impose some sort of
+ * Many {@code Executor} implementations impose some sort of
  * limitation on how and when tasks are scheduled.  The executor below
  * serializes the submission of tasks to a second executor,
  * illustrating a composite executor.
@@ -82,7 +82,7 @@
  *   }
  * }}</pre>
  *
- * The <tt>Executor</tt> implementations provided in this package
+ * The {@code Executor} implementations provided in this package
  * implement {@link ExecutorService}, which is a more extensive
  * interface.  The {@link ThreadPoolExecutor} class provides an
  * extensible thread pool implementation. The {@link Executors} class
@@ -101,11 +101,11 @@
     /**
      * Executes the given command at some time in the future.  The command
      * may execute in a new thread, in a pooled thread, or in the calling
-     * thread, at the discretion of the <tt>Executor</tt> implementation.
+     * thread, at the discretion of the {@code Executor} implementation.
      *
      * @param command the runnable task
      * @throws RejectedExecutionException if this task cannot be
-     * accepted for execution.
+     * accepted for execution
      * @throws NullPointerException if command is null
      */
     void execute(Runnable command);
diff --git a/luni/src/main/java/java/util/concurrent/ExecutorService.java b/luni/src/main/java/java/util/concurrent/ExecutorService.java
index a33ceec..4599f59 100644
--- a/luni/src/main/java/java/util/concurrent/ExecutorService.java
+++ b/luni/src/main/java/java/util/concurrent/ExecutorService.java
@@ -17,21 +17,21 @@
  * methods that can produce a {@link Future} for tracking progress of
  * one or more asynchronous tasks.
  *
- * <p> An <tt>ExecutorService</tt> can be shut down, which will cause
+ * <p>An {@code ExecutorService} can be shut down, which will cause
  * it to reject new tasks.  Two different methods are provided for
- * shutting down an <tt>ExecutorService</tt>. The {@link #shutdown}
+ * shutting down an {@code ExecutorService}. The {@link #shutdown}
  * method will allow previously submitted tasks to execute before
  * terminating, while the {@link #shutdownNow} method prevents waiting
  * tasks from starting and attempts to stop currently executing tasks.
  * Upon termination, an executor has no tasks actively executing, no
  * tasks awaiting execution, and no new tasks can be submitted.  An
- * unused <tt>ExecutorService</tt> should be shut down to allow
+ * unused {@code ExecutorService} should be shut down to allow
  * reclamation of its resources.
  *
- * <p> Method <tt>submit</tt> extends base method {@link
+ * <p>Method {@code submit} extends base method {@link
  * Executor#execute} by creating and returning a {@link Future} that
  * can be used to cancel execution and/or wait for completion.
- * Methods <tt>invokeAny</tt> and <tt>invokeAll</tt> perform the most
+ * Methods {@code invokeAny} and {@code invokeAll} perform the most
  * commonly useful forms of bulk execution, executing a collection of
  * tasks and then waiting for at least one, or all, to
  * complete. (Class {@link ExecutorCompletionService} can be used to
@@ -76,9 +76,9 @@
  *   }
  * }}</pre>
  *
- * The following method shuts down an <tt>ExecutorService</tt> in two phases,
- * first by calling <tt>shutdown</tt> to reject incoming tasks, and then
- * calling <tt>shutdownNow</tt>, if necessary, to cancel any lingering tasks:
+ * The following method shuts down an {@code ExecutorService} in two phases,
+ * first by calling {@code shutdown} to reject incoming tasks, and then
+ * calling {@code shutdownNow}, if necessary, to cancel any lingering tasks:
  *
  *  <pre> {@code
  * void shutdownAndAwaitTermination(ExecutorService pool) {
@@ -141,18 +141,18 @@
     List<Runnable> shutdownNow();
 
     /**
-     * Returns <tt>true</tt> if this executor has been shut down.
+     * Returns {@code true} if this executor has been shut down.
      *
-     * @return <tt>true</tt> if this executor has been shut down
+     * @return {@code true} if this executor has been shut down
      */
     boolean isShutdown();
 
     /**
-     * Returns <tt>true</tt> if all tasks have completed following shut down.
-     * Note that <tt>isTerminated</tt> is never <tt>true</tt> unless
-     * either <tt>shutdown</tt> or <tt>shutdownNow</tt> was called first.
+     * Returns {@code true} if all tasks have completed following shut down.
+     * Note that {@code isTerminated} is never {@code true} unless
+     * either {@code shutdown} or {@code shutdownNow} was called first.
      *
-     * @return <tt>true</tt> if all tasks have completed following shut down
+     * @return {@code true} if all tasks have completed following shut down
      */
     boolean isTerminated();
 
@@ -163,26 +163,25 @@
      *
      * @param timeout the maximum time to wait
      * @param unit the time unit of the timeout argument
-     * @return <tt>true</tt> if this executor terminated and
-     *         <tt>false</tt> if the timeout elapsed before termination
+     * @return {@code true} if this executor terminated and
+     *         {@code false} if the timeout elapsed before termination
      * @throws InterruptedException if interrupted while waiting
      */
     boolean awaitTermination(long timeout, TimeUnit unit)
         throws InterruptedException;
 
-
     /**
      * Submits a value-returning task for execution and returns a
      * Future representing the pending results of the task. The
-     * Future's <tt>get</tt> method will return the task's result upon
+     * Future's {@code get} method will return the task's result upon
      * successful completion.
      *
      * <p>
      * If you would like to immediately block waiting
      * for a task, you can use constructions of the form
-     * <tt>result = exec.submit(aCallable).get();</tt>
+     * {@code result = exec.submit(aCallable).get();}
      *
-     * <p> Note: The {@link Executors} class includes a set of methods
+     * <p>Note: The {@link Executors} class includes a set of methods
      * that can convert some other common closure-like objects,
      * for example, {@link java.security.PrivilegedAction} to
      * {@link Callable} form so they can be submitted.
@@ -197,7 +196,7 @@
 
     /**
      * Submits a Runnable task for execution and returns a Future
-     * representing that task. The Future's <tt>get</tt> method will
+     * representing that task. The Future's {@code get} method will
      * return the given result upon successful completion.
      *
      * @param task the task to submit
@@ -211,8 +210,8 @@
 
     /**
      * Submits a Runnable task for execution and returns a Future
-     * representing that task. The Future's <tt>get</tt> method will
-     * return <tt>null</tt> upon <em>successful</em> completion.
+     * representing that task. The Future's {@code get} method will
+     * return {@code null} upon <em>successful</em> completion.
      *
      * @param task the task to submit
      * @return a Future representing pending completion of the task
@@ -225,7 +224,7 @@
     /**
      * Executes the given tasks, returning a list of Futures holding
      * their status and results when all complete.
-     * {@link Future#isDone} is <tt>true</tt> for each
+     * {@link Future#isDone} is {@code true} for each
      * element of the returned list.
      * Note that a <em>completed</em> task could have
      * terminated either normally or by throwing an exception.
@@ -233,16 +232,15 @@
      * collection is modified while this operation is in progress.
      *
      * @param tasks the collection of tasks
-     * @return A list of Futures representing the tasks, in the same
+     * @return a list of Futures representing the tasks, in the same
      *         sequential order as produced by the iterator for the
-     *         given task list, each of which has completed.
+     *         given task list, each of which has completed
      * @throws InterruptedException if interrupted while waiting, in
-     *         which case unfinished tasks are cancelled.
-     * @throws NullPointerException if tasks or any of its elements are <tt>null</tt>
+     *         which case unfinished tasks are cancelled
+     * @throws NullPointerException if tasks or any of its elements are {@code null}
      * @throws RejectedExecutionException if any task cannot be
      *         scheduled for execution
      */
-
     <T> List<Future<T>> invokeAll(Collection<? extends Callable<T>> tasks)
         throws InterruptedException;
 
@@ -250,7 +248,7 @@
      * Executes the given tasks, returning a list of Futures holding
      * their status and results
      * when all complete or the timeout expires, whichever happens first.
-     * {@link Future#isDone} is <tt>true</tt> for each
+     * {@link Future#isDone} is {@code true} for each
      * element of the returned list.
      * Upon return, tasks that have not completed are cancelled.
      * Note that a <em>completed</em> task could have
@@ -269,7 +267,7 @@
      * @throws InterruptedException if interrupted while waiting, in
      *         which case unfinished tasks are cancelled
      * @throws NullPointerException if tasks, any of its elements, or
-     *         unit are <tt>null</tt>
+     *         unit are {@code null}
      * @throws RejectedExecutionException if any task cannot be scheduled
      *         for execution
      */
@@ -289,7 +287,7 @@
      * @return the result returned by one of the tasks
      * @throws InterruptedException if interrupted while waiting
      * @throws NullPointerException if tasks or any element task
-     *         subject to execution is <tt>null</tt>
+     *         subject to execution is {@code null}
      * @throws IllegalArgumentException if tasks is empty
      * @throws ExecutionException if no task successfully completes
      * @throws RejectedExecutionException if tasks cannot be scheduled
@@ -310,10 +308,10 @@
      * @param tasks the collection of tasks
      * @param timeout the maximum time to wait
      * @param unit the time unit of the timeout argument
-     * @return the result returned by one of the tasks.
+     * @return the result returned by one of the tasks
      * @throws InterruptedException if interrupted while waiting
      * @throws NullPointerException if tasks, or unit, or any element
-     *         task subject to execution is <tt>null</tt>
+     *         task subject to execution is {@code null}
      * @throws TimeoutException if the given timeout elapses before
      *         any task successfully completes
      * @throws ExecutionException if no task successfully completes
diff --git a/luni/src/main/java/java/util/concurrent/Executors.java b/luni/src/main/java/java/util/concurrent/Executors.java
index b4f03ba..53c68fc 100644
--- a/luni/src/main/java/java/util/concurrent/Executors.java
+++ b/luni/src/main/java/java/util/concurrent/Executors.java
@@ -34,7 +34,7 @@
  *        that sets newly created threads to a known state.
  *   <li> Methods that create and return a {@link Callable}
  *        out of other closure-like forms, so they can be used
- *        in execution methods requiring <tt>Callable</tt>.
+ *        in execution methods requiring {@code Callable}.
  * </ul>
  *
  * @since 1.5
@@ -45,7 +45,7 @@
     /**
      * Creates a thread pool that reuses a fixed number of threads
      * operating off a shared unbounded queue.  At any point, at most
-     * <tt>nThreads</tt> threads will be active processing tasks.
+     * {@code nThreads} threads will be active processing tasks.
      * If additional tasks are submitted when all threads are active,
      * they will wait in the queue until a thread is available.
      * If any thread terminates due to a failure during execution
@@ -64,10 +64,48 @@
     }
 
     /**
+     * Creates a thread pool that maintains enough threads to support
+     * the given parallelism level, and may use multiple queues to
+     * reduce contention. The parallelism level corresponds to the
+     * maximum number of threads actively engaged in, or available to
+     * engage in, task processing. The actual number of threads may
+     * grow and shrink dynamically. A work-stealing pool makes no
+     * guarantees about the order in which submitted tasks are
+     * executed.
+     *
+     * @param parallelism the targeted parallelism level
+     * @return the newly created thread pool
+     * @throws IllegalArgumentException if {@code parallelism <= 0}
+     * @since 1.8
+     * @hide
+     */
+    public static ExecutorService newWorkStealingPool(int parallelism) {
+        return new ForkJoinPool
+            (parallelism,
+             ForkJoinPool.defaultForkJoinWorkerThreadFactory,
+             null, true);
+    }
+
+    /**
+     * Creates a work-stealing thread pool using all
+     * {@link Runtime#availableProcessors available processors}
+     * as its target parallelism level.
+     * @return the newly created thread pool
+     * @since 1.8
+     * @hide
+     */
+    public static ExecutorService newWorkStealingPool() {
+        return new ForkJoinPool
+            (Runtime.getRuntime().availableProcessors(),
+             ForkJoinPool.defaultForkJoinWorkerThreadFactory,
+             null, true);
+    }
+
+    /**
      * Creates a thread pool that reuses a fixed number of threads
      * operating off a shared unbounded queue, using the provided
      * ThreadFactory to create new threads when needed.  At any point,
-     * at most <tt>nThreads</tt> threads will be active processing
+     * at most {@code nThreads} threads will be active processing
      * tasks.  If additional tasks are submitted when all threads are
      * active, they will wait in the queue until a thread is
      * available.  If any thread terminates due to a failure during
@@ -97,7 +135,7 @@
      * subsequent tasks.)  Tasks are guaranteed to execute
      * sequentially, and no more than one task will be active at any
      * given time. Unlike the otherwise equivalent
-     * <tt>newFixedThreadPool(1)</tt> the returned executor is
+     * {@code newFixedThreadPool(1)} the returned executor is
      * guaranteed not to be reconfigurable to use additional threads.
      *
      * @return the newly created single-threaded Executor
@@ -113,7 +151,7 @@
      * Creates an Executor that uses a single worker thread operating
      * off an unbounded queue, and uses the provided ThreadFactory to
      * create a new thread when needed. Unlike the otherwise
-     * equivalent <tt>newFixedThreadPool(1, threadFactory)</tt> the
+     * equivalent {@code newFixedThreadPool(1, threadFactory)} the
      * returned executor is guaranteed not to be reconfigurable to use
      * additional threads.
      *
@@ -136,7 +174,7 @@
      * will reuse previously constructed threads when they are
      * available.  These pools will typically improve the performance
      * of programs that execute many short-lived asynchronous tasks.
-     * Calls to <tt>execute</tt> will reuse previously constructed
+     * Calls to {@code execute} will reuse previously constructed
      * threads if available. If no existing thread is available, a new
      * thread will be created and added to the pool. Threads that have
      * not been used for sixty seconds are terminated and removed from
@@ -178,7 +216,7 @@
      * subsequent tasks.)  Tasks are guaranteed to execute
      * sequentially, and no more than one task will be active at any
      * given time. Unlike the otherwise equivalent
-     * <tt>newScheduledThreadPool(1)</tt> the returned executor is
+     * {@code newScheduledThreadPool(1)} the returned executor is
      * guaranteed not to be reconfigurable to use additional threads.
      * @return the newly created scheduled executor
      */
@@ -195,7 +233,7 @@
      * place if needed to execute subsequent tasks.)  Tasks are
      * guaranteed to execute sequentially, and no more than one task
      * will be active at any given time. Unlike the otherwise
-     * equivalent <tt>newScheduledThreadPool(1, threadFactory)</tt>
+     * equivalent {@code newScheduledThreadPool(1, threadFactory)}
      * the returned executor is guaranteed not to be reconfigurable to
      * use additional threads.
      * @param threadFactory the factory to use when creating new
@@ -212,7 +250,7 @@
      * Creates a thread pool that can schedule commands to run after a
      * given delay, or to execute periodically.
      * @param corePoolSize the number of threads to keep in the pool,
-     * even if they are idle.
+     * even if they are idle
      * @return a newly created scheduled thread pool
      * @throws IllegalArgumentException if {@code corePoolSize < 0}
      */
@@ -224,9 +262,9 @@
      * Creates a thread pool that can schedule commands to run after a
      * given delay, or to execute periodically.
      * @param corePoolSize the number of threads to keep in the pool,
-     * even if they are idle.
+     * even if they are idle
      * @param threadFactory the factory to use when the executor
-     * creates a new thread.
+     * creates a new thread
      * @return a newly created scheduled thread pool
      * @throws IllegalArgumentException if {@code corePoolSize < 0}
      * @throws NullPointerException if threadFactory is null
@@ -236,7 +274,6 @@
         return new ScheduledThreadPoolExecutor(corePoolSize, threadFactory);
     }
 
-
     /**
      * Returns an object that delegates all defined {@link
      * ExecutorService} methods to the given executor, but not any
@@ -244,7 +281,7 @@
      * casts. This provides a way to safely "freeze" configuration and
      * disallow tuning of a given concrete implementation.
      * @param executor the underlying implementation
-     * @return an <tt>ExecutorService</tt> instance
+     * @return an {@code ExecutorService} instance
      * @throws NullPointerException if executor null
      */
     public static ExecutorService unconfigurableExecutorService(ExecutorService executor) {
@@ -260,7 +297,7 @@
      * casts. This provides a way to safely "freeze" configuration and
      * disallow tuning of a given concrete implementation.
      * @param executor the underlying implementation
-     * @return a <tt>ScheduledExecutorService</tt> instance
+     * @return a {@code ScheduledExecutorService} instance
      * @throws NullPointerException if executor null
      */
     public static ScheduledExecutorService unconfigurableScheduledExecutorService(ScheduledExecutorService executor) {
@@ -274,7 +311,7 @@
      * This factory creates all new threads used by an Executor in the
      * same {@link ThreadGroup}. Each new
      * thread is created as a non-daemon thread with priority set to
-     * the smaller of <tt>Thread.NORM_PRIORITY</tt> and the maximum
+     * the smaller of {@code Thread.NORM_PRIORITY} and the maximum
      * priority permitted in the thread group.  New threads have names
      * accessible via {@link Thread#getName} of
      * <em>pool-N-thread-M</em>, where <em>N</em> is the sequence
@@ -297,7 +334,7 @@
      * Returns a {@link Callable} object that, when
      * called, runs the given task and returns the given result.  This
      * can be useful when applying methods requiring a
-     * <tt>Callable</tt> to an otherwise resultless action.
+     * {@code Callable} to an otherwise resultless action.
      * @param task the task to run
      * @param result the result to return
      * @return a callable object
@@ -311,7 +348,7 @@
 
     /**
      * Returns a {@link Callable} object that, when
-     * called, runs the given task and returns <tt>null</tt>.
+     * called, runs the given task and returns {@code null}.
      * @param task the task to run
      * @return a callable object
      * @throws NullPointerException if task null
@@ -361,12 +398,7 @@
     }
 
     /**
-     * Returns a {@link Callable} object that will, when
-     * called, execute the given <tt>callable</tt> under the current
-     * with the current context class loader as the context class loader.
-     *
-     * @return a callable object
-     * @throws NullPointerException if callable null
+     * Legacy security code; do not use.
      */
     public static <T> Callable<T> privilegedCallableUsingCurrentClassLoader(Callable<T> callable) {
         if (callable == null)
@@ -451,18 +483,17 @@
                 return AccessController.doPrivileged(
                     new PrivilegedExceptionAction<T>() {
                         public T run() throws Exception {
-                            ClassLoader savedcl = null;
                             Thread t = Thread.currentThread();
-                            try {
-                                ClassLoader cl = t.getContextClassLoader();
-                                if (ccl != cl) {
-                                    t.setContextClassLoader(ccl);
-                                    savedcl = cl;
-                                }
+                            ClassLoader cl = t.getContextClassLoader();
+                            if (ccl == cl) {
                                 return task.call();
-                            } finally {
-                                if (savedcl != null)
-                                    t.setContextClassLoader(savedcl);
+                            } else {
+                                t.setContextClassLoader(ccl);
+                                try {
+                                    return task.call();
+                                } finally {
+                                    t.setContextClassLoader(cl);
+                                }
                             }
                         }
                     }, acc);
@@ -609,21 +640,20 @@
             super(executor);
             e = executor;
         }
-        public ScheduledFuture<?> schedule(Runnable command, long delay,  TimeUnit unit) {
+        public ScheduledFuture<?> schedule(Runnable command, long delay, TimeUnit unit) {
             return e.schedule(command, delay, unit);
         }
         public <V> ScheduledFuture<V> schedule(Callable<V> callable, long delay, TimeUnit unit) {
             return e.schedule(callable, delay, unit);
         }
-        public ScheduledFuture<?> scheduleAtFixedRate(Runnable command, long initialDelay,  long period, TimeUnit unit) {
+        public ScheduledFuture<?> scheduleAtFixedRate(Runnable command, long initialDelay, long period, TimeUnit unit) {
             return e.scheduleAtFixedRate(command, initialDelay, period, unit);
         }
-        public ScheduledFuture<?> scheduleWithFixedDelay(Runnable command, long initialDelay,  long delay, TimeUnit unit) {
+        public ScheduledFuture<?> scheduleWithFixedDelay(Runnable command, long initialDelay, long delay, TimeUnit unit) {
             return e.scheduleWithFixedDelay(command, initialDelay, delay, unit);
         }
     }
 
-
     /** Cannot instantiate. */
     private Executors() {}
 }
diff --git a/luni/src/main/java/java/util/concurrent/ForkJoinPool.java b/luni/src/main/java/java/util/concurrent/ForkJoinPool.java
index ee15ac8..87ffff3 100644
--- a/luni/src/main/java/java/util/concurrent/ForkJoinPool.java
+++ b/luni/src/main/java/java/util/concurrent/ForkJoinPool.java
@@ -11,7 +11,6 @@
 import java.util.Collection;
 import java.util.Collections;
 import java.util.List;
-import java.util.Random;
 import java.util.concurrent.AbstractExecutorService;
 import java.util.concurrent.Callable;
 import java.util.concurrent.ExecutorService;
@@ -19,15 +18,6 @@
 import java.util.concurrent.RejectedExecutionException;
 import java.util.concurrent.RunnableFuture;
 import java.util.concurrent.TimeUnit;
-import java.util.concurrent.atomic.AtomicInteger;
-import java.util.concurrent.locks.LockSupport;
-import java.util.concurrent.locks.ReentrantLock;
-import java.util.concurrent.locks.Condition;
-import libcore.util.SneakyThrow;
-
-// BEGIN android-note
-// removed security manager docs
-// END android-note
 
 /**
  * An {@link ExecutorService} for running {@link ForkJoinTask}s.
@@ -38,21 +28,31 @@
  * <p>A {@code ForkJoinPool} differs from other kinds of {@link
  * ExecutorService} mainly by virtue of employing
  * <em>work-stealing</em>: all threads in the pool attempt to find and
- * execute subtasks created by other active tasks (eventually blocking
- * waiting for work if none exist). This enables efficient processing
- * when most tasks spawn other subtasks (as do most {@code
- * ForkJoinTask}s). When setting <em>asyncMode</em> to true in
- * constructors, {@code ForkJoinPool}s may also be appropriate for use
- * with event-style tasks that are never joined.
+ * execute tasks submitted to the pool and/or created by other active
+ * tasks (eventually blocking waiting for work if none exist). This
+ * enables efficient processing when most tasks spawn other subtasks
+ * (as do most {@code ForkJoinTask}s), as well as when many small
+ * tasks are submitted to the pool from external clients.  Especially
+ * when setting <em>asyncMode</em> to true in constructors, {@code
+ * ForkJoinPool}s may also be appropriate for use with event-style
+ * tasks that are never joined.
  *
- * <p>A {@code ForkJoinPool} is constructed with a given target
- * parallelism level; by default, equal to the number of available
- * processors. The pool attempts to maintain enough active (or
- * available) threads by dynamically adding, suspending, or resuming
- * internal worker threads, even if some tasks are stalled waiting to
- * join others. However, no such adjustments are guaranteed in the
- * face of blocked IO or other unmanaged synchronization. The nested
- * {@link ManagedBlocker} interface enables extension of the kinds of
+ * <p>A static {@link #commonPool()} is available and appropriate for
+ * most applications. The common pool is used by any ForkJoinTask that
+ * is not explicitly submitted to a specified pool. Using the common
+ * pool normally reduces resource usage (its threads are slowly
+ * reclaimed during periods of non-use, and reinstated upon subsequent
+ * use).
+ *
+ * <p>For applications that require separate or custom pools, a {@code
+ * ForkJoinPool} may be constructed with a given target parallelism
+ * level; by default, equal to the number of available processors. The
+ * pool attempts to maintain enough active (or available) threads by
+ * dynamically adding, suspending, or resuming internal worker
+ * threads, even if some tasks are stalled waiting to join
+ * others. However, no such adjustments are guaranteed in the face of
+ * blocked I/O or other unmanaged synchronization. The nested {@link
+ * ManagedBlocker} interface enables extension of the kinds of
  * synchronization accommodated.
  *
  * <p>In addition to execution and lifecycle control methods, this
@@ -62,16 +62,17 @@
  * {@link #toString} returns indications of pool state in a
  * convenient form for informal monitoring.
  *
- * <p> As is the case with other ExecutorServices, there are three
- * main task execution methods summarized in the following
- * table. These are designed to be used by clients not already engaged
- * in fork/join computations in the current pool.  The main forms of
- * these methods accept instances of {@code ForkJoinTask}, but
- * overloaded forms also allow mixed execution of plain {@code
+ * <p>As is the case with other ExecutorServices, there are three
+ * main task execution methods summarized in the following table.
+ * These are designed to be used primarily by clients not already
+ * engaged in fork/join computations in the current pool.  The main
+ * forms of these methods accept instances of {@code ForkJoinTask},
+ * but overloaded forms also allow mixed execution of plain {@code
  * Runnable}- or {@code Callable}- based activities as well.  However,
- * tasks that are already executing in a pool should normally
- * <em>NOT</em> use these pool execution methods, but instead use the
- * within-computation forms listed in the table.
+ * tasks that are already executing in a pool should normally instead
+ * use the within-computation forms listed in the table unless using
+ * async event-style tasks that are not usually joined, in which case
+ * there is little difference among choice of methods.
  *
  * <table BORDER CELLPADDING=3 CELLSPACING=1>
  *  <tr>
@@ -96,22 +97,16 @@
  *  </tr>
  * </table>
  *
- * <p><b>Sample Usage.</b> Normally a single {@code ForkJoinPool} is
- * used for all parallel task execution in a program or subsystem.
- * Otherwise, use would not usually outweigh the construction and
- * bookkeeping overhead of creating a large set of threads. For
- * example, a common pool could be used for the {@code SortTasks}
- * illustrated in {@link RecursiveAction}. Because {@code
- * ForkJoinPool} uses threads in {@linkplain java.lang.Thread#isDaemon
- * daemon} mode, there is typically no need to explicitly {@link
- * #shutdown} such a pool upon program exit.
- *
- *  <pre> {@code
- * static final ForkJoinPool mainPool = new ForkJoinPool();
- * ...
- * public void sort(long[] array) {
- *   mainPool.invoke(new SortTask(array, 0, array.length));
- * }}</pre>
+ * <p>The common pool is by default constructed with default
+ * parameters, but these may be controlled by setting three {@link
+ * System#getProperty system properties} with prefix {@code
+ * java.util.concurrent.ForkJoinPool.common}: {@code parallelism} --
+ * an integer greater than zero, {@code threadFactory} -- the class
+ * name of a {@link ForkJoinWorkerThreadFactory}, and {@code
+ * exceptionHandler} -- the class name of a {@link
+ * java.lang.Thread.UncaughtExceptionHandler
+ * Thread.UncaughtExceptionHandler}. Upon any error in establishing
+ * these settings, default parameters are used.
  *
  * <p><b>Implementation notes</b>: This implementation restricts the
  * maximum number of running threads to 32767. Attempts to create
@@ -131,214 +126,388 @@
     /*
      * Implementation Overview
      *
-     * This class provides the central bookkeeping and control for a
-     * set of worker threads: Submissions from non-FJ threads enter
-     * into a submission queue. Workers take these tasks and typically
-     * split them into subtasks that may be stolen by other workers.
-     * Preference rules give first priority to processing tasks from
-     * their own queues (LIFO or FIFO, depending on mode), then to
-     * randomized FIFO steals of tasks in other worker queues, and
-     * lastly to new submissions.
+     * This class and its nested classes provide the main
+     * functionality and control for a set of worker threads:
+     * Submissions from non-FJ threads enter into submission queues.
+     * Workers take these tasks and typically split them into subtasks
+     * that may be stolen by other workers.  Preference rules give
+     * first priority to processing tasks from their own queues (LIFO
+     * or FIFO, depending on mode), then to randomized FIFO steals of
+     * tasks in other queues.
+     *
+     * WorkQueues
+     * ==========
+     *
+     * Most operations occur within work-stealing queues (in nested
+     * class WorkQueue).  These are special forms of Deques that
+     * support only three of the four possible end-operations -- push,
+     * pop, and poll (aka steal), under the further constraints that
+     * push and pop are called only from the owning thread (or, as
+     * extended here, under a lock), while poll may be called from
+     * other threads.  (If you are unfamiliar with them, you probably
+     * want to read Herlihy and Shavit's book "The Art of
+     * Multiprocessor programming", chapter 16 describing these in
+     * more detail before proceeding.)  The main work-stealing queue
+     * design is roughly similar to those in the papers "Dynamic
+     * Circular Work-Stealing Deque" by Chase and Lev, SPAA 2005
+     * (http://research.sun.com/scalable/pubs/index.html) and
+     * "Idempotent work stealing" by Michael, Saraswat, and Vechev,
+     * PPoPP 2009 (http://portal.acm.org/citation.cfm?id=1504186).
+     * The main differences ultimately stem from GC requirements that
+     * we null out taken slots as soon as we can, to maintain as small
+     * a footprint as possible even in programs generating huge
+     * numbers of tasks. To accomplish this, we shift the CAS
+     * arbitrating pop vs poll (steal) from being on the indices
+     * ("base" and "top") to the slots themselves.  So, both a
+     * successful pop and poll mainly entail a CAS of a slot from
+     * non-null to null.  Because we rely on CASes of references, we
+     * do not need tag bits on base or top.  They are simple ints as
+     * used in any circular array-based queue (see for example
+     * ArrayDeque).  Updates to the indices must still be ordered in a
+     * way that guarantees that top == base means the queue is empty,
+     * but otherwise may err on the side of possibly making the queue
+     * appear nonempty when a push, pop, or poll have not fully
+     * committed. Note that this means that the poll operation,
+     * considered individually, is not wait-free. One thief cannot
+     * successfully continue until another in-progress one (or, if
+     * previously empty, a push) completes.  However, in the
+     * aggregate, we ensure at least probabilistic non-blockingness.
+     * If an attempted steal fails, a thief always chooses a different
+     * random victim target to try next. So, in order for one thief to
+     * progress, it suffices for any in-progress poll or new push on
+     * any empty queue to complete. (This is why we normally use
+     * method pollAt and its variants that try once at the apparent
+     * base index, else consider alternative actions, rather than
+     * method poll.)
+     *
+     * This approach also enables support of a user mode in which local
+     * task processing is in FIFO, not LIFO order, simply by using
+     * poll rather than pop.  This can be useful in message-passing
+     * frameworks in which tasks are never joined.  However neither
+     * mode considers affinities, loads, cache localities, etc, so
+     * rarely provide the best possible performance on a given
+     * machine, but portably provide good throughput by averaging over
+     * these factors.  (Further, even if we did try to use such
+     * information, we do not usually have a basis for exploiting it.
+     * For example, some sets of tasks profit from cache affinities,
+     * but others are harmed by cache pollution effects.)
+     *
+     * WorkQueues are also used in a similar way for tasks submitted
+     * to the pool. We cannot mix these tasks in the same queues used
+     * for work-stealing (this would contaminate lifo/fifo
+     * processing). Instead, we randomly associate submission queues
+     * with submitting threads, using a form of hashing.  The
+     * ThreadLocal Submitter class contains a value initially used as
+     * a hash code for choosing existing queues, but may be randomly
+     * repositioned upon contention with other submitters.  In
+     * essence, submitters act like workers except that they are
+     * restricted to executing local tasks that they submitted (or in
+     * the case of CountedCompleters, others with the same root task).
+     * However, because most shared/external queue operations are more
+     * expensive than internal, and because, at steady state, external
+     * submitters will compete for CPU with workers, ForkJoinTask.join
+     * and related methods disable them from repeatedly helping to
+     * process tasks if all workers are active.  Insertion of tasks in
+     * shared mode requires a lock (mainly to protect in the case of
+     * resizing) but we use only a simple spinlock (using bits in
+     * field qlock), because submitters encountering a busy queue move
+     * on to try or create other queues -- they block only when
+     * creating and registering new queues.
+     *
+     * Management
+     * ==========
      *
      * The main throughput advantages of work-stealing stem from
      * decentralized control -- workers mostly take tasks from
      * themselves or each other. We cannot negate this in the
      * implementation of other management responsibilities. The main
      * tactic for avoiding bottlenecks is packing nearly all
-     * essentially atomic control state into a single 64bit volatile
-     * variable ("ctl"). This variable is read on the order of 10-100
-     * times as often as it is modified (always via CAS). (There is
-     * some additional control state, for example variable "shutdown"
-     * for which we can cope with uncoordinated updates.)  This
-     * streamlines synchronization and control at the expense of messy
-     * constructions needed to repack status bits upon updates.
-     * Updates tend not to contend with each other except during
-     * bursts while submitted tasks begin or end.  In some cases when
-     * they do contend, threads can instead do something else
-     * (usually, scan for tasks) until contention subsides.
+     * essentially atomic control state into two volatile variables
+     * that are by far most often read (not written) as status and
+     * consistency checks.
      *
-     * To enable packing, we restrict maximum parallelism to (1<<15)-1
-     * (which is far in excess of normal operating range) to allow
-     * ids, counts, and their negations (used for thresholding) to fit
-     * into 16bit fields.
+     * Field "ctl" contains 64 bits holding all the information needed
+     * to atomically decide to add, inactivate, enqueue (on an event
+     * queue), dequeue, and/or re-activate workers.  To enable this
+     * packing, we restrict maximum parallelism to (1<<15)-1 (which is
+     * far in excess of normal operating range) to allow ids, counts,
+     * and their negations (used for thresholding) to fit into 16bit
+     * fields.
      *
-     * Recording Workers.  Workers are recorded in the "workers" array
-     * that is created upon pool construction and expanded if (rarely)
-     * necessary.  This is an array as opposed to some other data
-     * structure to support index-based random steals by workers.
-     * Updates to the array recording new workers and unrecording
-     * terminated ones are protected from each other by a seqLock
-     * (scanGuard) but the array is otherwise concurrently readable,
-     * and accessed directly by workers. To simplify index-based
-     * operations, the array size is always a power of two, and all
-     * readers must tolerate null slots. To avoid flailing during
-     * start-up, the array is presized to hold twice #parallelism
-     * workers (which is unlikely to need further resizing during
-     * execution). But to avoid dealing with so many null slots,
-     * variable scanGuard includes a mask for the nearest power of two
-     * that contains all current workers.  All worker thread creation
-     * is on-demand, triggered by task submissions, replacement of
-     * terminated workers, and/or compensation for blocked
-     * workers. However, all other support code is set up to work with
-     * other policies.  To ensure that we do not hold on to worker
-     * references that would prevent GC, ALL accesses to workers are
-     * via indices into the workers array (which is one source of some
-     * of the messy code constructions here). In essence, the workers
-     * array serves as a weak reference mechanism. Thus for example
-     * the wait queue field of ctl stores worker indices, not worker
-     * references.  Access to the workers in associated methods (for
-     * example signalWork) must both index-check and null-check the
-     * IDs. All such accesses ignore bad IDs by returning out early
-     * from what they are doing, since this can only be associated
-     * with termination, in which case it is OK to give up.
+     * Field "plock" is a form of sequence lock with a saturating
+     * shutdown bit (similarly for per-queue "qlocks"), mainly
+     * protecting updates to the workQueues array, as well as to
+     * enable shutdown.  When used as a lock, it is normally only very
+     * briefly held, so is nearly always available after at most a
+     * brief spin, but we use a monitor-based backup strategy to
+     * block when needed.
      *
-     * All uses of the workers array, as well as queue arrays, check
-     * that the array is non-null (even if previously non-null). This
-     * allows nulling during termination, which is currently not
-     * necessary, but remains an option for resource-revocation-based
-     * shutdown schemes.
+     * Recording WorkQueues.  WorkQueues are recorded in the
+     * "workQueues" array that is created upon first use and expanded
+     * if necessary.  Updates to the array while recording new workers
+     * and unrecording terminated ones are protected from each other
+     * by a lock but the array is otherwise concurrently readable, and
+     * accessed directly.  To simplify index-based operations, the
+     * array size is always a power of two, and all readers must
+     * tolerate null slots. Worker queues are at odd indices. Shared
+     * (submission) queues are at even indices, up to a maximum of 64
+     * slots, to limit growth even if array needs to expand to add
+     * more workers. Grouping them together in this way simplifies and
+     * speeds up task scanning.
      *
-     * Wait Queuing. Unlike HPC work-stealing frameworks, we cannot
+     * All worker thread creation is on-demand, triggered by task
+     * submissions, replacement of terminated workers, and/or
+     * compensation for blocked workers. However, all other support
+     * code is set up to work with other policies.  To ensure that we
+     * do not hold on to worker references that would prevent GC, ALL
+     * accesses to workQueues are via indices into the workQueues
+     * array (which is one source of some of the messy code
+     * constructions here). In essence, the workQueues array serves as
+     * a weak reference mechanism. Thus for example the wait queue
+     * field of ctl stores indices, not references.  Access to the
+     * workQueues in associated methods (for example signalWork) must
+     * both index-check and null-check the IDs. All such accesses
+     * ignore bad IDs by returning out early from what they are doing,
+     * since this can only be associated with termination, in which
+     * case it is OK to give up.  All uses of the workQueues array
+     * also check that it is non-null (even if previously
+     * non-null). This allows nulling during termination, which is
+     * currently not necessary, but remains an option for
+     * resource-revocation-based shutdown schemes. It also helps
+     * reduce JIT issuance of uncommon-trap code, which tends to
+     * unnecessarily complicate control flow in some methods.
+     *
+     * Event Queuing. Unlike HPC work-stealing frameworks, we cannot
      * let workers spin indefinitely scanning for tasks when none can
      * be found immediately, and we cannot start/resume workers unless
      * there appear to be tasks available.  On the other hand, we must
      * quickly prod them into action when new tasks are submitted or
-     * generated.  We park/unpark workers after placing in an event
-     * wait queue when they cannot find work. This "queue" is actually
-     * a simple Treiber stack, headed by the "id" field of ctl, plus a
-     * 15bit counter value to both wake up waiters (by advancing their
-     * count) and avoid ABA effects. Successors are held in worker
-     * field "nextWait".  Queuing deals with several intrinsic races,
-     * mainly that a task-producing thread can miss seeing (and
+     * generated. In many usages, ramp-up time to activate workers is
+     * the main limiting factor in overall performance (this is
+     * compounded at program start-up by JIT compilation and
+     * allocation). So we try to streamline this as much as possible.
+     * We park/unpark workers after placing in an event wait queue
+     * when they cannot find work. This "queue" is actually a simple
+     * Treiber stack, headed by the "id" field of ctl, plus a 15bit
+     * counter value (that reflects the number of times a worker has
+     * been inactivated) to avoid ABA effects (we need only as many
+     * version numbers as worker threads). Successors are held in
+     * field WorkQueue.nextWait.  Queuing deals with several intrinsic
+     * races, mainly that a task-producing thread can miss seeing (and
      * signalling) another thread that gave up looking for work but
      * has not yet entered the wait queue. We solve this by requiring
-     * a full sweep of all workers both before (in scan()) and after
-     * (in tryAwaitWork()) a newly waiting worker is added to the wait
-     * queue. During a rescan, the worker might release some other
-     * queued worker rather than itself, which has the same net
-     * effect. Because enqueued workers may actually be rescanning
-     * rather than waiting, we set and clear the "parked" field of
-     * ForkJoinWorkerThread to reduce unnecessary calls to unpark.
-     * (Use of the parked field requires a secondary recheck to avoid
-     * missed signals.)
+     * a full sweep of all workers (via repeated calls to method
+     * scan()) both before and after a newly waiting worker is added
+     * to the wait queue. During a rescan, the worker might release
+     * some other queued worker rather than itself, which has the same
+     * net effect. Because enqueued workers may actually be rescanning
+     * rather than waiting, we set and clear the "parker" field of
+     * WorkQueues to reduce unnecessary calls to unpark.  (This
+     * requires a secondary recheck to avoid missed signals.)  Note
+     * the unusual conventions about Thread.interrupts surrounding
+     * parking and other blocking: Because interrupts are used solely
+     * to alert threads to check termination, which is checked anyway
+     * upon blocking, we clear status (using Thread.interrupted)
+     * before any call to park, so that park does not immediately
+     * return due to status being set via some other unrelated call to
+     * interrupt in user code.
      *
      * Signalling.  We create or wake up workers only when there
      * appears to be at least one task they might be able to find and
-     * execute.  When a submission is added or another worker adds a
-     * task to a queue that previously had two or fewer tasks, they
-     * signal waiting workers (or trigger creation of new ones if
-     * fewer than the given parallelism level -- see signalWork).
-     * These primary signals are buttressed by signals during rescans
-     * as well as those performed when a worker steals a task and
-     * notices that there are more tasks too; together these cover the
-     * signals needed in cases when more than two tasks are pushed
-     * but untaken.
+     * execute. However, many other threads may notice the same task
+     * and each signal to wake up a thread that might take it. So in
+     * general, pools will be over-signalled.  When a submission is
+     * added or another worker adds a task to a queue that has fewer
+     * than two tasks, they signal waiting workers (or trigger
+     * creation of new ones if fewer than the given parallelism level
+     * -- signalWork), and may leave a hint to the unparked worker to
+     * help signal others upon wakeup).  These primary signals are
+     * buttressed by others (see method helpSignal) whenever other
+     * threads scan for work or do not have a task to process.  On
+     * most platforms, signalling (unpark) overhead time is noticeably
+     * long, and the time between signalling a thread and it actually
+     * making progress can be very noticeably long, so it is worth
+     * offloading these delays from critical paths as much as
+     * possible.
      *
      * Trimming workers. To release resources after periods of lack of
      * use, a worker starting to wait when the pool is quiescent will
-     * time out and terminate if the pool has remained quiescent for
-     * SHRINK_RATE nanosecs. This will slowly propagate, eventually
-     * terminating all workers after long periods of non-use.
+     * time out and terminate if the pool has remained quiescent for a
+     * given period -- a short period if there are more threads than
+     * parallelism, longer as the number of threads decreases. This
+     * will slowly propagate, eventually terminating all workers after
+     * periods of non-use.
      *
-     * Submissions. External submissions are maintained in an
-     * array-based queue that is structured identically to
-     * ForkJoinWorkerThread queues except for the use of
-     * submissionLock in method addSubmission. Unlike the case for
-     * worker queues, multiple external threads can add new
-     * submissions, so adding requires a lock.
+     * Shutdown and Termination. A call to shutdownNow atomically sets
+     * a plock bit and then (non-atomically) sets each worker's
+     * qlock status, cancels all unprocessed tasks, and wakes up
+     * all waiting workers.  Detecting whether termination should
+     * commence after a non-abrupt shutdown() call requires more work
+     * and bookkeeping. We need consensus about quiescence (i.e., that
+     * there is no more work). The active count provides a primary
+     * indication but non-abrupt shutdown still requires a rechecking
+     * scan for any workers that are inactive but not queued.
      *
-     * Compensation. Beyond work-stealing support and lifecycle
-     * control, the main responsibility of this framework is to take
-     * actions when one worker is waiting to join a task stolen (or
-     * always held by) another.  Because we are multiplexing many
-     * tasks on to a pool of workers, we can't just let them block (as
-     * in Thread.join).  We also cannot just reassign the joiner's
-     * run-time stack with another and replace it later, which would
-     * be a form of "continuation", that even if possible is not
-     * necessarily a good idea since we sometimes need both an
-     * unblocked task and its continuation to progress. Instead we
-     * combine two tactics:
+     * Joining Tasks
+     * =============
+     *
+     * Any of several actions may be taken when one worker is waiting
+     * to join a task stolen (or always held) by another.  Because we
+     * are multiplexing many tasks on to a pool of workers, we can't
+     * just let them block (as in Thread.join).  We also cannot just
+     * reassign the joiner's run-time stack with another and replace
+     * it later, which would be a form of "continuation", that even if
+     * possible is not necessarily a good idea since we sometimes need
+     * both an unblocked task and its continuation to progress.
+     * Instead we combine two tactics:
      *
      *   Helping: Arranging for the joiner to execute some task that it
-     *      would be running if the steal had not occurred.  Method
-     *      ForkJoinWorkerThread.joinTask tracks joining->stealing
-     *      links to try to find such a task.
+     *      would be running if the steal had not occurred.
      *
      *   Compensating: Unless there are already enough live threads,
-     *      method tryPreBlock() may create or re-activate a spare
-     *      thread to compensate for blocked joiners until they
-     *      unblock.
+     *      method tryCompensate() may create or re-activate a spare
+     *      thread to compensate for blocked joiners until they unblock.
+     *
+     * A third form (implemented in tryRemoveAndExec) amounts to
+     * helping a hypothetical compensator: If we can readily tell that
+     * a possible action of a compensator is to steal and execute the
+     * task being joined, the joining thread can do so directly,
+     * without the need for a compensation thread (although at the
+     * expense of larger run-time stacks, but the tradeoff is
+     * typically worthwhile).
      *
      * The ManagedBlocker extension API can't use helping so relies
      * only on compensation in method awaitBlocker.
      *
+     * The algorithm in tryHelpStealer entails a form of "linear"
+     * helping: Each worker records (in field currentSteal) the most
+     * recent task it stole from some other worker. Plus, it records
+     * (in field currentJoin) the task it is currently actively
+     * joining. Method tryHelpStealer uses these markers to try to
+     * find a worker to help (i.e., steal back a task from and execute
+     * it) that could hasten completion of the actively joined task.
+     * In essence, the joiner executes a task that would be on its own
+     * local deque had the to-be-joined task not been stolen. This may
+     * be seen as a conservative variant of the approach in Wagner &
+     * Calder "Leapfrogging: a portable technique for implementing
+     * efficient futures" SIGPLAN Notices, 1993
+     * (http://portal.acm.org/citation.cfm?id=155354). It differs in
+     * that: (1) We only maintain dependency links across workers upon
+     * steals, rather than use per-task bookkeeping.  This sometimes
+     * requires a linear scan of workQueues array to locate stealers,
+     * but often doesn't because stealers leave hints (that may become
+     * stale/wrong) of where to locate them.  It is only a hint
+     * because a worker might have had multiple steals and the hint
+     * records only one of them (usually the most current).  Hinting
+     * isolates cost to when it is needed, rather than adding to
+     * per-task overhead.  (2) It is "shallow", ignoring nesting and
+     * potentially cyclic mutual steals.  (3) It is intentionally
+     * racy: field currentJoin is updated only while actively joining,
+     * which means that we miss links in the chain during long-lived
+     * tasks, GC stalls etc (which is OK since blocking in such cases
+     * is usually a good idea).  (4) We bound the number of attempts
+     * to find work (see MAX_HELP) and fall back to suspending the
+     * worker and if necessary replacing it with another.
+     *
+     * Helping actions for CountedCompleters are much simpler: Method
+     * helpComplete can take and execute any task with the same root
+     * as the task being waited on. However, this still entails some
+     * traversal of completer chains, so is less efficient than using
+     * CountedCompleters without explicit joins.
+     *
      * It is impossible to keep exactly the target parallelism number
      * of threads running at any given time.  Determining the
      * existence of conservatively safe helping targets, the
      * availability of already-created spares, and the apparent need
-     * to create new spares are all racy and require heuristic
-     * guidance, so we rely on multiple retries of each.  Currently,
-     * in keeping with on-demand signalling policy, we compensate only
-     * if blocking would leave less than one active (non-waiting,
-     * non-blocked) worker. Additionally, to avoid some false alarms
-     * due to GC, lagging counters, system activity, etc, compensated
-     * blocking for joins is only attempted after rechecks stabilize
-     * (retries are interspersed with Thread.yield, for good
-     * citizenship).  The variable blockedCount, incremented before
-     * blocking and decremented after, is sometimes needed to
-     * distinguish cases of waiting for work vs blocking on joins or
-     * other managed sync. Both cases are equivalent for most pool
-     * control, so we can update non-atomically. (Additionally,
-     * contention on blockedCount alleviates some contention on ctl).
+     * to create new spares are all racy, so we rely on multiple
+     * retries of each.  Compensation in the apparent absence of
+     * helping opportunities is challenging to control on JVMs, where
+     * GC and other activities can stall progress of tasks that in
+     * turn stall out many other dependent tasks, without us being
+     * able to determine whether they will ever require compensation.
+     * Even though work-stealing otherwise encounters little
+     * degradation in the presence of more threads than cores,
+     * aggressively adding new threads in such cases entails risk of
+     * unwanted positive feedback control loops in which more threads
+     * cause more dependent stalls (as well as delayed progress of
+     * unblocked threads to the point that we know they are available)
+     * leading to more situations requiring more threads, and so
+     * on. This aspect of control can be seen as an (analytically
+     * intractable) game with an opponent that may choose the worst
+     * (for us) active thread to stall at any time.  We take several
+     * precautions to bound losses (and thus bound gains), mainly in
+     * methods tryCompensate and awaitJoin.
      *
-     * Shutdown and Termination. A call to shutdownNow atomically sets
-     * the ctl stop bit and then (non-atomically) sets each workers
-     * "terminate" status, cancels all unprocessed tasks, and wakes up
-     * all waiting workers.  Detecting whether termination should
-     * commence after a non-abrupt shutdown() call requires more work
-     * and bookkeeping. We need consensus about quiescence (i.e., that
-     * there is no more work) which is reflected in active counts so
-     * long as there are no current blockers, as well as possible
-     * re-evaluations during independent changes in blocking or
-     * quiescing workers.
+     * Common Pool
+     * ===========
      *
-     * Style notes: There is a lot of representation-level coupling
-     * among classes ForkJoinPool, ForkJoinWorkerThread, and
-     * ForkJoinTask.  Most fields of ForkJoinWorkerThread maintain
-     * data structures managed by ForkJoinPool, so are directly
-     * accessed.  Conversely we allow access to "workers" array by
-     * workers, and direct access to ForkJoinTask.status by both
-     * ForkJoinPool and ForkJoinWorkerThread.  There is little point
+     * The static commonPool always exists after static
+     * initialization.  Since it (or any other created pool) need
+     * never be used, we minimize initial construction overhead and
+     * footprint to the setup of about a dozen fields, with no nested
+     * allocation. Most bootstrapping occurs within method
+     * fullExternalPush during the first submission to the pool.
+     *
+     * When external threads submit to the common pool, they can
+     * perform some subtask processing (see externalHelpJoin and
+     * related methods).  We do not need to record whether these
+     * submissions are to the common pool -- if not, externalHelpJoin
+     * returns quickly (at the most helping to signal some common pool
+     * workers). These submitters would otherwise be blocked waiting
+     * for completion, so the extra effort (with liberally sprinkled
+     * task status checks) in inapplicable cases amounts to an odd
+     * form of limited spin-wait before blocking in ForkJoinTask.join.
+     *
+     * Style notes
+     * ===========
+     *
+     * There is a lot of representation-level coupling among classes
+     * ForkJoinPool, ForkJoinWorkerThread, and ForkJoinTask.  The
+     * fields of WorkQueue maintain data structures managed by
+     * ForkJoinPool, so are directly accessed.  There is little point
      * trying to reduce this, since any associated future changes in
      * representations will need to be accompanied by algorithmic
-     * changes anyway. All together, these low-level implementation
-     * choices produce as much as a factor of 4 performance
-     * improvement compared to naive implementations, and enable the
-     * processing of billions of tasks per second, at the expense of
-     * some ugliness.
+     * changes anyway. Several methods intrinsically sprawl because
+     * they must accumulate sets of consistent reads of volatiles held
+     * in local variables.  Methods signalWork() and scan() are the
+     * main bottlenecks, so are especially heavily
+     * micro-optimized/mangled.  There are lots of inline assignments
+     * (of form "while ((local = field) != 0)") which are usually the
+     * simplest way to ensure the required read orderings (which are
+     * sometimes critical). This leads to a "C"-like style of listing
+     * declarations of these locals at the heads of methods or blocks.
+     * There are several occurrences of the unusual "do {} while
+     * (!cas...)"  which is the simplest way to force an update of a
+     * CAS'ed variable. There are also other coding oddities (including
+     * several unnecessary-looking hoisted null checks) that help
+     * some methods perform reasonably even when interpreted (not
+     * compiled).
      *
-     * Methods signalWork() and scan() are the main bottlenecks so are
-     * especially heavily micro-optimized/mangled.  There are lots of
-     * inline assignments (of form "while ((local = field) != 0)")
-     * which are usually the simplest way to ensure the required read
-     * orderings (which are sometimes critical). This leads to a
-     * "C"-like style of listing declarations of these locals at the
-     * heads of methods or blocks.  There are several occurrences of
-     * the unusual "do {} while (!cas...)"  which is the simplest way
-     * to force an update of a CAS'ed variable. There are also other
-     * coding oddities that help some methods perform reasonably even
-     * when interpreted (not compiled).
-     *
-     * The order of declarations in this file is: (1) declarations of
-     * statics (2) fields (along with constants used when unpacking
-     * some of them), listed in an order that tends to reduce
-     * contention among them a bit under most JVMs.  (3) internal
-     * control methods (4) callbacks and other support for
-     * ForkJoinTask and ForkJoinWorkerThread classes, (5) exported
-     * methods (plus a few little helpers). (6) static block
-     * initializing all statics in a minimally dependent order.
+     * The order of declarations in this file is:
+     * (1) Static utility functions
+     * (2) Nested (static) classes
+     * (3) Static fields
+     * (4) Fields, along with constants used when unpacking some of them
+     * (5) Internal control methods
+     * (6) Callbacks and other support for ForkJoinTask methods
+     * (7) Exported methods
+     * (8) Static block initializing statics in minimally dependent order
      */
 
+    // Static utilities
+
+    /**
+     * If there is a security manager, makes sure caller has
+     * permission to modify threads.
+     */
+    private static void checkPermission() {
+        SecurityManager security = System.getSecurityManager();
+        if (security != null)
+            security.checkPermission(modifyThreadPermission);
+    }
+
+    // Nested classes
+
     /**
      * Factory for creating new {@link ForkJoinWorkerThread}s.
      * A {@code ForkJoinWorkerThreadFactory} must be defined and used
@@ -359,14 +528,525 @@
      * Default ForkJoinWorkerThreadFactory implementation; creates a
      * new ForkJoinWorkerThread.
      */
-    static class DefaultForkJoinWorkerThreadFactory
+    static final class DefaultForkJoinWorkerThreadFactory
         implements ForkJoinWorkerThreadFactory {
-        public ForkJoinWorkerThread newThread(ForkJoinPool pool) {
+        public final ForkJoinWorkerThread newThread(ForkJoinPool pool) {
             return new ForkJoinWorkerThread(pool);
         }
     }
 
     /**
+     * Per-thread records for threads that submit to pools. Currently
+     * holds only pseudo-random seed / index that is used to choose
+     * submission queues in method externalPush. In the future, this may
+     * also incorporate a means to implement different task rejection
+     * and resubmission policies.
+     *
+     * Seeds for submitters and workers/workQueues work in basically
+     * the same way but are initialized and updated using slightly
+     * different mechanics. Both are initialized using the same
+     * approach as in class ThreadLocal, where successive values are
+     * unlikely to collide with previous values. Seeds are then
+     * randomly modified upon collisions using xorshifts, which
+     * requires a non-zero seed.
+     */
+    static final class Submitter {
+        int seed;
+        Submitter(int s) { seed = s; }
+    }
+
+    /**
+     * Class for artificial tasks that are used to replace the target
+     * of local joins if they are removed from an interior queue slot
+     * in WorkQueue.tryRemoveAndExec. We don't need the proxy to
+     * actually do anything beyond having a unique identity.
+     */
+    static final class EmptyTask extends ForkJoinTask<Void> {
+        private static final long serialVersionUID = -7721805057305804111L;
+        EmptyTask() { status = ForkJoinTask.NORMAL; } // force done
+        public final Void getRawResult() { return null; }
+        public final void setRawResult(Void x) {}
+        public final boolean exec() { return true; }
+    }
+
+    /**
+     * Queues supporting work-stealing as well as external task
+     * submission. See above for main rationale and algorithms.
+     * Implementation relies heavily on "Unsafe" intrinsics
+     * and selective use of "volatile":
+     *
+     * Field "base" is the index (mod array.length) of the least valid
+     * queue slot, which is always the next position to steal (poll)
+     * from if nonempty. Reads and writes require volatile orderings
+     * but not CAS, because updates are only performed after slot
+     * CASes.
+     *
+     * Field "top" is the index (mod array.length) of the next queue
+     * slot to push to or pop from. It is written only by owner thread
+     * for push, or under lock for external/shared push, and accessed
+     * by other threads only after reading (volatile) base.  Both top
+     * and base are allowed to wrap around on overflow, but (top -
+     * base) (or more commonly -(base - top) to force volatile read of
+     * base before top) still estimates size. The lock ("qlock") is
+     * forced to -1 on termination, causing all further lock attempts
+     * to fail. (Note: we don't need CAS for termination state because
+     * upon pool shutdown, all shared-queues will stop being used
+     * anyway.)  Nearly all lock bodies are set up so that exceptions
+     * within lock bodies are "impossible" (modulo JVM errors that
+     * would cause failure anyway.)
+     *
+     * The array slots are read and written using the emulation of
+     * volatiles/atomics provided by Unsafe. Insertions must in
+     * general use putOrderedObject as a form of releasing store to
+     * ensure that all writes to the task object are ordered before
+     * its publication in the queue.  All removals entail a CAS to
+     * null.  The array is always a power of two. To ensure safety of
+     * Unsafe array operations, all accesses perform explicit null
+     * checks and implicit bounds checks via power-of-two masking.
+     *
+     * In addition to basic queuing support, this class contains
+     * fields described elsewhere to control execution. It turns out
+     * to work better memory-layout-wise to include them in this class
+     * rather than a separate class.
+     *
+     * Performance on most platforms is very sensitive to placement of
+     * instances of both WorkQueues and their arrays -- we absolutely
+     * do not want multiple WorkQueue instances or multiple queue
+     * arrays sharing cache lines. (It would be best for queue objects
+     * and their arrays to share, but there is nothing available to
+     * help arrange that).  Unfortunately, because they are recorded
+     * in a common array, WorkQueue instances are often moved to be
+     * adjacent by garbage collectors. To reduce impact, we use field
+     * padding that works OK on common platforms; this effectively
+     * trades off slightly slower average field access for the sake of
+     * avoiding really bad worst-case access. (Until better JVM
+     * support is in place, this padding is dependent on transient
+     * properties of JVM field layout rules.) We also take care in
+     * allocating, sizing and resizing the array. Non-shared queue
+     * arrays are initialized by workers before use. Others are
+     * allocated on first use.
+     */
+    static final class WorkQueue {
+        /**
+         * Capacity of work-stealing queue array upon initialization.
+         * Must be a power of two; at least 4, but should be larger to
+         * reduce or eliminate cacheline sharing among queues.
+         * Currently, it is much larger, as a partial workaround for
+         * the fact that JVMs often place arrays in locations that
+         * share GC bookkeeping (especially cardmarks) such that
+         * per-write accesses encounter serious memory contention.
+         */
+        static final int INITIAL_QUEUE_CAPACITY = 1 << 13;
+
+        /**
+         * Maximum size for queue arrays. Must be a power of two less
+         * than or equal to 1 << (31 - width of array entry) to ensure
+         * lack of wraparound of index calculations, but defined to a
+         * value a bit less than this to help users trap runaway
+         * programs before saturating systems.
+         */
+        static final int MAXIMUM_QUEUE_CAPACITY = 1 << 26; // 64M
+
+        // Heuristic padding to ameliorate unfortunate memory placements
+        volatile long pad00, pad01, pad02, pad03, pad04, pad05, pad06;
+
+        int seed;                  // for random scanning; initialize nonzero
+        volatile int eventCount;   // encoded inactivation count; < 0 if inactive
+        int nextWait;              // encoded record of next event waiter
+        int hint;                  // steal or signal hint (index)
+        int poolIndex;             // index of this queue in pool (or 0)
+        final int mode;            // 0: lifo, > 0: fifo, < 0: shared
+        int nsteals;               // number of steals
+        volatile int qlock;        // 1: locked, -1: terminate; else 0
+        volatile int base;         // index of next slot for poll
+        int top;                   // index of next slot for push
+        ForkJoinTask<?>[] array;   // the elements (initially unallocated)
+        final ForkJoinPool pool;   // the containing pool (may be null)
+        final ForkJoinWorkerThread owner; // owning thread or null if shared
+        volatile Thread parker;    // == owner during call to park; else null
+        volatile ForkJoinTask<?> currentJoin;  // task being joined in awaitJoin
+        ForkJoinTask<?> currentSteal; // current non-local task being executed
+
+        volatile Object pad10, pad11, pad12, pad13, pad14, pad15, pad16, pad17;
+        volatile Object pad18, pad19, pad1a, pad1b, pad1c, pad1d;
+
+        WorkQueue(ForkJoinPool pool, ForkJoinWorkerThread owner, int mode,
+                  int seed) {
+            this.pool = pool;
+            this.owner = owner;
+            this.mode = mode;
+            this.seed = seed;
+            // Place indices in the center of array (that is not yet allocated)
+            base = top = INITIAL_QUEUE_CAPACITY >>> 1;
+        }
+
+        /**
+         * Returns the approximate number of tasks in the queue.
+         */
+        final int queueSize() {
+            int n = base - top;       // non-owner callers must read base first
+            return (n >= 0) ? 0 : -n; // ignore transient negative
+        }
+
+       /**
+         * Provides a more accurate estimate of whether this queue has
+         * any tasks than does queueSize, by checking whether a
+         * near-empty queue has at least one unclaimed task.
+         */
+        final boolean isEmpty() {
+            ForkJoinTask<?>[] a; int m, s;
+            int n = base - (s = top);
+            return (n >= 0 ||
+                    (n == -1 &&
+                     ((a = array) == null ||
+                      (m = a.length - 1) < 0 ||
+                      U.getObject
+                      (a, (long)((m & (s - 1)) << ASHIFT) + ABASE) == null)));
+        }
+
+        /**
+         * Pushes a task. Call only by owner in unshared queues.  (The
+         * shared-queue version is embedded in method externalPush.)
+         *
+         * @param task the task. Caller must ensure non-null.
+         * @throws RejectedExecutionException if array cannot be resized
+         */
+        final void push(ForkJoinTask<?> task) {
+            ForkJoinTask<?>[] a; ForkJoinPool p;
+            int s = top, m, n;
+            if ((a = array) != null) {    // ignore if queue removed
+                int j = (((m = a.length - 1) & s) << ASHIFT) + ABASE;
+                U.putOrderedObject(a, j, task);
+                if ((n = (top = s + 1) - base) <= 2) {
+                    if ((p = pool) != null)
+                        p.signalWork(this);
+                }
+                else if (n >= m)
+                    growArray();
+            }
+        }
+
+       /**
+         * Initializes or doubles the capacity of array. Call either
+         * by owner or with lock held -- it is OK for base, but not
+         * top, to move while resizings are in progress.
+         */
+        final ForkJoinTask<?>[] growArray() {
+            ForkJoinTask<?>[] oldA = array;
+            int size = oldA != null ? oldA.length << 1 : INITIAL_QUEUE_CAPACITY;
+            if (size > MAXIMUM_QUEUE_CAPACITY)
+                throw new RejectedExecutionException("Queue capacity exceeded");
+            int oldMask, t, b;
+            ForkJoinTask<?>[] a = array = new ForkJoinTask<?>[size];
+            if (oldA != null && (oldMask = oldA.length - 1) >= 0 &&
+                (t = top) - (b = base) > 0) {
+                int mask = size - 1;
+                do {
+                    ForkJoinTask<?> x;
+                    int oldj = ((b & oldMask) << ASHIFT) + ABASE;
+                    int j    = ((b &    mask) << ASHIFT) + ABASE;
+                    x = (ForkJoinTask<?>)U.getObjectVolatile(oldA, oldj);
+                    if (x != null &&
+                        U.compareAndSwapObject(oldA, oldj, x, null))
+                        U.putObjectVolatile(a, j, x);
+                } while (++b != t);
+            }
+            return a;
+        }
+
+        /**
+         * Takes next task, if one exists, in LIFO order.  Call only
+         * by owner in unshared queues.
+         */
+        final ForkJoinTask<?> pop() {
+            ForkJoinTask<?>[] a; ForkJoinTask<?> t; int m;
+            if ((a = array) != null && (m = a.length - 1) >= 0) {
+                for (int s; (s = top - 1) - base >= 0;) {
+                    long j = ((m & s) << ASHIFT) + ABASE;
+                    if ((t = (ForkJoinTask<?>)U.getObject(a, j)) == null)
+                        break;
+                    if (U.compareAndSwapObject(a, j, t, null)) {
+                        top = s;
+                        return t;
+                    }
+                }
+            }
+            return null;
+        }
+
+        /**
+         * Takes a task in FIFO order if b is base of queue and a task
+         * can be claimed without contention. Specialized versions
+         * appear in ForkJoinPool methods scan and tryHelpStealer.
+         */
+        final ForkJoinTask<?> pollAt(int b) {
+            ForkJoinTask<?> t; ForkJoinTask<?>[] a;
+            if ((a = array) != null) {
+                int j = (((a.length - 1) & b) << ASHIFT) + ABASE;
+                if ((t = (ForkJoinTask<?>)U.getObjectVolatile(a, j)) != null &&
+                    base == b &&
+                    U.compareAndSwapObject(a, j, t, null)) {
+                    base = b + 1;
+                    return t;
+                }
+            }
+            return null;
+        }
+
+        /**
+         * Takes next task, if one exists, in FIFO order.
+         */
+        final ForkJoinTask<?> poll() {
+            ForkJoinTask<?>[] a; int b; ForkJoinTask<?> t;
+            while ((b = base) - top < 0 && (a = array) != null) {
+                int j = (((a.length - 1) & b) << ASHIFT) + ABASE;
+                t = (ForkJoinTask<?>)U.getObjectVolatile(a, j);
+                if (t != null) {
+                    if (base == b &&
+                        U.compareAndSwapObject(a, j, t, null)) {
+                        base = b + 1;
+                        return t;
+                    }
+                }
+                else if (base == b) {
+                    if (b + 1 == top)
+                        break;
+                    Thread.yield(); // wait for lagging update (very rare)
+                }
+            }
+            return null;
+        }
+
+        /**
+         * Takes next task, if one exists, in order specified by mode.
+         */
+        final ForkJoinTask<?> nextLocalTask() {
+            return mode == 0 ? pop() : poll();
+        }
+
+        /**
+         * Returns next task, if one exists, in order specified by mode.
+         */
+        final ForkJoinTask<?> peek() {
+            ForkJoinTask<?>[] a = array; int m;
+            if (a == null || (m = a.length - 1) < 0)
+                return null;
+            int i = mode == 0 ? top - 1 : base;
+            int j = ((i & m) << ASHIFT) + ABASE;
+            return (ForkJoinTask<?>)U.getObjectVolatile(a, j);
+        }
+
+        /**
+         * Pops the given task only if it is at the current top.
+         * (A shared version is available only via FJP.tryExternalUnpush)
+         */
+        final boolean tryUnpush(ForkJoinTask<?> t) {
+            ForkJoinTask<?>[] a; int s;
+            if ((a = array) != null && (s = top) != base &&
+                U.compareAndSwapObject
+                (a, (((a.length - 1) & --s) << ASHIFT) + ABASE, t, null)) {
+                top = s;
+                return true;
+            }
+            return false;
+        }
+
+        /**
+         * Removes and cancels all known tasks, ignoring any exceptions.
+         */
+        final void cancelAll() {
+            ForkJoinTask.cancelIgnoringExceptions(currentJoin);
+            ForkJoinTask.cancelIgnoringExceptions(currentSteal);
+            for (ForkJoinTask<?> t; (t = poll()) != null; )
+                ForkJoinTask.cancelIgnoringExceptions(t);
+        }
+
+        /**
+         * Computes next value for random probes.  Scans don't require
+         * a very high quality generator, but also not a crummy one.
+         * Marsaglia xor-shift is cheap and works well enough.  Note:
+         * This is manually inlined in its usages in ForkJoinPool to
+         * avoid writes inside busy scan loops.
+         */
+        final int nextSeed() {
+            int r = seed;
+            r ^= r << 13;
+            r ^= r >>> 17;
+            return seed = r ^= r << 5;
+        }
+
+        // Specialized execution methods
+
+        /**
+         * Pops and runs tasks until empty.
+         */
+        private void popAndExecAll() {
+            // A bit faster than repeated pop calls
+            ForkJoinTask<?>[] a; int m, s; long j; ForkJoinTask<?> t;
+            while ((a = array) != null && (m = a.length - 1) >= 0 &&
+                   (s = top - 1) - base >= 0 &&
+                   (t = ((ForkJoinTask<?>)
+                         U.getObject(a, j = ((m & s) << ASHIFT) + ABASE)))
+                   != null) {
+                if (U.compareAndSwapObject(a, j, t, null)) {
+                    top = s;
+                    t.doExec();
+                }
+            }
+        }
+
+        /**
+         * Polls and runs tasks until empty.
+         */
+        private void pollAndExecAll() {
+            for (ForkJoinTask<?> t; (t = poll()) != null;)
+                t.doExec();
+        }
+
+        /**
+         * If present, removes from queue and executes the given task,
+         * or any other cancelled task. Returns (true) on any CAS
+         * or consistency check failure so caller can retry.
+         *
+         * @return false if no progress can be made, else true
+         */
+        final boolean tryRemoveAndExec(ForkJoinTask<?> task) {
+            boolean stat = true, removed = false, empty = true;
+            ForkJoinTask<?>[] a; int m, s, b, n;
+            if ((a = array) != null && (m = a.length - 1) >= 0 &&
+                (n = (s = top) - (b = base)) > 0) {
+                for (ForkJoinTask<?> t;;) {           // traverse from s to b
+                    int j = ((--s & m) << ASHIFT) + ABASE;
+                    t = (ForkJoinTask<?>)U.getObjectVolatile(a, j);
+                    if (t == null)                    // inconsistent length
+                        break;
+                    else if (t == task) {
+                        if (s + 1 == top) {           // pop
+                            if (!U.compareAndSwapObject(a, j, task, null))
+                                break;
+                            top = s;
+                            removed = true;
+                        }
+                        else if (base == b)           // replace with proxy
+                            removed = U.compareAndSwapObject(a, j, task,
+                                                             new EmptyTask());
+                        break;
+                    }
+                    else if (t.status >= 0)
+                        empty = false;
+                    else if (s + 1 == top) {          // pop and throw away
+                        if (U.compareAndSwapObject(a, j, t, null))
+                            top = s;
+                        break;
+                    }
+                    if (--n == 0) {
+                        if (!empty && base == b)
+                            stat = false;
+                        break;
+                    }
+                }
+            }
+            if (removed)
+                task.doExec();
+            return stat;
+        }
+
+        /**
+         * Polls for and executes the given task or any other task in
+         * its CountedCompleter computation.
+         */
+        final boolean pollAndExecCC(ForkJoinTask<?> root) {
+            ForkJoinTask<?>[] a; int b; Object o;
+            outer: while ((b = base) - top < 0 && (a = array) != null) {
+                long j = (((a.length - 1) & b) << ASHIFT) + ABASE;
+                if ((o = U.getObject(a, j)) == null ||
+                    !(o instanceof CountedCompleter))
+                    break;
+                for (CountedCompleter<?> t = (CountedCompleter<?>)o, r = t;;) {
+                    if (r == root) {
+                        if (base == b &&
+                            U.compareAndSwapObject(a, j, t, null)) {
+                            base = b + 1;
+                            t.doExec();
+                            return true;
+                        }
+                        else
+                            break; // restart
+                    }
+                    if ((r = r.completer) == null)
+                        break outer; // not part of root computation
+                }
+            }
+            return false;
+        }
+
+        /**
+         * Executes a top-level task and any local tasks remaining
+         * after execution.
+         */
+        final void runTask(ForkJoinTask<?> t) {
+            if (t != null) {
+                (currentSteal = t).doExec();
+                currentSteal = null;
+                ++nsteals;
+                if (base - top < 0) {       // process remaining local tasks
+                    if (mode == 0)
+                        popAndExecAll();
+                    else
+                        pollAndExecAll();
+                }
+            }
+        }
+
+        /**
+         * Executes a non-top-level (stolen) task.
+         */
+        final void runSubtask(ForkJoinTask<?> t) {
+            if (t != null) {
+                ForkJoinTask<?> ps = currentSteal;
+                (currentSteal = t).doExec();
+                currentSteal = ps;
+            }
+        }
+
+        /**
+         * Returns true if owned and not known to be blocked.
+         */
+        final boolean isApparentlyUnblocked() {
+            Thread wt; Thread.State s;
+            return (eventCount >= 0 &&
+                    (wt = owner) != null &&
+                    (s = wt.getState()) != Thread.State.BLOCKED &&
+                    s != Thread.State.WAITING &&
+                    s != Thread.State.TIMED_WAITING);
+        }
+
+        // Unsafe mechanics
+        private static final sun.misc.Unsafe U;
+        private static final long QLOCK;
+        private static final int ABASE;
+        private static final int ASHIFT;
+        static {
+            try {
+                U = sun.misc.Unsafe.getUnsafe();
+                Class<?> k = WorkQueue.class;
+                Class<?> ak = ForkJoinTask[].class;
+                QLOCK = U.objectFieldOffset
+                    (k.getDeclaredField("qlock"));
+                ABASE = U.arrayBaseOffset(ak);
+                int scale = U.arrayIndexScale(ak);
+                if ((scale & (scale - 1)) != 0)
+                    throw new Error("data type scale not a power of two");
+                ASHIFT = 31 - Integer.numberOfLeadingZeros(scale);
+            } catch (Exception e) {
+                throw new Error(e);
+            }
+        }
+    }
+
+    // static fields (initialized in static initializer below)
+
+    /**
      * Creates a new ForkJoinWorkerThread. This factory is used unless
      * overridden in ForkJoinPool constructors.
      */
@@ -374,107 +1054,93 @@
         defaultForkJoinWorkerThreadFactory;
 
     /**
+     * Per-thread submission bookkeeping. Shared across all pools
+     * to reduce ThreadLocal pollution and because random motion
+     * to avoid contention in one pool is likely to hold for others.
+     * Lazily initialized on first submission (but null-checked
+     * in other contexts to avoid unnecessary initialization).
+     */
+    static final ThreadLocal<Submitter> submitters;
+
+    /**
      * Permission required for callers of methods that may start or
      * kill threads.
      */
     private static final RuntimePermission modifyThreadPermission;
 
     /**
-     * If there is a security manager, makes sure caller has
-     * permission to modify threads.
+     * Common (static) pool. Non-null for public use unless a static
+     * construction exception, but internal usages null-check on use
+     * to paranoically avoid potential initialization circularities
+     * as well as to simplify generated code.
      */
-    private static void checkPermission() {
-        SecurityManager security = System.getSecurityManager();
-        if (security != null)
-            security.checkPermission(modifyThreadPermission);
+    static final ForkJoinPool commonPool;
+
+    /**
+     * Common pool parallelism. Must equal commonPool.parallelism.
+     */
+    static final int commonPoolParallelism;
+
+    /**
+     * Sequence number for creating workerNamePrefix.
+     */
+    private static int poolNumberSequence;
+
+    /**
+     * Returns the next sequence number. We don't expect this to
+     * ever contend, so use simple builtin sync.
+     */
+    private static final synchronized int nextPoolId() {
+        return ++poolNumberSequence;
     }
 
-    /**
-     * Generator for assigning sequence numbers as pool names.
-     */
-    private static final AtomicInteger poolNumberGenerator;
+    // static constants
 
     /**
-     * Generator for initial random seeds for worker victim
-     * selection. This is used only to create initial seeds. Random
-     * steals use a cheaper xorshift generator per steal attempt. We
-     * don't expect much contention on seedGenerator, so just use a
-     * plain Random.
+     * Initial timeout value (in nanoseconds) for the thread
+     * triggering quiescence to park waiting for new work. On timeout,
+     * the thread will instead try to shrink the number of
+     * workers. The value should be large enough to avoid overly
+     * aggressive shrinkage during most transient stalls (long GCs
+     * etc).
      */
-    static final Random workerSeedGenerator;
+    private static final long IDLE_TIMEOUT      = 2000L * 1000L * 1000L; // 2sec
 
     /**
-     * Array holding all worker threads in the pool.  Initialized upon
-     * construction. Array size must be a power of two.  Updates and
-     * replacements are protected by scanGuard, but the array is
-     * always kept in a consistent enough state to be randomly
-     * accessed without locking by workers performing work-stealing,
-     * as well as other traversal-based methods in this class, so long
-     * as reads memory-acquire by first reading ctl. All readers must
-     * tolerate that some array slots may be null.
+     * Timeout value when there are more threads than parallelism level
      */
-    ForkJoinWorkerThread[] workers;
+    private static final long FAST_IDLE_TIMEOUT =  200L * 1000L * 1000L;
 
     /**
-     * Initial size for submission queue array. Must be a power of
-     * two.  In many applications, these always stay small so we use a
-     * small initial cap.
+     * Tolerance for idle timeouts, to cope with timer undershoots
      */
-    private static final int INITIAL_QUEUE_CAPACITY = 8;
+    private static final long TIMEOUT_SLOP = 2000000L; // 20ms
 
     /**
-     * Maximum size for submission queue array. Must be a power of two
-     * less than or equal to 1 << (31 - width of array entry) to
-     * ensure lack of index wraparound, but is capped at a lower
-     * value to help users trap runaway computations.
+     * The maximum stolen->joining link depth allowed in method
+     * tryHelpStealer.  Must be a power of two.  Depths for legitimate
+     * chains are unbounded, but we use a fixed constant to avoid
+     * (otherwise unchecked) cycles and to bound staleness of
+     * traversal parameters at the expense of sometimes blocking when
+     * we could be helping.
      */
-    private static final int MAXIMUM_QUEUE_CAPACITY = 1 << 24; // 16M
+    private static final int MAX_HELP = 64;
 
     /**
-     * Array serving as submission queue. Initialized upon construction.
+     * Increment for seed generators. See class ThreadLocal for
+     * explanation.
      */
-    private ForkJoinTask<?>[] submissionQueue;
+    private static final int SEED_INCREMENT = 0x61c88647;
 
-    /**
-     * Lock protecting submissions array for addSubmission
-     */
-    private final ReentrantLock submissionLock;
-
-    /**
-     * Condition for awaitTermination, using submissionLock for
-     * convenience.
-     */
-    private final Condition termination;
-
-    /**
-     * Creation factory for worker threads.
-     */
-    private final ForkJoinWorkerThreadFactory factory;
-
-    /**
-     * The uncaught exception handler used when any worker abruptly
-     * terminates.
-     */
-    final Thread.UncaughtExceptionHandler ueh;
-
-    /**
-     * Prefix for assigning names to worker threads
-     */
-    private final String workerNamePrefix;
-
-    /**
-     * Sum of per-thread steal counts, updated only when threads are
-     * idle or terminating.
-     */
-    private volatile long stealCount;
-
-    /**
-     * Main pool control -- a long packed with:
+    /*
+     * Bits and masks for control variables
+     *
+     * Field ctl is a long packed with:
      * AC: Number of active running workers minus target parallelism (16 bits)
      * TC: Number of total workers minus target parallelism (16 bits)
      * ST: true if pool is terminating (1 bit)
      * EC: the wait count of top waiting thread (15 bits)
-     * ID: ~poolIndex of top of Treiber stack of waiting threads (16 bits)
+     * ID: poolIndex of top of Treiber stack of waiters (16 bits)
      *
      * When convenient, we can extract the upper 32 bits of counts and
      * the lower 32 bits of queue state, u = (int)(ctl >>> 32) and e =
@@ -483,13 +1149,26 @@
      * parallelism and the positionings of fields makes it possible to
      * perform the most common checks via sign tests of fields: When
      * ac is negative, there are not enough active workers, when tc is
-     * negative, there are not enough total workers, when id is
-     * negative, there is at least one waiting worker, and when e is
+     * negative, there are not enough total workers, and when e is
      * negative, the pool is terminating.  To deal with these possibly
      * negative fields, we use casts in and out of "short" and/or
      * signed shifts to maintain signedness.
+     *
+     * When a thread is queued (inactivated), its eventCount field is
+     * set negative, which is the only way to tell if a worker is
+     * prevented from executing tasks, even though it must continue to
+     * scan for them to avoid queuing races. Note however that
+     * eventCount updates lag releases so usage requires care.
+     *
+     * Field plock is an int packed with:
+     * SHUTDOWN: true if shutdown is enabled (1 bit)
+     * SEQ:  a sequence lock, with PL_LOCK bit set if locked (30 bits)
+     * SIGNAL: set when threads may be waiting on the lock (1 bit)
+     *
+     * The sequence number enables simple consistency checks:
+     * Staleness of read-only operations on the workQueues array can
+     * be checked by comparing plock before vs after the reads.
      */
-    volatile long ctl;
 
     // bit positions/shifts for fields
     private static final int  AC_SHIFT   = 48;
@@ -498,8 +1177,10 @@
     private static final int  EC_SHIFT   = 16;
 
     // bounds
-    private static final int  MAX_ID     = 0x7fff;  // max poolIndex
-    private static final int  SMASK      = 0xffff;  // mask short bits
+    private static final int  SMASK      = 0xffff;  // short bits
+    private static final int  MAX_CAP    = 0x7fff;  // max #workers - 1
+    private static final int  EVENMASK   = 0xfffe;  // even short bits
+    private static final int  SQMASK     = 0x007e;  // max 64 (even) slots
     private static final int  SHORT_SIGN = 1 << 15;
     private static final int  INT_SIGN   = 1 << 31;
 
@@ -521,648 +1202,648 @@
     private static final int  UTC_UNIT   = 1 << UTC_SHIFT;
 
     // masks and units for dealing with e = (int)ctl
-    private static final int  E_MASK     = 0x7fffffff; // no STOP_BIT
-    private static final int  EC_UNIT    = 1 << EC_SHIFT;
+    private static final int E_MASK      = 0x7fffffff; // no STOP_BIT
+    private static final int E_SEQ       = 1 << EC_SHIFT;
+
+    // plock bits
+    private static final int SHUTDOWN    = 1 << 31;
+    private static final int PL_LOCK     = 2;
+    private static final int PL_SIGNAL   = 1;
+    private static final int PL_SPINS    = 1 << 8;
+
+    // access mode for WorkQueue
+    static final int LIFO_QUEUE          =  0;
+    static final int FIFO_QUEUE          =  1;
+    static final int SHARED_QUEUE        = -1;
+
+    // bounds for #steps in scan loop -- must be power 2 minus 1
+    private static final int MIN_SCAN    = 0x1ff;   // cover estimation slop
+    private static final int MAX_SCAN    = 0x1ffff; // 4 * max workers
+
+    // Instance fields
+
+    /*
+     * Field layout of this class tends to matter more than one would
+     * like. Runtime layout order is only loosely related to
+     * declaration order and may differ across JVMs, but the following
+     * empirically works OK on current JVMs.
+     */
+
+    // Heuristic padding to ameliorate unfortunate memory placements
+    volatile long pad00, pad01, pad02, pad03, pad04, pad05, pad06;
+
+    volatile long stealCount;                  // collects worker counts
+    volatile long ctl;                         // main pool control
+    volatile int plock;                        // shutdown status and seqLock
+    volatile int indexSeed;                    // worker/submitter index seed
+    final int config;                          // mode and parallelism level
+    WorkQueue[] workQueues;                    // main registry
+    final ForkJoinWorkerThreadFactory factory;
+    final Thread.UncaughtExceptionHandler ueh; // per-worker UEH
+    final String workerNamePrefix;             // to create worker name string
+
+    volatile Object pad10, pad11, pad12, pad13, pad14, pad15, pad16, pad17;
+    volatile Object pad18, pad19, pad1a, pad1b;
 
     /**
-     * The target parallelism level.
+     * Acquires the plock lock to protect worker array and related
+     * updates. This method is called only if an initial CAS on plock
+     * fails. This acts as a spinlock for normal cases, but falls back
+     * to builtin monitor to block when (rarely) needed. This would be
+     * a terrible idea for a highly contended lock, but works fine as
+     * a more conservative alternative to a pure spinlock.
      */
-    final int parallelism;
-
-    /**
-     * Index (mod submission queue length) of next element to take
-     * from submission queue. Usage is identical to that for
-     * per-worker queues -- see ForkJoinWorkerThread internal
-     * documentation.
-     */
-    volatile int queueBase;
-
-    /**
-     * Index (mod submission queue length) of next element to add
-     * in submission queue. Usage is identical to that for
-     * per-worker queues -- see ForkJoinWorkerThread internal
-     * documentation.
-     */
-    int queueTop;
-
-    /**
-     * True when shutdown() has been called.
-     */
-    volatile boolean shutdown;
-
-    /**
-     * True if use local fifo, not default lifo, for local polling.
-     * Read by, and replicated by ForkJoinWorkerThreads.
-     */
-    final boolean locallyFifo;
-
-    /**
-     * The number of threads in ForkJoinWorkerThreads.helpQuiescePool.
-     * When non-zero, suppresses automatic shutdown when active
-     * counts become zero.
-     */
-    volatile int quiescerCount;
-
-    /**
-     * The number of threads blocked in join.
-     */
-    volatile int blockedCount;
-
-    /**
-     * Counter for worker Thread names (unrelated to their poolIndex)
-     */
-    private volatile int nextWorkerNumber;
-
-    /**
-     * The index for the next created worker. Accessed under scanGuard.
-     */
-    private int nextWorkerIndex;
-
-    /**
-     * SeqLock and index masking for updates to workers array.  Locked
-     * when SG_UNIT is set. Unlocking clears bit by adding
-     * SG_UNIT. Staleness of read-only operations can be checked by
-     * comparing scanGuard to value before the reads. The low 16 bits
-     * (i.e, anding with SMASK) hold (the smallest power of two
-     * covering all worker indices, minus one, and is used to avoid
-     * dealing with large numbers of null slots when the workers array
-     * is overallocated.
-     */
-    volatile int scanGuard;
-
-    private static final int SG_UNIT = 1 << 16;
-
-    /**
-     * The wakeup interval (in nanoseconds) for a worker waiting for a
-     * task when the pool is quiescent to instead try to shrink the
-     * number of workers.  The exact value does not matter too
-     * much. It must be short enough to release resources during
-     * sustained periods of idleness, but not so short that threads
-     * are continually re-created.
-     */
-    private static final long SHRINK_RATE =
-        4L * 1000L * 1000L * 1000L; // 4 seconds
-
-    /**
-     * Top-level loop for worker threads: On each step: if the
-     * previous step swept through all queues and found no tasks, or
-     * there are excess threads, then possibly blocks. Otherwise,
-     * scans for and, if found, executes a task. Returns when pool
-     * and/or worker terminate.
-     *
-     * @param w the worker
-     */
-    final void work(ForkJoinWorkerThread w) {
-        boolean swept = false;                // true on empty scans
-        long c;
-        while (!w.terminate && (int)(c = ctl) >= 0) {
-            int a;                            // active count
-            if (!swept && (a = (int)(c >> AC_SHIFT)) <= 0)
-                swept = scan(w, a);
-            else if (tryAwaitWork(w, c))
-                swept = false;
+    private int acquirePlock() {
+        int spins = PL_SPINS, r = 0, ps, nps;
+        for (;;) {
+            if (((ps = plock) & PL_LOCK) == 0 &&
+                U.compareAndSwapInt(this, PLOCK, ps, nps = ps + PL_LOCK))
+                return nps;
+            else if (r == 0) { // randomize spins if possible
+                Thread t = Thread.currentThread(); WorkQueue w; Submitter z;
+                if ((t instanceof ForkJoinWorkerThread) &&
+                    (w = ((ForkJoinWorkerThread)t).workQueue) != null)
+                    r = w.seed;
+                else if ((z = submitters.get()) != null)
+                    r = z.seed;
+                else
+                    r = 1;
+            }
+            else if (spins >= 0) {
+                r ^= r << 1; r ^= r >>> 3; r ^= r << 10; // xorshift
+                if (r >= 0)
+                    --spins;
+            }
+            else if (U.compareAndSwapInt(this, PLOCK, ps, ps | PL_SIGNAL)) {
+                synchronized (this) {
+                    if ((plock & PL_SIGNAL) != 0) {
+                        try {
+                            wait();
+                        } catch (InterruptedException ie) {
+                            try {
+                                Thread.currentThread().interrupt();
+                            } catch (SecurityException ignore) {
+                            }
+                        }
+                    }
+                    else
+                        notifyAll();
+                }
+            }
         }
     }
 
-    // Signalling
+    /**
+     * Unlocks and signals any thread waiting for plock. Called only
+     * when CAS of seq value for unlock fails.
+     */
+    private void releasePlock(int ps) {
+        plock = ps;
+        synchronized (this) { notifyAll(); }
+    }
 
     /**
-     * Wakes up or creates a worker.
+     * Performs secondary initialization, called when plock is zero.
+     * Creates workQueue array and sets plock to a valid value.  The
+     * lock body must be exception-free (so no try/finally) so we
+     * optimistically allocate new array outside the lock and throw
+     * away if (very rarely) not needed. (A similar tactic is used in
+     * fullExternalPush.)  Because the plock seq value can eventually
+     * wrap around zero, this method harmlessly fails to reinitialize
+     * if workQueues exists, while still advancing plock.
+     *
+     * Additionally tries to create the first worker.
      */
-    final void signalWork() {
-        /*
-         * The while condition is true if: (there is are too few total
-         * workers OR there is at least one waiter) AND (there are too
-         * few active workers OR the pool is terminating).  The value
-         * of e distinguishes the remaining cases: zero (no waiters)
-         * for create, negative if terminating (in which case do
-         * nothing), else release a waiter. The secondary checks for
-         * release (non-null array etc) can fail if the pool begins
-         * terminating after the test, and don't impose any added cost
-         * because JVMs must perform null and bounds checks anyway.
-         */
-        long c; int e, u;
-        while ((((e = (int)(c = ctl)) | (u = (int)(c >>> 32))) &
-                (INT_SIGN|SHORT_SIGN)) == (INT_SIGN|SHORT_SIGN) && e >= 0) {
-            if (e > 0) {                         // release a waiting worker
-                int i; ForkJoinWorkerThread w; ForkJoinWorkerThread[] ws;
-                if ((ws = workers) == null ||
-                    (i = ~e & SMASK) >= ws.length ||
-                    (w = ws[i]) == null)
-                    break;
-                long nc = (((long)(w.nextWait & E_MASK)) |
-                           ((long)(u + UAC_UNIT) << 32));
-                if (w.eventCount == e &&
-                    UNSAFE.compareAndSwapLong(this, ctlOffset, c, nc)) {
-                    w.eventCount = (e + EC_UNIT) & E_MASK;
-                    if (w.parked)
-                        UNSAFE.unpark(w);
-                    break;
+    private void initWorkers() {
+        WorkQueue[] ws, nws; int ps;
+        int p = config & SMASK;        // find power of two table size
+        int n = (p > 1) ? p - 1 : 1;   // ensure at least 2 slots
+        n |= n >>> 1; n |= n >>> 2; n |= n >>> 4; n |= n >>> 8; n |= n >>> 16;
+        n = (n + 1) << 1;
+        if ((ws = workQueues) == null || ws.length == 0)
+            nws = new WorkQueue[n];
+        else
+            nws = null;
+        if (((ps = plock) & PL_LOCK) != 0 ||
+            !U.compareAndSwapInt(this, PLOCK, ps, ps += PL_LOCK))
+            ps = acquirePlock();
+        if (((ws = workQueues) == null || ws.length == 0) && nws != null)
+            workQueues = nws;
+        int nps = (ps & SHUTDOWN) | ((ps + PL_LOCK) & ~SHUTDOWN);
+        if (!U.compareAndSwapInt(this, PLOCK, ps, nps))
+            releasePlock(nps);
+        tryAddWorker();
+    }
+
+    /**
+     * Tries to create and start one worker if fewer than target
+     * parallelism level exist. Adjusts counts etc on failure.
+     */
+    private void tryAddWorker() {
+        long c; int u;
+        while ((u = (int)((c = ctl) >>> 32)) < 0 &&
+               (u & SHORT_SIGN) != 0 && (int)c == 0) {
+            long nc = (long)(((u + UTC_UNIT) & UTC_MASK) |
+                             ((u + UAC_UNIT) & UAC_MASK)) << 32;
+            if (U.compareAndSwapLong(this, CTL, c, nc)) {
+                ForkJoinWorkerThreadFactory fac;
+                Throwable ex = null;
+                ForkJoinWorkerThread wt = null;
+                try {
+                    if ((fac = factory) != null &&
+                        (wt = fac.newThread(this)) != null) {
+                        wt.start();
+                        break;
+                    }
+                } catch (Throwable e) {
+                    ex = e;
                 }
-            }
-            else if (UNSAFE.compareAndSwapLong
-                     (this, ctlOffset, c,
-                      (long)(((u + UTC_UNIT) & UTC_MASK) |
-                             ((u + UAC_UNIT) & UAC_MASK)) << 32)) {
-                addWorker();
+                deregisterWorker(wt, ex);
                 break;
             }
         }
     }
 
-    /**
-     * Variant of signalWork to help release waiters on rescans.
-     * Tries once to release a waiter if active count < 0.
-     *
-     * @return false if failed due to contention, else true
-     */
-    private boolean tryReleaseWaiter() {
-        long c; int e, i; ForkJoinWorkerThread w; ForkJoinWorkerThread[] ws;
-        if ((e = (int)(c = ctl)) > 0 &&
-            (int)(c >> AC_SHIFT) < 0 &&
-            (ws = workers) != null &&
-            (i = ~e & SMASK) < ws.length &&
-            (w = ws[i]) != null) {
-            long nc = ((long)(w.nextWait & E_MASK) |
-                       ((c + AC_UNIT) & (AC_MASK|TC_MASK)));
-            if (w.eventCount != e ||
-                !UNSAFE.compareAndSwapLong(this, ctlOffset, c, nc))
-                return false;
-            w.eventCount = (e + EC_UNIT) & E_MASK;
-            if (w.parked)
-                UNSAFE.unpark(w);
-        }
-        return true;
-    }
-
-    // Scanning for tasks
+    //  Registering and deregistering workers
 
     /**
-     * Scans for and, if found, executes one task. Scans start at a
-     * random index of workers array, and randomly select the first
-     * (2*#workers)-1 probes, and then, if all empty, resort to 2
-     * circular sweeps, which is necessary to check quiescence. and
-     * taking a submission only if no stealable tasks were found.  The
-     * steal code inside the loop is a specialized form of
-     * ForkJoinWorkerThread.deqTask, followed bookkeeping to support
-     * helpJoinTask and signal propagation. The code for submission
-     * queues is almost identical. On each steal, the worker completes
-     * not only the task, but also all local tasks that this task may
-     * have generated. On detecting staleness or contention when
-     * trying to take a task, this method returns without finishing
-     * sweep, which allows global state rechecks before retry.
+     * Callback from ForkJoinWorkerThread to establish and record its
+     * WorkQueue. To avoid scanning bias due to packing entries in
+     * front of the workQueues array, we treat the array as a simple
+     * power-of-two hash table using per-thread seed as hash,
+     * expanding as needed.
      *
-     * @param w the worker
-     * @param a the number of active workers
-     * @return true if swept all queues without finding a task
+     * @param wt the worker thread
+     * @return the worker's queue
      */
-    private boolean scan(ForkJoinWorkerThread w, int a) {
-        int g = scanGuard; // mask 0 avoids useless scans if only one active
-        int m = (parallelism == 1 - a && blockedCount == 0) ? 0 : g & SMASK;
-        ForkJoinWorkerThread[] ws = workers;
-        if (ws == null || ws.length <= m)         // staleness check
-            return false;
-        for (int r = w.seed, k = r, j = -(m + m); j <= m + m; ++j) {
-            ForkJoinTask<?> t; ForkJoinTask<?>[] q; int b, i;
-            ForkJoinWorkerThread v = ws[k & m];
-            if (v != null && (b = v.queueBase) != v.queueTop &&
-                (q = v.queue) != null && (i = (q.length - 1) & b) >= 0) {
-                long u = (i << ASHIFT) + ABASE;
-                if ((t = q[i]) != null && v.queueBase == b &&
-                    UNSAFE.compareAndSwapObject(q, u, t, null)) {
-                    int d = (v.queueBase = b + 1) - v.queueTop;
-                    v.stealHint = w.poolIndex;
-                    if (d != 0)
-                        signalWork();             // propagate if nonempty
-                    w.execTask(t);
-                }
-                r ^= r << 13; r ^= r >>> 17; w.seed = r ^ (r << 5);
-                return false;                     // store next seed
-            }
-            else if (j < 0) {                     // xorshift
-                r ^= r << 13; r ^= r >>> 17; k = r ^= r << 5;
-            }
-            else
-                ++k;
-        }
-        if (scanGuard != g)                       // staleness check
-            return false;
-        else {                                    // try to take submission
-            ForkJoinTask<?> t; ForkJoinTask<?>[] q; int b, i;
-            if ((b = queueBase) != queueTop &&
-                (q = submissionQueue) != null &&
-                (i = (q.length - 1) & b) >= 0) {
-                long u = (i << ASHIFT) + ABASE;
-                if ((t = q[i]) != null && queueBase == b &&
-                    UNSAFE.compareAndSwapObject(q, u, t, null)) {
-                    queueBase = b + 1;
-                    w.execTask(t);
-                }
-                return false;
-            }
-            return true;                         // all queues empty
-        }
-    }
-
-    /**
-     * Tries to enqueue worker w in wait queue and await change in
-     * worker's eventCount.  If the pool is quiescent and there is
-     * more than one worker, possibly terminates worker upon exit.
-     * Otherwise, before blocking, rescans queues to avoid missed
-     * signals.  Upon finding work, releases at least one worker
-     * (which may be the current worker). Rescans restart upon
-     * detected staleness or failure to release due to
-     * contention. Note the unusual conventions about Thread.interrupt
-     * here and elsewhere: Because interrupts are used solely to alert
-     * threads to check termination, which is checked here anyway, we
-     * clear status (using Thread.interrupted) before any call to
-     * park, so that park does not immediately return due to status
-     * being set via some other unrelated call to interrupt in user
-     * code.
-     *
-     * @param w the calling worker
-     * @param c the ctl value on entry
-     * @return true if waited or another thread was released upon enq
-     */
-    private boolean tryAwaitWork(ForkJoinWorkerThread w, long c) {
-        int v = w.eventCount;
-        w.nextWait = (int)c;                      // w's successor record
-        long nc = (long)(v & E_MASK) | ((c - AC_UNIT) & (AC_MASK|TC_MASK));
-        if (ctl != c || !UNSAFE.compareAndSwapLong(this, ctlOffset, c, nc)) {
-            long d = ctl; // return true if lost to a deq, to force scan
-            return (int)d != (int)c && (d & AC_MASK) >= (c & AC_MASK);
-        }
-        for (int sc = w.stealCount; sc != 0;) {   // accumulate stealCount
-            long s = stealCount;
-            if (UNSAFE.compareAndSwapLong(this, stealCountOffset, s, s + sc))
-                sc = w.stealCount = 0;
-            else if (w.eventCount != v)
-                return true;                      // update next time
-        }
-        if ((!shutdown || !tryTerminate(false)) &&
-            (int)c != 0 && parallelism + (int)(nc >> AC_SHIFT) == 0 &&
-            blockedCount == 0 && quiescerCount == 0)
-            idleAwaitWork(w, nc, c, v);           // quiescent
-        for (boolean rescanned = false;;) {
-            if (w.eventCount != v)
-                return true;
-            if (!rescanned) {
-                int g = scanGuard, m = g & SMASK;
-                ForkJoinWorkerThread[] ws = workers;
-                if (ws != null && m < ws.length) {
-                    rescanned = true;
-                    for (int i = 0; i <= m; ++i) {
-                        ForkJoinWorkerThread u = ws[i];
-                        if (u != null) {
-                            if (u.queueBase != u.queueTop &&
-                                !tryReleaseWaiter())
-                                rescanned = false; // contended
-                            if (w.eventCount != v)
-                                return true;
+    final WorkQueue registerWorker(ForkJoinWorkerThread wt) {
+        Thread.UncaughtExceptionHandler handler; WorkQueue[] ws; int s, ps;
+        wt.setDaemon(true);
+        if ((handler = ueh) != null)
+            wt.setUncaughtExceptionHandler(handler);
+        do {} while (!U.compareAndSwapInt(this, INDEXSEED, s = indexSeed,
+                                          s += SEED_INCREMENT) ||
+                     s == 0); // skip 0
+        WorkQueue w = new WorkQueue(this, wt, config >>> 16, s);
+        if (((ps = plock) & PL_LOCK) != 0 ||
+            !U.compareAndSwapInt(this, PLOCK, ps, ps += PL_LOCK))
+            ps = acquirePlock();
+        int nps = (ps & SHUTDOWN) | ((ps + PL_LOCK) & ~SHUTDOWN);
+        try {
+            if ((ws = workQueues) != null) {    // skip if shutting down
+                int n = ws.length, m = n - 1;
+                int r = (s << 1) | 1;           // use odd-numbered indices
+                if (ws[r &= m] != null) {       // collision
+                    int probes = 0;             // step by approx half size
+                    int step = (n <= 4) ? 2 : ((n >>> 1) & EVENMASK) + 2;
+                    while (ws[r = (r + step) & m] != null) {
+                        if (++probes >= n) {
+                            workQueues = ws = Arrays.copyOf(ws, n <<= 1);
+                            m = n - 1;
+                            probes = 0;
                         }
                     }
                 }
-                if (scanGuard != g ||              // stale
-                    (queueBase != queueTop && !tryReleaseWaiter()))
-                    rescanned = false;
-                if (!rescanned)
-                    Thread.yield();                // reduce contention
-                else
-                    Thread.interrupted();          // clear before park
+                w.eventCount = w.poolIndex = r; // volatile write orders
+                ws[r] = w;
             }
-            else {
-                w.parked = true;                   // must recheck
-                if (w.eventCount != v) {
-                    w.parked = false;
-                    return true;
-                }
-                LockSupport.park(this);
-                rescanned = w.parked = false;
-            }
+        } finally {
+            if (!U.compareAndSwapInt(this, PLOCK, ps, nps))
+                releasePlock(nps);
         }
+        wt.setName(workerNamePrefix.concat(Integer.toString(w.poolIndex)));
+        return w;
     }
 
     /**
-     * If inactivating worker w has caused pool to become
-     * quiescent, check for pool termination, and wait for event
-     * for up to SHRINK_RATE nanosecs (rescans are unnecessary in
-     * this case because quiescence reflects consensus about lack
-     * of work). On timeout, if ctl has not changed, terminate the
-     * worker. Upon its termination (see deregisterWorker), it may
-     * wake up another worker to possibly repeat this process.
+     * Final callback from terminating worker, as well as upon failure
+     * to construct or start a worker.  Removes record of worker from
+     * array, and adjusts counts. If pool is shutting down, tries to
+     * complete termination.
      *
-     * @param w the calling worker
-     * @param currentCtl the ctl value after enqueuing w
-     * @param prevCtl the ctl value if w terminated
-     * @param v the eventCount w awaits change
+     * @param wt the worker thread or null if construction failed
+     * @param ex the exception causing failure, or null if none
      */
-    private void idleAwaitWork(ForkJoinWorkerThread w, long currentCtl,
-                               long prevCtl, int v) {
-        if (w.eventCount == v) {
-            if (shutdown)
-                tryTerminate(false);
-            ForkJoinTask.helpExpungeStaleExceptions(); // help clean weak refs
-            while (ctl == currentCtl) {
-                long startTime = System.nanoTime();
-                w.parked = true;
-                if (w.eventCount == v)             // must recheck
-                    LockSupport.parkNanos(this, SHRINK_RATE);
-                w.parked = false;
-                if (w.eventCount != v)
-                    break;
-                else if (System.nanoTime() - startTime <
-                         SHRINK_RATE - (SHRINK_RATE / 10)) // timing slop
-                    Thread.interrupted();          // spurious wakeup
-                else if (UNSAFE.compareAndSwapLong(this, ctlOffset,
-                                                   currentCtl, prevCtl)) {
-                    w.terminate = true;            // restore previous
-                    w.eventCount = ((int)currentCtl + EC_UNIT) & E_MASK;
+    final void deregisterWorker(ForkJoinWorkerThread wt, Throwable ex) {
+        WorkQueue w = null;
+        if (wt != null && (w = wt.workQueue) != null) {
+            int ps;
+            w.qlock = -1;                // ensure set
+            long ns = w.nsteals, sc;     // collect steal count
+            do {} while (!U.compareAndSwapLong(this, STEALCOUNT,
+                                               sc = stealCount, sc + ns));
+            if (((ps = plock) & PL_LOCK) != 0 ||
+                !U.compareAndSwapInt(this, PLOCK, ps, ps += PL_LOCK))
+                ps = acquirePlock();
+            int nps = (ps & SHUTDOWN) | ((ps + PL_LOCK) & ~SHUTDOWN);
+            try {
+                int idx = w.poolIndex;
+                WorkQueue[] ws = workQueues;
+                if (ws != null && idx >= 0 && idx < ws.length && ws[idx] == w)
+                    ws[idx] = null;
+            } finally {
+                if (!U.compareAndSwapInt(this, PLOCK, ps, nps))
+                    releasePlock(nps);
+            }
+        }
+
+        long c;                          // adjust ctl counts
+        do {} while (!U.compareAndSwapLong
+                     (this, CTL, c = ctl, (((c - AC_UNIT) & AC_MASK) |
+                                           ((c - TC_UNIT) & TC_MASK) |
+                                           (c & ~(AC_MASK|TC_MASK)))));
+
+        if (!tryTerminate(false, false) && w != null && w.array != null) {
+            w.cancelAll();               // cancel remaining tasks
+            WorkQueue[] ws; WorkQueue v; Thread p; int u, i, e;
+            while ((u = (int)((c = ctl) >>> 32)) < 0 && (e = (int)c) >= 0) {
+                if (e > 0) {             // activate or create replacement
+                    if ((ws = workQueues) == null ||
+                        (i = e & SMASK) >= ws.length ||
+                        (v = ws[i]) != null)
+                        break;
+                    long nc = (((long)(v.nextWait & E_MASK)) |
+                               ((long)(u + UAC_UNIT) << 32));
+                    if (v.eventCount != (e | INT_SIGN))
+                        break;
+                    if (U.compareAndSwapLong(this, CTL, c, nc)) {
+                        v.eventCount = (e + E_SEQ) & E_MASK;
+                        if ((p = v.parker) != null)
+                            U.unpark(p);
+                        break;
+                    }
+                }
+                else {
+                    if ((short)u < 0)
+                        tryAddWorker();
                     break;
                 }
             }
         }
+        if (ex == null)                     // help clean refs on way out
+            ForkJoinTask.helpExpungeStaleExceptions();
+        else                                // rethrow
+            ForkJoinTask.rethrow(ex);
     }
 
     // Submissions
 
     /**
-     * Enqueues the given task in the submissionQueue.  Same idea as
-     * ForkJoinWorkerThread.pushTask except for use of submissionLock.
+     * Unless shutting down, adds the given task to a submission queue
+     * at submitter's current queue index (modulo submission
+     * range). Only the most common path is directly handled in this
+     * method. All others are relayed to fullExternalPush.
      *
-     * @param t the task
+     * @param task the task. Caller must ensure non-null.
      */
-    private void addSubmission(ForkJoinTask<?> t) {
-        final ReentrantLock lock = this.submissionLock;
-        lock.lock();
-        try {
-            ForkJoinTask<?>[] q; int s, m;
-            if ((q = submissionQueue) != null) {    // ignore if queue removed
-                long u = (((s = queueTop) & (m = q.length-1)) << ASHIFT)+ABASE;
-                UNSAFE.putOrderedObject(q, u, t);
-                queueTop = s + 1;
-                if (s - queueBase == m)
-                    growSubmissionQueue();
+    final void externalPush(ForkJoinTask<?> task) {
+        WorkQueue[] ws; WorkQueue q; Submitter z; int m; ForkJoinTask<?>[] a;
+        if ((z = submitters.get()) != null && plock > 0 &&
+            (ws = workQueues) != null && (m = (ws.length - 1)) >= 0 &&
+            (q = ws[m & z.seed & SQMASK]) != null &&
+            U.compareAndSwapInt(q, QLOCK, 0, 1)) { // lock
+            int b = q.base, s = q.top, n, an;
+            if ((a = q.array) != null && (an = a.length) > (n = s + 1 - b)) {
+                int j = (((an - 1) & s) << ASHIFT) + ABASE;
+                U.putOrderedObject(a, j, task);
+                q.top = s + 1;                     // push on to deque
+                q.qlock = 0;
+                if (n <= 2)
+                    signalWork(q);
+                return;
             }
-        } finally {
-            lock.unlock();
+            q.qlock = 0;
         }
-        signalWork();
+        fullExternalPush(task);
     }
 
-    //  (pollSubmission is defined below with exported methods)
-
     /**
-     * Creates or doubles submissionQueue array.
-     * Basically identical to ForkJoinWorkerThread version.
+     * Full version of externalPush. This method is called, among
+     * other times, upon the first submission of the first task to the
+     * pool, so must perform secondary initialization (via
+     * initWorkers). It also detects first submission by an external
+     * thread by looking up its ThreadLocal, and creates a new shared
+     * queue if the one at index if empty or contended. The plock lock
+     * body must be exception-free (so no try/finally) so we
+     * optimistically allocate new queues outside the lock and throw
+     * them away if (very rarely) not needed.
      */
-    private void growSubmissionQueue() {
-        ForkJoinTask<?>[] oldQ = submissionQueue;
-        int size = oldQ != null ? oldQ.length << 1 : INITIAL_QUEUE_CAPACITY;
-        if (size > MAXIMUM_QUEUE_CAPACITY)
-            throw new RejectedExecutionException("Queue capacity exceeded");
-        if (size < INITIAL_QUEUE_CAPACITY)
-            size = INITIAL_QUEUE_CAPACITY;
-        ForkJoinTask<?>[] q = submissionQueue = new ForkJoinTask<?>[size];
-        int mask = size - 1;
-        int top = queueTop;
-        int oldMask;
-        if (oldQ != null && (oldMask = oldQ.length - 1) >= 0) {
-            for (int b = queueBase; b != top; ++b) {
-                long u = ((b & oldMask) << ASHIFT) + ABASE;
-                Object x = UNSAFE.getObjectVolatile(oldQ, u);
-                if (x != null && UNSAFE.compareAndSwapObject(oldQ, u, x, null))
-                    UNSAFE.putObjectVolatile
-                        (q, ((b & mask) << ASHIFT) + ABASE, x);
+    private void fullExternalPush(ForkJoinTask<?> task) {
+        int r = 0; // random index seed
+        for (Submitter z = submitters.get();;) {
+            WorkQueue[] ws; WorkQueue q; int ps, m, k;
+            if (z == null) {
+                if (U.compareAndSwapInt(this, INDEXSEED, r = indexSeed,
+                                        r += SEED_INCREMENT) && r != 0)
+                    submitters.set(z = new Submitter(r));
             }
-        }
-    }
-
-    // Blocking support
-
-    /**
-     * Tries to increment blockedCount, decrement active count
-     * (sometimes implicitly) and possibly release or create a
-     * compensating worker in preparation for blocking. Fails
-     * on contention or termination.
-     *
-     * @return true if the caller can block, else should recheck and retry
-     */
-    private boolean tryPreBlock() {
-        int b = blockedCount;
-        if (UNSAFE.compareAndSwapInt(this, blockedCountOffset, b, b + 1)) {
-            int pc = parallelism;
-            do {
-                ForkJoinWorkerThread[] ws; ForkJoinWorkerThread w;
-                int e, ac, tc, i;
-                long c = ctl;
-                int u = (int)(c >>> 32);
-                if ((e = (int)c) < 0) {
-                                                 // skip -- terminating
-                }
-                else if ((ac = (u >> UAC_SHIFT)) <= 0 && e != 0 &&
-                         (ws = workers) != null &&
-                         (i = ~e & SMASK) < ws.length &&
-                         (w = ws[i]) != null) {
-                    long nc = ((long)(w.nextWait & E_MASK) |
-                               (c & (AC_MASK|TC_MASK)));
-                    if (w.eventCount == e &&
-                        UNSAFE.compareAndSwapLong(this, ctlOffset, c, nc)) {
-                        w.eventCount = (e + EC_UNIT) & E_MASK;
-                        if (w.parked)
-                            UNSAFE.unpark(w);
-                        return true;             // release an idle worker
-                    }
-                }
-                else if ((tc = (short)(u >>> UTC_SHIFT)) >= 0 && ac + pc > 1) {
-                    long nc = ((c - AC_UNIT) & AC_MASK) | (c & ~AC_MASK);
-                    if (UNSAFE.compareAndSwapLong(this, ctlOffset, c, nc))
-                        return true;             // no compensation needed
-                }
-                else if (tc + pc < MAX_ID) {
-                    long nc = ((c + TC_UNIT) & TC_MASK) | (c & ~TC_MASK);
-                    if (UNSAFE.compareAndSwapLong(this, ctlOffset, c, nc)) {
-                        addWorker();
-                        return true;            // create a replacement
-                    }
-                }
-                // try to back out on any failure and let caller retry
-            } while (!UNSAFE.compareAndSwapInt(this, blockedCountOffset,
-                                               b = blockedCount, b - 1));
-        }
-        return false;
-    }
-
-    /**
-     * Decrements blockedCount and increments active count.
-     */
-    private void postBlock() {
-        long c;
-        do {} while (!UNSAFE.compareAndSwapLong(this, ctlOffset,  // no mask
-                                                c = ctl, c + AC_UNIT));
-        int b;
-        do {} while (!UNSAFE.compareAndSwapInt(this, blockedCountOffset,
-                                               b = blockedCount, b - 1));
-    }
-
-    /**
-     * Possibly blocks waiting for the given task to complete, or
-     * cancels the task if terminating.  Fails to wait if contended.
-     *
-     * @param joinMe the task
-     */
-    final void tryAwaitJoin(ForkJoinTask<?> joinMe) {
-        Thread.interrupted(); // clear interrupts before checking termination
-        if (joinMe.status >= 0) {
-            if (tryPreBlock()) {
-                joinMe.tryAwaitDone(0L);
-                postBlock();
+            else if (r == 0) {               // move to a different index
+                r = z.seed;
+                r ^= r << 13;                // same xorshift as WorkQueues
+                r ^= r >>> 17;
+                z.seed = r ^ (r << 5);
             }
-            else if ((ctl & STOP_BIT) != 0L)
-                joinMe.cancelIgnoringExceptions();
-        }
-    }
-
-    /**
-     * Possibly blocks the given worker waiting for joinMe to
-     * complete or timeout.
-     *
-     * @param joinMe the task
-     * @param nanos the wait time for underlying Object.wait
-     */
-    final void timedAwaitJoin(ForkJoinTask<?> joinMe, long nanos) {
-        while (joinMe.status >= 0) {
-            Thread.interrupted();
-            if ((ctl & STOP_BIT) != 0L) {
-                joinMe.cancelIgnoringExceptions();
-                break;
-            }
-            if (tryPreBlock()) {
-                long last = System.nanoTime();
-                while (joinMe.status >= 0) {
-                    long millis = TimeUnit.NANOSECONDS.toMillis(nanos);
-                    if (millis <= 0)
-                        break;
-                    joinMe.tryAwaitDone(millis);
-                    if (joinMe.status < 0)
-                        break;
-                    if ((ctl & STOP_BIT) != 0L) {
-                        joinMe.cancelIgnoringExceptions();
-                        break;
-                    }
-                    long now = System.nanoTime();
-                    nanos -= now - last;
-                    last = now;
-                }
-                postBlock();
-                break;
-            }
-        }
-    }
-
-    /**
-     * If necessary, compensates for blocker, and blocks.
-     */
-    private void awaitBlocker(ManagedBlocker blocker)
-        throws InterruptedException {
-        while (!blocker.isReleasable()) {
-            if (tryPreBlock()) {
-                try {
-                    do {} while (!blocker.isReleasable() && !blocker.block());
-                } finally {
-                    postBlock();
-                }
-                break;
-            }
-        }
-    }
-
-    // Creating, registering and deregistring workers
-
-    /**
-     * Tries to create and start a worker; minimally rolls back counts
-     * on failure.
-     */
-    private void addWorker() {
-        Throwable ex = null;
-        ForkJoinWorkerThread t = null;
-        try {
-            t = factory.newThread(this);
-        } catch (Throwable e) {
-            ex = e;
-        }
-        if (t == null) {  // null or exceptional factory return
-            long c;       // adjust counts
-            do {} while (!UNSAFE.compareAndSwapLong
-                         (this, ctlOffset, c = ctl,
-                          (((c - AC_UNIT) & AC_MASK) |
-                           ((c - TC_UNIT) & TC_MASK) |
-                           (c & ~(AC_MASK|TC_MASK)))));
-            // Propagate exception if originating from an external caller
-            if (!tryTerminate(false) && ex != null &&
-                !(Thread.currentThread() instanceof ForkJoinWorkerThread))
-                SneakyThrow.sneakyThrow(ex); // android-changed
-        }
-        else
-            t.start();
-    }
-
-    /**
-     * Callback from ForkJoinWorkerThread constructor to assign a
-     * public name
-     */
-    final String nextWorkerName() {
-        for (int n;;) {
-            if (UNSAFE.compareAndSwapInt(this, nextWorkerNumberOffset,
-                                         n = nextWorkerNumber, ++n))
-                return workerNamePrefix + n;
-        }
-    }
-
-    /**
-     * Callback from ForkJoinWorkerThread constructor to
-     * determine its poolIndex and record in workers array.
-     *
-     * @param w the worker
-     * @return the worker's pool index
-     */
-    final int registerWorker(ForkJoinWorkerThread w) {
-        /*
-         * In the typical case, a new worker acquires the lock, uses
-         * next available index and returns quickly.  Since we should
-         * not block callers (ultimately from signalWork or
-         * tryPreBlock) waiting for the lock needed to do this, we
-         * instead help release other workers while waiting for the
-         * lock.
-         */
-        for (int g;;) {
-            ForkJoinWorkerThread[] ws;
-            if (((g = scanGuard) & SG_UNIT) == 0 &&
-                UNSAFE.compareAndSwapInt(this, scanGuardOffset,
-                                         g, g | SG_UNIT)) {
-                int k = nextWorkerIndex;
-                try {
-                    if ((ws = workers) != null) { // ignore on shutdown
-                        int n = ws.length;
-                        if (k < 0 || k >= n || ws[k] != null) {
-                            for (k = 0; k < n && ws[k] != null; ++k)
-                                ;
-                            if (k == n)
-                                ws = workers = Arrays.copyOf(ws, n << 1);
+            else if ((ps = plock) < 0)
+                throw new RejectedExecutionException();
+            else if (ps == 0 || (ws = workQueues) == null ||
+                     (m = ws.length - 1) < 0)
+                initWorkers();
+            else if ((q = ws[k = r & m & SQMASK]) != null) {
+                if (q.qlock == 0 && U.compareAndSwapInt(q, QLOCK, 0, 1)) {
+                    ForkJoinTask<?>[] a = q.array;
+                    int s = q.top;
+                    boolean submitted = false;
+                    try {                      // locked version of push
+                        if ((a != null && a.length > s + 1 - q.base) ||
+                            (a = q.growArray()) != null) {   // must presize
+                            int j = (((a.length - 1) & s) << ASHIFT) + ABASE;
+                            U.putOrderedObject(a, j, task);
+                            q.top = s + 1;
+                            submitted = true;
                         }
-                        ws[k] = w;
-                        nextWorkerIndex = k + 1;
-                        int m = g & SMASK;
-                        g = (k > m) ? ((m << 1) + 1) & SMASK : g + (SG_UNIT<<1);
+                    } finally {
+                        q.qlock = 0;  // unlock
                     }
-                } finally {
-                    scanGuard = g;
+                    if (submitted) {
+                        signalWork(q);
+                        return;
+                    }
                 }
-                return k;
+                r = 0; // move on failure
             }
-            else if ((ws = workers) != null) { // help release others
-                for (ForkJoinWorkerThread u : ws) {
-                    if (u != null && u.queueBase != u.queueTop) {
-                        if (tryReleaseWaiter())
+            else if (((ps = plock) & PL_LOCK) == 0) { // create new queue
+                q = new WorkQueue(this, null, SHARED_QUEUE, r);
+                if (((ps = plock) & PL_LOCK) != 0 ||
+                    !U.compareAndSwapInt(this, PLOCK, ps, ps += PL_LOCK))
+                    ps = acquirePlock();
+                if ((ws = workQueues) != null && k < ws.length && ws[k] == null)
+                    ws[k] = q;
+                int nps = (ps & SHUTDOWN) | ((ps + PL_LOCK) & ~SHUTDOWN);
+                if (!U.compareAndSwapInt(this, PLOCK, ps, nps))
+                    releasePlock(nps);
+            }
+            else
+                r = 0; // try elsewhere while lock held
+        }
+    }
+
+    // Maintaining ctl counts
+
+    /**
+     * Increments active count; mainly called upon return from blocking.
+     */
+    final void incrementActiveCount() {
+        long c;
+        do {} while (!U.compareAndSwapLong(this, CTL, c = ctl, c + AC_UNIT));
+    }
+
+    /**
+     * Tries to create or activate a worker if too few are active.
+     *
+     * @param q the (non-null) queue holding tasks to be signalled
+     */
+    final void signalWork(WorkQueue q) {
+        int hint = q.poolIndex;
+        long c; int e, u, i, n; WorkQueue[] ws; WorkQueue w; Thread p;
+        while ((u = (int)((c = ctl) >>> 32)) < 0) {
+            if ((e = (int)c) > 0) {
+                if ((ws = workQueues) != null && ws.length > (i = e & SMASK) &&
+                    (w = ws[i]) != null && w.eventCount == (e | INT_SIGN)) {
+                    long nc = (((long)(w.nextWait & E_MASK)) |
+                               ((long)(u + UAC_UNIT) << 32));
+                    if (U.compareAndSwapLong(this, CTL, c, nc)) {
+                        w.hint = hint;
+                        w.eventCount = (e + E_SEQ) & E_MASK;
+                        if ((p = w.parker) != null)
+                            U.unpark(p);
+                        break;
+                    }
+                    if (q.top - q.base <= 0)
+                        break;
+                }
+                else
+                    break;
+            }
+            else {
+                if ((short)u < 0)
+                    tryAddWorker();
+                break;
+            }
+        }
+    }
+
+    // Scanning for tasks
+
+    /**
+     * Top-level runloop for workers, called by ForkJoinWorkerThread.run.
+     */
+    final void runWorker(WorkQueue w) {
+        w.growArray(); // allocate queue
+        do { w.runTask(scan(w)); } while (w.qlock >= 0);
+    }
+
+    /**
+     * Scans for and, if found, returns one task, else possibly
+     * inactivates the worker. This method operates on single reads of
+     * volatile state and is designed to be re-invoked continuously,
+     * in part because it returns upon detecting inconsistencies,
+     * contention, or state changes that indicate possible success on
+     * re-invocation.
+     *
+     * The scan searches for tasks across queues (starting at a random
+     * index, and relying on registerWorker to irregularly scatter
+     * them within array to avoid bias), checking each at least twice.
+     * The scan terminates upon either finding a non-empty queue, or
+     * completing the sweep. If the worker is not inactivated, it
+     * takes and returns a task from this queue. Otherwise, if not
+     * activated, it signals workers (that may include itself) and
+     * returns so caller can retry. Also returns for true if the
+     * worker array may have changed during an empty scan.  On failure
+     * to find a task, we take one of the following actions, after
+     * which the caller will retry calling this method unless
+     * terminated.
+     *
+     * * If pool is terminating, terminate the worker.
+     *
+     * * If not already enqueued, try to inactivate and enqueue the
+     * worker on wait queue. Or, if inactivating has caused the pool
+     * to be quiescent, relay to idleAwaitWork to possibly shrink
+     * pool.
+     *
+     * * If already enqueued and none of the above apply, possibly
+     * park awaiting signal, else lingering to help scan and signal.
+     *
+     * * If a non-empty queue discovered or left as a hint,
+     * help wake up other workers before return.
+     *
+     * @param w the worker (via its WorkQueue)
+     * @return a task or null if none found
+     */
+    private final ForkJoinTask<?> scan(WorkQueue w) {
+        WorkQueue[] ws; int m;
+        int ps = plock;                          // read plock before ws
+        if (w != null && (ws = workQueues) != null && (m = ws.length - 1) >= 0) {
+            int ec = w.eventCount;               // ec is negative if inactive
+            int r = w.seed; r ^= r << 13; r ^= r >>> 17; w.seed = r ^= r << 5;
+            w.hint = -1;                         // update seed and clear hint
+            int j = ((m + m + 1) | MIN_SCAN) & MAX_SCAN;
+            do {
+                WorkQueue q; ForkJoinTask<?>[] a; int b;
+                if ((q = ws[(r + j) & m]) != null && (b = q.base) - q.top < 0 &&
+                    (a = q.array) != null) {     // probably nonempty
+                    int i = (((a.length - 1) & b) << ASHIFT) + ABASE;
+                    ForkJoinTask<?> t = (ForkJoinTask<?>)
+                        U.getObjectVolatile(a, i);
+                    if (q.base == b && ec >= 0 && t != null &&
+                        U.compareAndSwapObject(a, i, t, null)) {
+                        if ((q.base = b + 1) - q.top < 0)
+                            signalWork(q);
+                        return t;                // taken
+                    }
+                    else if ((ec < 0 || j < m) && (int)(ctl >> AC_SHIFT) <= 0) {
+                        w.hint = (r + j) & m;    // help signal below
+                        break;                   // cannot take
+                    }
+                }
+            } while (--j >= 0);
+
+            int h, e, ns; long c, sc; WorkQueue q;
+            if ((ns = w.nsteals) != 0) {
+                if (U.compareAndSwapLong(this, STEALCOUNT,
+                                         sc = stealCount, sc + ns))
+                    w.nsteals = 0;               // collect steals and rescan
+            }
+            else if (plock != ps)                // consistency check
+                ;                                // skip
+            else if ((e = (int)(c = ctl)) < 0)
+                w.qlock = -1;                    // pool is terminating
+            else {
+                if ((h = w.hint) < 0) {
+                    if (ec >= 0) {               // try to enqueue/inactivate
+                        long nc = (((long)ec |
+                                    ((c - AC_UNIT) & (AC_MASK|TC_MASK))));
+                        w.nextWait = e;          // link and mark inactive
+                        w.eventCount = ec | INT_SIGN;
+                        if (ctl != c || !U.compareAndSwapLong(this, CTL, c, nc))
+                            w.eventCount = ec;   // unmark on CAS failure
+                        else if ((int)(c >> AC_SHIFT) == 1 - (config & SMASK))
+                            idleAwaitWork(w, nc, c);
+                    }
+                    else if (w.eventCount < 0 && !tryTerminate(false, false) &&
+                             ctl == c) {         // block
+                        Thread wt = Thread.currentThread();
+                        Thread.interrupted();    // clear status
+                        U.putObject(wt, PARKBLOCKER, this);
+                        w.parker = wt;           // emulate LockSupport.park
+                        if (w.eventCount < 0)    // recheck
+                            U.park(false, 0L);
+                        w.parker = null;
+                        U.putObject(wt, PARKBLOCKER, null);
+                    }
+                }
+                if ((h >= 0 || (h = w.hint) >= 0) &&
+                    (ws = workQueues) != null && h < ws.length &&
+                    (q = ws[h]) != null) {      // signal others before retry
+                    WorkQueue v; Thread p; int u, i, s;
+                    for (int n = (config & SMASK) >>> 1;;) {
+                        int idleCount = (w.eventCount < 0) ? 0 : -1;
+                        if (((s = idleCount - q.base + q.top) <= n &&
+                             (n = s) <= 0) ||
+                            (u = (int)((c = ctl) >>> 32)) >= 0 ||
+                            (e = (int)c) <= 0 || m < (i = e & SMASK) ||
+                            (v = ws[i]) == null)
+                            break;
+                        long nc = (((long)(v.nextWait & E_MASK)) |
+                                   ((long)(u + UAC_UNIT) << 32));
+                        if (v.eventCount != (e | INT_SIGN) ||
+                            !U.compareAndSwapLong(this, CTL, c, nc))
+                            break;
+                        v.hint = h;
+                        v.eventCount = (e + E_SEQ) & E_MASK;
+                        if ((p = v.parker) != null)
+                            U.unpark(p);
+                        if (--n <= 0)
+                            break;
+                    }
+                }
+            }
+        }
+        return null;
+    }
+
+    /**
+     * If inactivating worker w has caused the pool to become
+     * quiescent, checks for pool termination, and, so long as this is
+     * not the only worker, waits for event for up to a given
+     * duration.  On timeout, if ctl has not changed, terminates the
+     * worker, which will in turn wake up another worker to possibly
+     * repeat this process.
+     *
+     * @param w the calling worker
+     * @param currentCtl the ctl value triggering possible quiescence
+     * @param prevCtl the ctl value to restore if thread is terminated
+     */
+    private void idleAwaitWork(WorkQueue w, long currentCtl, long prevCtl) {
+        if (w != null && w.eventCount < 0 &&
+            !tryTerminate(false, false) && (int)prevCtl != 0) {
+            int dc = -(short)(currentCtl >>> TC_SHIFT);
+            long parkTime = dc < 0 ? FAST_IDLE_TIMEOUT: (dc + 1) * IDLE_TIMEOUT;
+            long deadline = System.nanoTime() + parkTime - TIMEOUT_SLOP;
+            Thread wt = Thread.currentThread();
+            while (ctl == currentCtl) {
+                Thread.interrupted();  // timed variant of version in scan()
+                U.putObject(wt, PARKBLOCKER, this);
+                w.parker = wt;
+                if (ctl == currentCtl)
+                    U.park(false, parkTime);
+                w.parker = null;
+                U.putObject(wt, PARKBLOCKER, null);
+                if (ctl != currentCtl)
+                    break;
+                if (deadline - System.nanoTime() <= 0L &&
+                    U.compareAndSwapLong(this, CTL, currentCtl, prevCtl)) {
+                    w.eventCount = (w.eventCount + E_SEQ) | E_MASK;
+                    w.qlock = -1;   // shrink
+                    break;
+                }
+            }
+        }
+    }
+
+    /**
+     * Scans through queues looking for work while joining a task; if
+     * any present, signals. May return early if more signalling is
+     * detectably unneeded.
+     *
+     * @param task return early if done
+     * @param origin an index to start scan
+     */
+    private void helpSignal(ForkJoinTask<?> task, int origin) {
+        WorkQueue[] ws; WorkQueue w; Thread p; long c; int m, u, e, i, s;
+        if (task != null && task.status >= 0 &&
+            (u = (int)(ctl >>> 32)) < 0 && (u >> UAC_SHIFT) < 0 &&
+            (ws = workQueues) != null && (m = ws.length - 1) >= 0) {
+            outer: for (int k = origin, j = m; j >= 0; --j) {
+                WorkQueue q = ws[k++ & m];
+                for (int n = m;;) { // limit to at most m signals
+                    if (task.status < 0)
+                        break outer;
+                    if (q == null ||
+                        ((s = -q.base + q.top) <= n && (n = s) <= 0))
+                        break;
+                    if ((u = (int)((c = ctl) >>> 32)) >= 0 ||
+                        (e = (int)c) <= 0 || m < (i = e & SMASK) ||
+                        (w = ws[i]) == null)
+                        break outer;
+                    long nc = (((long)(w.nextWait & E_MASK)) |
+                               ((long)(u + UAC_UNIT) << 32));
+                    if (w.eventCount != (e | INT_SIGN))
+                        break outer;
+                    if (U.compareAndSwapLong(this, CTL, c, nc)) {
+                        w.eventCount = (e + E_SEQ) & E_MASK;
+                        if ((p = w.parker) != null)
+                            U.unpark(p);
+                        if (--n <= 0)
                             break;
                     }
                 }
@@ -1171,201 +1852,630 @@
     }
 
     /**
-     * Final callback from terminating worker.  Removes record of
-     * worker from array, and adjusts counts. If pool is shutting
-     * down, tries to complete termination.
+     * Tries to locate and execute tasks for a stealer of the given
+     * task, or in turn one of its stealers, Traces currentSteal ->
+     * currentJoin links looking for a thread working on a descendant
+     * of the given task and with a non-empty queue to steal back and
+     * execute tasks from. The first call to this method upon a
+     * waiting join will often entail scanning/search, (which is OK
+     * because the joiner has nothing better to do), but this method
+     * leaves hints in workers to speed up subsequent calls. The
+     * implementation is very branchy to cope with potential
+     * inconsistencies or loops encountering chains that are stale,
+     * unknown, or so long that they are likely cyclic.
      *
-     * @param w the worker
+     * @param joiner the joining worker
+     * @param task the task to join
+     * @return 0 if no progress can be made, negative if task
+     * known complete, else positive
      */
-    final void deregisterWorker(ForkJoinWorkerThread w, Throwable ex) {
-        int idx = w.poolIndex;
-        int sc = w.stealCount;
-        int steps = 0;
-        // Remove from array, adjust worker counts and collect steal count.
-        // We can intermix failed removes or adjusts with steal updates
-        do {
-            long s, c;
-            int g;
-            if (steps == 0 && ((g = scanGuard) & SG_UNIT) == 0 &&
-                UNSAFE.compareAndSwapInt(this, scanGuardOffset,
-                                         g, g |= SG_UNIT)) {
-                ForkJoinWorkerThread[] ws = workers;
-                if (ws != null && idx >= 0 &&
-                    idx < ws.length && ws[idx] == w)
-                    ws[idx] = null;    // verify
-                nextWorkerIndex = idx;
-                scanGuard = g + SG_UNIT;
-                steps = 1;
-            }
-            if (steps == 1 &&
-                UNSAFE.compareAndSwapLong(this, ctlOffset, c = ctl,
-                                          (((c - AC_UNIT) & AC_MASK) |
-                                           ((c - TC_UNIT) & TC_MASK) |
-                                           (c & ~(AC_MASK|TC_MASK)))))
-                steps = 2;
-            if (sc != 0 &&
-                UNSAFE.compareAndSwapLong(this, stealCountOffset,
-                                          s = stealCount, s + sc))
-                sc = 0;
-        } while (steps != 2 || sc != 0);
-        if (!tryTerminate(false)) {
-            if (ex != null)   // possibly replace if died abnormally
-                signalWork();
-            else
-                tryReleaseWaiter();
-        }
-    }
-
-    // Shutdown and termination
-
-    /**
-     * Possibly initiates and/or completes termination.
-     *
-     * @param now if true, unconditionally terminate, else only
-     * if shutdown and empty queue and no active workers
-     * @return true if now terminating or terminated
-     */
-    private boolean tryTerminate(boolean now) {
-        long c;
-        while (((c = ctl) & STOP_BIT) == 0) {
-            if (!now) {
-                if ((int)(c >> AC_SHIFT) != -parallelism)
-                    return false;
-                if (!shutdown || blockedCount != 0 || quiescerCount != 0 ||
-                    queueBase != queueTop) {
-                    if (ctl == c) // staleness check
-                        return false;
-                    continue;
-                }
-            }
-            if (UNSAFE.compareAndSwapLong(this, ctlOffset, c, c | STOP_BIT))
-                startTerminating();
-        }
-        if ((short)(c >>> TC_SHIFT) == -parallelism) { // signal when 0 workers
-            final ReentrantLock lock = this.submissionLock;
-            lock.lock();
-            try {
-                termination.signalAll();
-            } finally {
-                lock.unlock();
-            }
-        }
-        return true;
-    }
-
-    /**
-     * Runs up to three passes through workers: (0) Setting
-     * termination status for each worker, followed by wakeups up to
-     * queued workers; (1) helping cancel tasks; (2) interrupting
-     * lagging threads (likely in external tasks, but possibly also
-     * blocked in joins).  Each pass repeats previous steps because of
-     * potential lagging thread creation.
-     */
-    private void startTerminating() {
-        cancelSubmissions();
-        for (int pass = 0; pass < 3; ++pass) {
-            ForkJoinWorkerThread[] ws = workers;
-            if (ws != null) {
-                for (ForkJoinWorkerThread w : ws) {
-                    if (w != null) {
-                        w.terminate = true;
-                        if (pass > 0) {
-                            w.cancelTasks();
-                            if (pass > 1 && !w.isInterrupted()) {
-                                try {
-                                    w.interrupt();
-                                } catch (SecurityException ignore) {
-                                }
+    private int tryHelpStealer(WorkQueue joiner, ForkJoinTask<?> task) {
+        int stat = 0, steps = 0;                    // bound to avoid cycles
+        if (joiner != null && task != null) {       // hoist null checks
+            restart: for (;;) {
+                ForkJoinTask<?> subtask = task;     // current target
+                for (WorkQueue j = joiner, v;;) {   // v is stealer of subtask
+                    WorkQueue[] ws; int m, s, h;
+                    if ((s = task.status) < 0) {
+                        stat = s;
+                        break restart;
+                    }
+                    if ((ws = workQueues) == null || (m = ws.length - 1) <= 0)
+                        break restart;              // shutting down
+                    if ((v = ws[h = (j.hint | 1) & m]) == null ||
+                        v.currentSteal != subtask) {
+                        for (int origin = h;;) {    // find stealer
+                            if (((h = (h + 2) & m) & 15) == 1 &&
+                                (subtask.status < 0 || j.currentJoin != subtask))
+                                continue restart;   // occasional staleness check
+                            if ((v = ws[h]) != null &&
+                                v.currentSteal == subtask) {
+                                j.hint = h;        // save hint
+                                break;
+                            }
+                            if (h == origin)
+                                break restart;      // cannot find stealer
+                        }
+                    }
+                    for (;;) { // help stealer or descend to its stealer
+                        ForkJoinTask[] a;  int b;
+                        if (subtask.status < 0)     // surround probes with
+                            continue restart;       //   consistency checks
+                        if ((b = v.base) - v.top < 0 && (a = v.array) != null) {
+                            int i = (((a.length - 1) & b) << ASHIFT) + ABASE;
+                            ForkJoinTask<?> t =
+                                (ForkJoinTask<?>)U.getObjectVolatile(a, i);
+                            if (subtask.status < 0 || j.currentJoin != subtask ||
+                                v.currentSteal != subtask)
+                                continue restart;   // stale
+                            stat = 1;               // apparent progress
+                            if (t != null && v.base == b &&
+                                U.compareAndSwapObject(a, i, t, null)) {
+                                v.base = b + 1;     // help stealer
+                                joiner.runSubtask(t);
+                            }
+                            else if (v.base == b && ++steps == MAX_HELP)
+                                break restart;      // v apparently stalled
+                        }
+                        else {                      // empty -- try to descend
+                            ForkJoinTask<?> next = v.currentJoin;
+                            if (subtask.status < 0 || j.currentJoin != subtask ||
+                                v.currentSteal != subtask)
+                                continue restart;   // stale
+                            else if (next == null || ++steps == MAX_HELP)
+                                break restart;      // dead-end or maybe cyclic
+                            else {
+                                subtask = next;
+                                j = v;
+                                break;
                             }
                         }
                     }
                 }
-                terminateWaiters();
             }
         }
+        return stat;
+    }
+
+    /**
+     * Analog of tryHelpStealer for CountedCompleters. Tries to steal
+     * and run tasks within the target's computation.
+     *
+     * @param task the task to join
+     * @param mode if shared, exit upon completing any task
+     * if all workers are active
+     */
+    private int helpComplete(ForkJoinTask<?> task, int mode) {
+        WorkQueue[] ws; WorkQueue q; int m, n, s, u;
+        if (task != null && (ws = workQueues) != null &&
+            (m = ws.length - 1) >= 0) {
+            for (int j = 1, origin = j;;) {
+                if ((s = task.status) < 0)
+                    return s;
+                if ((q = ws[j & m]) != null && q.pollAndExecCC(task)) {
+                    origin = j;
+                    if (mode == SHARED_QUEUE &&
+                        ((u = (int)(ctl >>> 32)) >= 0 || (u >> UAC_SHIFT) >= 0))
+                        break;
+                }
+                else if ((j = (j + 2) & m) == origin)
+                    break;
+            }
+        }
+        return 0;
+    }
+
+    /**
+     * Tries to decrement active count (sometimes implicitly) and
+     * possibly release or create a compensating worker in preparation
+     * for blocking. Fails on contention or termination. Otherwise,
+     * adds a new thread if no idle workers are available and pool
+     * may become starved.
+     */
+    final boolean tryCompensate() {
+        int pc = config & SMASK, e, i, tc; long c;
+        WorkQueue[] ws; WorkQueue w; Thread p;
+        if ((ws = workQueues) != null && (e = (int)(c = ctl)) >= 0) {
+            if (e != 0 && (i = e & SMASK) < ws.length &&
+                (w = ws[i]) != null && w.eventCount == (e | INT_SIGN)) {
+                long nc = ((long)(w.nextWait & E_MASK) |
+                           (c & (AC_MASK|TC_MASK)));
+                if (U.compareAndSwapLong(this, CTL, c, nc)) {
+                    w.eventCount = (e + E_SEQ) & E_MASK;
+                    if ((p = w.parker) != null)
+                        U.unpark(p);
+                    return true;   // replace with idle worker
+                }
+            }
+            else if ((tc = (short)(c >>> TC_SHIFT)) >= 0 &&
+                     (int)(c >> AC_SHIFT) + pc > 1) {
+                long nc = ((c - AC_UNIT) & AC_MASK) | (c & ~AC_MASK);
+                if (U.compareAndSwapLong(this, CTL, c, nc))
+                    return true;   // no compensation
+            }
+            else if (tc + pc < MAX_CAP) {
+                long nc = ((c + TC_UNIT) & TC_MASK) | (c & ~TC_MASK);
+                if (U.compareAndSwapLong(this, CTL, c, nc)) {
+                    ForkJoinWorkerThreadFactory fac;
+                    Throwable ex = null;
+                    ForkJoinWorkerThread wt = null;
+                    try {
+                        if ((fac = factory) != null &&
+                            (wt = fac.newThread(this)) != null) {
+                            wt.start();
+                            return true;
+                        }
+                    } catch (Throwable rex) {
+                        ex = rex;
+                    }
+                    deregisterWorker(wt, ex); // clean up and return false
+                }
+            }
+        }
+        return false;
+    }
+
+    /**
+     * Helps and/or blocks until the given task is done.
+     *
+     * @param joiner the joining worker
+     * @param task the task
+     * @return task status on exit
+     */
+    final int awaitJoin(WorkQueue joiner, ForkJoinTask<?> task) {
+        int s = 0;
+        if (joiner != null && task != null && (s = task.status) >= 0) {
+            ForkJoinTask<?> prevJoin = joiner.currentJoin;
+            joiner.currentJoin = task;
+            do {} while ((s = task.status) >= 0 && !joiner.isEmpty() &&
+                         joiner.tryRemoveAndExec(task)); // process local tasks
+            if (s >= 0 && (s = task.status) >= 0) {
+                helpSignal(task, joiner.poolIndex);
+                if ((s = task.status) >= 0 &&
+                    (task instanceof CountedCompleter))
+                    s = helpComplete(task, LIFO_QUEUE);
+            }
+            while (s >= 0 && (s = task.status) >= 0) {
+                if ((!joiner.isEmpty() ||           // try helping
+                     (s = tryHelpStealer(joiner, task)) == 0) &&
+                    (s = task.status) >= 0) {
+                    helpSignal(task, joiner.poolIndex);
+                    if ((s = task.status) >= 0 && tryCompensate()) {
+                        if (task.trySetSignal() && (s = task.status) >= 0) {
+                            synchronized (task) {
+                                if (task.status >= 0) {
+                                    try {                // see ForkJoinTask
+                                        task.wait();     //  for explanation
+                                    } catch (InterruptedException ie) {
+                                    }
+                                }
+                                else
+                                    task.notifyAll();
+                            }
+                        }
+                        long c;                          // re-activate
+                        do {} while (!U.compareAndSwapLong
+                                     (this, CTL, c = ctl, c + AC_UNIT));
+                    }
+                }
+            }
+            joiner.currentJoin = prevJoin;
+        }
+        return s;
+    }
+
+    /**
+     * Stripped-down variant of awaitJoin used by timed joins. Tries
+     * to help join only while there is continuous progress. (Caller
+     * will then enter a timed wait.)
+     *
+     * @param joiner the joining worker
+     * @param task the task
+     */
+    final void helpJoinOnce(WorkQueue joiner, ForkJoinTask<?> task) {
+        int s;
+        if (joiner != null && task != null && (s = task.status) >= 0) {
+            ForkJoinTask<?> prevJoin = joiner.currentJoin;
+            joiner.currentJoin = task;
+            do {} while ((s = task.status) >= 0 && !joiner.isEmpty() &&
+                         joiner.tryRemoveAndExec(task));
+            if (s >= 0 && (s = task.status) >= 0) {
+                helpSignal(task, joiner.poolIndex);
+                if ((s = task.status) >= 0 &&
+                    (task instanceof CountedCompleter))
+                    s = helpComplete(task, LIFO_QUEUE);
+            }
+            if (s >= 0 && joiner.isEmpty()) {
+                do {} while (task.status >= 0 &&
+                             tryHelpStealer(joiner, task) > 0);
+            }
+            joiner.currentJoin = prevJoin;
+        }
     }
 
     /**
-     * Polls and cancels all submissions. Called only during termination.
+     * Returns a (probably) non-empty steal queue, if one is found
+     * during a random, then cyclic scan, else null.  This method must
+     * be retried by caller if, by the time it tries to use the queue,
+     * it is empty.
+     * @param r a (random) seed for scanning
      */
-    private void cancelSubmissions() {
-        while (queueBase != queueTop) {
-            ForkJoinTask<?> task = pollSubmission();
-            if (task != null) {
-                try {
-                    task.cancel(false);
-                } catch (Throwable ignore) {
+    private WorkQueue findNonEmptyStealQueue(int r) {
+        for (WorkQueue[] ws;;) {
+            int ps = plock, m, n;
+            if ((ws = workQueues) == null || (m = ws.length - 1) < 1)
+                return null;
+            for (int j = (m + 1) << 2; ;) {
+                WorkQueue q = ws[(((r + j) << 1) | 1) & m];
+                if (q != null && (n = q.base - q.top) < 0) {
+                    if (n < -1)
+                        signalWork(q);
+                    return q;
+                }
+                else if (--j < 0) {
+                    if (plock == ps)
+                        return null;
+                    break;
                 }
             }
         }
     }
 
     /**
-     * Tries to set the termination status of waiting workers, and
-     * then wakes them up (after which they will terminate).
+     * Runs tasks until {@code isQuiescent()}. We piggyback on
+     * active count ctl maintenance, but rather than blocking
+     * when tasks cannot be found, we rescan until all others cannot
+     * find tasks either.
      */
-    private void terminateWaiters() {
-        ForkJoinWorkerThread[] ws = workers;
-        if (ws != null) {
-            ForkJoinWorkerThread w; long c; int i, e;
-            int n = ws.length;
-            while ((i = ~(e = (int)(c = ctl)) & SMASK) < n &&
-                   (w = ws[i]) != null && w.eventCount == (e & E_MASK)) {
-                if (UNSAFE.compareAndSwapLong(this, ctlOffset, c,
-                                              (long)(w.nextWait & E_MASK) |
-                                              ((c + AC_UNIT) & AC_MASK) |
-                                              (c & (TC_MASK|STOP_BIT)))) {
-                    w.terminate = true;
-                    w.eventCount = e + EC_UNIT;
-                    if (w.parked)
-                        UNSAFE.unpark(w);
+    final void helpQuiescePool(WorkQueue w) {
+        for (boolean active = true;;) {
+            ForkJoinTask<?> localTask; // exhaust local queue
+            while ((localTask = w.nextLocalTask()) != null)
+                localTask.doExec();
+            // Similar to loop in scan(), but ignoring submissions
+            WorkQueue q = findNonEmptyStealQueue(w.nextSeed());
+            if (q != null) {
+                ForkJoinTask<?> t; int b;
+                if (!active) {      // re-establish active count
+                    long c;
+                    active = true;
+                    do {} while (!U.compareAndSwapLong
+                                 (this, CTL, c = ctl, c + AC_UNIT));
+                }
+                if ((b = q.base) - q.top < 0 && (t = q.pollAt(b)) != null)
+                    w.runSubtask(t);
+            }
+            else {
+                long c;
+                if (active) {       // decrement active count without queuing
+                    active = false;
+                    do {} while (!U.compareAndSwapLong
+                                 (this, CTL, c = ctl, c -= AC_UNIT));
+                }
+                else
+                    c = ctl;        // re-increment on exit
+                if ((int)(c >> AC_SHIFT) + (config & SMASK) == 0) {
+                    do {} while (!U.compareAndSwapLong
+                                 (this, CTL, c = ctl, c + AC_UNIT));
+                    break;
                 }
             }
         }
     }
 
-    // misc ForkJoinWorkerThread support
-
     /**
-     * Increments or decrements quiescerCount. Needed only to prevent
-     * triggering shutdown if a worker is transiently inactive while
-     * checking quiescence.
+     * Gets and removes a local or stolen task for the given worker.
      *
-     * @param delta 1 for increment, -1 for decrement
+     * @return a task, if available
      */
-    final void addQuiescerCount(int delta) {
-        int c;
-        do {} while (!UNSAFE.compareAndSwapInt(this, quiescerCountOffset,
-                                               c = quiescerCount, c + delta));
+    final ForkJoinTask<?> nextTaskFor(WorkQueue w) {
+        for (ForkJoinTask<?> t;;) {
+            WorkQueue q; int b;
+            if ((t = w.nextLocalTask()) != null)
+                return t;
+            if ((q = findNonEmptyStealQueue(w.nextSeed())) == null)
+                return null;
+            if ((b = q.base) - q.top < 0 && (t = q.pollAt(b)) != null)
+                return t;
+        }
     }
 
     /**
-     * Directly increments or decrements active count without queuing.
-     * This method is used to transiently assert inactivation while
-     * checking quiescence.
+     * Returns a cheap heuristic guide for task partitioning when
+     * programmers, frameworks, tools, or languages have little or no
+     * idea about task granularity.  In essence by offering this
+     * method, we ask users only about tradeoffs in overhead vs
+     * expected throughput and its variance, rather than how finely to
+     * partition tasks.
      *
-     * @param delta 1 for increment, -1 for decrement
+     * In a steady state strict (tree-structured) computation, each
+     * thread makes available for stealing enough tasks for other
+     * threads to remain active. Inductively, if all threads play by
+     * the same rules, each thread should make available only a
+     * constant number of tasks.
+     *
+     * The minimum useful constant is just 1. But using a value of 1
+     * would require immediate replenishment upon each steal to
+     * maintain enough tasks, which is infeasible.  Further,
+     * partitionings/granularities of offered tasks should minimize
+     * steal rates, which in general means that threads nearer the top
+     * of computation tree should generate more than those nearer the
+     * bottom. In perfect steady state, each thread is at
+     * approximately the same level of computation tree. However,
+     * producing extra tasks amortizes the uncertainty of progress and
+     * diffusion assumptions.
+     *
+     * So, users will want to use values larger (but not much larger)
+     * than 1 to both smooth over transient shortages and hedge
+     * against uneven progress; as traded off against the cost of
+     * extra task overhead. We leave the user to pick a threshold
+     * value to compare with the results of this call to guide
+     * decisions, but recommend values such as 3.
+     *
+     * When all threads are active, it is on average OK to estimate
+     * surplus strictly locally. In steady-state, if one thread is
+     * maintaining say 2 surplus tasks, then so are others. So we can
+     * just use estimated queue length.  However, this strategy alone
+     * leads to serious mis-estimates in some non-steady-state
+     * conditions (ramp-up, ramp-down, other stalls). We can detect
+     * many of these by further considering the number of "idle"
+     * threads, that are known to have zero queued tasks, so
+     * compensate by a factor of (#idle/#active) threads.
+     *
+     * Note: The approximation of #busy workers as #active workers is
+     * not very good under current signalling scheme, and should be
+     * improved.
      */
-    final void addActiveCount(int delta) {
-        long d = (long)delta << AC_SHIFT;
-        long c;
-        do {} while (!UNSAFE.compareAndSwapLong(this, ctlOffset,
-                                                c = ctl, c + d));
+    static int getSurplusQueuedTaskCount() {
+        Thread t; ForkJoinWorkerThread wt; ForkJoinPool pool; WorkQueue q;
+        if (((t = Thread.currentThread()) instanceof ForkJoinWorkerThread)) {
+            int p = (pool = (wt = (ForkJoinWorkerThread)t).pool).config & SMASK;
+            int n = (q = wt.workQueue).top - q.base;
+            int a = (int)(pool.ctl >> AC_SHIFT) + p;
+            return n - (a > (p >>>= 1) ? 0 :
+                        a > (p >>>= 1) ? 1 :
+                        a > (p >>>= 1) ? 2 :
+                        a > (p >>>= 1) ? 4 :
+                        8);
+        }
+        return 0;
+    }
+
+    //  Termination
+
+    /**
+     * Possibly initiates and/or completes termination.  The caller
+     * triggering termination runs three passes through workQueues:
+     * (0) Setting termination status, followed by wakeups of queued
+     * workers; (1) cancelling all tasks; (2) interrupting lagging
+     * threads (likely in external tasks, but possibly also blocked in
+     * joins).  Each pass repeats previous steps because of potential
+     * lagging thread creation.
+     *
+     * @param now if true, unconditionally terminate, else only
+     * if no work and no active workers
+     * @param enable if true, enable shutdown when next possible
+     * @return true if now terminating or terminated
+     */
+    private boolean tryTerminate(boolean now, boolean enable) {
+        if (this == commonPool)                     // cannot shut down
+            return false;
+        for (long c;;) {
+            if (((c = ctl) & STOP_BIT) != 0) {      // already terminating
+                if ((short)(c >>> TC_SHIFT) == -(config & SMASK)) {
+                    synchronized (this) {
+                        notifyAll();                // signal when 0 workers
+                    }
+                }
+                return true;
+            }
+            if (plock >= 0) {                       // not yet enabled
+                int ps;
+                if (!enable)
+                    return false;
+                if (((ps = plock) & PL_LOCK) != 0 ||
+                    !U.compareAndSwapInt(this, PLOCK, ps, ps += PL_LOCK))
+                    ps = acquirePlock();
+                if (!U.compareAndSwapInt(this, PLOCK, ps, SHUTDOWN))
+                    releasePlock(SHUTDOWN);
+            }
+            if (!now) {                             // check if idle & no tasks
+                if ((int)(c >> AC_SHIFT) != -(config & SMASK) ||
+                    hasQueuedSubmissions())
+                    return false;
+                // Check for unqueued inactive workers. One pass suffices.
+                WorkQueue[] ws = workQueues; WorkQueue w;
+                if (ws != null) {
+                    for (int i = 1; i < ws.length; i += 2) {
+                        if ((w = ws[i]) != null && w.eventCount >= 0)
+                            return false;
+                    }
+                }
+            }
+            if (U.compareAndSwapLong(this, CTL, c, c | STOP_BIT)) {
+                for (int pass = 0; pass < 3; ++pass) {
+                    WorkQueue[] ws = workQueues;
+                    if (ws != null) {
+                        WorkQueue w; Thread wt;
+                        int n = ws.length;
+                        for (int i = 0; i < n; ++i) {
+                            if ((w = ws[i]) != null) {
+                                w.qlock = -1;
+                                if (pass > 0) {
+                                    w.cancelAll();
+                                    if (pass > 1 && (wt = w.owner) != null) {
+                                        if (!wt.isInterrupted()) {
+                                            try {
+                                                wt.interrupt();
+                                            } catch (SecurityException ignore) {
+                                            }
+                                        }
+                                        U.unpark(wt);
+                                    }
+                                }
+                            }
+                        }
+                        // Wake up workers parked on event queue
+                        int i, e; long cc; Thread p;
+                        while ((e = (int)(cc = ctl) & E_MASK) != 0 &&
+                               (i = e & SMASK) < n &&
+                               (w = ws[i]) != null) {
+                            long nc = ((long)(w.nextWait & E_MASK) |
+                                       ((cc + AC_UNIT) & AC_MASK) |
+                                       (cc & (TC_MASK|STOP_BIT)));
+                            if (w.eventCount == (e | INT_SIGN) &&
+                                U.compareAndSwapLong(this, CTL, cc, nc)) {
+                                w.eventCount = (e + E_SEQ) & E_MASK;
+                                w.qlock = -1;
+                                if ((p = w.parker) != null)
+                                    U.unpark(p);
+                            }
+                        }
+                    }
+                }
+            }
+        }
+    }
+
+    // external operations on common pool
+
+    /**
+     * Returns common pool queue for a thread that has submitted at
+     * least one task.
+     */
+    static WorkQueue commonSubmitterQueue() {
+        ForkJoinPool p; WorkQueue[] ws; int m; Submitter z;
+        return ((z = submitters.get()) != null &&
+                (p = commonPool) != null &&
+                (ws = p.workQueues) != null &&
+                (m = ws.length - 1) >= 0) ?
+            ws[m & z.seed & SQMASK] : null;
     }
 
     /**
-     * Returns the approximate (non-atomic) number of idle threads per
-     * active thread.
+     * Tries to pop the given task from submitter's queue in common pool.
      */
-    final int idlePerActive() {
-        // Approximate at powers of two for small values, saturate past 4
-        int p = parallelism;
-        int a = p + (int)(ctl >> AC_SHIFT);
-        return (a > (p >>>= 1) ? 0 :
-                a > (p >>>= 1) ? 1 :
-                a > (p >>>= 1) ? 2 :
-                a > (p >>>= 1) ? 4 :
-                8);
+    static boolean tryExternalUnpush(ForkJoinTask<?> t) {
+        ForkJoinPool p; WorkQueue[] ws; WorkQueue q; Submitter z;
+        ForkJoinTask<?>[] a;  int m, s;
+        if (t != null &&
+            (z = submitters.get()) != null &&
+            (p = commonPool) != null &&
+            (ws = p.workQueues) != null &&
+            (m = ws.length - 1) >= 0 &&
+            (q = ws[m & z.seed & SQMASK]) != null &&
+            (s = q.top) != q.base &&
+            (a = q.array) != null) {
+            long j = (((a.length - 1) & (s - 1)) << ASHIFT) + ABASE;
+            if (U.getObject(a, j) == t &&
+                U.compareAndSwapInt(q, QLOCK, 0, 1)) {
+                if (q.array == a && q.top == s && // recheck
+                    U.compareAndSwapObject(a, j, t, null)) {
+                    q.top = s - 1;
+                    q.qlock = 0;
+                    return true;
+                }
+                q.qlock = 0;
+            }
+        }
+        return false;
+    }
+
+    /**
+     * Tries to pop and run local tasks within the same computation
+     * as the given root. On failure, tries to help complete from
+     * other queues via helpComplete.
+     */
+    private void externalHelpComplete(WorkQueue q, ForkJoinTask<?> root) {
+        ForkJoinTask<?>[] a; int m;
+        if (q != null && (a = q.array) != null && (m = (a.length - 1)) >= 0 &&
+            root != null && root.status >= 0) {
+            for (;;) {
+                int s, u; Object o; CountedCompleter<?> task = null;
+                if ((s = q.top) - q.base > 0) {
+                    long j = ((m & (s - 1)) << ASHIFT) + ABASE;
+                    if ((o = U.getObject(a, j)) != null &&
+                        (o instanceof CountedCompleter)) {
+                        CountedCompleter<?> t = (CountedCompleter<?>)o, r = t;
+                        do {
+                            if (r == root) {
+                                if (U.compareAndSwapInt(q, QLOCK, 0, 1)) {
+                                    if (q.array == a && q.top == s &&
+                                        U.compareAndSwapObject(a, j, t, null)) {
+                                        q.top = s - 1;
+                                        task = t;
+                                    }
+                                    q.qlock = 0;
+                                }
+                                break;
+                            }
+                        } while ((r = r.completer) != null);
+                    }
+                }
+                if (task != null)
+                    task.doExec();
+                if (root.status < 0 ||
+                    (u = (int)(ctl >>> 32)) >= 0 || (u >> UAC_SHIFT) >= 0)
+                    break;
+                if (task == null) {
+                    helpSignal(root, q.poolIndex);
+                    if (root.status >= 0)
+                        helpComplete(root, SHARED_QUEUE);
+                    break;
+                }
+            }
+        }
+    }
+
+    /**
+     * Tries to help execute or signal availability of the given task
+     * from submitter's queue in common pool.
+     */
+    static void externalHelpJoin(ForkJoinTask<?> t) {
+        // Some hard-to-avoid overlap with tryExternalUnpush
+        ForkJoinPool p; WorkQueue[] ws; WorkQueue q, w; Submitter z;
+        ForkJoinTask<?>[] a;  int m, s, n;
+        if (t != null &&
+            (z = submitters.get()) != null &&
+            (p = commonPool) != null &&
+            (ws = p.workQueues) != null &&
+            (m = ws.length - 1) >= 0 &&
+            (q = ws[m & z.seed & SQMASK]) != null &&
+            (a = q.array) != null) {
+            int am = a.length - 1;
+            if ((s = q.top) != q.base) {
+                long j = ((am & (s - 1)) << ASHIFT) + ABASE;
+                if (U.getObject(a, j) == t &&
+                    U.compareAndSwapInt(q, QLOCK, 0, 1)) {
+                    if (q.array == a && q.top == s &&
+                        U.compareAndSwapObject(a, j, t, null)) {
+                        q.top = s - 1;
+                        q.qlock = 0;
+                        t.doExec();
+                    }
+                    else
+                        q.qlock = 0;
+                }
+            }
+            if (t.status >= 0) {
+                if (t instanceof CountedCompleter)
+                    p.externalHelpComplete(q, t);
+                else
+                    p.helpSignal(t, q.poolIndex);
+            }
+        }
+    }
+
+    /**
+     * Restricted version of helpQuiescePool for external callers
+     */
+    static void externalHelpQuiescePool() {
+        ForkJoinPool p; ForkJoinTask<?> t; WorkQueue q; int b;
+        if ((p = commonPool) != null &&
+            (q = p.findNonEmptyStealQueue(1)) != null &&
+            (b = q.base) - q.top < 0 &&
+            (t = q.pollAt(b)) != null)
+            t.doExec();
     }
 
     // Exported methods
@@ -1379,7 +2489,7 @@
      * no UncaughtExceptionHandler, and non-async LIFO processing mode.
      */
     public ForkJoinPool() {
-        this(Runtime.getRuntime().availableProcessors(),
+        this(Math.min(MAX_CAP, Runtime.getRuntime().availableProcessors()),
              defaultForkJoinWorkerThreadFactory, null, false);
     }
 
@@ -1424,31 +2534,46 @@
         checkPermission();
         if (factory == null)
             throw new NullPointerException();
-        if (parallelism <= 0 || parallelism > MAX_ID)
+        if (parallelism <= 0 || parallelism > MAX_CAP)
             throw new IllegalArgumentException();
-        this.parallelism = parallelism;
         this.factory = factory;
         this.ueh = handler;
-        this.locallyFifo = asyncMode;
+        this.config = parallelism | (asyncMode ? (FIFO_QUEUE << 16) : 0);
         long np = (long)(-parallelism); // offset ctl counts
         this.ctl = ((np << AC_SHIFT) & AC_MASK) | ((np << TC_SHIFT) & TC_MASK);
-        this.submissionQueue = new ForkJoinTask<?>[INITIAL_QUEUE_CAPACITY];
-        // initialize workers array with room for 2*parallelism if possible
-        int n = parallelism << 1;
-        if (n >= MAX_ID)
-            n = MAX_ID;
-        else { // See Hackers Delight, sec 3.2, where n < (1 << 16)
-            n |= n >>> 1; n |= n >>> 2; n |= n >>> 4; n |= n >>> 8;
-        }
-        workers = new ForkJoinWorkerThread[n + 1];
-        this.submissionLock = new ReentrantLock();
-        this.termination = submissionLock.newCondition();
+        int pn = nextPoolId();
         StringBuilder sb = new StringBuilder("ForkJoinPool-");
-        sb.append(poolNumberGenerator.incrementAndGet());
+        sb.append(Integer.toString(pn));
         sb.append("-worker-");
         this.workerNamePrefix = sb.toString();
     }
 
+    /**
+     * Constructor for common pool, suitable only for static initialization.
+     * Basically the same as above, but uses smallest possible initial footprint.
+     */
+    ForkJoinPool(int parallelism, long ctl,
+                 ForkJoinWorkerThreadFactory factory,
+                 Thread.UncaughtExceptionHandler handler) {
+        this.config = parallelism;
+        this.ctl = ctl;
+        this.factory = factory;
+        this.ueh = handler;
+        this.workerNamePrefix = "ForkJoinPool.commonPool-worker-";
+    }
+
+    /**
+     * Returns the common pool instance.
+     *
+     * @return the common pool instance
+     * @since 1.8
+     * @hide
+     */
+    public static ForkJoinPool commonPool() {
+        // assert commonPool != null : "static init error";
+        return commonPool;
+    }
+
     // Execution methods
 
     /**
@@ -1468,34 +2593,10 @@
      *         scheduled for execution
      */
     public <T> T invoke(ForkJoinTask<T> task) {
-        Thread t = Thread.currentThread();
         if (task == null)
             throw new NullPointerException();
-        if (shutdown)
-            throw new RejectedExecutionException();
-        if ((t instanceof ForkJoinWorkerThread) &&
-            ((ForkJoinWorkerThread)t).pool == this)
-            return task.invoke();  // bypass submit if in same pool
-        else {
-            addSubmission(task);
-            return task.join();
-        }
-    }
-
-    /**
-     * Unless terminating, forks task if within an ongoing FJ
-     * computation in the current pool, else submits as external task.
-     */
-    private <T> void forkOrSubmit(ForkJoinTask<T> task) {
-        ForkJoinWorkerThread w;
-        Thread t = Thread.currentThread();
-        if (shutdown)
-            throw new RejectedExecutionException();
-        if ((t instanceof ForkJoinWorkerThread) &&
-            (w = (ForkJoinWorkerThread)t).pool == this)
-            w.pushTask(task);
-        else
-            addSubmission(task);
+        externalPush(task);
+        return task.join();
     }
 
     /**
@@ -1509,7 +2610,7 @@
     public void execute(ForkJoinTask<?> task) {
         if (task == null)
             throw new NullPointerException();
-        forkOrSubmit(task);
+        externalPush(task);
     }
 
     // AbstractExecutorService methods
@@ -1526,8 +2627,8 @@
         if (task instanceof ForkJoinTask<?>) // avoid re-wrap
             job = (ForkJoinTask<?>) task;
         else
-            job = ForkJoinTask.adapt(task, null);
-        forkOrSubmit(job);
+            job = new ForkJoinTask.AdaptedRunnableAction(task);
+        externalPush(job);
     }
 
     /**
@@ -1542,7 +2643,7 @@
     public <T> ForkJoinTask<T> submit(ForkJoinTask<T> task) {
         if (task == null)
             throw new NullPointerException();
-        forkOrSubmit(task);
+        externalPush(task);
         return task;
     }
 
@@ -1552,10 +2653,8 @@
      *         scheduled for execution
      */
     public <T> ForkJoinTask<T> submit(Callable<T> task) {
-        if (task == null)
-            throw new NullPointerException();
-        ForkJoinTask<T> job = ForkJoinTask.adapt(task);
-        forkOrSubmit(job);
+        ForkJoinTask<T> job = new ForkJoinTask.AdaptedCallable<T>(task);
+        externalPush(job);
         return job;
     }
 
@@ -1565,10 +2664,8 @@
      *         scheduled for execution
      */
     public <T> ForkJoinTask<T> submit(Runnable task, T result) {
-        if (task == null)
-            throw new NullPointerException();
-        ForkJoinTask<T> job = ForkJoinTask.adapt(task, result);
-        forkOrSubmit(job);
+        ForkJoinTask<T> job = new ForkJoinTask.AdaptedRunnable<T>(task, result);
+        externalPush(job);
         return job;
     }
 
@@ -1584,8 +2681,8 @@
         if (task instanceof ForkJoinTask<?>) // avoid re-wrap
             job = (ForkJoinTask<?>) task;
         else
-            job = ForkJoinTask.adapt(task, null);
-        forkOrSubmit(job);
+            job = new ForkJoinTask.AdaptedRunnableAction(task);
+        externalPush(job);
         return job;
     }
 
@@ -1594,25 +2691,27 @@
      * @throws RejectedExecutionException {@inheritDoc}
      */
     public <T> List<Future<T>> invokeAll(Collection<? extends Callable<T>> tasks) {
-        ArrayList<ForkJoinTask<T>> forkJoinTasks =
-            new ArrayList<ForkJoinTask<T>>(tasks.size());
-        for (Callable<T> task : tasks)
-            forkJoinTasks.add(ForkJoinTask.adapt(task));
-        invoke(new InvokeAll<T>(forkJoinTasks));
+        // In previous versions of this class, this method constructed
+        // a task to run ForkJoinTask.invokeAll, but now external
+        // invocation of multiple tasks is at least as efficient.
+        ArrayList<Future<T>> futures = new ArrayList<Future<T>>(tasks.size());
 
-        @SuppressWarnings({"unchecked", "rawtypes"})
-            List<Future<T>> futures = (List<Future<T>>) (List) forkJoinTasks;
-        return futures;
-    }
-
-    static final class InvokeAll<T> extends RecursiveAction {
-        final ArrayList<ForkJoinTask<T>> tasks;
-        InvokeAll(ArrayList<ForkJoinTask<T>> tasks) { this.tasks = tasks; }
-        public void compute() {
-            try { invokeAll(tasks); }
-            catch (Exception ignore) {}
+        boolean done = false;
+        try {
+            for (Callable<T> t : tasks) {
+                ForkJoinTask<T> f = new ForkJoinTask.AdaptedCallable<T>(t);
+                futures.add(f);
+                externalPush(f);
+            }
+            for (int i = 0, size = futures.size(); i < size; i++)
+                ((ForkJoinTask<?>)futures.get(i)).quietlyJoin();
+            done = true;
+            return futures;
+        } finally {
+            if (!done)
+                for (int i = 0, size = futures.size(); i < size; i++)
+                    futures.get(i).cancel(false);
         }
-        private static final long serialVersionUID = -7914297376763021607L;
     }
 
     /**
@@ -1640,7 +2739,18 @@
      * @return the targeted parallelism level of this pool
      */
     public int getParallelism() {
-        return parallelism;
+        return config & SMASK;
+    }
+
+    /**
+     * Returns the targeted parallelism level of the common pool.
+     *
+     * @return the targeted parallelism level of the common pool
+     * @since 1.8
+     * @hide
+     */
+    public static int getCommonPoolParallelism() {
+        return commonPoolParallelism;
     }
 
     /**
@@ -1652,7 +2762,7 @@
      * @return the number of worker threads
      */
     public int getPoolSize() {
-        return parallelism + (short)(ctl >>> TC_SHIFT);
+        return (config & SMASK) + (short)(ctl >>> TC_SHIFT);
     }
 
     /**
@@ -1662,7 +2772,7 @@
      * @return {@code true} if this pool uses async mode
      */
     public boolean getAsyncMode() {
-        return locallyFifo;
+        return (config >>> 16) == FIFO_QUEUE;
     }
 
     /**
@@ -1674,8 +2784,15 @@
      * @return the number of worker threads
      */
     public int getRunningThreadCount() {
-        int r = parallelism + (int)(ctl >> AC_SHIFT);
-        return (r <= 0) ? 0 : r; // suppress momentarily negative values
+        int rc = 0;
+        WorkQueue[] ws; WorkQueue w;
+        if ((ws = workQueues) != null) {
+            for (int i = 1; i < ws.length; i += 2) {
+                if ((w = ws[i]) != null && w.isApparentlyUnblocked())
+                    ++rc;
+            }
+        }
+        return rc;
     }
 
     /**
@@ -1686,7 +2803,7 @@
      * @return the number of active threads
      */
     public int getActiveThreadCount() {
-        int r = parallelism + (int)(ctl >> AC_SHIFT) + blockedCount;
+        int r = (config & SMASK) + (int)(ctl >> AC_SHIFT);
         return (r <= 0) ? 0 : r; // suppress momentarily negative values
     }
 
@@ -1702,7 +2819,7 @@
      * @return {@code true} if all threads are currently idle
      */
     public boolean isQuiescent() {
-        return parallelism + (int)(ctl >> AC_SHIFT) + blockedCount == 0;
+        return (int)(ctl >> AC_SHIFT) + (config & SMASK) == 0;
     }
 
     /**
@@ -1717,7 +2834,15 @@
      * @return the number of steals
      */
     public long getStealCount() {
-        return stealCount;
+        long count = stealCount;
+        WorkQueue[] ws; WorkQueue w;
+        if ((ws = workQueues) != null) {
+            for (int i = 1; i < ws.length; i += 2) {
+                if ((w = ws[i]) != null)
+                    count += w.nsteals;
+            }
+        }
+        return count;
     }
 
     /**
@@ -1732,12 +2857,12 @@
      */
     public long getQueuedTaskCount() {
         long count = 0;
-        ForkJoinWorkerThread[] ws;
-        if ((short)(ctl >>> TC_SHIFT) > -parallelism &&
-            (ws = workers) != null) {
-            for (ForkJoinWorkerThread w : ws)
-                if (w != null)
-                    count -= w.queueBase - w.queueTop; // must read base first
+        WorkQueue[] ws; WorkQueue w;
+        if ((ws = workQueues) != null) {
+            for (int i = 1; i < ws.length; i += 2) {
+                if ((w = ws[i]) != null)
+                    count += w.queueSize();
+            }
         }
         return count;
     }
@@ -1750,7 +2875,15 @@
      * @return the number of queued submissions
      */
     public int getQueuedSubmissionCount() {
-        return -queueBase + queueTop;
+        int count = 0;
+        WorkQueue[] ws; WorkQueue w;
+        if ((ws = workQueues) != null) {
+            for (int i = 0; i < ws.length; i += 2) {
+                if ((w = ws[i]) != null)
+                    count += w.queueSize();
+            }
+        }
+        return count;
     }
 
     /**
@@ -1760,7 +2893,14 @@
      * @return {@code true} if there are any queued submissions
      */
     public boolean hasQueuedSubmissions() {
-        return queueBase != queueTop;
+        WorkQueue[] ws; WorkQueue w;
+        if ((ws = workQueues) != null) {
+            for (int i = 0; i < ws.length; i += 2) {
+                if ((w = ws[i]) != null && !w.isEmpty())
+                    return true;
+            }
+        }
+        return false;
     }
 
     /**
@@ -1771,16 +2911,11 @@
      * @return the next submission, or {@code null} if none
      */
     protected ForkJoinTask<?> pollSubmission() {
-        ForkJoinTask<?> t; ForkJoinTask<?>[] q; int b, i;
-        while ((b = queueBase) != queueTop &&
-               (q = submissionQueue) != null &&
-               (i = (q.length - 1) & b) >= 0) {
-            long u = (i << ASHIFT) + ABASE;
-            if ((t = q[i]) != null &&
-                queueBase == b &&
-                UNSAFE.compareAndSwapObject(q, u, t, null)) {
-                queueBase = b + 1;
-                return t;
+        WorkQueue[] ws; WorkQueue w; ForkJoinTask<?> t;
+        if ((ws = workQueues) != null) {
+            for (int i = 0; i < ws.length; i += 2) {
+                if ((w = ws[i]) != null && (t = w.poll()) != null)
+                    return t;
             }
         }
         return null;
@@ -1805,20 +2940,17 @@
      */
     protected int drainTasksTo(Collection<? super ForkJoinTask<?>> c) {
         int count = 0;
-        while (queueBase != queueTop) {
-            ForkJoinTask<?> t = pollSubmission();
-            if (t != null) {
-                c.add(t);
-                ++count;
+        WorkQueue[] ws; WorkQueue w; ForkJoinTask<?> t;
+        if ((ws = workQueues) != null) {
+            for (int i = 0; i < ws.length; ++i) {
+                if ((w = ws[i]) != null) {
+                    while ((t = w.poll()) != null) {
+                        c.add(t);
+                        ++count;
+                    }
+                }
             }
         }
-        ForkJoinWorkerThread[] ws;
-        if ((short)(ctl >>> TC_SHIFT) > -parallelism &&
-            (ws = workers) != null) {
-            for (ForkJoinWorkerThread w : ws)
-                if (w != null)
-                    count += w.drainTasksTo(c);
-        }
         return count;
     }
 
@@ -1830,21 +2962,36 @@
      * @return a string identifying this pool, as well as its state
      */
     public String toString() {
-        long st = getStealCount();
-        long qt = getQueuedTaskCount();
-        long qs = getQueuedSubmissionCount();
-        int pc = parallelism;
+        // Use a single pass through workQueues to collect counts
+        long qt = 0L, qs = 0L; int rc = 0;
+        long st = stealCount;
         long c = ctl;
+        WorkQueue[] ws; WorkQueue w;
+        if ((ws = workQueues) != null) {
+            for (int i = 0; i < ws.length; ++i) {
+                if ((w = ws[i]) != null) {
+                    int size = w.queueSize();
+                    if ((i & 1) == 0)
+                        qs += size;
+                    else {
+                        qt += size;
+                        st += w.nsteals;
+                        if (w.isApparentlyUnblocked())
+                            ++rc;
+                    }
+                }
+            }
+        }
+        int pc = (config & SMASK);
         int tc = pc + (short)(c >>> TC_SHIFT);
-        int rc = pc + (int)(c >> AC_SHIFT);
-        if (rc < 0) // ignore transient negative
-            rc = 0;
-        int ac = rc + blockedCount;
+        int ac = pc + (int)(c >> AC_SHIFT);
+        if (ac < 0) // ignore transient negative
+            ac = 0;
         String level;
         if ((c & STOP_BIT) != 0)
             level = (tc == 0) ? "Terminated" : "Terminating";
         else
-            level = shutdown ? "Shutting down" : "Running";
+            level = plock < 0 ? "Shutting down" : "Running";
         return super.toString() +
             "[" + level +
             ", parallelism = " + pc +
@@ -1858,34 +3005,41 @@
     }
 
     /**
-     * Initiates an orderly shutdown in which previously submitted
-     * tasks are executed, but no new tasks will be accepted.
-     * Invocation has no additional effect if already shut down.
-     * Tasks that are in the process of being submitted concurrently
-     * during the course of this method may or may not be rejected.
+     * Possibly initiates an orderly shutdown in which previously
+     * submitted tasks are executed, but no new tasks will be
+     * accepted. Invocation has no effect on execution state if this
+     * is the {@link #commonPool()}, and no additional effect if
+     * already shut down.  Tasks that are in the process of being
+     * submitted concurrently during the course of this method may or
+     * may not be rejected.
+     *
+     * @throws SecurityException if a security manager exists and
+     *         the caller is not permitted to modify threads
+     *         because it does not hold {@link
+     *         java.lang.RuntimePermission}{@code ("modifyThread")}
      */
     public void shutdown() {
         checkPermission();
-        shutdown = true;
-        tryTerminate(false);
+        tryTerminate(false, true);
     }
 
     /**
-     * Attempts to cancel and/or stop all tasks, and reject all
-     * subsequently submitted tasks.  Tasks that are in the process of
-     * being submitted or executed concurrently during the course of
-     * this method may or may not be rejected. This method cancels
-     * both existing and unexecuted tasks, in order to permit
-     * termination in the presence of task dependencies. So the method
-     * always returns an empty list (unlike the case for some other
-     * Executors).
+     * Possibly attempts to cancel and/or stop all tasks, and reject
+     * all subsequently submitted tasks.  Invocation has no effect on
+     * execution state if this is the {@link #commonPool()}, and no
+     * additional effect if already shut down. Otherwise, tasks that
+     * are in the process of being submitted or executed concurrently
+     * during the course of this method may or may not be
+     * rejected. This method cancels both existing and unexecuted
+     * tasks, in order to permit termination in the presence of task
+     * dependencies. So the method always returns an empty list
+     * (unlike the case for some other Executors).
      *
      * @return an empty list
      */
     public List<Runnable> shutdownNow() {
         checkPermission();
-        shutdown = true;
-        tryTerminate(true);
+        tryTerminate(true, true);
         return Collections.emptyList();
     }
 
@@ -1897,7 +3051,7 @@
     public boolean isTerminated() {
         long c = ctl;
         return ((c & STOP_BIT) != 0L &&
-                (short)(c >>> TC_SHIFT) == -parallelism);
+                (short)(c >>> TC_SHIFT) == -(config & SMASK));
     }
 
     /**
@@ -1905,7 +3059,7 @@
      * commenced but not yet completed.  This method may be useful for
      * debugging. A return of {@code true} reported a sufficient
      * period after shutdown may indicate that submitted tasks have
-     * ignored or suppressed interruption, or are waiting for IO,
+     * ignored or suppressed interruption, or are waiting for I/O,
      * causing this executor not to properly terminate. (See the
      * advisory notes for class {@link ForkJoinTask} stating that
      * tasks should not normally entail blocking operations.  But if
@@ -1916,14 +3070,7 @@
     public boolean isTerminating() {
         long c = ctl;
         return ((c & STOP_BIT) != 0L &&
-                (short)(c >>> TC_SHIFT) != -parallelism);
-    }
-
-    /**
-     * Returns true if terminating or terminated. Used by ForkJoinWorkerThread.
-     */
-    final boolean isAtLeastTerminating() {
-        return (ctl & STOP_BIT) != 0L;
+                (short)(c >>> TC_SHIFT) != -(config & SMASK));
     }
 
     /**
@@ -1932,13 +3079,15 @@
      * @return {@code true} if this pool has been shut down
      */
     public boolean isShutdown() {
-        return shutdown;
+        return plock < 0;
     }
 
     /**
-     * Blocks until all tasks have completed execution after a shutdown
-     * request, or the timeout occurs, or the current thread is
-     * interrupted, whichever happens first.
+     * Blocks until all tasks have completed execution after a
+     * shutdown request, or the timeout occurs, or the current thread
+     * is interrupted, whichever happens first. Note that the {@link
+     * #commonPool()} never terminates until program shutdown so
+     * this method will always time out.
      *
      * @param timeout the maximum time to wait
      * @param unit the time unit of the timeout argument
@@ -1949,18 +3098,21 @@
     public boolean awaitTermination(long timeout, TimeUnit unit)
         throws InterruptedException {
         long nanos = unit.toNanos(timeout);
-        final ReentrantLock lock = this.submissionLock;
-        lock.lock();
-        try {
+        if (isTerminated())
+            return true;
+        if (nanos <= 0L)
+            return false;
+        long deadline = System.nanoTime() + nanos;
+        synchronized (this) {
             for (;;) {
                 if (isTerminated())
                     return true;
-                if (nanos <= 0)
+                if (nanos <= 0L)
                     return false;
-                nanos = termination.awaitNanos(nanos);
+                long millis = TimeUnit.NANOSECONDS.toMillis(nanos);
+                wait(millis > 0L ? millis : 1L);
+                nanos = deadline - System.nanoTime();
             }
-        } finally {
-            lock.unlock();
         }
     }
 
@@ -2061,11 +3213,35 @@
         throws InterruptedException {
         Thread t = Thread.currentThread();
         if (t instanceof ForkJoinWorkerThread) {
-            ForkJoinWorkerThread w = (ForkJoinWorkerThread) t;
-            w.pool.awaitBlocker(blocker);
+            ForkJoinPool p = ((ForkJoinWorkerThread)t).pool;
+            while (!blocker.isReleasable()) { // variant of helpSignal
+                WorkQueue[] ws; WorkQueue q; int m, u;
+                if ((ws = p.workQueues) != null && (m = ws.length - 1) >= 0) {
+                    for (int i = 0; i <= m; ++i) {
+                        if (blocker.isReleasable())
+                            return;
+                        if ((q = ws[i]) != null && q.base - q.top < 0) {
+                            p.signalWork(q);
+                            if ((u = (int)(p.ctl >>> 32)) >= 0 ||
+                                (u >> UAC_SHIFT) >= 0)
+                                break;
+                        }
+                    }
+                }
+                if (p.tryCompensate()) {
+                    try {
+                        do {} while (!blocker.isReleasable() &&
+                                     !blocker.block());
+                    } finally {
+                        p.incrementActiveCount();
+                    }
+                    break;
+                }
+            }
         }
         else {
-            do {} while (!blocker.isReleasable() && !blocker.block());
+            do {} while (!blocker.isReleasable() &&
+                         !blocker.block());
         }
     }
 
@@ -2074,54 +3250,93 @@
     // implement RunnableFuture.
 
     protected <T> RunnableFuture<T> newTaskFor(Runnable runnable, T value) {
-        return (RunnableFuture<T>) ForkJoinTask.adapt(runnable, value);
+        return new ForkJoinTask.AdaptedRunnable<T>(runnable, value);
     }
 
     protected <T> RunnableFuture<T> newTaskFor(Callable<T> callable) {
-        return (RunnableFuture<T>) ForkJoinTask.adapt(callable);
+        return new ForkJoinTask.AdaptedCallable<T>(callable);
     }
 
     // Unsafe mechanics
-    private static final sun.misc.Unsafe UNSAFE;
-    private static final long ctlOffset;
-    private static final long stealCountOffset;
-    private static final long blockedCountOffset;
-    private static final long quiescerCountOffset;
-    private static final long scanGuardOffset;
-    private static final long nextWorkerNumberOffset;
-    private static final long ABASE;
+    private static final sun.misc.Unsafe U;
+    private static final long CTL;
+    private static final long PARKBLOCKER;
+    private static final int ABASE;
     private static final int ASHIFT;
+    private static final long STEALCOUNT;
+    private static final long PLOCK;
+    private static final long INDEXSEED;
+    private static final long QLOCK;
 
     static {
-        poolNumberGenerator = new AtomicInteger();
-        workerSeedGenerator = new Random();
-        modifyThreadPermission = new RuntimePermission("modifyThread");
-        defaultForkJoinWorkerThreadFactory =
-            new DefaultForkJoinWorkerThreadFactory();
+        // initialize field offsets for CAS etc
         try {
-            UNSAFE = sun.misc.Unsafe.getUnsafe();
+            U = sun.misc.Unsafe.getUnsafe();
             Class<?> k = ForkJoinPool.class;
-            ctlOffset = UNSAFE.objectFieldOffset
+            CTL = U.objectFieldOffset
                 (k.getDeclaredField("ctl"));
-            stealCountOffset = UNSAFE.objectFieldOffset
+            STEALCOUNT = U.objectFieldOffset
                 (k.getDeclaredField("stealCount"));
-            blockedCountOffset = UNSAFE.objectFieldOffset
-                (k.getDeclaredField("blockedCount"));
-            quiescerCountOffset = UNSAFE.objectFieldOffset
-                (k.getDeclaredField("quiescerCount"));
-            scanGuardOffset = UNSAFE.objectFieldOffset
-                (k.getDeclaredField("scanGuard"));
-            nextWorkerNumberOffset = UNSAFE.objectFieldOffset
-                (k.getDeclaredField("nextWorkerNumber"));
+            PLOCK = U.objectFieldOffset
+                (k.getDeclaredField("plock"));
+            INDEXSEED = U.objectFieldOffset
+                (k.getDeclaredField("indexSeed"));
+            Class<?> tk = Thread.class;
+            PARKBLOCKER = U.objectFieldOffset
+                (tk.getDeclaredField("parkBlocker"));
+            Class<?> wk = WorkQueue.class;
+            QLOCK = U.objectFieldOffset
+                (wk.getDeclaredField("qlock"));
+            Class<?> ak = ForkJoinTask[].class;
+            ABASE = U.arrayBaseOffset(ak);
+            int scale = U.arrayIndexScale(ak);
+            if ((scale & (scale - 1)) != 0)
+                throw new Error("data type scale not a power of two");
+            ASHIFT = 31 - Integer.numberOfLeadingZeros(scale);
         } catch (Exception e) {
             throw new Error(e);
         }
-        Class<?> a = ForkJoinTask[].class;
-        ABASE = UNSAFE.arrayBaseOffset(a);
-        int s = UNSAFE.arrayIndexScale(a);
-        if ((s & (s-1)) != 0)
-            throw new Error("data type scale not a power of two");
-        ASHIFT = 31 - Integer.numberOfLeadingZeros(s);
+
+        submitters = new ThreadLocal<Submitter>();
+        ForkJoinWorkerThreadFactory fac = defaultForkJoinWorkerThreadFactory =
+            new DefaultForkJoinWorkerThreadFactory();
+        modifyThreadPermission = new RuntimePermission("modifyThread");
+
+        /*
+         * Establish common pool parameters.  For extra caution,
+         * computations to set up common pool state are here; the
+         * constructor just assigns these values to fields.
+         */
+
+        int par = 0;
+        Thread.UncaughtExceptionHandler handler = null;
+        try {  // TBD: limit or report ignored exceptions?
+            String pp = System.getProperty
+                ("java.util.concurrent.ForkJoinPool.common.parallelism");
+            String hp = System.getProperty
+                ("java.util.concurrent.ForkJoinPool.common.exceptionHandler");
+            String fp = System.getProperty
+                ("java.util.concurrent.ForkJoinPool.common.threadFactory");
+            if (fp != null)
+                fac = ((ForkJoinWorkerThreadFactory)ClassLoader.
+                       getSystemClassLoader().loadClass(fp).newInstance());
+            if (hp != null)
+                handler = ((Thread.UncaughtExceptionHandler)ClassLoader.
+                           getSystemClassLoader().loadClass(hp).newInstance());
+            if (pp != null)
+                par = Integer.parseInt(pp);
+        } catch (Exception ignore) {
+        }
+
+        if (par <= 0)
+            par = Runtime.getRuntime().availableProcessors();
+        if (par > MAX_CAP)
+            par = MAX_CAP;
+        commonPoolParallelism = par;
+        long np = (long)(-par); // precompute initial ctl value
+        long ct = ((np << AC_SHIFT) & AC_MASK) | ((np << TC_SHIFT) & TC_MASK);
+
+        commonPool = new ForkJoinPool(par, ct, fac, handler);
     }
 
 }
diff --git a/luni/src/main/java/java/util/concurrent/ForkJoinTask.java b/luni/src/main/java/java/util/concurrent/ForkJoinTask.java
index 86a29d7..818788e 100644
--- a/luni/src/main/java/java/util/concurrent/ForkJoinTask.java
+++ b/luni/src/main/java/java/util/concurrent/ForkJoinTask.java
@@ -22,7 +22,6 @@
 import java.util.concurrent.TimeoutException;
 import java.util.concurrent.locks.ReentrantLock;
 import java.lang.reflect.Constructor;
-import libcore.util.SneakyThrow;
 
 /**
  * Abstract base class for tasks that run within a {@link ForkJoinPool}.
@@ -31,46 +30,59 @@
  * subtasks may be hosted by a small number of actual threads in a
  * ForkJoinPool, at the price of some usage limitations.
  *
- * <p>A "main" {@code ForkJoinTask} begins execution when submitted
- * to a {@link ForkJoinPool}.  Once started, it will usually in turn
- * start other subtasks.  As indicated by the name of this class,
- * many programs using {@code ForkJoinTask} employ only methods
- * {@link #fork} and {@link #join}, or derivatives such as {@link
+ * <p>A "main" {@code ForkJoinTask} begins execution when it is
+ * explicitly submitted to a {@link ForkJoinPool}, or, if not already
+ * engaged in a ForkJoin computation, commenced in the {@link
+ * ForkJoinPool#commonPool()} via {@link #fork}, {@link #invoke}, or
+ * related methods.  Once started, it will usually in turn start other
+ * subtasks.  As indicated by the name of this class, many programs
+ * using {@code ForkJoinTask} employ only methods {@link #fork} and
+ * {@link #join}, or derivatives such as {@link
  * #invokeAll(ForkJoinTask...) invokeAll}.  However, this class also
  * provides a number of other methods that can come into play in
- * advanced usages, as well as extension mechanics that allow
- * support of new forms of fork/join processing.
+ * advanced usages, as well as extension mechanics that allow support
+ * of new forms of fork/join processing.
  *
  * <p>A {@code ForkJoinTask} is a lightweight form of {@link Future}.
  * The efficiency of {@code ForkJoinTask}s stems from a set of
  * restrictions (that are only partially statically enforceable)
- * reflecting their intended use as computational tasks calculating
- * pure functions or operating on purely isolated objects.  The
- * primary coordination mechanisms are {@link #fork}, that arranges
+ * reflecting their main use as computational tasks calculating pure
+ * functions or operating on purely isolated objects.  The primary
+ * coordination mechanisms are {@link #fork}, that arranges
  * asynchronous execution, and {@link #join}, that doesn't proceed
  * until the task's result has been computed.  Computations should
- * avoid {@code synchronized} methods or blocks, and should minimize
- * other blocking synchronization apart from joining other tasks or
- * using synchronizers such as Phasers that are advertised to
- * cooperate with fork/join scheduling. Tasks should also not perform
- * blocking IO, and should ideally access variables that are
- * completely independent of those accessed by other running
- * tasks. Minor breaches of these restrictions, for example using
- * shared output streams, may be tolerable in practice, but frequent
- * use may result in poor performance, and the potential to
- * indefinitely stall if the number of threads not waiting for IO or
- * other external synchronization becomes exhausted. This usage
- * restriction is in part enforced by not permitting checked
- * exceptions such as {@code IOExceptions} to be thrown. However,
- * computations may still encounter unchecked exceptions, that are
- * rethrown to callers attempting to join them. These exceptions may
- * additionally include {@link RejectedExecutionException} stemming
- * from internal resource exhaustion, such as failure to allocate
- * internal task queues. Rethrown exceptions behave in the same way as
- * regular exceptions, but, when possible, contain stack traces (as
- * displayed for example using {@code ex.printStackTrace()}) of both
- * the thread that initiated the computation as well as the thread
- * actually encountering the exception; minimally only the latter.
+ * ideally avoid {@code synchronized} methods or blocks, and should
+ * minimize other blocking synchronization apart from joining other
+ * tasks or using synchronizers such as Phasers that are advertised to
+ * cooperate with fork/join scheduling. Subdividable tasks should also
+ * not perform blocking I/O, and should ideally access variables that
+ * are completely independent of those accessed by other running
+ * tasks. These guidelines are loosely enforced by not permitting
+ * checked exceptions such as {@code IOExceptions} to be
+ * thrown. However, computations may still encounter unchecked
+ * exceptions, that are rethrown to callers attempting to join
+ * them. These exceptions may additionally include {@link
+ * RejectedExecutionException} stemming from internal resource
+ * exhaustion, such as failure to allocate internal task
+ * queues. Rethrown exceptions behave in the same way as regular
+ * exceptions, but, when possible, contain stack traces (as displayed
+ * for example using {@code ex.printStackTrace()}) of both the thread
+ * that initiated the computation as well as the thread actually
+ * encountering the exception; minimally only the latter.
+ *
+ * <p>It is possible to define and use ForkJoinTasks that may block,
+ * but doing do requires three further considerations: (1) Completion
+ * of few if any <em>other</em> tasks should be dependent on a task
+ * that blocks on external synchronization or I/O. Event-style async
+ * tasks that are never joined (for example, those subclassing {@link
+ * CountedCompleter}) often fall into this category.  (2) To minimize
+ * resource impact, tasks should be small; ideally performing only the
+ * (possibly) blocking action. (3) Unless the {@link
+ * ForkJoinPool.ManagedBlocker} API is used, or the number of possibly
+ * blocked tasks is known to be less than the pool's {@link
+ * ForkJoinPool#getParallelism} level, the pool cannot guarantee that
+ * enough threads will be available to ensure progress or good
+ * performance.
  *
  * <p>The primary method for awaiting completion and extracting
  * results of a task is {@link #join}, but there are several variants:
@@ -86,6 +98,13 @@
  * performs the most common form of parallel invocation: forking a set
  * of tasks and joining them all.
  *
+ * <p>In the most typical usages, a fork-join pair act like a call
+ * (fork) and return (join) from a parallel recursive function. As is
+ * the case with other forms of recursive calls, returns (joins)
+ * should be performed innermost-first. For example, {@code a.fork();
+ * b.fork(); b.join(); a.join();} is likely to be substantially more
+ * efficient than joining {@code a} before {@code b}.
+ *
  * <p>The execution status of tasks may be queried at several levels
  * of detail: {@link #isDone} is true if a task completed in any way
  * (including the case where a task was cancelled without executing);
@@ -101,18 +120,13 @@
  * <p>The ForkJoinTask class is not usually directly subclassed.
  * Instead, you subclass one of the abstract classes that support a
  * particular style of fork/join processing, typically {@link
- * RecursiveAction} for computations that do not return results, or
- * {@link RecursiveTask} for those that do.  Normally, a concrete
- * ForkJoinTask subclass declares fields comprising its parameters,
- * established in a constructor, and then defines a {@code compute}
- * method that somehow uses the control methods supplied by this base
- * class. While these methods have {@code public} access (to allow
- * instances of different task subclasses to call each other's
- * methods), some of them may only be called from within other
- * ForkJoinTasks (as may be determined using method {@link
- * #inForkJoinPool}).  Attempts to invoke them in other contexts
- * result in exceptions or errors, possibly including
- * {@code ClassCastException}.
+ * RecursiveAction} for most computations that do not return results,
+ * {@link RecursiveTask} for those that do, and {@link
+ * CountedCompleter} for those in which completed actions trigger
+ * other actions.  Normally, a concrete ForkJoinTask subclass declares
+ * fields comprising its parameters, established in a constructor, and
+ * then defines a {@code compute} method that somehow uses the control
+ * methods supplied by this base class.
  *
  * <p>Method {@link #join} and its variants are appropriate for use
  * only when completion dependencies are acyclic; that is, the
@@ -122,7 +136,17 @@
  * supports other methods and techniques (for example the use of
  * {@link Phaser}, {@link #helpQuiesce}, and {@link #complete}) that
  * may be of use in constructing custom subclasses for problems that
- * are not statically structured as DAGs.
+ * are not statically structured as DAGs. To support such usages a
+ * ForkJoinTask may be atomically <em>tagged</em> with a {@code short}
+ * value using {@link #setForkJoinTaskTag} or {@link
+ * #compareAndSetForkJoinTaskTag} and checked using {@link
+ * #getForkJoinTaskTag}. The ForkJoinTask implementation does not use
+ * these {@code protected} methods or tags for any purpose, but they
+ * may be of use in the construction of specialized subclasses.  For
+ * example, parallel graph traversals can use the supplied methods to
+ * avoid revisiting nodes/tasks that have already been processed.
+ * (Method names for tagging are bulky in part to encourage definition
+ * of methods that reflect their usage patterns.)
  *
  * <p>Most base support methods are {@code final}, to prevent
  * overriding of implementations that are intrinsically tied to the
@@ -177,29 +201,36 @@
      * The status field holds run control status bits packed into a
      * single int to minimize footprint and to ensure atomicity (via
      * CAS).  Status is initially zero, and takes on nonnegative
-     * values until completed, upon which status holds value
-     * NORMAL, CANCELLED, or EXCEPTIONAL. Tasks undergoing blocking
-     * waits by other threads have the SIGNAL bit set.  Completion of
-     * a stolen task with SIGNAL set awakens any waiters via
-     * notifyAll. Even though suboptimal for some purposes, we use
-     * basic builtin wait/notify to take advantage of "monitor
-     * inflation" in JVMs that we would otherwise need to emulate to
-     * avoid adding further per-task bookkeeping overhead.  We want
-     * these monitors to be "fat", i.e., not use biasing or thin-lock
-     * techniques, so use some odd coding idioms that tend to avoid
-     * them.
+     * values until completed, upon which status (anded with
+     * DONE_MASK) holds value NORMAL, CANCELLED, or EXCEPTIONAL. Tasks
+     * undergoing blocking waits by other threads have the SIGNAL bit
+     * set.  Completion of a stolen task with SIGNAL set awakens any
+     * waiters via notifyAll. Even though suboptimal for some
+     * purposes, we use basic builtin wait/notify to take advantage of
+     * "monitor inflation" in JVMs that we would otherwise need to
+     * emulate to avoid adding further per-task bookkeeping overhead.
+     * We want these monitors to be "fat", i.e., not use biasing or
+     * thin-lock techniques, so use some odd coding idioms that tend
+     * to avoid them, mainly by arranging that every synchronized
+     * block performs a wait, notifyAll or both.
+     *
+     * These control bits occupy only (some of) the upper half (16
+     * bits) of status field. The lower bits are used for user-defined
+     * tags.
      */
 
     /** The run status of this task */
     volatile int status; // accessed directly by pool and workers
-    private static final int NORMAL      = -1;
-    private static final int CANCELLED   = -2;
-    private static final int EXCEPTIONAL = -3;
-    private static final int SIGNAL      =  1;
+    static final int DONE_MASK   = 0xf0000000;  // mask out non-completion bits
+    static final int NORMAL      = 0xf0000000;  // must be negative
+    static final int CANCELLED   = 0xc0000000;  // must be < NORMAL
+    static final int EXCEPTIONAL = 0x80000000;  // must be < CANCELLED
+    static final int SIGNAL      = 0x00010000;  // must be >= 1 << 16
+    static final int SMASK       = 0x0000ffff;  // short bits for tags
 
     /**
-     * Marks completion and wakes up threads waiting to join this task,
-     * also clearing signal request bits.
+     * Marks completion and wakes up threads waiting to join this
+     * task.
      *
      * @param completion one of NORMAL, CANCELLED, EXCEPTIONAL
      * @return completion status on exit
@@ -208,8 +239,8 @@
         for (int s;;) {
             if ((s = status) < 0)
                 return s;
-            if (UNSAFE.compareAndSwapInt(this, statusOffset, s, completion)) {
-                if (s != 0)
+            if (U.compareAndSwapInt(this, STATUS, s, s | completion)) {
+                if ((s >>> 16) != 0)
                     synchronized (this) { notifyAll(); }
                 return completion;
             }
@@ -217,27 +248,36 @@
     }
 
     /**
-     * Tries to block a worker thread until completed or timed out.
-     * Uses Object.wait time argument conventions.
-     * May fail on contention or interrupt.
+     * Primary execution method for stolen tasks. Unless done, calls
+     * exec and records status if completed, but doesn't wait for
+     * completion otherwise.
      *
-     * @param millis if > 0, wait time.
+     * @return status on exit from this method
      */
-    final void tryAwaitDone(long millis) {
-        int s;
-        try {
-            if (((s = status) > 0 ||
-                 (s == 0 &&
-                  UNSAFE.compareAndSwapInt(this, statusOffset, 0, SIGNAL))) &&
-                status > 0) {
-                synchronized (this) {
-                    if (status > 0)
-                        wait(millis);
-                }
+    final int doExec() {
+        int s; boolean completed;
+        if ((s = status) >= 0) {
+            try {
+                completed = exec();
+            } catch (Throwable rex) {
+                return setExceptionalCompletion(rex);
             }
-        } catch (InterruptedException ie) {
-            // caller must check termination
+            if (completed)
+                s = setCompletion(NORMAL);
         }
+        return s;
+    }
+
+    /**
+     * Tries to set SIGNAL status unless already completed. Used by
+     * ForkJoinPool. Other variants are directly incorporated into
+     * externalAwaitDone etc.
+     *
+     * @return true if successful
+     */
+    final boolean trySetSignal() {
+        int s = status;
+        return s >= 0 && U.compareAndSwapInt(this, STATUS, s, s | SIGNAL);
     }
 
     /**
@@ -246,113 +286,78 @@
      */
     private int externalAwaitDone() {
         int s;
-        if ((s = status) >= 0) {
-            boolean interrupted = false;
-            synchronized (this) {
-                while ((s = status) >= 0) {
-                    if (s == 0)
-                        UNSAFE.compareAndSwapInt(this, statusOffset,
-                                                 0, SIGNAL);
-                    else {
+        ForkJoinPool.externalHelpJoin(this);
+        boolean interrupted = false;
+        while ((s = status) >= 0) {
+            if (U.compareAndSwapInt(this, STATUS, s, s | SIGNAL)) {
+                synchronized (this) {
+                    if (status >= 0) {
                         try {
                             wait();
                         } catch (InterruptedException ie) {
                             interrupted = true;
                         }
                     }
+                    else
+                        notifyAll();
                 }
             }
-            if (interrupted)
-                Thread.currentThread().interrupt();
         }
+        if (interrupted)
+            Thread.currentThread().interrupt();
         return s;
     }
 
     /**
-     * Blocks a non-worker-thread until completion or interruption or timeout.
+     * Blocks a non-worker-thread until completion or interruption.
      */
-    private int externalInterruptibleAwaitDone(long millis)
-        throws InterruptedException {
+    private int externalInterruptibleAwaitDone() throws InterruptedException {
         int s;
         if (Thread.interrupted())
             throw new InterruptedException();
-        if ((s = status) >= 0) {
-            synchronized (this) {
-                while ((s = status) >= 0) {
-                    if (s == 0)
-                        UNSAFE.compareAndSwapInt(this, statusOffset,
-                                                 0, SIGNAL);
-                    else {
-                        wait(millis);
-                        if (millis > 0L)
-                            break;
-                    }
+        ForkJoinPool.externalHelpJoin(this);
+        while ((s = status) >= 0) {
+            if (U.compareAndSwapInt(this, STATUS, s, s | SIGNAL)) {
+                synchronized (this) {
+                    if (status >= 0)
+                        wait();
+                    else
+                        notifyAll();
                 }
             }
         }
         return s;
     }
 
-    /**
-     * Primary execution method for stolen tasks. Unless done, calls
-     * exec and records status if completed, but doesn't wait for
-     * completion otherwise.
-     */
-    final void doExec() {
-        if (status >= 0) {
-            boolean completed;
-            try {
-                completed = exec();
-            } catch (Throwable rex) {
-                setExceptionalCompletion(rex);
-                return;
-            }
-            if (completed)
-                setCompletion(NORMAL); // must be outside try block
-        }
-    }
 
     /**
-     * Primary mechanics for join, get, quietlyJoin.
+     * Implementation for join, get, quietlyJoin. Directly handles
+     * only cases of already-completed, external wait, and
+     * unfork+exec.  Others are relayed to ForkJoinPool.awaitJoin.
+     *
      * @return status upon completion
      */
     private int doJoin() {
-        Thread t; ForkJoinWorkerThread w; int s; boolean completed;
-        if ((t = Thread.currentThread()) instanceof ForkJoinWorkerThread) {
-            if ((s = status) < 0)
-                return s;
-            if ((w = (ForkJoinWorkerThread)t).unpushTask(this)) {
-                try {
-                    completed = exec();
-                } catch (Throwable rex) {
-                    return setExceptionalCompletion(rex);
-                }
-                if (completed)
-                    return setCompletion(NORMAL);
-            }
-            return w.joinTask(this);
-        }
-        else
-            return externalAwaitDone();
+        int s; Thread t; ForkJoinWorkerThread wt; ForkJoinPool.WorkQueue w;
+        return (s = status) < 0 ? s :
+            ((t = Thread.currentThread()) instanceof ForkJoinWorkerThread) ?
+            (w = (wt = (ForkJoinWorkerThread)t).workQueue).
+            tryUnpush(this) && (s = doExec()) < 0 ? s :
+            wt.pool.awaitJoin(w, this) :
+            externalAwaitDone();
     }
 
     /**
-     * Primary mechanics for invoke, quietlyInvoke.
+     * Implementation for invoke, quietlyInvoke.
+     *
      * @return status upon completion
      */
     private int doInvoke() {
-        int s; boolean completed;
-        if ((s = status) < 0)
-            return s;
-        try {
-            completed = exec();
-        } catch (Throwable rex) {
-            return setExceptionalCompletion(rex);
-        }
-        if (completed)
-            return setCompletion(NORMAL);
-        else
-            return doJoin();
+        int s; Thread t; ForkJoinWorkerThread wt;
+        return (s = doExec()) < 0 ? s :
+            ((t = Thread.currentThread()) instanceof ForkJoinWorkerThread) ?
+            (wt = (ForkJoinWorkerThread)t).pool.awaitJoin(wt.workQueue, this) :
+            externalAwaitDone();
     }
 
     // Exception table support
@@ -387,7 +392,7 @@
      * any ForkJoinPool will call helpExpungeStaleExceptions when its
      * pool becomes isQuiescent.
      */
-    static final class ExceptionNode extends WeakReference<ForkJoinTask<?>>{
+    static final class ExceptionNode extends WeakReference<ForkJoinTask<?>> {
         final Throwable ex;
         ExceptionNode next;
         final long thrower;  // use id not ref to avoid weak cycles
@@ -400,34 +405,71 @@
     }
 
     /**
-     * Records exception and sets exceptional completion.
+     * Records exception and sets status.
+     *
+     * @return status on exit
+     */
+    final int recordExceptionalCompletion(Throwable ex) {
+        int s;
+        if ((s = status) >= 0) {
+            int h = System.identityHashCode(this);
+            final ReentrantLock lock = exceptionTableLock;
+            lock.lock();
+            try {
+                expungeStaleExceptions();
+                ExceptionNode[] t = exceptionTable;
+                int i = h & (t.length - 1);
+                for (ExceptionNode e = t[i]; ; e = e.next) {
+                    if (e == null) {
+                        t[i] = new ExceptionNode(this, ex, t[i]);
+                        break;
+                    }
+                    if (e.get() == this) // already present
+                        break;
+                }
+            } finally {
+                lock.unlock();
+            }
+            s = setCompletion(EXCEPTIONAL);
+        }
+        return s;
+    }
+
+    /**
+     * Records exception and possibly propagates.
      *
      * @return status on exit
      */
     private int setExceptionalCompletion(Throwable ex) {
-        int h = System.identityHashCode(this);
-        final ReentrantLock lock = exceptionTableLock;
-        lock.lock();
-        try {
-            expungeStaleExceptions();
-            ExceptionNode[] t = exceptionTable;
-            int i = h & (t.length - 1);
-            for (ExceptionNode e = t[i]; ; e = e.next) {
-                if (e == null) {
-                    t[i] = new ExceptionNode(this, ex, t[i]);
-                    break;
-                }
-                if (e.get() == this) // already present
-                    break;
-            }
-        } finally {
-            lock.unlock();
-        }
-        return setCompletion(EXCEPTIONAL);
+        int s = recordExceptionalCompletion(ex);
+        if ((s & DONE_MASK) == EXCEPTIONAL)
+            internalPropagateException(ex);
+        return s;
     }
 
     /**
-     * Removes exception node and clears status
+     * Hook for exception propagation support for tasks with completers.
+     */
+    void internalPropagateException(Throwable ex) {
+    }
+
+    /**
+     * Cancels, ignoring any exceptions thrown by cancel. Used during
+     * worker and pool shutdown. Cancel is spec'ed not to throw any
+     * exceptions, but if it does anyway, we have no recourse during
+     * shutdown, so guard against this case.
+     */
+    static final void cancelIgnoringExceptions(ForkJoinTask<?> t) {
+        if (t != null && t.status >= 0) {
+            try {
+                t.cancel(false);
+            } catch (Throwable ignore) {
+            }
+        }
+    }
+
+    /**
+     * Removes exception node and clears status.
      */
     private void clearExceptionalCompletion() {
         int h = System.identityHashCode(this);
@@ -472,7 +514,7 @@
      * @return the exception, or null if none
      */
     private Throwable getThrowableException() {
-        if (status != EXCEPTIONAL)
+        if ((status & DONE_MASK) != EXCEPTIONAL)
             return null;
         int h = System.identityHashCode(this);
         ExceptionNode e;
@@ -490,7 +532,7 @@
         Throwable ex;
         if (e == null || (ex = e.ex) == null)
             return null;
-        if (e.thrower != Thread.currentThread().getId()) {
+        if (false && e.thrower != Thread.currentThread().getId()) {
             Class<? extends Throwable> ec = ex.getClass();
             try {
                 Constructor<?> noArgCtor = null;
@@ -557,41 +599,61 @@
     }
 
     /**
-     * Report the result of invoke or join; called only upon
-     * non-normal return of internal versions.
+     * A version of "sneaky throw" to relay exceptions
      */
-    private V reportResult() {
-        int s; Throwable ex;
-        if ((s = status) == CANCELLED)
+    static void rethrow(final Throwable ex) {
+        if (ex != null) {
+            if (ex instanceof Error)
+                throw (Error)ex;
+            if (ex instanceof RuntimeException)
+                throw (RuntimeException)ex;
+            throw uncheckedThrowable(ex, RuntimeException.class);
+        }
+    }
+
+    /**
+     * The sneaky part of sneaky throw, relying on generics
+     * limitations to evade compiler complaints about rethrowing
+     * unchecked exceptions
+     */
+    @SuppressWarnings("unchecked") static <T extends Throwable>
+        T uncheckedThrowable(final Throwable t, final Class<T> c) {
+        return (T)t; // rely on vacuous cast
+    }
+
+    /**
+     * Throws exception, if any, associated with the given status.
+     */
+    private void reportException(int s) {
+        if (s == CANCELLED)
             throw new CancellationException();
-        if (s == EXCEPTIONAL && (ex = getThrowableException()) != null)
-            SneakyThrow.sneakyThrow(ex); // android-changed
-        return getRawResult();
+        if (s == EXCEPTIONAL)
+            rethrow(getThrowableException());
     }
 
     // public methods
 
     /**
-     * Arranges to asynchronously execute this task.  While it is not
-     * necessarily enforced, it is a usage error to fork a task more
-     * than once unless it has completed and been reinitialized.
-     * Subsequent modifications to the state of this task or any data
-     * it operates on are not necessarily consistently observable by
-     * any thread other than the one executing it unless preceded by a
-     * call to {@link #join} or related methods, or a call to {@link
-     * #isDone} returning {@code true}.
-     *
-     * <p>This method may be invoked only from within {@code
-     * ForkJoinPool} computations (as may be determined using method
-     * {@link #inForkJoinPool}).  Attempts to invoke in other contexts
-     * result in exceptions or errors, possibly including {@code
-     * ClassCastException}.
+     * Arranges to asynchronously execute this task in the pool the
+     * current task is running in, if applicable, or using the {@link
+     * ForkJoinPool#commonPool()} if not {@link #inForkJoinPool}.  While
+     * it is not necessarily enforced, it is a usage error to fork a
+     * task more than once unless it has completed and been
+     * reinitialized.  Subsequent modifications to the state of this
+     * task or any data it operates on are not necessarily
+     * consistently observable by any thread other than the one
+     * executing it unless preceded by a call to {@link #join} or
+     * related methods, or a call to {@link #isDone} returning {@code
+     * true}.
      *
      * @return {@code this}, to simplify usage
      */
     public final ForkJoinTask<V> fork() {
-        ((ForkJoinWorkerThread) Thread.currentThread())
-            .pushTask(this);
+        Thread t;
+        if ((t = Thread.currentThread()) instanceof ForkJoinWorkerThread)
+            ((ForkJoinWorkerThread)t).workQueue.push(this);
+        else
+            ForkJoinPool.commonPool.externalPush(this);
         return this;
     }
 
@@ -607,10 +669,10 @@
      * @return the computed result
      */
     public final V join() {
-        if (doJoin() != NORMAL)
-            return reportResult();
-        else
-            return getRawResult();
+        int s;
+        if ((s = doJoin() & DONE_MASK) != NORMAL)
+            reportException(s);
+        return getRawResult();
     }
 
     /**
@@ -622,10 +684,10 @@
      * @return the computed result
      */
     public final V invoke() {
-        if (doInvoke() != NORMAL)
-            return reportResult();
-        else
-            return getRawResult();
+        int s;
+        if ((s = doInvoke() & DONE_MASK) != NORMAL)
+            reportException(s);
+        return getRawResult();
     }
 
     /**
@@ -641,20 +703,17 @@
      * cancelled, completed normally or exceptionally, or left
      * unprocessed.
      *
-     * <p>This method may be invoked only from within {@code
-     * ForkJoinPool} computations (as may be determined using method
-     * {@link #inForkJoinPool}).  Attempts to invoke in other contexts
-     * result in exceptions or errors, possibly including {@code
-     * ClassCastException}.
-     *
      * @param t1 the first task
      * @param t2 the second task
      * @throws NullPointerException if any task is null
      */
     public static void invokeAll(ForkJoinTask<?> t1, ForkJoinTask<?> t2) {
+        int s1, s2;
         t2.fork();
-        t1.invoke();
-        t2.join();
+        if ((s1 = t1.doInvoke() & DONE_MASK) != NORMAL)
+            t1.reportException(s1);
+        if ((s2 = t2.doJoin() & DONE_MASK) != NORMAL)
+            t2.reportException(s2);
     }
 
     /**
@@ -669,12 +728,6 @@
      * related methods to check if they have been cancelled, completed
      * normally or exceptionally, or left unprocessed.
      *
-     * <p>This method may be invoked only from within {@code
-     * ForkJoinPool} computations (as may be determined using method
-     * {@link #inForkJoinPool}).  Attempts to invoke in other contexts
-     * result in exceptions or errors, possibly including {@code
-     * ClassCastException}.
-     *
      * @param tasks the tasks
      * @throws NullPointerException if any task is null
      */
@@ -702,7 +755,7 @@
             }
         }
         if (ex != null)
-            SneakyThrow.sneakyThrow(ex); // android-changed
+            rethrow(ex);
     }
 
     /**
@@ -718,15 +771,11 @@
      * cancelled, completed normally or exceptionally, or left
      * unprocessed.
      *
-     * <p>This method may be invoked only from within {@code
-     * ForkJoinPool} computations (as may be determined using method
-     * {@link #inForkJoinPool}).  Attempts to invoke in other contexts
-     * result in exceptions or errors, possibly including {@code
-     * ClassCastException}.
-     *
      * @param tasks the collection of tasks
      * @return the tasks argument, to simplify usage
      * @throws NullPointerException if tasks or any element are null
+
+     * @hide
      */
     public static <T extends ForkJoinTask<?>> Collection<T> invokeAll(Collection<T> tasks) {
         if (!(tasks instanceof RandomAccess) || !(tasks instanceof List<?>)) {
@@ -759,7 +808,7 @@
             }
         }
         if (ex != null)
-            SneakyThrow.sneakyThrow(ex); // android-changed
+            rethrow(ex);
         return tasks;
     }
 
@@ -791,20 +840,7 @@
      * @return {@code true} if this task is now cancelled
      */
     public boolean cancel(boolean mayInterruptIfRunning) {
-        return setCompletion(CANCELLED) == CANCELLED;
-    }
-
-    /**
-     * Cancels, ignoring any exceptions thrown by cancel. Used during
-     * worker and pool shutdown. Cancel is spec'ed not to throw any
-     * exceptions, but if it does anyway, we have no recourse during
-     * shutdown, so guard against this case.
-     */
-    final void cancelIgnoringExceptions() {
-        try {
-            cancel(false);
-        } catch (Throwable ignore) {
-        }
+        return (setCompletion(CANCELLED) & DONE_MASK) == CANCELLED;
     }
 
     public final boolean isDone() {
@@ -812,7 +848,7 @@
     }
 
     public final boolean isCancelled() {
-        return status == CANCELLED;
+        return (status & DONE_MASK) == CANCELLED;
     }
 
     /**
@@ -832,7 +868,7 @@
      * exception and was not cancelled
      */
     public final boolean isCompletedNormally() {
-        return status == NORMAL;
+        return (status & DONE_MASK) == NORMAL;
     }
 
     /**
@@ -843,7 +879,7 @@
      * @return the exception, or {@code null} if none
      */
     public final Throwable getException() {
-        int s = status;
+        int s = status & DONE_MASK;
         return ((s >= NORMAL)    ? null :
                 (s == CANCELLED) ? new CancellationException() :
                 getThrowableException());
@@ -893,6 +929,19 @@
     }
 
     /**
+     * Completes this task normally without setting a value. The most
+     * recent value established by {@link #setRawResult} (or {@code
+     * null} by default) will be returned as the result of subsequent
+     * invocations of {@code join} and related operations.
+     *
+     * @since 1.8
+     * @hide
+     */
+    public final void quietlyComplete() {
+        setCompletion(NORMAL);
+    }
+
+    /**
      * Waits if necessary for the computation to complete, and then
      * retrieves its result.
      *
@@ -905,9 +954,9 @@
      */
     public final V get() throws InterruptedException, ExecutionException {
         int s = (Thread.currentThread() instanceof ForkJoinWorkerThread) ?
-            doJoin() : externalInterruptibleAwaitDone(0L);
+            doJoin() : externalInterruptibleAwaitDone();
         Throwable ex;
-        if (s == CANCELLED)
+        if ((s &= DONE_MASK) == CANCELLED)
             throw new CancellationException();
         if (s == EXCEPTIONAL && (ex = getThrowableException()) != null)
             throw new ExecutionException(ex);
@@ -930,32 +979,63 @@
      */
     public final V get(long timeout, TimeUnit unit)
         throws InterruptedException, ExecutionException, TimeoutException {
-        Thread t = Thread.currentThread();
-        if (t instanceof ForkJoinWorkerThread) {
-            ForkJoinWorkerThread w = (ForkJoinWorkerThread) t;
-            long nanos = unit.toNanos(timeout);
-            if (status >= 0) {
-                boolean completed = false;
-                if (w.unpushTask(this)) {
-                    try {
-                        completed = exec();
-                    } catch (Throwable rex) {
-                        setExceptionalCompletion(rex);
+        if (Thread.interrupted())
+            throw new InterruptedException();
+        // Messy in part because we measure in nanosecs, but wait in millisecs
+        int s; long ms;
+        long ns = unit.toNanos(timeout);
+        if ((s = status) >= 0 && ns > 0L) {
+            long deadline = System.nanoTime() + ns;
+            ForkJoinPool p = null;
+            ForkJoinPool.WorkQueue w = null;
+            Thread t = Thread.currentThread();
+            if (t instanceof ForkJoinWorkerThread) {
+                ForkJoinWorkerThread wt = (ForkJoinWorkerThread)t;
+                p = wt.pool;
+                w = wt.workQueue;
+                p.helpJoinOnce(w, this); // no retries on failure
+            }
+            else
+                ForkJoinPool.externalHelpJoin(this);
+            boolean canBlock = false;
+            boolean interrupted = false;
+            try {
+                while ((s = status) >= 0) {
+                    if (w != null && w.qlock < 0)
+                        cancelIgnoringExceptions(this);
+                    else if (!canBlock) {
+                        if (p == null || p.tryCompensate())
+                            canBlock = true;
+                    }
+                    else {
+                        if ((ms = TimeUnit.NANOSECONDS.toMillis(ns)) > 0L &&
+                            U.compareAndSwapInt(this, STATUS, s, s | SIGNAL)) {
+                            synchronized (this) {
+                                if (status >= 0) {
+                                    try {
+                                        wait(ms);
+                                    } catch (InterruptedException ie) {
+                                        if (p == null)
+                                            interrupted = true;
+                                    }
+                                }
+                                else
+                                    notifyAll();
+                            }
+                        }
+                        if ((s = status) < 0 || interrupted ||
+                            (ns = deadline - System.nanoTime()) <= 0L)
+                            break;
                     }
                 }
-                if (completed)
-                    setCompletion(NORMAL);
-                else if (status >= 0 && nanos > 0)
-                    w.pool.timedAwaitJoin(this, nanos);
+            } finally {
+                if (p != null && canBlock)
+                    p.incrementActiveCount();
             }
+            if (interrupted)
+                throw new InterruptedException();
         }
-        else {
-            long millis = unit.toMillis(timeout);
-            if (millis > 0)
-                externalInterruptibleAwaitDone(millis);
-        }
-        int s = status;
-        if (s != NORMAL) {
+        if ((s &= DONE_MASK) != NORMAL) {
             Throwable ex;
             if (s == CANCELLED)
                 throw new CancellationException();
@@ -992,16 +1072,15 @@
      * be of use in designs in which many tasks are forked, but none
      * are explicitly joined, instead executing them until all are
      * processed.
-     *
-     * <p>This method may be invoked only from within {@code
-     * ForkJoinPool} computations (as may be determined using method
-     * {@link #inForkJoinPool}).  Attempts to invoke in other contexts
-     * result in exceptions or errors, possibly including {@code
-     * ClassCastException}.
      */
     public static void helpQuiesce() {
-        ((ForkJoinWorkerThread) Thread.currentThread())
-            .helpQuiescePool();
+        Thread t;
+        if ((t = Thread.currentThread()) instanceof ForkJoinWorkerThread) {
+            ForkJoinWorkerThread wt = (ForkJoinWorkerThread)t;
+            wt.pool.helpQuiescePool(wt.workQueue);
+        }
+        else
+            ForkJoinPool.externalHelpQuiescePool();
     }
 
     /**
@@ -1021,7 +1100,7 @@
      * setRawResult(null)}.
      */
     public void reinitialize() {
-        if (status == EXCEPTIONAL)
+        if ((status & DONE_MASK) == EXCEPTIONAL)
             clearExceptionalCompletion();
         else
             status = 0;
@@ -1054,23 +1133,19 @@
 
     /**
      * Tries to unschedule this task for execution. This method will
-     * typically succeed if this task is the most recently forked task
-     * by the current thread, and has not commenced executing in
-     * another thread.  This method may be useful when arranging
-     * alternative local processing of tasks that could have been, but
-     * were not, stolen.
-     *
-     * <p>This method may be invoked only from within {@code
-     * ForkJoinPool} computations (as may be determined using method
-     * {@link #inForkJoinPool}).  Attempts to invoke in other contexts
-     * result in exceptions or errors, possibly including {@code
-     * ClassCastException}.
+     * typically (but is not guaranteed to) succeed if this task is
+     * the most recently forked task by the current thread, and has
+     * not commenced executing in another thread.  This method may be
+     * useful when arranging alternative local processing of tasks
+     * that could have been, but were not, stolen.
      *
      * @return {@code true} if unforked
      */
     public boolean tryUnfork() {
-        return ((ForkJoinWorkerThread) Thread.currentThread())
-            .unpushTask(this);
+        Thread t;
+        return (((t = Thread.currentThread()) instanceof ForkJoinWorkerThread) ?
+                ((ForkJoinWorkerThread)t).workQueue.tryUnpush(this) :
+                ForkJoinPool.tryExternalUnpush(this));
     }
 
     /**
@@ -1079,40 +1154,32 @@
      * value may be useful for heuristic decisions about whether to
      * fork other tasks.
      *
-     * <p>This method may be invoked only from within {@code
-     * ForkJoinPool} computations (as may be determined using method
-     * {@link #inForkJoinPool}).  Attempts to invoke in other contexts
-     * result in exceptions or errors, possibly including {@code
-     * ClassCastException}.
-     *
      * @return the number of tasks
      */
     public static int getQueuedTaskCount() {
-        return ((ForkJoinWorkerThread) Thread.currentThread())
-            .getQueueSize();
+        Thread t; ForkJoinPool.WorkQueue q;
+        if ((t = Thread.currentThread()) instanceof ForkJoinWorkerThread)
+            q = ((ForkJoinWorkerThread)t).workQueue;
+        else
+            q = ForkJoinPool.commonSubmitterQueue();
+        return (q == null) ? 0 : q.queueSize();
     }
 
     /**
      * Returns an estimate of how many more locally queued tasks are
      * held by the current worker thread than there are other worker
-     * threads that might steal them.  This value may be useful for
+     * threads that might steal them, or zero if this thread is not
+     * operating in a ForkJoinPool. This value may be useful for
      * heuristic decisions about whether to fork other tasks. In many
      * usages of ForkJoinTasks, at steady state, each worker should
      * aim to maintain a small constant surplus (for example, 3) of
      * tasks, and to process computations locally if this threshold is
      * exceeded.
      *
-     * <p>This method may be invoked only from within {@code
-     * ForkJoinPool} computations (as may be determined using method
-     * {@link #inForkJoinPool}).  Attempts to invoke in other contexts
-     * result in exceptions or errors, possibly including {@code
-     * ClassCastException}.
-     *
      * @return the surplus number of tasks, which may be negative
      */
     public static int getSurplusQueuedTaskCount() {
-        return ((ForkJoinWorkerThread) Thread.currentThread())
-            .getEstimatedSurplusTaskCount();
+        return ForkJoinPool.getSurplusQueuedTaskCount();
     }
 
     // Extension methods
@@ -1138,15 +1205,18 @@
     protected abstract void setRawResult(V value);
 
     /**
-     * Immediately performs the base action of this task.  This method
-     * is designed to support extensions, and should not in general be
-     * called otherwise. The return value controls whether this task
-     * is considered to be done normally. It may return false in
+     * Immediately performs the base action of this task and returns
+     * true if, upon return from this method, this task is guaranteed
+     * to have completed normally. This method may return false
+     * otherwise, to indicate that this task is not necessarily
+     * complete (or is not known to be complete), for example in
      * asynchronous actions that require explicit invocations of
-     * {@link #complete} to become joinable. It may also throw an
-     * (unchecked) exception to indicate abnormal exit.
+     * completion methods. This method may also throw an (unchecked)
+     * exception to indicate abnormal exit. This method is designed to
+     * support extensions, and should not in general be called
+     * otherwise.
      *
-     * @return {@code true} if completed normally
+     * @return {@code true} if this task is known to have completed normally
      */
     protected abstract boolean exec();
 
@@ -1160,59 +1230,105 @@
      * primarily to support extensions, and is unlikely to be useful
      * otherwise.
      *
-     * <p>This method may be invoked only from within {@code
-     * ForkJoinPool} computations (as may be determined using method
-     * {@link #inForkJoinPool}).  Attempts to invoke in other contexts
-     * result in exceptions or errors, possibly including {@code
-     * ClassCastException}.
-     *
      * @return the next task, or {@code null} if none are available
      */
     protected static ForkJoinTask<?> peekNextLocalTask() {
-        return ((ForkJoinWorkerThread) Thread.currentThread())
-            .peekTask();
+        Thread t; ForkJoinPool.WorkQueue q;
+        if ((t = Thread.currentThread()) instanceof ForkJoinWorkerThread)
+            q = ((ForkJoinWorkerThread)t).workQueue;
+        else
+            q = ForkJoinPool.commonSubmitterQueue();
+        return (q == null) ? null : q.peek();
     }
 
     /**
      * Unschedules and returns, without executing, the next task
-     * queued by the current thread but not yet executed.  This method
-     * is designed primarily to support extensions, and is unlikely to
-     * be useful otherwise.
-     *
-     * <p>This method may be invoked only from within {@code
-     * ForkJoinPool} computations (as may be determined using method
-     * {@link #inForkJoinPool}).  Attempts to invoke in other contexts
-     * result in exceptions or errors, possibly including {@code
-     * ClassCastException}.
+     * queued by the current thread but not yet executed, if the
+     * current thread is operating in a ForkJoinPool.  This method is
+     * designed primarily to support extensions, and is unlikely to be
+     * useful otherwise.
      *
      * @return the next task, or {@code null} if none are available
      */
     protected static ForkJoinTask<?> pollNextLocalTask() {
-        return ((ForkJoinWorkerThread) Thread.currentThread())
-            .pollLocalTask();
+        Thread t;
+        return ((t = Thread.currentThread()) instanceof ForkJoinWorkerThread) ?
+            ((ForkJoinWorkerThread)t).workQueue.nextLocalTask() :
+            null;
     }
 
     /**
-     * Unschedules and returns, without executing, the next task
+     * If the current thread is operating in a ForkJoinPool,
+     * unschedules and returns, without executing, the next task
      * queued by the current thread but not yet executed, if one is
      * available, or if not available, a task that was forked by some
      * other thread, if available. Availability may be transient, so a
-     * {@code null} result does not necessarily imply quiescence
-     * of the pool this task is operating in.  This method is designed
+     * {@code null} result does not necessarily imply quiescence of
+     * the pool this task is operating in.  This method is designed
      * primarily to support extensions, and is unlikely to be useful
      * otherwise.
      *
-     * <p>This method may be invoked only from within {@code
-     * ForkJoinPool} computations (as may be determined using method
-     * {@link #inForkJoinPool}).  Attempts to invoke in other contexts
-     * result in exceptions or errors, possibly including {@code
-     * ClassCastException}.
-     *
      * @return a task, or {@code null} if none are available
      */
     protected static ForkJoinTask<?> pollTask() {
-        return ((ForkJoinWorkerThread) Thread.currentThread())
-            .pollTask();
+        Thread t; ForkJoinWorkerThread wt;
+        return ((t = Thread.currentThread()) instanceof ForkJoinWorkerThread) ?
+            (wt = (ForkJoinWorkerThread)t).pool.nextTaskFor(wt.workQueue) :
+            null;
+    }
+
+    // tag operations
+
+    /**
+     * Returns the tag for this task.
+     *
+     * @return the tag for this task
+     * @since 1.8
+     * @hide
+     */
+    public final short getForkJoinTaskTag() {
+        return (short)status;
+    }
+
+    /**
+     * Atomically sets the tag value for this task.
+     *
+     * @param tag the tag value
+     * @return the previous value of the tag
+     * @since 1.8
+     * @hide
+     */
+    public final short setForkJoinTaskTag(short tag) {
+        for (int s;;) {
+            if (U.compareAndSwapInt(this, STATUS, s = status,
+                                    (s & ~SMASK) | (tag & SMASK)))
+                return (short)s;
+        }
+    }
+
+    /**
+     * Atomically conditionally sets the tag value for this task.
+     * Among other applications, tags can be used as visit markers
+     * in tasks operating on graphs, as in methods that check: {@code
+     * if (task.compareAndSetForkJoinTaskTag((short)0, (short)1))}
+     * before processing, otherwise exiting because the node has
+     * already been visited.
+     *
+     * @param e the expected tag value
+     * @param tag the new tag value
+     * @return true if successful; i.e., the current value was
+     * equal to e and is now tag.
+     * @since 1.8
+     * @hide
+     */
+    public final boolean compareAndSetForkJoinTaskTag(short e, short tag) {
+        for (int s;;) {
+            if ((short)(s = status) != e)
+                return false;
+            if (U.compareAndSwapInt(this, STATUS, s,
+                                    (s & ~SMASK) | (tag & SMASK)))
+                return true;
+        }
     }
 
     /**
@@ -1223,21 +1339,33 @@
     static final class AdaptedRunnable<T> extends ForkJoinTask<T>
         implements RunnableFuture<T> {
         final Runnable runnable;
-        final T resultOnCompletion;
         T result;
         AdaptedRunnable(Runnable runnable, T result) {
             if (runnable == null) throw new NullPointerException();
             this.runnable = runnable;
-            this.resultOnCompletion = result;
+            this.result = result; // OK to set this even before completion
         }
-        public T getRawResult() { return result; }
-        public void setRawResult(T v) { result = v; }
-        public boolean exec() {
-            runnable.run();
-            result = resultOnCompletion;
-            return true;
+        public final T getRawResult() { return result; }
+        public final void setRawResult(T v) { result = v; }
+        public final boolean exec() { runnable.run(); return true; }
+        public final void run() { invoke(); }
+        private static final long serialVersionUID = 5232453952276885070L;
+    }
+
+    /**
+     * Adaptor for Runnables without results
+     */
+    static final class AdaptedRunnableAction extends ForkJoinTask<Void>
+        implements RunnableFuture<Void> {
+        final Runnable runnable;
+        AdaptedRunnableAction(Runnable runnable) {
+            if (runnable == null) throw new NullPointerException();
+            this.runnable = runnable;
         }
-        public void run() { invoke(); }
+        public final Void getRawResult() { return null; }
+        public final void setRawResult(Void v) { }
+        public final boolean exec() { runnable.run(); return true; }
+        public final void run() { invoke(); }
         private static final long serialVersionUID = 5232453952276885070L;
     }
 
@@ -1252,9 +1380,9 @@
             if (callable == null) throw new NullPointerException();
             this.callable = callable;
         }
-        public T getRawResult() { return result; }
-        public void setRawResult(T v) { result = v; }
-        public boolean exec() {
+        public final T getRawResult() { return result; }
+        public final void setRawResult(T v) { result = v; }
+        public final boolean exec() {
             try {
                 result = callable.call();
                 return true;
@@ -1266,7 +1394,7 @@
                 throw new RuntimeException(ex);
             }
         }
-        public void run() { invoke(); }
+        public final void run() { invoke(); }
         private static final long serialVersionUID = 2838392045355241008L;
     }
 
@@ -1279,7 +1407,7 @@
      * @return the task
      */
     public static ForkJoinTask<?> adapt(Runnable runnable) {
-        return new AdaptedRunnable<Void>(runnable, null);
+        return new AdaptedRunnableAction(runnable);
     }
 
     /**
@@ -1313,11 +1441,10 @@
     private static final long serialVersionUID = -7721805057305804111L;
 
     /**
-     * Saves the state to a stream (that is, serializes it).
+     * Saves this task to a stream (that is, serializes it).
      *
      * @serialData the current run status and the exception thrown
      * during execution, or {@code null} if none
-     * @param s the stream
      */
     private void writeObject(java.io.ObjectOutputStream s)
         throws java.io.IOException {
@@ -1326,9 +1453,7 @@
     }
 
     /**
-     * Reconstitutes the instance from a stream (that is, deserializes it).
-     *
-     * @param s the stream
+     * Reconstitutes this task from a stream (that is, deserializes it).
      */
     private void readObject(java.io.ObjectInputStream s)
         throws java.io.IOException, ClassNotFoundException {
@@ -1339,16 +1464,18 @@
     }
 
     // Unsafe mechanics
-    private static final sun.misc.Unsafe UNSAFE;
-    private static final long statusOffset;
+    private static final sun.misc.Unsafe U;
+    private static final long STATUS;
+
     static {
         exceptionTableLock = new ReentrantLock();
         exceptionTableRefQueue = new ReferenceQueue<Object>();
         exceptionTable = new ExceptionNode[EXCEPTION_MAP_CAPACITY];
         try {
-            UNSAFE = sun.misc.Unsafe.getUnsafe();
-            statusOffset = UNSAFE.objectFieldOffset
-                (ForkJoinTask.class.getDeclaredField("status"));
+            U = sun.misc.Unsafe.getUnsafe();
+            Class<?> k = ForkJoinTask.class;
+            STATUS = U.objectFieldOffset
+                (k.getDeclaredField("status"));
         } catch (Exception e) {
             throw new Error(e);
         }
diff --git a/luni/src/main/java/java/util/concurrent/ForkJoinWorkerThread.java b/luni/src/main/java/java/util/concurrent/ForkJoinWorkerThread.java
index d99ffe9..f31763c 100644
--- a/luni/src/main/java/java/util/concurrent/ForkJoinWorkerThread.java
+++ b/luni/src/main/java/java/util/concurrent/ForkJoinWorkerThread.java
@@ -6,10 +6,6 @@
 
 package java.util.concurrent;
 
-import java.util.Collection;
-import java.util.concurrent.RejectedExecutionException;
-import libcore.util.SneakyThrow;
-
 /**
  * A thread managed by a {@link ForkJoinPool}, which executes
  * {@link ForkJoinTask}s.
@@ -27,238 +23,20 @@
  */
 public class ForkJoinWorkerThread extends Thread {
     /*
-     * Overview:
-     *
      * ForkJoinWorkerThreads are managed by ForkJoinPools and perform
-     * ForkJoinTasks. This class includes bookkeeping in support of
-     * worker activation, suspension, and lifecycle control described
-     * in more detail in the internal documentation of class
-     * ForkJoinPool. And as described further below, this class also
-     * includes special-cased support for some ForkJoinTask
-     * methods. But the main mechanics involve work-stealing:
+     * ForkJoinTasks. For explanation, see the internal documentation
+     * of class ForkJoinPool.
      *
-     * Work-stealing queues are special forms of Deques that support
-     * only three of the four possible end-operations -- push, pop,
-     * and deq (aka steal), under the further constraints that push
-     * and pop are called only from the owning thread, while deq may
-     * be called from other threads.  (If you are unfamiliar with
-     * them, you probably want to read Herlihy and Shavit's book "The
-     * Art of Multiprocessor programming", chapter 16 describing these
-     * in more detail before proceeding.)  The main work-stealing
-     * queue design is roughly similar to those in the papers "Dynamic
-     * Circular Work-Stealing Deque" by Chase and Lev, SPAA 2005
-     * (http://research.sun.com/scalable/pubs/index.html) and
-     * "Idempotent work stealing" by Michael, Saraswat, and Vechev,
-     * PPoPP 2009 (http://portal.acm.org/citation.cfm?id=1504186).
-     * The main differences ultimately stem from gc requirements that
-     * we null out taken slots as soon as we can, to maintain as small
-     * a footprint as possible even in programs generating huge
-     * numbers of tasks. To accomplish this, we shift the CAS
-     * arbitrating pop vs deq (steal) from being on the indices
-     * ("queueBase" and "queueTop") to the slots themselves (mainly
-     * via method "casSlotNull()"). So, both a successful pop and deq
-     * mainly entail a CAS of a slot from non-null to null.  Because
-     * we rely on CASes of references, we do not need tag bits on
-     * queueBase or queueTop.  They are simple ints as used in any
-     * circular array-based queue (see for example ArrayDeque).
-     * Updates to the indices must still be ordered in a way that
-     * guarantees that queueTop == queueBase means the queue is empty,
-     * but otherwise may err on the side of possibly making the queue
-     * appear nonempty when a push, pop, or deq have not fully
-     * committed. Note that this means that the deq operation,
-     * considered individually, is not wait-free. One thief cannot
-     * successfully continue until another in-progress one (or, if
-     * previously empty, a push) completes.  However, in the
-     * aggregate, we ensure at least probabilistic non-blockingness.
-     * If an attempted steal fails, a thief always chooses a different
-     * random victim target to try next. So, in order for one thief to
-     * progress, it suffices for any in-progress deq or new push on
-     * any empty queue to complete.
-     *
-     * This approach also enables support for "async mode" where local
-     * task processing is in FIFO, not LIFO order; simply by using a
-     * version of deq rather than pop when locallyFifo is true (as set
-     * by the ForkJoinPool).  This allows use in message-passing
-     * frameworks in which tasks are never joined.  However neither
-     * mode considers affinities, loads, cache localities, etc, so
-     * rarely provide the best possible performance on a given
-     * machine, but portably provide good throughput by averaging over
-     * these factors.  (Further, even if we did try to use such
-     * information, we do not usually have a basis for exploiting
-     * it. For example, some sets of tasks profit from cache
-     * affinities, but others are harmed by cache pollution effects.)
-     *
-     * When a worker would otherwise be blocked waiting to join a
-     * task, it first tries a form of linear helping: Each worker
-     * records (in field currentSteal) the most recent task it stole
-     * from some other worker. Plus, it records (in field currentJoin)
-     * the task it is currently actively joining. Method joinTask uses
-     * these markers to try to find a worker to help (i.e., steal back
-     * a task from and execute it) that could hasten completion of the
-     * actively joined task. In essence, the joiner executes a task
-     * that would be on its own local deque had the to-be-joined task
-     * not been stolen. This may be seen as a conservative variant of
-     * the approach in Wagner & Calder "Leapfrogging: a portable
-     * technique for implementing efficient futures" SIGPLAN Notices,
-     * 1993 (http://portal.acm.org/citation.cfm?id=155354). It differs
-     * in that: (1) We only maintain dependency links across workers
-     * upon steals, rather than use per-task bookkeeping.  This may
-     * require a linear scan of workers array to locate stealers, but
-     * usually doesn't because stealers leave hints (that may become
-     * stale/wrong) of where to locate them. This isolates cost to
-     * when it is needed, rather than adding to per-task overhead.
-     * (2) It is "shallow", ignoring nesting and potentially cyclic
-     * mutual steals.  (3) It is intentionally racy: field currentJoin
-     * is updated only while actively joining, which means that we
-     * miss links in the chain during long-lived tasks, GC stalls etc
-     * (which is OK since blocking in such cases is usually a good
-     * idea).  (4) We bound the number of attempts to find work (see
-     * MAX_HELP) and fall back to suspending the worker and if
-     * necessary replacing it with another.
-     *
-     * Efficient implementation of these algorithms currently relies
-     * on an uncomfortable amount of "Unsafe" mechanics. To maintain
-     * correct orderings, reads and writes of variable queueBase
-     * require volatile ordering.  Variable queueTop need not be
-     * volatile because non-local reads always follow those of
-     * queueBase.  Similarly, because they are protected by volatile
-     * queueBase reads, reads of the queue array and its slots by
-     * other threads do not need volatile load semantics, but writes
-     * (in push) require store order and CASes (in pop and deq)
-     * require (volatile) CAS semantics.  (Michael, Saraswat, and
-     * Vechev's algorithm has similar properties, but without support
-     * for nulling slots.)  Since these combinations aren't supported
-     * using ordinary volatiles, the only way to accomplish these
-     * efficiently is to use direct Unsafe calls. (Using external
-     * AtomicIntegers and AtomicReferenceArrays for the indices and
-     * array is significantly slower because of memory locality and
-     * indirection effects.)
-     *
-     * Further, performance on most platforms is very sensitive to
-     * placement and sizing of the (resizable) queue array.  Even
-     * though these queues don't usually become all that big, the
-     * initial size must be large enough to counteract cache
-     * contention effects across multiple queues (especially in the
-     * presence of GC cardmarking). Also, to improve thread-locality,
-     * queues are initialized after starting.
+     * This class just maintains links to its pool and WorkQueue.  The
+     * pool field is set immediately upon construction, but the
+     * workQueue field is not set until a call to registerWorker
+     * completes. This leads to a visibility race, that is tolerated
+     * by requiring that the workQueue field is only accessed by the
+     * owning thread.
      */
 
-    /**
-     * Mask for pool indices encoded as shorts
-     */
-    private static final int  SMASK  = 0xffff;
-
-    /**
-     * Capacity of work-stealing queue array upon initialization.
-     * Must be a power of two. Initial size must be at least 4, but is
-     * padded to minimize cache effects.
-     */
-    private static final int INITIAL_QUEUE_CAPACITY = 1 << 13;
-
-    /**
-     * Maximum size for queue array. Must be a power of two
-     * less than or equal to 1 << (31 - width of array entry) to
-     * ensure lack of index wraparound, but is capped at a lower
-     * value to help users trap runaway computations.
-     */
-    private static final int MAXIMUM_QUEUE_CAPACITY = 1 << 24; // 16M
-
-    /**
-     * The work-stealing queue array. Size must be a power of two.
-     * Initialized when started (as opposed to when constructed), to
-     * improve memory locality.
-     */
-    ForkJoinTask<?>[] queue;
-
-    /**
-     * The pool this thread works in. Accessed directly by ForkJoinTask.
-     */
-    final ForkJoinPool pool;
-
-    /**
-     * Index (mod queue.length) of next queue slot to push to or pop
-     * from. It is written only by owner thread, and accessed by other
-     * threads only after reading (volatile) queueBase.  Both queueTop
-     * and queueBase are allowed to wrap around on overflow, but
-     * (queueTop - queueBase) still estimates size.
-     */
-    int queueTop;
-
-    /**
-     * Index (mod queue.length) of least valid queue slot, which is
-     * always the next position to steal from if nonempty.
-     */
-    volatile int queueBase;
-
-    /**
-     * The index of most recent stealer, used as a hint to avoid
-     * traversal in method helpJoinTask. This is only a hint because a
-     * worker might have had multiple steals and this only holds one
-     * of them (usually the most current). Declared non-volatile,
-     * relying on other prevailing sync to keep reasonably current.
-     */
-    int stealHint;
-
-    /**
-     * Index of this worker in pool array. Set once by pool before
-     * running, and accessed directly by pool to locate this worker in
-     * its workers array.
-     */
-    final int poolIndex;
-
-    /**
-     * Encoded record for pool task waits. Usages are always
-     * surrounded by volatile reads/writes
-     */
-    int nextWait;
-
-    /**
-     * Complement of poolIndex, offset by count of entries of task
-     * waits. Accessed by ForkJoinPool to manage event waiters.
-     */
-    volatile int eventCount;
-
-    /**
-     * Seed for random number generator for choosing steal victims.
-     * Uses Marsaglia xorshift. Must be initialized as nonzero.
-     */
-    int seed;
-
-    /**
-     * Number of steals. Directly accessed (and reset) by pool when
-     * idle.
-     */
-    int stealCount;
-
-    /**
-     * True if this worker should or did terminate
-     */
-    volatile boolean terminate;
-
-    /**
-     * Set to true before LockSupport.park; false on return
-     */
-    volatile boolean parked;
-
-    /**
-     * True if use local fifo, not default lifo, for local polling.
-     * Shadows value from ForkJoinPool.
-     */
-    final boolean locallyFifo;
-
-    /**
-     * The task most recently stolen from another worker (or
-     * submission queue).  All uses are surrounded by enough volatile
-     * reads/writes to maintain as non-volatile.
-     */
-    ForkJoinTask<?> currentSteal;
-
-    /**
-     * The task currently being joined, set only when actively trying
-     * to help other stealers in helpJoinTask. All uses are surrounded
-     * by enough volatile reads/writes to maintain as non-volatile.
-     */
-    ForkJoinTask<?> currentJoin;
+    final ForkJoinPool pool;                // the pool this thread works in
+    final ForkJoinPool.WorkQueue workQueue; // work-stealing mechanics
 
     /**
      * Creates a ForkJoinWorkerThread operating in the given pool.
@@ -267,20 +45,12 @@
      * @throws NullPointerException if pool is null
      */
     protected ForkJoinWorkerThread(ForkJoinPool pool) {
-        super(pool.nextWorkerName());
+        // Use a placeholder until a useful name can be set in registerWorker
+        super("aForkJoinWorkerThread");
         this.pool = pool;
-        int k = pool.registerWorker(this);
-        poolIndex = k;
-        eventCount = ~k & SMASK; // clear wait count
-        locallyFifo = pool.locallyFifo;
-        Thread.UncaughtExceptionHandler ueh = pool.ueh;
-        if (ueh != null)
-            setUncaughtExceptionHandler(ueh);
-        setDaemon(true);
+        this.workQueue = pool.registerWorker(this);
     }
 
-    // Public methods
-
     /**
      * Returns the pool hosting this thread.
      *
@@ -300,28 +70,9 @@
      * @return the index number
      */
     public int getPoolIndex() {
-        return poolIndex;
+        return workQueue.poolIndex;
     }
 
-    // Randomization
-
-    /**
-     * Computes next value for random victim probes and backoffs.
-     * Scans don't require a very high quality generator, but also not
-     * a crummy one.  Marsaglia xor-shift is cheap and works well
-     * enough.  Note: This is manually inlined in FJP.scan() to avoid
-     * writes inside busy loops.
-     */
-    private int nextSeed() {
-        int r = seed;
-        r ^= r << 13;
-        r ^= r >>> 17;
-        r ^= r << 5;
-        return seed = r;
-    }
-
-    // Run State management
-
     /**
      * Initializes internal state after construction but before
      * processing any tasks. If you override this method, you must
@@ -332,9 +83,6 @@
      * processing tasks.
      */
     protected void onStart() {
-        queue = new ForkJoinTask<?>[INITIAL_QUEUE_CAPACITY];
-        int r = ForkJoinPool.workerSeedGenerator.nextInt();
-        seed = (r == 0) ? 1 : r; //  must be nonzero
     }
 
     /**
@@ -346,17 +94,6 @@
      * to an unrecoverable error, or {@code null} if completed normally
      */
     protected void onTermination(Throwable exception) {
-        try {
-            terminate = true;
-            cancelTasks();
-            pool.deregisterWorker(this, exception);
-        } catch (Throwable ex) {        // Shouldn't ever happen
-            if (exception == null)      // but if so, at least rethrown
-                exception = ex;
-        } finally {
-            if (exception != null)
-                SneakyThrow.sneakyThrow(exception); // android-changed
-        }
     }
 
     /**
@@ -368,603 +105,18 @@
         Throwable exception = null;
         try {
             onStart();
-            pool.work(this);
+            pool.runWorker(workQueue);
         } catch (Throwable ex) {
             exception = ex;
         } finally {
-            onTermination(exception);
-        }
-    }
-
-    /*
-     * Intrinsics-based atomic writes for queue slots. These are
-     * basically the same as methods in AtomicReferenceArray, but
-     * specialized for (1) ForkJoinTask elements (2) requirement that
-     * nullness and bounds checks have already been performed by
-     * callers and (3) effective offsets are known not to overflow
-     * from int to long (because of MAXIMUM_QUEUE_CAPACITY). We don't
-     * need corresponding version for reads: plain array reads are OK
-     * because they are protected by other volatile reads and are
-     * confirmed by CASes.
-     *
-     * Most uses don't actually call these methods, but instead
-     * contain inlined forms that enable more predictable
-     * optimization.  We don't define the version of write used in
-     * pushTask at all, but instead inline there a store-fenced array
-     * slot write.
-     *
-     * Also in most methods, as a performance (not correctness) issue,
-     * we'd like to encourage compilers not to arbitrarily postpone
-     * setting queueTop after writing slot.  Currently there is no
-     * intrinsic for arranging this, but using Unsafe putOrderedInt
-     * may be a preferable strategy on some compilers even though its
-     * main effect is a pre-, not post- fence. To simplify possible
-     * changes, the option is left in comments next to the associated
-     * assignments.
-     */
-
-    /**
-     * CASes slot i of array q from t to null. Caller must ensure q is
-     * non-null and index is in range.
-     */
-    private static final boolean casSlotNull(ForkJoinTask<?>[] q, int i,
-                                             ForkJoinTask<?> t) {
-        return UNSAFE.compareAndSwapObject(q, (i << ASHIFT) + ABASE, t, null);
-    }
-
-    /**
-     * Performs a volatile write of the given task at given slot of
-     * array q.  Caller must ensure q is non-null and index is in
-     * range. This method is used only during resets and backouts.
-     */
-    private static final void writeSlot(ForkJoinTask<?>[] q, int i,
-                                        ForkJoinTask<?> t) {
-        UNSAFE.putObjectVolatile(q, (i << ASHIFT) + ABASE, t);
-    }
-
-    // queue methods
-
-    /**
-     * Pushes a task. Call only from this thread.
-     *
-     * @param t the task. Caller must ensure non-null.
-     */
-    final void pushTask(ForkJoinTask<?> t) {
-        ForkJoinTask<?>[] q; int s, m;
-        if ((q = queue) != null) {    // ignore if queue removed
-            long u = (((s = queueTop) & (m = q.length - 1)) << ASHIFT) + ABASE;
-            UNSAFE.putOrderedObject(q, u, t);
-            queueTop = s + 1;         // or use putOrderedInt
-            if ((s -= queueBase) <= 2)
-                pool.signalWork();
-            else if (s == m)
-                growQueue();
-        }
-    }
-
-    /**
-     * Creates or doubles queue array.  Transfers elements by
-     * emulating steals (deqs) from old array and placing, oldest
-     * first, into new array.
-     */
-    private void growQueue() {
-        ForkJoinTask<?>[] oldQ = queue;
-        int size = oldQ != null ? oldQ.length << 1 : INITIAL_QUEUE_CAPACITY;
-        if (size > MAXIMUM_QUEUE_CAPACITY)
-            throw new RejectedExecutionException("Queue capacity exceeded");
-        if (size < INITIAL_QUEUE_CAPACITY)
-            size = INITIAL_QUEUE_CAPACITY;
-        ForkJoinTask<?>[] q = queue = new ForkJoinTask<?>[size];
-        int mask = size - 1;
-        int top = queueTop;
-        int oldMask;
-        if (oldQ != null && (oldMask = oldQ.length - 1) >= 0) {
-            for (int b = queueBase; b != top; ++b) {
-                long u = ((b & oldMask) << ASHIFT) + ABASE;
-                Object x = UNSAFE.getObjectVolatile(oldQ, u);
-                if (x != null && UNSAFE.compareAndSwapObject(oldQ, u, x, null))
-                    UNSAFE.putObjectVolatile
-                        (q, ((b & mask) << ASHIFT) + ABASE, x);
+            try {
+                onTermination(exception);
+            } catch (Throwable ex) {
+                if (exception == null)
+                    exception = ex;
+            } finally {
+                pool.deregisterWorker(this, exception);
             }
         }
     }
-
-    /**
-     * Tries to take a task from the base of the queue, failing if
-     * empty or contended. Note: Specializations of this code appear
-     * in locallyDeqTask and elsewhere.
-     *
-     * @return a task, or null if none or contended
-     */
-    final ForkJoinTask<?> deqTask() {
-        ForkJoinTask<?> t; ForkJoinTask<?>[] q; int b, i;
-        if (queueTop != (b = queueBase) &&
-            (q = queue) != null && // must read q after b
-            (i = (q.length - 1) & b) >= 0 &&
-            (t = q[i]) != null && queueBase == b &&
-            UNSAFE.compareAndSwapObject(q, (i << ASHIFT) + ABASE, t, null)) {
-            queueBase = b + 1;
-            return t;
-        }
-        return null;
-    }
-
-    /**
-     * Tries to take a task from the base of own queue.  Called only
-     * by this thread.
-     *
-     * @return a task, or null if none
-     */
-    final ForkJoinTask<?> locallyDeqTask() {
-        ForkJoinTask<?> t; int m, b, i;
-        ForkJoinTask<?>[] q = queue;
-        if (q != null && (m = q.length - 1) >= 0) {
-            while (queueTop != (b = queueBase)) {
-                if ((t = q[i = m & b]) != null &&
-                    queueBase == b &&
-                    UNSAFE.compareAndSwapObject(q, (i << ASHIFT) + ABASE,
-                                                t, null)) {
-                    queueBase = b + 1;
-                    return t;
-                }
-            }
-        }
-        return null;
-    }
-
-    /**
-     * Returns a popped task, or null if empty.
-     * Called only by this thread.
-     */
-    private ForkJoinTask<?> popTask() {
-        int m;
-        ForkJoinTask<?>[] q = queue;
-        if (q != null && (m = q.length - 1) >= 0) {
-            for (int s; (s = queueTop) != queueBase;) {
-                int i = m & --s;
-                long u = (i << ASHIFT) + ABASE; // raw offset
-                ForkJoinTask<?> t = q[i];
-                if (t == null)   // lost to stealer
-                    break;
-                if (UNSAFE.compareAndSwapObject(q, u, t, null)) {
-                    queueTop = s; // or putOrderedInt
-                    return t;
-                }
-            }
-        }
-        return null;
-    }
-
-    /**
-     * Specialized version of popTask to pop only if topmost element
-     * is the given task. Called only by this thread.
-     *
-     * @param t the task. Caller must ensure non-null.
-     */
-    final boolean unpushTask(ForkJoinTask<?> t) {
-        ForkJoinTask<?>[] q;
-        int s;
-        if ((q = queue) != null && (s = queueTop) != queueBase &&
-            UNSAFE.compareAndSwapObject
-            (q, (((q.length - 1) & --s) << ASHIFT) + ABASE, t, null)) {
-            queueTop = s; // or putOrderedInt
-            return true;
-        }
-        return false;
-    }
-
-    /**
-     * Returns next task, or null if empty or contended.
-     */
-    final ForkJoinTask<?> peekTask() {
-        int m;
-        ForkJoinTask<?>[] q = queue;
-        if (q == null || (m = q.length - 1) < 0)
-            return null;
-        int i = locallyFifo ? queueBase : (queueTop - 1);
-        return q[i & m];
-    }
-
-    // Support methods for ForkJoinPool
-
-    /**
-     * Runs the given task, plus any local tasks until queue is empty
-     */
-    final void execTask(ForkJoinTask<?> t) {
-        currentSteal = t;
-        for (;;) {
-            if (t != null)
-                t.doExec();
-            if (queueTop == queueBase)
-                break;
-            t = locallyFifo ? locallyDeqTask() : popTask();
-        }
-        ++stealCount;
-        currentSteal = null;
-    }
-
-    /**
-     * Removes and cancels all tasks in queue.  Can be called from any
-     * thread.
-     */
-    final void cancelTasks() {
-        ForkJoinTask<?> cj = currentJoin; // try to cancel ongoing tasks
-        if (cj != null && cj.status >= 0)
-            cj.cancelIgnoringExceptions();
-        ForkJoinTask<?> cs = currentSteal;
-        if (cs != null && cs.status >= 0)
-            cs.cancelIgnoringExceptions();
-        while (queueBase != queueTop) {
-            ForkJoinTask<?> t = deqTask();
-            if (t != null)
-                t.cancelIgnoringExceptions();
-        }
-    }
-
-    /**
-     * Drains tasks to given collection c.
-     *
-     * @return the number of tasks drained
-     */
-    final int drainTasksTo(Collection<? super ForkJoinTask<?>> c) {
-        int n = 0;
-        while (queueBase != queueTop) {
-            ForkJoinTask<?> t = deqTask();
-            if (t != null) {
-                c.add(t);
-                ++n;
-            }
-        }
-        return n;
-    }
-
-    // Support methods for ForkJoinTask
-
-    /**
-     * Returns an estimate of the number of tasks in the queue.
-     */
-    final int getQueueSize() {
-        return queueTop - queueBase;
-    }
-
-    /**
-     * Gets and removes a local task.
-     *
-     * @return a task, if available
-     */
-    final ForkJoinTask<?> pollLocalTask() {
-        return locallyFifo ? locallyDeqTask() : popTask();
-    }
-
-    /**
-     * Gets and removes a local or stolen task.
-     *
-     * @return a task, if available
-     */
-    final ForkJoinTask<?> pollTask() {
-        ForkJoinWorkerThread[] ws;
-        ForkJoinTask<?> t = pollLocalTask();
-        if (t != null || (ws = pool.workers) == null)
-            return t;
-        int n = ws.length; // cheap version of FJP.scan
-        int steps = n << 1;
-        int r = nextSeed();
-        int i = 0;
-        while (i < steps) {
-            ForkJoinWorkerThread w = ws[(i++ + r) & (n - 1)];
-            if (w != null && w.queueBase != w.queueTop && w.queue != null) {
-                if ((t = w.deqTask()) != null)
-                    return t;
-                i = 0;
-            }
-        }
-        return null;
-    }
-
-    /**
-     * The maximum stolen->joining link depth allowed in helpJoinTask,
-     * as well as the maximum number of retries (allowing on average
-     * one staleness retry per level) per attempt to instead try
-     * compensation.  Depths for legitimate chains are unbounded, but
-     * we use a fixed constant to avoid (otherwise unchecked) cycles
-     * and bound staleness of traversal parameters at the expense of
-     * sometimes blocking when we could be helping.
-     */
-    private static final int MAX_HELP = 16;
-
-    /**
-     * Possibly runs some tasks and/or blocks, until joinMe is done.
-     *
-     * @param joinMe the task to join
-     * @return completion status on exit
-     */
-    final int joinTask(ForkJoinTask<?> joinMe) {
-        ForkJoinTask<?> prevJoin = currentJoin;
-        currentJoin = joinMe;
-        for (int s, retries = MAX_HELP;;) {
-            if ((s = joinMe.status) < 0) {
-                currentJoin = prevJoin;
-                return s;
-            }
-            if (retries > 0) {
-                if (queueTop != queueBase) {
-                    if (!localHelpJoinTask(joinMe))
-                        retries = 0;           // cannot help
-                }
-                else if (retries == MAX_HELP >>> 1) {
-                    --retries;                 // check uncommon case
-                    if (tryDeqAndExec(joinMe) >= 0)
-                        Thread.yield();        // for politeness
-                }
-                else
-                    retries = helpJoinTask(joinMe) ? MAX_HELP : retries - 1;
-            }
-            else {
-                retries = MAX_HELP;           // restart if not done
-                pool.tryAwaitJoin(joinMe);
-            }
-        }
-    }
-
-    /**
-     * If present, pops and executes the given task, or any other
-     * cancelled task
-     *
-     * @return false if any other non-cancelled task exists in local queue
-     */
-    private boolean localHelpJoinTask(ForkJoinTask<?> joinMe) {
-        int s, i; ForkJoinTask<?>[] q; ForkJoinTask<?> t;
-        if ((s = queueTop) != queueBase && (q = queue) != null &&
-            (i = (q.length - 1) & --s) >= 0 &&
-            (t = q[i]) != null) {
-            if (t != joinMe && t.status >= 0)
-                return false;
-            if (UNSAFE.compareAndSwapObject
-                (q, (i << ASHIFT) + ABASE, t, null)) {
-                queueTop = s;           // or putOrderedInt
-                t.doExec();
-            }
-        }
-        return true;
-    }
-
-    /**
-     * Tries to locate and execute tasks for a stealer of the given
-     * task, or in turn one of its stealers, Traces
-     * currentSteal->currentJoin links looking for a thread working on
-     * a descendant of the given task and with a non-empty queue to
-     * steal back and execute tasks from.  The implementation is very
-     * branchy to cope with potential inconsistencies or loops
-     * encountering chains that are stale, unknown, or of length
-     * greater than MAX_HELP links.  All of these cases are dealt with
-     * by just retrying by caller.
-     *
-     * @param joinMe the task to join
-     * @return true if ran a task
-     */
-    private boolean helpJoinTask(ForkJoinTask<?> joinMe) {
-        boolean helped = false;
-        int m = pool.scanGuard & SMASK;
-        ForkJoinWorkerThread[] ws = pool.workers;
-        if (ws != null && ws.length > m && joinMe.status >= 0) {
-            int levels = MAX_HELP;              // remaining chain length
-            ForkJoinTask<?> task = joinMe;      // base of chain
-            outer:for (ForkJoinWorkerThread thread = this;;) {
-                // Try to find v, the stealer of task, by first using hint
-                ForkJoinWorkerThread v = ws[thread.stealHint & m];
-                if (v == null || v.currentSteal != task) {
-                    for (int j = 0; ;) {        // search array
-                        if ((v = ws[j]) != null && v.currentSteal == task) {
-                            thread.stealHint = j;
-                            break;              // save hint for next time
-                        }
-                        if (++j > m)
-                            break outer;        // can't find stealer
-                    }
-                }
-                // Try to help v, using specialized form of deqTask
-                for (;;) {
-                    ForkJoinTask<?>[] q; int b, i;
-                    if (joinMe.status < 0)
-                        break outer;
-                    if ((b = v.queueBase) == v.queueTop ||
-                        (q = v.queue) == null ||
-                        (i = (q.length-1) & b) < 0)
-                        break;                  // empty
-                    long u = (i << ASHIFT) + ABASE;
-                    ForkJoinTask<?> t = q[i];
-                    if (task.status < 0)
-                        break outer;            // stale
-                    if (t != null && v.queueBase == b &&
-                        UNSAFE.compareAndSwapObject(q, u, t, null)) {
-                        v.queueBase = b + 1;
-                        v.stealHint = poolIndex;
-                        ForkJoinTask<?> ps = currentSteal;
-                        currentSteal = t;
-                        t.doExec();
-                        currentSteal = ps;
-                        helped = true;
-                    }
-                }
-                // Try to descend to find v's stealer
-                ForkJoinTask<?> next = v.currentJoin;
-                if (--levels > 0 && task.status >= 0 &&
-                    next != null && next != task) {
-                    task = next;
-                    thread = v;
-                }
-                else
-                    break;  // max levels, stale, dead-end, or cyclic
-            }
-        }
-        return helped;
-    }
-
-    /**
-     * Performs an uncommon case for joinTask: If task t is at base of
-     * some workers queue, steals and executes it.
-     *
-     * @param t the task
-     * @return t's status
-     */
-    private int tryDeqAndExec(ForkJoinTask<?> t) {
-        int m = pool.scanGuard & SMASK;
-        ForkJoinWorkerThread[] ws = pool.workers;
-        if (ws != null && ws.length > m && t.status >= 0) {
-            for (int j = 0; j <= m; ++j) {
-                ForkJoinTask<?>[] q; int b, i;
-                ForkJoinWorkerThread v = ws[j];
-                if (v != null &&
-                    (b = v.queueBase) != v.queueTop &&
-                    (q = v.queue) != null &&
-                    (i = (q.length - 1) & b) >= 0 &&
-                    q[i] ==  t) {
-                    long u = (i << ASHIFT) + ABASE;
-                    if (v.queueBase == b &&
-                        UNSAFE.compareAndSwapObject(q, u, t, null)) {
-                        v.queueBase = b + 1;
-                        v.stealHint = poolIndex;
-                        ForkJoinTask<?> ps = currentSteal;
-                        currentSteal = t;
-                        t.doExec();
-                        currentSteal = ps;
-                    }
-                    break;
-                }
-            }
-        }
-        return t.status;
-    }
-
-    /**
-     * Implements ForkJoinTask.getSurplusQueuedTaskCount().  Returns
-     * an estimate of the number of tasks, offset by a function of
-     * number of idle workers.
-     *
-     * This method provides a cheap heuristic guide for task
-     * partitioning when programmers, frameworks, tools, or languages
-     * have little or no idea about task granularity.  In essence by
-     * offering this method, we ask users only about tradeoffs in
-     * overhead vs expected throughput and its variance, rather than
-     * how finely to partition tasks.
-     *
-     * In a steady state strict (tree-structured) computation, each
-     * thread makes available for stealing enough tasks for other
-     * threads to remain active. Inductively, if all threads play by
-     * the same rules, each thread should make available only a
-     * constant number of tasks.
-     *
-     * The minimum useful constant is just 1. But using a value of 1
-     * would require immediate replenishment upon each steal to
-     * maintain enough tasks, which is infeasible.  Further,
-     * partitionings/granularities of offered tasks should minimize
-     * steal rates, which in general means that threads nearer the top
-     * of computation tree should generate more than those nearer the
-     * bottom. In perfect steady state, each thread is at
-     * approximately the same level of computation tree. However,
-     * producing extra tasks amortizes the uncertainty of progress and
-     * diffusion assumptions.
-     *
-     * So, users will want to use values larger, but not much larger
-     * than 1 to both smooth over transient shortages and hedge
-     * against uneven progress; as traded off against the cost of
-     * extra task overhead. We leave the user to pick a threshold
-     * value to compare with the results of this call to guide
-     * decisions, but recommend values such as 3.
-     *
-     * When all threads are active, it is on average OK to estimate
-     * surplus strictly locally. In steady-state, if one thread is
-     * maintaining say 2 surplus tasks, then so are others. So we can
-     * just use estimated queue length (although note that (queueTop -
-     * queueBase) can be an overestimate because of stealers lagging
-     * increments of queueBase).  However, this strategy alone leads
-     * to serious mis-estimates in some non-steady-state conditions
-     * (ramp-up, ramp-down, other stalls). We can detect many of these
-     * by further considering the number of "idle" threads, that are
-     * known to have zero queued tasks, so compensate by a factor of
-     * (#idle/#active) threads.
-     */
-    final int getEstimatedSurplusTaskCount() {
-        return queueTop - queueBase - pool.idlePerActive();
-    }
-
-    /**
-     * Runs tasks until {@code pool.isQuiescent()}. We piggyback on
-     * pool's active count ctl maintenance, but rather than blocking
-     * when tasks cannot be found, we rescan until all others cannot
-     * find tasks either. The bracketing by pool quiescerCounts
-     * updates suppresses pool auto-shutdown mechanics that could
-     * otherwise prematurely terminate the pool because all threads
-     * appear to be inactive.
-     */
-    final void helpQuiescePool() {
-        boolean active = true;
-        ForkJoinTask<?> ps = currentSteal; // to restore below
-        ForkJoinPool p = pool;
-        p.addQuiescerCount(1);
-        for (;;) {
-            ForkJoinWorkerThread[] ws = p.workers;
-            ForkJoinWorkerThread v = null;
-            int n;
-            if (queueTop != queueBase)
-                v = this;
-            else if (ws != null && (n = ws.length) > 1) {
-                ForkJoinWorkerThread w;
-                int r = nextSeed(); // cheap version of FJP.scan
-                int steps = n << 1;
-                for (int i = 0; i < steps; ++i) {
-                    if ((w = ws[(i + r) & (n - 1)]) != null &&
-                        w.queueBase != w.queueTop) {
-                        v = w;
-                        break;
-                    }
-                }
-            }
-            if (v != null) {
-                ForkJoinTask<?> t;
-                if (!active) {
-                    active = true;
-                    p.addActiveCount(1);
-                }
-                if ((t = (v != this) ? v.deqTask() :
-                     locallyFifo ? locallyDeqTask() : popTask()) != null) {
-                    currentSteal = t;
-                    t.doExec();
-                    currentSteal = ps;
-                }
-            }
-            else {
-                if (active) {
-                    active = false;
-                    p.addActiveCount(-1);
-                }
-                if (p.isQuiescent()) {
-                    p.addActiveCount(1);
-                    p.addQuiescerCount(-1);
-                    break;
-                }
-            }
-        }
-    }
-
-    // Unsafe mechanics
-    private static final sun.misc.Unsafe UNSAFE;
-    private static final long ABASE;
-    private static final int ASHIFT;
-
-    static {
-        int s;
-        try {
-            UNSAFE = sun.misc.Unsafe.getUnsafe();
-            Class<?> a = ForkJoinTask[].class;
-            ABASE = UNSAFE.arrayBaseOffset(a);
-            s = UNSAFE.arrayIndexScale(a);
-        } catch (Exception e) {
-            throw new Error(e);
-        }
-        if ((s & (s-1)) != 0)
-            throw new Error("data type scale not a power of two");
-        ASHIFT = 31 - Integer.numberOfLeadingZeros(s);
-    }
-
 }
diff --git a/luni/src/main/java/java/util/concurrent/Future.java b/luni/src/main/java/java/util/concurrent/Future.java
index 6a37729..32e8145 100644
--- a/luni/src/main/java/java/util/concurrent/Future.java
+++ b/luni/src/main/java/java/util/concurrent/Future.java
@@ -7,19 +7,19 @@
 package java.util.concurrent;
 
 /**
- * A <tt>Future</tt> represents the result of an asynchronous
+ * A {@code Future} represents the result of an asynchronous
  * computation.  Methods are provided to check if the computation is
  * complete, to wait for its completion, and to retrieve the result of
  * the computation.  The result can only be retrieved using method
- * <tt>get</tt> when the computation has completed, blocking if
+ * {@code get} when the computation has completed, blocking if
  * necessary until it is ready.  Cancellation is performed by the
- * <tt>cancel</tt> method.  Additional methods are provided to
+ * {@code cancel} method.  Additional methods are provided to
  * determine if the task completed normally or was cancelled. Once a
  * computation has completed, the computation cannot be cancelled.
- * If you would like to use a <tt>Future</tt> for the sake
+ * If you would like to use a {@code Future} for the sake
  * of cancellability but not provide a usable result, you can
  * declare types of the form {@code Future<?>} and
- * return <tt>null</tt> as a result of the underlying task.
+ * return {@code null} as a result of the underlying task.
  *
  * <p>
  * <b>Sample Usage</b> (Note that the following classes are all
@@ -43,16 +43,16 @@
  *   }
  * }}</pre>
  *
- * The {@link FutureTask} class is an implementation of <tt>Future</tt> that
- * implements <tt>Runnable</tt>, and so may be executed by an <tt>Executor</tt>.
- * For example, the above construction with <tt>submit</tt> could be replaced by:
+ * The {@link FutureTask} class is an implementation of {@code Future} that
+ * implements {@code Runnable}, and so may be executed by an {@code Executor}.
+ * For example, the above construction with {@code submit} could be replaced by:
  *  <pre> {@code
- *     FutureTask<String> future =
- *       new FutureTask<String>(new Callable<String>() {
- *         public String call() {
- *           return searcher.search(target);
- *       }});
- *     executor.execute(future);}</pre>
+ * FutureTask<String> future =
+ *   new FutureTask<String>(new Callable<String>() {
+ *     public String call() {
+ *       return searcher.search(target);
+ *   }});
+ * executor.execute(future);}</pre>
  *
  * <p>Memory consistency effects: Actions taken by the asynchronous computation
  * <a href="package-summary.html#MemoryVisibility"> <i>happen-before</i></a>
@@ -62,7 +62,7 @@
  * @see Executor
  * @since 1.5
  * @author Doug Lea
- * @param <V> The result type returned by this Future's <tt>get</tt> method
+ * @param <V> The result type returned by this Future's {@code get} method
  */
 public interface Future<V> {
 
@@ -70,41 +70,41 @@
      * Attempts to cancel execution of this task.  This attempt will
      * fail if the task has already completed, has already been cancelled,
      * or could not be cancelled for some other reason. If successful,
-     * and this task has not started when <tt>cancel</tt> is called,
+     * and this task has not started when {@code cancel} is called,
      * this task should never run.  If the task has already started,
-     * then the <tt>mayInterruptIfRunning</tt> parameter determines
+     * then the {@code mayInterruptIfRunning} parameter determines
      * whether the thread executing this task should be interrupted in
      * an attempt to stop the task.
      *
      * <p>After this method returns, subsequent calls to {@link #isDone} will
-     * always return <tt>true</tt>.  Subsequent calls to {@link #isCancelled}
-     * will always return <tt>true</tt> if this method returned <tt>true</tt>.
+     * always return {@code true}.  Subsequent calls to {@link #isCancelled}
+     * will always return {@code true} if this method returned {@code true}.
      *
-     * @param mayInterruptIfRunning <tt>true</tt> if the thread executing this
+     * @param mayInterruptIfRunning {@code true} if the thread executing this
      * task should be interrupted; otherwise, in-progress tasks are allowed
      * to complete
-     * @return <tt>false</tt> if the task could not be cancelled,
+     * @return {@code false} if the task could not be cancelled,
      * typically because it has already completed normally;
-     * <tt>true</tt> otherwise
+     * {@code true} otherwise
      */
     boolean cancel(boolean mayInterruptIfRunning);
 
     /**
-     * Returns <tt>true</tt> if this task was cancelled before it completed
+     * Returns {@code true} if this task was cancelled before it completed
      * normally.
      *
-     * @return <tt>true</tt> if this task was cancelled before it completed
+     * @return {@code true} if this task was cancelled before it completed
      */
     boolean isCancelled();
 
     /**
-     * Returns <tt>true</tt> if this task completed.
+     * Returns {@code true} if this task completed.
      *
      * Completion may be due to normal termination, an exception, or
      * cancellation -- in all of these cases, this method will return
-     * <tt>true</tt>.
+     * {@code true}.
      *
-     * @return <tt>true</tt> if this task completed
+     * @return {@code true} if this task completed
      */
     boolean isDone();
 
diff --git a/luni/src/main/java/java/util/concurrent/FutureTask.java b/luni/src/main/java/java/util/concurrent/FutureTask.java
index d51881d..114fe49 100644
--- a/luni/src/main/java/java/util/concurrent/FutureTask.java
+++ b/luni/src/main/java/java/util/concurrent/FutureTask.java
@@ -83,12 +83,11 @@
      *
      * @param s completed state value
      */
+    @SuppressWarnings("unchecked")
     private V report(int s) throws ExecutionException {
         Object x = outcome;
-        if (s == NORMAL) {
-            @SuppressWarnings("unchecked") V v = (V)x;
-            return v;
-        }
+        if (s == NORMAL)
+            return (V)x;
         if (s >= CANCELLED)
             throw new CancellationException();
         throw new ExecutionException((Throwable)x);
@@ -134,19 +133,23 @@
     }
 
     public boolean cancel(boolean mayInterruptIfRunning) {
-        if (state != NEW)
+        if (!(state == NEW &&
+              UNSAFE.compareAndSwapInt(this, stateOffset, NEW,
+                  mayInterruptIfRunning ? INTERRUPTING : CANCELLED)))
             return false;
-        if (mayInterruptIfRunning) {
-            if (!UNSAFE.compareAndSwapInt(this, stateOffset, NEW, INTERRUPTING))
-                return false;
-            Thread t = runner;
-            if (t != null)
-                t.interrupt();
-            UNSAFE.putOrderedInt(this, stateOffset, INTERRUPTED); // final state
+        try {    // in case call to interrupt throws exception
+            if (mayInterruptIfRunning) {
+                try {
+                    Thread t = runner;
+                    if (t != null)
+                        t.interrupt();
+                } finally { // final state
+                    UNSAFE.putOrderedInt(this, stateOffset, INTERRUPTED);
+                }
+            }
+        } finally {
+            finishCompletion();
         }
-        else if (!UNSAFE.compareAndSwapInt(this, stateOffset, NEW, CANCELLED))
-            return false;
-        finishCompletion();
         return true;
     }
 
@@ -363,7 +366,7 @@
      */
     private int awaitDone(boolean timed, long nanos)
         throws InterruptedException {
-        long last = timed ? System.nanoTime() : 0L;
+        final long deadline = timed ? System.nanoTime() + nanos : 0L;
         WaitNode q = null;
         boolean queued = false;
         for (;;) {
@@ -378,18 +381,19 @@
                     q.thread = null;
                 return s;
             }
+            else if (s == COMPLETING) // cannot time out yet
+                Thread.yield();
             else if (q == null)
                 q = new WaitNode();
             else if (!queued)
                 queued = UNSAFE.compareAndSwapObject(this, waitersOffset,
                                                      q.next = waiters, q);
             else if (timed) {
-                long now = System.nanoTime();
-                if ((nanos -= (now - last)) <= 0L) {
+                nanos = deadline - System.nanoTime();
+                if (nanos <= 0L) {
                     removeWaiter(q);
                     return state;
                 }
-                last = now;
                 LockSupport.parkNanos(this, nanos);
             }
             else
diff --git a/luni/src/main/java/java/util/concurrent/LinkedBlockingDeque.java b/luni/src/main/java/java/util/concurrent/LinkedBlockingDeque.java
index 6f32c47..64b0bf1 100644
--- a/luni/src/main/java/java/util/concurrent/LinkedBlockingDeque.java
+++ b/luni/src/main/java/java/util/concurrent/LinkedBlockingDeque.java
@@ -21,7 +21,7 @@
  * An optionally-bounded {@linkplain BlockingDeque blocking deque} based on
  * linked nodes.
  *
- * <p> The optional capacity bound constructor argument serves as a
+ * <p>The optional capacity bound constructor argument serves as a
  * way to prevent excessive expansion. The capacity, if unspecified,
  * is equal to {@link Integer#MAX_VALUE}.  Linked nodes are
  * dynamically created upon each insertion unless this would bring the
@@ -44,7 +44,7 @@
  */
 public class LinkedBlockingDeque<E>
     extends AbstractQueue<E>
-    implements BlockingDeque<E>,  java.io.Serializable {
+    implements BlockingDeque<E>, java.io.Serializable {
 
     /*
      * Implemented as a simple doubly-linked list protected by a
@@ -1123,11 +1123,10 @@
     }
 
     /**
-     * Saves the state of this deque to a stream (that is, serializes it).
+     * Saves this deque to a stream (that is, serializes it).
      *
      * @serialData The capacity (int), followed by elements (each an
      * {@code Object}) in the proper order, followed by a null
-     * @param s the stream
      */
     private void writeObject(java.io.ObjectOutputStream s)
         throws java.io.IOException {
@@ -1148,8 +1147,6 @@
 
     /**
      * Reconstitutes this deque from a stream (that is, deserializes it).
-     *
-     * @param s the stream
      */
     private void readObject(java.io.ObjectInputStream s)
         throws java.io.IOException, ClassNotFoundException {
diff --git a/luni/src/main/java/java/util/concurrent/LinkedBlockingQueue.java b/luni/src/main/java/java/util/concurrent/LinkedBlockingQueue.java
index e8c9edb..0719828 100644
--- a/luni/src/main/java/java/util/concurrent/LinkedBlockingQueue.java
+++ b/luni/src/main/java/java/util/concurrent/LinkedBlockingQueue.java
@@ -31,7 +31,7 @@
  * Linked queues typically have higher throughput than array-based queues but
  * less predictable performance in most concurrent applications.
  *
- * <p> The optional capacity bound constructor argument serves as a
+ * <p>The optional capacity bound constructor argument serves as a
  * way to prevent excessive queue expansion. The capacity, if unspecified,
  * is equal to {@link Integer#MAX_VALUE}.  Linked nodes are
  * dynamically created upon each insertion unless this would bring the
@@ -44,7 +44,6 @@
  * @since 1.5
  * @author Doug Lea
  * @param <E> the type of elements held in this collection
- *
  */
 public class LinkedBlockingQueue<E> extends AbstractQueue<E>
         implements BlockingQueue<E>, java.io.Serializable {
@@ -188,7 +187,7 @@
     }
 
     /**
-     * Lock to prevent both puts and takes.
+     * Locks to prevent both puts and takes.
      */
     void fullyLock() {
         putLock.lock();
@@ -196,7 +195,7 @@
     }
 
     /**
-     * Unlock to allow both puts and takes.
+     * Unlocks to allow both puts and takes.
      */
     void fullyUnlock() {
         takeLock.unlock();
@@ -262,7 +261,6 @@
         }
     }
 
-
     // this doc comment is overridden to remove the reference to collections
     // greater in size than Integer.MAX_VALUE
     /**
@@ -335,7 +333,7 @@
      * necessary up to the specified wait time for space to become available.
      *
      * @return {@code true} if successful, or {@code false} if
-     *         the specified waiting time elapses before space is available.
+     *         the specified waiting time elapses before space is available
      * @throws InterruptedException {@inheritDoc}
      * @throws NullPointerException {@inheritDoc}
      */
@@ -401,7 +399,6 @@
         return c >= 0;
     }
 
-
     public E take() throws InterruptedException {
         E x;
         int c = -1;
@@ -756,6 +753,7 @@
          * item to hand out so that if hasNext() reports true, we will
          * still have it to return even if lost race with a take etc.
          */
+
         private Node<E> current;
         private Node<E> lastRet;
         private E currentElement;
@@ -830,12 +828,11 @@
     }
 
     /**
-     * Saves the state to a stream (that is, serializes it).
+     * Saves this queue to a stream (that is, serializes it).
      *
      * @serialData The capacity is emitted (int), followed by all of
      * its elements (each an {@code Object}) in the proper order,
      * followed by a null
-     * @param s the stream
      */
     private void writeObject(java.io.ObjectOutputStream s)
         throws java.io.IOException {
@@ -858,8 +855,6 @@
 
     /**
      * Reconstitutes this queue from a stream (that is, deserializes it).
-     *
-     * @param s the stream
      */
     private void readObject(java.io.ObjectInputStream s)
         throws java.io.IOException, ClassNotFoundException {
diff --git a/luni/src/main/java/java/util/concurrent/LinkedTransferQueue.java b/luni/src/main/java/java/util/concurrent/LinkedTransferQueue.java
index 2a3446e..cff5dbf 100644
--- a/luni/src/main/java/java/util/concurrent/LinkedTransferQueue.java
+++ b/luni/src/main/java/java/util/concurrent/LinkedTransferQueue.java
@@ -667,7 +667,7 @@
      * @return matched item, or e if unmatched on interrupt or timeout
      */
     private E awaitMatch(Node s, Node pred, E e, boolean timed, long nanos) {
-        long lastTime = timed ? System.nanoTime() : 0L;
+        final long deadline = timed ? System.nanoTime() + nanos : 0L;
         Thread w = Thread.currentThread();
         int spins = -1; // initialized after first item and cancel checks
         ThreadLocalRandom randomYields = null; // bound if needed
@@ -698,10 +698,9 @@
                 s.waiter = w;                 // request unpark then recheck
             }
             else if (timed) {
-                long now = System.nanoTime();
-                if ((nanos -= now - lastTime) > 0)
+                nanos = deadline - System.nanoTime();
+                if (nanos > 0L)
                     LockSupport.parkNanos(this, nanos);
-                lastTime = now;
             }
             else {
                 LockSupport.park(this);
@@ -980,7 +979,6 @@
         return false;
     }
 
-
     /**
      * Creates an initially empty {@code LinkedTransferQueue}.
      */
@@ -1267,11 +1265,10 @@
     }
 
     /**
-     * Saves the state to a stream (that is, serializes it).
+     * Saves this queue to a stream (that is, serializes it).
      *
      * @serialData All of the elements (each an {@code E}) in
      * the proper order, followed by a null
-     * @param s the stream
      */
     private void writeObject(java.io.ObjectOutputStream s)
         throws java.io.IOException {
@@ -1283,16 +1280,14 @@
     }
 
     /**
-     * Reconstitutes the Queue instance from a stream (that is,
-     * deserializes it).
-     *
-     * @param s the stream
+     * Reconstitutes this queue from a stream (that is, deserializes it).
      */
     private void readObject(java.io.ObjectInputStream s)
         throws java.io.IOException, ClassNotFoundException {
         s.defaultReadObject();
         for (;;) {
-            @SuppressWarnings("unchecked") E item = (E) s.readObject();
+            @SuppressWarnings("unchecked")
+            E item = (E) s.readObject();
             if (item == null)
                 break;
             else
diff --git a/luni/src/main/java/java/util/concurrent/Phaser.java b/luni/src/main/java/java/util/concurrent/Phaser.java
index 25ff743..a9adbe5 100644
--- a/luni/src/main/java/java/util/concurrent/Phaser.java
+++ b/luni/src/main/java/java/util/concurrent/Phaser.java
@@ -17,7 +17,7 @@
  * {@link java.util.concurrent.CountDownLatch CountDownLatch}
  * but supporting more flexible usage.
  *
- * <p> <b>Registration.</b> Unlike the case for other barriers, the
+ * <p><b>Registration.</b> Unlike the case for other barriers, the
  * number of parties <em>registered</em> to synchronize on a phaser
  * may vary over time.  Tasks may be registered at any time (using
  * methods {@link #register}, {@link #bulkRegister}, or forms of
@@ -30,7 +30,7 @@
  * (However, you can introduce such bookkeeping by subclassing this
  * class.)
  *
- * <p> <b>Synchronization.</b> Like a {@code CyclicBarrier}, a {@code
+ * <p><b>Synchronization.</b> Like a {@code CyclicBarrier}, a {@code
  * Phaser} may be repeatedly awaited.  Method {@link
  * #arriveAndAwaitAdvance} has effect analogous to {@link
  * java.util.concurrent.CyclicBarrier#await CyclicBarrier.await}. Each
@@ -74,7 +74,7 @@
  *
  * </ul>
  *
- * <p> <b>Termination.</b> A phaser may enter a <em>termination</em>
+ * <p><b>Termination.</b> A phaser may enter a <em>termination</em>
  * state, that may be checked using method {@link #isTerminated}. Upon
  * termination, all synchronization methods immediately return without
  * waiting for advance, as indicated by a negative return value.
@@ -89,7 +89,7 @@
  * also available to abruptly release waiting threads and allow them
  * to terminate.
  *
- * <p> <b>Tiering.</b> Phasers may be <em>tiered</em> (i.e.,
+ * <p><b>Tiering.</b> Phasers may be <em>tiered</em> (i.e.,
  * constructed in tree structures) to reduce contention. Phasers with
  * large numbers of parties that would otherwise experience heavy
  * synchronization contention costs may instead be set up so that
@@ -1065,7 +1065,7 @@
         final boolean timed;
         boolean wasInterrupted;
         long nanos;
-        long lastTime;
+        final long deadline;
         volatile Thread thread; // nulled to cancel wait
         QNode next;
 
@@ -1076,7 +1076,7 @@
             this.interruptible = interruptible;
             this.nanos = nanos;
             this.timed = timed;
-            this.lastTime = timed ? System.nanoTime() : 0L;
+            this.deadline = timed ? System.nanoTime() + nanos : 0L;
             thread = Thread.currentThread();
         }
 
@@ -1095,9 +1095,7 @@
             }
             if (timed) {
                 if (nanos > 0L) {
-                    long now = System.nanoTime();
-                    nanos -= now - lastTime;
-                    lastTime = now;
+                    nanos = deadline - System.nanoTime();
                 }
                 if (nanos <= 0L) {
                     thread = null;
@@ -1112,7 +1110,7 @@
                 return true;
             else if (!timed)
                 LockSupport.park(this);
-            else if (nanos > 0)
+            else if (nanos > 0L)
                 LockSupport.parkNanos(this, nanos);
             return isReleasable();
         }
diff --git a/luni/src/main/java/java/util/concurrent/PriorityBlockingQueue.java b/luni/src/main/java/java/util/concurrent/PriorityBlockingQueue.java
index 26c72eb..0f9e715 100644
--- a/luni/src/main/java/java/util/concurrent/PriorityBlockingQueue.java
+++ b/luni/src/main/java/java/util/concurrent/PriorityBlockingQueue.java
@@ -6,7 +6,8 @@
 
 package java.util.concurrent;
 
-import java.util.concurrent.locks.*;
+import java.util.concurrent.locks.Condition;
+import java.util.concurrent.locks.ReentrantLock;
 import java.util.*;
 
 // BEGIN android-note
@@ -65,6 +66,7 @@
  * @author Doug Lea
  * @param <E> the type of elements held in this collection
  */
+@SuppressWarnings("unchecked")
 public class PriorityBlockingQueue<E> extends AbstractQueue<E>
     implements BlockingQueue<E>, java.io.Serializable {
     private static final long serialVersionUID = 5595510919245408276L;
@@ -311,7 +313,6 @@
      * @param k the position to fill
      * @param x the item to insert
      * @param array the heap array
-     * @param n heap size
      */
     private static <T> void siftUpComparable(int k, T x, Object[] array) {
         Comparable<? super T> key = (Comparable<? super T>) x;
@@ -351,39 +352,43 @@
      */
     private static <T> void siftDownComparable(int k, T x, Object[] array,
                                                int n) {
-        Comparable<? super T> key = (Comparable<? super T>)x;
-        int half = n >>> 1;           // loop while a non-leaf
-        while (k < half) {
-            int child = (k << 1) + 1; // assume left child is least
-            Object c = array[child];
-            int right = child + 1;
-            if (right < n &&
-                ((Comparable<? super T>) c).compareTo((T) array[right]) > 0)
-                c = array[child = right];
-            if (key.compareTo((T) c) <= 0)
-                break;
-            array[k] = c;
-            k = child;
+        if (n > 0) {
+            Comparable<? super T> key = (Comparable<? super T>)x;
+            int half = n >>> 1;           // loop while a non-leaf
+            while (k < half) {
+                int child = (k << 1) + 1; // assume left child is least
+                Object c = array[child];
+                int right = child + 1;
+                if (right < n &&
+                    ((Comparable<? super T>) c).compareTo((T) array[right]) > 0)
+                    c = array[child = right];
+                if (key.compareTo((T) c) <= 0)
+                    break;
+                array[k] = c;
+                k = child;
+            }
+            array[k] = key;
         }
-        array[k] = key;
     }
 
     private static <T> void siftDownUsingComparator(int k, T x, Object[] array,
                                                     int n,
                                                     Comparator<? super T> cmp) {
-        int half = n >>> 1;
-        while (k < half) {
-            int child = (k << 1) + 1;
-            Object c = array[child];
-            int right = child + 1;
-            if (right < n && cmp.compare((T) c, (T) array[right]) > 0)
-                c = array[child = right];
-            if (cmp.compare(x, (T) c) <= 0)
-                break;
-            array[k] = c;
-            k = child;
+        if (n > 0) {
+            int half = n >>> 1;
+            while (k < half) {
+                int child = (k << 1) + 1;
+                Object c = array[child];
+                int right = child + 1;
+                if (right < n && cmp.compare((T) c, (T) array[right]) > 0)
+                    c = array[child = right];
+                if (cmp.compare(x, (T) c) <= 0)
+                    break;
+                array[k] = c;
+                k = child;
+            }
+            array[k] = x;
         }
-        array[k] = x;
     }
 
     /**
@@ -866,10 +871,11 @@
     }
 
     /**
-     * Saves the state to a stream (that is, serializes it).  For
-     * compatibility with previous version of this class,
-     * elements are first copied to a java.util.PriorityQueue,
-     * which is then serialized.
+     * Saves this queue to a stream (that is, serializes it).
+     *
+     * For compatibility with previous version of this class, elements
+     * are first copied to a java.util.PriorityQueue, which is then
+     * serialized.
      */
     private void writeObject(java.io.ObjectOutputStream s)
         throws java.io.IOException {
@@ -886,10 +892,7 @@
     }
 
     /**
-     * Reconstitutes the {@code PriorityBlockingQueue} instance from a stream
-     * (that is, deserializes it).
-     *
-     * @param s the stream
+     * Reconstitutes this queue from a stream (that is, deserializes it).
      */
     private void readObject(java.io.ObjectInputStream s)
         throws java.io.IOException, ClassNotFoundException {
diff --git a/luni/src/main/java/java/util/concurrent/RecursiveAction.java b/luni/src/main/java/java/util/concurrent/RecursiveAction.java
index 48066c9..8d666f6 100644
--- a/luni/src/main/java/java/util/concurrent/RecursiveAction.java
+++ b/luni/src/main/java/java/util/concurrent/RecursiveAction.java
@@ -34,7 +34,7 @@
  *     }
  *   }
  *   // implementation details follow:
- *   final static int THRESHOLD = 1000;
+ *   static final int THRESHOLD = 1000;
  *   void sortSequentially(int lo, int hi) {
  *     Arrays.sort(array, lo, hi);
  *   }
@@ -111,21 +111,21 @@
  *     int h = hi;
  *     Applyer right = null;
  *     while (h - l > 1 && getSurplusQueuedTaskCount() <= 3) {
- *        int mid = (l + h) >>> 1;
- *        right = new Applyer(array, mid, h, right);
- *        right.fork();
- *        h = mid;
+ *       int mid = (l + h) >>> 1;
+ *       right = new Applyer(array, mid, h, right);
+ *       right.fork();
+ *       h = mid;
  *     }
  *     double sum = atLeaf(l, h);
  *     while (right != null) {
- *        if (right.tryUnfork()) // directly calculate if not stolen
- *          sum += right.atLeaf(right.lo, right.hi);
+ *       if (right.tryUnfork()) // directly calculate if not stolen
+ *         sum += right.atLeaf(right.lo, right.hi);
  *       else {
- *          right.join();
- *          sum += right.result;
- *        }
- *        right = right.next;
- *      }
+ *         right.join();
+ *         sum += right.result;
+ *       }
+ *       right = right.next;
+ *     }
  *     result = sum;
  *   }
  * }}</pre>
diff --git a/luni/src/main/java/java/util/concurrent/RecursiveTask.java b/luni/src/main/java/java/util/concurrent/RecursiveTask.java
index 5e17280..421c9d3 100644
--- a/luni/src/main/java/java/util/concurrent/RecursiveTask.java
+++ b/luni/src/main/java/java/util/concurrent/RecursiveTask.java
@@ -17,7 +17,7 @@
  *   Fibonacci(int n) { this.n = n; }
  *   Integer compute() {
  *     if (n <= 1)
- *        return n;
+ *       return n;
  *     Fibonacci f1 = new Fibonacci(n - 1);
  *     f1.fork();
  *     Fibonacci f2 = new Fibonacci(n - 2);
diff --git a/luni/src/main/java/java/util/concurrent/RejectedExecutionException.java b/luni/src/main/java/java/util/concurrent/RejectedExecutionException.java
index f0005d1..c61365f 100644
--- a/luni/src/main/java/java/util/concurrent/RejectedExecutionException.java
+++ b/luni/src/main/java/java/util/concurrent/RejectedExecutionException.java
@@ -17,14 +17,14 @@
     private static final long serialVersionUID = -375805702767069545L;
 
     /**
-     * Constructs a <tt>RejectedExecutionException</tt> with no detail message.
+     * Constructs a {@code RejectedExecutionException} with no detail message.
      * The cause is not initialized, and may subsequently be
      * initialized by a call to {@link #initCause(Throwable) initCause}.
      */
     public RejectedExecutionException() { }
 
     /**
-     * Constructs a <tt>RejectedExecutionException</tt> with the
+     * Constructs a {@code RejectedExecutionException} with the
      * specified detail message. The cause is not initialized, and may
      * subsequently be initialized by a call to {@link
      * #initCause(Throwable) initCause}.
@@ -36,7 +36,7 @@
     }
 
     /**
-     * Constructs a <tt>RejectedExecutionException</tt> with the
+     * Constructs a {@code RejectedExecutionException} with the
      * specified detail message and cause.
      *
      * @param  message the detail message
@@ -48,10 +48,10 @@
     }
 
     /**
-     * Constructs a <tt>RejectedExecutionException</tt> with the
+     * Constructs a {@code RejectedExecutionException} with the
      * specified cause.  The detail message is set to {@code (cause ==
      * null ? null : cause.toString())} (which typically contains
-     * the class and detail message of <tt>cause</tt>).
+     * the class and detail message of {@code cause}).
      *
      * @param  cause the cause (which is saved for later retrieval by the
      *         {@link #getCause()} method)
diff --git a/luni/src/main/java/java/util/concurrent/RunnableFuture.java b/luni/src/main/java/java/util/concurrent/RunnableFuture.java
index 2d6d52c..ccd28e3 100644
--- a/luni/src/main/java/java/util/concurrent/RunnableFuture.java
+++ b/luni/src/main/java/java/util/concurrent/RunnableFuture.java
@@ -8,13 +8,13 @@
 
 /**
  * A {@link Future} that is {@link Runnable}. Successful execution of
- * the <tt>run</tt> method causes completion of the <tt>Future</tt>
+ * the {@code run} method causes completion of the {@code Future}
  * and allows access to its results.
  * @see FutureTask
  * @see Executor
  * @since 1.6
  * @author Doug Lea
- * @param <V> The result type returned by this Future's <tt>get</tt> method
+ * @param <V> The result type returned by this Future's {@code get} method
  */
 public interface RunnableFuture<V> extends Runnable, Future<V> {
     /**
diff --git a/luni/src/main/java/java/util/concurrent/RunnableScheduledFuture.java b/luni/src/main/java/java/util/concurrent/RunnableScheduledFuture.java
index fbb995c..7125233 100644
--- a/luni/src/main/java/java/util/concurrent/RunnableScheduledFuture.java
+++ b/luni/src/main/java/java/util/concurrent/RunnableScheduledFuture.java
@@ -8,13 +8,13 @@
 
 /**
  * A {@link ScheduledFuture} that is {@link Runnable}. Successful
- * execution of the <tt>run</tt> method causes completion of the
- * <tt>Future</tt> and allows access to its results.
+ * execution of the {@code run} method causes completion of the
+ * {@code Future} and allows access to its results.
  * @see FutureTask
  * @see Executor
  * @since 1.6
  * @author Doug Lea
- * @param <V> The result type returned by this Future's <tt>get</tt> method
+ * @param <V> The result type returned by this Future's {@code get} method
  */
 public interface RunnableScheduledFuture<V> extends RunnableFuture<V>, ScheduledFuture<V> {
 
diff --git a/luni/src/main/java/java/util/concurrent/ScheduledExecutorService.java b/luni/src/main/java/java/util/concurrent/ScheduledExecutorService.java
index 71e57ed..b978fae 100644
--- a/luni/src/main/java/java/util/concurrent/ScheduledExecutorService.java
+++ b/luni/src/main/java/java/util/concurrent/ScheduledExecutorService.java
@@ -10,26 +10,26 @@
  * An {@link ExecutorService} that can schedule commands to run after a given
  * delay, or to execute periodically.
  *
- * <p> The <tt>schedule</tt> methods create tasks with various delays
+ * <p>The {@code schedule} methods create tasks with various delays
  * and return a task object that can be used to cancel or check
- * execution. The <tt>scheduleAtFixedRate</tt> and
- * <tt>scheduleWithFixedDelay</tt> methods create and execute tasks
+ * execution. The {@code scheduleAtFixedRate} and
+ * {@code scheduleWithFixedDelay} methods create and execute tasks
  * that run periodically until cancelled.
  *
- * <p> Commands submitted using the {@link Executor#execute} and
- * {@link ExecutorService} <tt>submit</tt> methods are scheduled with
+ * <p>Commands submitted using the {@link Executor#execute} and
+ * {@link ExecutorService} {@code submit} methods are scheduled with
  * a requested delay of zero. Zero and negative delays (but not
- * periods) are also allowed in <tt>schedule</tt> methods, and are
+ * periods) are also allowed in {@code schedule} methods, and are
  * treated as requests for immediate execution.
  *
- * <p>All <tt>schedule</tt> methods accept <em>relative</em> delays and
+ * <p>All {@code schedule} methods accept <em>relative</em> delays and
  * periods as arguments, not absolute times or dates. It is a simple
  * matter to transform an absolute time represented as a {@link
  * java.util.Date} to the required form. For example, to schedule at
- * a certain future <tt>date</tt>, you can use: <tt>schedule(task,
+ * a certain future {@code date}, you can use: {@code schedule(task,
  * date.getTime() - System.currentTimeMillis(),
- * TimeUnit.MILLISECONDS)</tt>. Beware however that expiration of a
- * relative delay need not coincide with the current <tt>Date</tt> at
+ * TimeUnit.MILLISECONDS)}. Beware however that expiration of a
+ * relative delay need not coincide with the current {@code Date} at
  * which the task is enabled due to network time synchronization
  * protocols, clock drift, or other factors.
  *
@@ -72,8 +72,8 @@
      * @param delay the time from now to delay execution
      * @param unit the time unit of the delay parameter
      * @return a ScheduledFuture representing pending completion of
-     *         the task and whose <tt>get()</tt> method will return
-     *         <tt>null</tt> upon completion
+     *         the task and whose {@code get()} method will return
+     *         {@code null} upon completion
      * @throws RejectedExecutionException if the task cannot be
      *         scheduled for execution
      * @throws NullPointerException if command is null
@@ -100,8 +100,8 @@
      * Creates and executes a periodic action that becomes enabled first
      * after the given initial delay, and subsequently with the given
      * period; that is executions will commence after
-     * <tt>initialDelay</tt> then <tt>initialDelay+period</tt>, then
-     * <tt>initialDelay + 2 * period</tt>, and so on.
+     * {@code initialDelay} then {@code initialDelay+period}, then
+     * {@code initialDelay + 2 * period}, and so on.
      * If any execution of the task
      * encounters an exception, subsequent executions are suppressed.
      * Otherwise, the task will only terminate via cancellation or
@@ -114,7 +114,7 @@
      * @param period the period between successive executions
      * @param unit the time unit of the initialDelay and period parameters
      * @return a ScheduledFuture representing pending completion of
-     *         the task, and whose <tt>get()</tt> method will throw an
+     *         the task, and whose {@code get()} method will throw an
      *         exception upon cancellation
      * @throws RejectedExecutionException if the task cannot be
      *         scheduled for execution
@@ -141,7 +141,7 @@
      * execution and the commencement of the next
      * @param unit the time unit of the initialDelay and delay parameters
      * @return a ScheduledFuture representing pending completion of
-     *         the task, and whose <tt>get()</tt> method will throw an
+     *         the task, and whose {@code get()} method will throw an
      *         exception upon cancellation
      * @throws RejectedExecutionException if the task cannot be
      *         scheduled for execution
diff --git a/luni/src/main/java/java/util/concurrent/ScheduledThreadPoolExecutor.java b/luni/src/main/java/java/util/concurrent/ScheduledThreadPoolExecutor.java
index e41f0c3..a52351b 100644
--- a/luni/src/main/java/java/util/concurrent/ScheduledThreadPoolExecutor.java
+++ b/luni/src/main/java/java/util/concurrent/ScheduledThreadPoolExecutor.java
@@ -13,7 +13,6 @@
 
 // BEGIN android-note
 // omit class-level docs on setRemoveOnCancelPolicy()
-// removed security manager docs
 // END android-note
 
 /**
@@ -31,7 +30,7 @@
  * submission.
  *
  * <p>When a submitted task is cancelled before it is run, execution
- * is suppressed. Such a cancelled task is not
+ * is suppressed. By default, such a cancelled task is not
  * automatically removed from the work queue until its delay
  * elapses. While this enables further inspection and monitoring, it
  * may also cause unbounded retention of cancelled tasks.
@@ -54,7 +53,7 @@
  * without threads to handle tasks once they become eligible to run.
  *
  * <p><b>Extension notes:</b> This class overrides the
- * {@link ThreadPoolExecutor#execute execute} and
+ * {@link ThreadPoolExecutor#execute(Runnable) execute} and
  * {@link AbstractExecutorService#submit(Runnable) submit}
  * methods to generate internal {@link ScheduledFuture} objects to
  * control per-task delays and scheduling.  To preserve
@@ -196,7 +195,7 @@
         }
 
         /**
-         * Creates a one-shot action with given nanoTime-based trigger.
+         * Creates a one-shot action with given nanoTime-based trigger time.
          */
         ScheduledFutureTask(Callable<V> callable, long ns) {
             super(callable);
@@ -224,15 +223,14 @@
                 else
                     return 1;
             }
-            long diff = (getDelay(NANOSECONDS) -
-                         other.getDelay(NANOSECONDS));
+            long diff = getDelay(NANOSECONDS) - other.getDelay(NANOSECONDS);
             return (diff < 0) ? -1 : (diff > 0) ? 1 : 0;
         }
 
         /**
-         * Returns true if this is a periodic (not a one-shot) action.
+         * Returns {@code true} if this is a periodic (not a one-shot) action.
          *
-         * @return true if periodic
+         * @return {@code true} if periodic
          */
         public boolean isPeriodic() {
             return period != 0;
@@ -289,7 +287,7 @@
      * is shut down, rejects the task. Otherwise adds task to queue
      * and starts a thread, if necessary, to run it.  (We cannot
      * prestart the thread to run the task because the task (probably)
-     * shouldn't be run yet,) If the pool is shut down while the task
+     * shouldn't be run yet.)  If the pool is shut down while the task
      * is being added, cancel and remove it if required by state and
      * run-after-shutdown parameters.
      *
@@ -628,7 +626,7 @@
      * {@code false} when already shutdown.
      * This value is by default {@code false}.
      *
-     * @param value if {@code true}, continue after shutdown, else don't.
+     * @param value if {@code true}, continue after shutdown, else don't
      * @see #getContinueExistingPeriodicTasksAfterShutdownPolicy
      */
     public void setContinueExistingPeriodicTasksAfterShutdownPolicy(boolean value) {
@@ -660,7 +658,7 @@
      * {@code false} when already shutdown.
      * This value is by default {@code true}.
      *
-     * @param value if {@code true}, execute after shutdown, else don't.
+     * @param value if {@code true}, execute after shutdown, else don't
      * @see #getExecuteExistingDelayedTasksAfterShutdownPolicy
      */
     public void setExecuteExistingDelayedTasksAfterShutdownPolicy(boolean value) {
@@ -834,7 +832,7 @@
         private final Condition available = lock.newCondition();
 
         /**
-         * Set f's heapIndex if it is a ScheduledFutureTask.
+         * Sets f's heapIndex if it is a ScheduledFutureTask.
          */
         private void setIndex(RunnableScheduledFuture<?> f, int idx) {
             if (f instanceof ScheduledFutureTask)
@@ -842,7 +840,7 @@
         }
 
         /**
-         * Sift element added at bottom up to its heap-ordered spot.
+         * Sifts element added at bottom up to its heap-ordered spot.
          * Call only when holding lock.
          */
         private void siftUp(int k, RunnableScheduledFuture<?> key) {
@@ -860,7 +858,7 @@
         }
 
         /**
-         * Sift element added at top down to its heap-ordered spot.
+         * Sifts element added at top down to its heap-ordered spot.
          * Call only when holding lock.
          */
         private void siftDown(int k, RunnableScheduledFuture<?> key) {
@@ -882,7 +880,7 @@
         }
 
         /**
-         * Resize the heap array.  Call only when holding lock.
+         * Resizes the heap array.  Call only when holding lock.
          */
         private void grow() {
             int oldCapacity = queue.length;
@@ -893,7 +891,7 @@
         }
 
         /**
-         * Find index of given object, or -1 if absent
+         * Finds index of given object, or -1 if absent.
          */
         private int indexOf(Object x) {
             if (x != null) {
@@ -1054,7 +1052,8 @@
                         long delay = first.getDelay(NANOSECONDS);
                         if (delay <= 0)
                             return finishPoll(first);
-                        else if (leader != null)
+                        first = null; // don't retain ref while waiting
+                        if (leader != null)
                             available.await();
                         else {
                             Thread thisThread = Thread.currentThread();
@@ -1094,6 +1093,7 @@
                             return finishPoll(first);
                         if (nanos <= 0)
                             return null;
+                        first = null; // don't retain ref while waiting
                         if (nanos < delay || leader != null)
                             nanos = available.awaitNanos(nanos);
                         else {
@@ -1134,7 +1134,7 @@
         }
 
         /**
-         * Return first element only if it is expired.
+         * Returns first element only if it is expired.
          * Used only by drainTo.  Call only when holding lock.
          */
         private RunnableScheduledFuture<?> peekExpired() {
diff --git a/luni/src/main/java/java/util/concurrent/Semaphore.java b/luni/src/main/java/java/util/concurrent/Semaphore.java
index bf2524c..9ee18a8 100644
--- a/luni/src/main/java/java/util/concurrent/Semaphore.java
+++ b/luni/src/main/java/java/util/concurrent/Semaphore.java
@@ -5,8 +5,8 @@
  */
 
 package java.util.concurrent;
-import java.util.*;
-import java.util.concurrent.locks.*;
+import java.util.Collection;
+import java.util.concurrent.locks.AbstractQueuedSynchronizer;
 
 /**
  * A counting semaphore.  Conceptually, a semaphore maintains a set of
@@ -79,13 +79,13 @@
  * exclusion lock.  This is more commonly known as a <em>binary
  * semaphore</em>, because it only has two states: one permit
  * available, or zero permits available.  When used in this way, the
- * binary semaphore has the property (unlike many {@link Lock}
+ * binary semaphore has the property (unlike many {@link java.util.concurrent.locks.Lock}
  * implementations), that the &quot;lock&quot; can be released by a
  * thread other than the owner (as semaphores have no notion of
  * ownership).  This can be useful in some specialized contexts, such
  * as deadlock recovery.
  *
- * <p> The constructor for this class optionally accepts a
+ * <p>The constructor for this class optionally accepts a
  * <em>fairness</em> parameter. When set false, this class makes no
  * guarantees about the order in which threads acquire permits. In
  * particular, <em>barging</em> is permitted, that is, a thread
@@ -123,9 +123,7 @@
  *
  * @since 1.5
  * @author Doug Lea
- *
  */
-
 public class Semaphore implements java.io.Serializable {
     private static final long serialVersionUID = -3222578661600680210L;
     /** All mechanics via AbstractQueuedSynchronizer subclass */
@@ -461,7 +459,6 @@
      *
      * @param permits the number of permits to acquire
      * @throws IllegalArgumentException if {@code permits} is negative
-     *
      */
     public void acquireUninterruptibly(int permits) {
         if (permits < 0) throw new IllegalArgumentException();
diff --git a/luni/src/main/java/java/util/concurrent/SynchronousQueue.java b/luni/src/main/java/java/util/concurrent/SynchronousQueue.java
index b05ae0a..ea6b3d1 100644
--- a/luni/src/main/java/java/util/concurrent/SynchronousQueue.java
+++ b/luni/src/main/java/java/util/concurrent/SynchronousQueue.java
@@ -6,7 +6,8 @@
  */
 
 package java.util.concurrent;
-import java.util.concurrent.locks.*;
+import java.util.concurrent.locks.LockSupport;
+import java.util.concurrent.locks.ReentrantLock;
 import java.util.*;
 
 // BEGIN android-note
@@ -18,17 +19,17 @@
  * operation must wait for a corresponding remove operation by another
  * thread, and vice versa.  A synchronous queue does not have any
  * internal capacity, not even a capacity of one.  You cannot
- * <tt>peek</tt> at a synchronous queue because an element is only
+ * {@code peek} at a synchronous queue because an element is only
  * present when you try to remove it; you cannot insert an element
  * (using any method) unless another thread is trying to remove it;
  * you cannot iterate as there is nothing to iterate.  The
  * <em>head</em> of the queue is the element that the first queued
  * inserting thread is trying to add to the queue; if there is no such
  * queued thread then no element is available for removal and
- * <tt>poll()</tt> will return <tt>null</tt>.  For purposes of other
- * <tt>Collection</tt> methods (for example <tt>contains</tt>), a
- * <tt>SynchronousQueue</tt> acts as an empty collection.  This queue
- * does not permit <tt>null</tt> elements.
+ * {@code poll()} will return {@code null}.  For purposes of other
+ * {@code Collection} methods (for example {@code contains}), a
+ * {@code SynchronousQueue} acts as an empty collection.  This queue
+ * does not permit {@code null} elements.
  *
  * <p>Synchronous queues are similar to rendezvous channels used in
  * CSP and Ada. They are well suited for handoff designs, in which an
@@ -36,10 +37,10 @@
  * in another thread in order to hand it some information, event, or
  * task.
  *
- * <p> This class supports an optional fairness policy for ordering
+ * <p>This class supports an optional fairness policy for ordering
  * waiting producer and consumer threads.  By default, this ordering
  * is not guaranteed. However, a queue constructed with fairness set
- * to <tt>true</tt> grants threads access in FIFO order.
+ * to {@code true} grants threads access in FIFO order.
  *
  * <p>This class and its iterator implement all of the
  * <em>optional</em> methods of the {@link Collection} and {@link
@@ -133,7 +134,7 @@
     /**
      * Shared internal API for dual stacks and queues.
      */
-    abstract static class Transferer {
+    abstract static class Transferer<E> {
         /**
          * Performs a put or take.
          *
@@ -147,7 +148,7 @@
          *         the caller can distinguish which of these occurred
          *         by checking Thread.interrupted.
          */
-        abstract Object transfer(Object e, boolean timed, long nanos);
+        abstract E transfer(E e, boolean timed, long nanos);
     }
 
     /** The number of CPUs, for spin control */
@@ -176,7 +177,7 @@
     static final long spinForTimeoutThreshold = 1000L;
 
     /** Dual stack */
-    static final class TransferStack extends Transferer {
+    static final class TransferStack<E> extends Transferer<E> {
         /*
          * This extends Scherer-Scott dual stack algorithm, differing,
          * among other ways, by using "covering" nodes rather than
@@ -193,7 +194,7 @@
         /** Node is fulfilling another unfulfilled DATA or REQUEST */
         static final int FULFILLING = 2;
 
-        /** Return true if m has fulfilling bit set */
+        /** Returns true if m has fulfilling bit set. */
         static boolean isFulfilling(int m) { return (m & FULFILLING) != 0; }
 
         /** Node class for TransferStacks. */
@@ -292,7 +293,8 @@
         /**
          * Puts or takes an item.
          */
-        Object transfer(Object e, boolean timed, long nanos) {
+        @SuppressWarnings("unchecked")
+        E transfer(E e, boolean timed, long nanos) {
             /*
              * Basic algorithm is to loop trying one of three actions:
              *
@@ -333,7 +335,7 @@
                         }
                         if ((h = head) != null && h.next == s)
                             casHead(h, s.next);     // help s's fulfiller
-                        return (mode == REQUEST) ? m.item : s.item;
+                        return (E) ((mode == REQUEST) ? m.item : s.item);
                     }
                 } else if (!isFulfilling(h.mode)) { // try to fulfill
                     if (h.isCancelled())            // already cancelled
@@ -349,7 +351,7 @@
                             SNode mn = m.next;
                             if (m.tryMatch(s)) {
                                 casHead(s, mn);     // pop both s and m
-                                return (mode == REQUEST) ? m.item : s.item;
+                                return (E) ((mode == REQUEST) ? m.item : s.item);
                             } else                  // lost match
                                 s.casNext(m, mn);   // help unlink
                         }
@@ -400,7 +402,7 @@
              * and don't wait at all, so are trapped in transfer
              * method rather than calling awaitFulfill.
              */
-            long lastTime = timed ? System.nanoTime() : 0;
+            final long deadline = timed ? System.nanoTime() + nanos : 0L;
             Thread w = Thread.currentThread();
             int spins = (shouldSpin(s) ?
                          (timed ? maxTimedSpins : maxUntimedSpins) : 0);
@@ -411,10 +413,8 @@
                 if (m != null)
                     return m;
                 if (timed) {
-                    long now = System.nanoTime();
-                    nanos -= now - lastTime;
-                    lastTime = now;
-                    if (nanos <= 0) {
+                    nanos = deadline - System.nanoTime();
+                    if (nanos <= 0L) {
                         s.tryCancel();
                         continue;
                     }
@@ -492,7 +492,7 @@
     }
 
     /** Dual Queue */
-    static final class TransferQueue extends Transferer {
+    static final class TransferQueue<E> extends Transferer<E> {
         /*
          * This extends Scherer-Scott dual queue algorithm, differing,
          * among other ways, by using modes within nodes rather than
@@ -570,7 +570,7 @@
         /**
          * Reference to a cancelled node that might not yet have been
          * unlinked from queue because it was the last inserted node
-         * when it cancelled.
+         * when it was cancelled.
          */
         transient volatile QNode cleanMe;
 
@@ -609,7 +609,8 @@
         /**
          * Puts or takes an item.
          */
-        Object transfer(Object e, boolean timed, long nanos) {
+        @SuppressWarnings("unchecked")
+        E transfer(E e, boolean timed, long nanos) {
             /* Basic algorithm is to loop trying to take either of
              * two actions:
              *
@@ -672,7 +673,7 @@
                             s.item = s;
                         s.waiter = null;
                     }
-                    return (x != null) ? x : e;
+                    return (x != null) ? (E)x : e;
 
                 } else {                            // complementary-mode
                     QNode m = h.next;               // node to fulfill
@@ -689,7 +690,7 @@
 
                     advanceHead(h, m);              // successfully fulfilled
                     LockSupport.unpark(m.waiter);
-                    return (x != null) ? x : e;
+                    return (x != null) ? (E)x : e;
                 }
             }
         }
@@ -703,9 +704,9 @@
          * @param nanos timeout value
          * @return matched item, or s if cancelled
          */
-        Object awaitFulfill(QNode s, Object e, boolean timed, long nanos) {
+        Object awaitFulfill(QNode s, E e, boolean timed, long nanos) {
             /* Same idea as TransferStack.awaitFulfill */
-            long lastTime = timed ? System.nanoTime() : 0;
+            final long deadline = timed ? System.nanoTime() + nanos : 0L;
             Thread w = Thread.currentThread();
             int spins = ((head.next == s) ?
                          (timed ? maxTimedSpins : maxUntimedSpins) : 0);
@@ -716,10 +717,8 @@
                 if (x != e)
                     return x;
                 if (timed) {
-                    long now = System.nanoTime();
-                    nanos -= now - lastTime;
-                    lastTime = now;
-                    if (nanos <= 0) {
+                    nanos = deadline - System.nanoTime();
+                    if (nanos <= 0L) {
                         s.tryCancel(e);
                         continue;
                     }
@@ -816,23 +815,23 @@
      * isn't a noticeable performance penalty for using volatile
      * instead of final here.
      */
-    private transient volatile Transferer transferer;
+    private transient volatile Transferer<E> transferer;
 
     /**
-     * Creates a <tt>SynchronousQueue</tt> with nonfair access policy.
+     * Creates a {@code SynchronousQueue} with nonfair access policy.
      */
     public SynchronousQueue() {
         this(false);
     }
 
     /**
-     * Creates a <tt>SynchronousQueue</tt> with the specified fairness policy.
+     * Creates a {@code SynchronousQueue} with the specified fairness policy.
      *
      * @param fair if true, waiting threads contend in FIFO order for
      *        access; otherwise the order is unspecified.
      */
     public SynchronousQueue(boolean fair) {
-        transferer = fair ? new TransferQueue() : new TransferStack();
+        transferer = fair ? new TransferQueue<E>() : new TransferStack<E>();
     }
 
     /**
@@ -842,9 +841,9 @@
      * @throws InterruptedException {@inheritDoc}
      * @throws NullPointerException {@inheritDoc}
      */
-    public void put(E o) throws InterruptedException {
-        if (o == null) throw new NullPointerException();
-        if (transferer.transfer(o, false, 0) == null) {
+    public void put(E e) throws InterruptedException {
+        if (e == null) throw new NullPointerException();
+        if (transferer.transfer(e, false, 0) == null) {
             Thread.interrupted();
             throw new InterruptedException();
         }
@@ -854,15 +853,15 @@
      * Inserts the specified element into this queue, waiting if necessary
      * up to the specified wait time for another thread to receive it.
      *
-     * @return <tt>true</tt> if successful, or <tt>false</tt> if the
-     *         specified waiting time elapses before a consumer appears.
+     * @return {@code true} if successful, or {@code false} if the
+     *         specified waiting time elapses before a consumer appears
      * @throws InterruptedException {@inheritDoc}
      * @throws NullPointerException {@inheritDoc}
      */
-    public boolean offer(E o, long timeout, TimeUnit unit)
+    public boolean offer(E e, long timeout, TimeUnit unit)
         throws InterruptedException {
-        if (o == null) throw new NullPointerException();
-        if (transferer.transfer(o, true, unit.toNanos(timeout)) != null)
+        if (e == null) throw new NullPointerException();
+        if (transferer.transfer(e, true, unit.toNanos(timeout)) != null)
             return true;
         if (!Thread.interrupted())
             return false;
@@ -874,8 +873,8 @@
      * waiting to receive it.
      *
      * @param e the element to add
-     * @return <tt>true</tt> if the element was added to this queue, else
-     *         <tt>false</tt>
+     * @return {@code true} if the element was added to this queue, else
+     *         {@code false}
      * @throws NullPointerException if the specified element is null
      */
     public boolean offer(E e) {
@@ -891,9 +890,9 @@
      * @throws InterruptedException {@inheritDoc}
      */
     public E take() throws InterruptedException {
-        Object e = transferer.transfer(null, false, 0);
+        E e = transferer.transfer(null, false, 0);
         if (e != null)
-            return (E)e;
+            return e;
         Thread.interrupted();
         throw new InterruptedException();
     }
@@ -903,14 +902,14 @@
      * if necessary up to the specified wait time, for another thread
      * to insert it.
      *
-     * @return the head of this queue, or <tt>null</tt> if the
-     *         specified waiting time elapses before an element is present.
+     * @return the head of this queue, or {@code null} if the
+     *         specified waiting time elapses before an element is present
      * @throws InterruptedException {@inheritDoc}
      */
     public E poll(long timeout, TimeUnit unit) throws InterruptedException {
-        Object e = transferer.transfer(null, true, unit.toNanos(timeout));
+        E e = transferer.transfer(null, true, unit.toNanos(timeout));
         if (e != null || !Thread.interrupted())
-            return (E)e;
+            return e;
         throw new InterruptedException();
     }
 
@@ -918,18 +917,18 @@
      * Retrieves and removes the head of this queue, if another thread
      * is currently making an element available.
      *
-     * @return the head of this queue, or <tt>null</tt> if no
-     *         element is available.
+     * @return the head of this queue, or {@code null} if no
+     *         element is available
      */
     public E poll() {
-        return (E)transferer.transfer(null, true, 0);
+        return transferer.transfer(null, true, 0);
     }
 
     /**
-     * Always returns <tt>true</tt>.
-     * A <tt>SynchronousQueue</tt> has no internal capacity.
+     * Always returns {@code true}.
+     * A {@code SynchronousQueue} has no internal capacity.
      *
-     * @return <tt>true</tt>
+     * @return {@code true}
      */
     public boolean isEmpty() {
         return true;
@@ -937,9 +936,9 @@
 
     /**
      * Always returns zero.
-     * A <tt>SynchronousQueue</tt> has no internal capacity.
+     * A {@code SynchronousQueue} has no internal capacity.
      *
-     * @return zero.
+     * @return zero
      */
     public int size() {
         return 0;
@@ -947,9 +946,9 @@
 
     /**
      * Always returns zero.
-     * A <tt>SynchronousQueue</tt> has no internal capacity.
+     * A {@code SynchronousQueue} has no internal capacity.
      *
-     * @return zero.
+     * @return zero
      */
     public int remainingCapacity() {
         return 0;
@@ -957,80 +956,80 @@
 
     /**
      * Does nothing.
-     * A <tt>SynchronousQueue</tt> has no internal capacity.
+     * A {@code SynchronousQueue} has no internal capacity.
      */
     public void clear() {
     }
 
     /**
-     * Always returns <tt>false</tt>.
-     * A <tt>SynchronousQueue</tt> has no internal capacity.
+     * Always returns {@code false}.
+     * A {@code SynchronousQueue} has no internal capacity.
      *
      * @param o the element
-     * @return <tt>false</tt>
+     * @return {@code false}
      */
     public boolean contains(Object o) {
         return false;
     }
 
     /**
-     * Always returns <tt>false</tt>.
-     * A <tt>SynchronousQueue</tt> has no internal capacity.
+     * Always returns {@code false}.
+     * A {@code SynchronousQueue} has no internal capacity.
      *
      * @param o the element to remove
-     * @return <tt>false</tt>
+     * @return {@code false}
      */
     public boolean remove(Object o) {
         return false;
     }
 
     /**
-     * Returns <tt>false</tt> unless the given collection is empty.
-     * A <tt>SynchronousQueue</tt> has no internal capacity.
+     * Returns {@code false} unless the given collection is empty.
+     * A {@code SynchronousQueue} has no internal capacity.
      *
      * @param c the collection
-     * @return <tt>false</tt> unless given collection is empty
+     * @return {@code false} unless given collection is empty
      */
     public boolean containsAll(Collection<?> c) {
         return c.isEmpty();
     }
 
     /**
-     * Always returns <tt>false</tt>.
-     * A <tt>SynchronousQueue</tt> has no internal capacity.
+     * Always returns {@code false}.
+     * A {@code SynchronousQueue} has no internal capacity.
      *
      * @param c the collection
-     * @return <tt>false</tt>
+     * @return {@code false}
      */
     public boolean removeAll(Collection<?> c) {
         return false;
     }
 
     /**
-     * Always returns <tt>false</tt>.
-     * A <tt>SynchronousQueue</tt> has no internal capacity.
+     * Always returns {@code false}.
+     * A {@code SynchronousQueue} has no internal capacity.
      *
      * @param c the collection
-     * @return <tt>false</tt>
+     * @return {@code false}
      */
     public boolean retainAll(Collection<?> c) {
         return false;
     }
 
     /**
-     * Always returns <tt>null</tt>.
-     * A <tt>SynchronousQueue</tt> does not return elements
+     * Always returns {@code null}.
+     * A {@code SynchronousQueue} does not return elements
      * unless actively waited on.
      *
-     * @return <tt>null</tt>
+     * @return {@code null}
      */
     public E peek() {
         return null;
     }
 
     /**
-     * Returns an empty iterator in which <tt>hasNext</tt> always returns
-     * <tt>false</tt>.
+     * Returns an empty iterator in which {@code hasNext} always returns
+     * {@code false}.
      *
      * @return an empty iterator
      */
@@ -1058,7 +1057,7 @@
     }
 
     /**
-     * Sets the zeroeth element of the specified array to <tt>null</tt>
+     * Sets the zeroeth element of the specified array to {@code null}
      * (if the array has non-zero length) and returns it.
      *
      * @param a the array
@@ -1117,6 +1116,7 @@
      * object is ever serialized or deserialized.
      */
 
+    @SuppressWarnings("serial")
     static class WaitQueue implements java.io.Serializable { }
     static class LifoWaitQueue extends WaitQueue {
         private static final long serialVersionUID = -3633113410248163686L;
@@ -1129,9 +1129,7 @@
     private WaitQueue waitingConsumers;
 
     /**
-     * Saves the state to a stream (that is, serializes it).
-     *
-     * @param s the stream
+     * Saves this queue to a stream (that is, serializes it).
      */
     private void writeObject(java.io.ObjectOutputStream s)
         throws java.io.IOException {
@@ -1149,13 +1147,16 @@
         s.defaultWriteObject();
     }
 
+    /**
+     * Reconstitutes this queue from a stream (that is, deserializes it).
+     */
     private void readObject(final java.io.ObjectInputStream s)
         throws java.io.IOException, ClassNotFoundException {
         s.defaultReadObject();
         if (waitingProducers instanceof FifoWaitQueue)
-            transferer = new TransferQueue();
+            transferer = new TransferQueue<E>();
         else
-            transferer = new TransferStack();
+            transferer = new TransferStack<E>();
     }
 
     // Unsafe mechanics
diff --git a/luni/src/main/java/java/util/concurrent/ThreadPoolExecutor.java b/luni/src/main/java/java/util/concurrent/ThreadPoolExecutor.java
index 331e225..9586293 100644
--- a/luni/src/main/java/java/util/concurrent/ThreadPoolExecutor.java
+++ b/luni/src/main/java/java/util/concurrent/ThreadPoolExecutor.java
@@ -5,8 +5,10 @@
  */
 
 package java.util.concurrent;
-import java.util.concurrent.locks.*;
-import java.util.concurrent.atomic.*;
+import java.util.concurrent.locks.AbstractQueuedSynchronizer;
+import java.util.concurrent.locks.Condition;
+import java.util.concurrent.locks.ReentrantLock;
+import java.util.concurrent.atomic.AtomicInteger;
 import java.util.*;
 
 // BEGIN android-note
@@ -48,22 +50,23 @@
  * corePoolSize (see {@link #getCorePoolSize}) and
  * maximumPoolSize (see {@link #getMaximumPoolSize}).
  *
- * When a new task is submitted in method {@link #execute}, and fewer
- * than corePoolSize threads are running, a new thread is created to
- * handle the request, even if other worker threads are idle.  If
- * there are more than corePoolSize but less than maximumPoolSize
- * threads running, a new thread will be created only if the queue is
- * full.  By setting corePoolSize and maximumPoolSize the same, you
- * create a fixed-size thread pool. By setting maximumPoolSize to an
- * essentially unbounded value such as {@code Integer.MAX_VALUE}, you
- * allow the pool to accommodate an arbitrary number of concurrent
- * tasks. Most typically, core and maximum pool sizes are set only
- * upon construction, but they may also be changed dynamically using
- * {@link #setCorePoolSize} and {@link #setMaximumPoolSize}. </dd>
+ * When a new task is submitted in method {@link #execute(Runnable)},
+ * and fewer than corePoolSize threads are running, a new thread is
+ * created to handle the request, even if other worker threads are
+ * idle.  If there are more than corePoolSize but less than
+ * maximumPoolSize threads running, a new thread will be created only
+ * if the queue is full.  By setting corePoolSize and maximumPoolSize
+ * the same, you create a fixed-size thread pool. By setting
+ * maximumPoolSize to an essentially unbounded value such as {@code
+ * Integer.MAX_VALUE}, you allow the pool to accommodate an arbitrary
+ * number of concurrent tasks. Most typically, core and maximum pool
+ * sizes are set only upon construction, but they may also be changed
+ * dynamically using {@link #setCorePoolSize} and {@link
+ * #setMaximumPoolSize}. </dd>
  *
  * <dt>On-demand construction</dt>
  *
- * <dd> By default, even core threads are initially created and
+ * <dd>By default, even core threads are initially created and
  * started only when new tasks arrive, but this can be overridden
  * dynamically using method {@link #prestartCoreThread} or {@link
  * #prestartAllCoreThreads}.  You probably want to prestart threads if
@@ -85,17 +88,17 @@
  *
  * <dd>If the pool currently has more than corePoolSize threads,
  * excess threads will be terminated if they have been idle for more
- * than the keepAliveTime (see {@link #getKeepAliveTime}). This
- * provides a means of reducing resource consumption when the pool is
- * not being actively used. If the pool becomes more active later, new
- * threads will be constructed. This parameter can also be changed
- * dynamically using method {@link #setKeepAliveTime}. Using a value
- * of {@code Long.MAX_VALUE} {@link TimeUnit#NANOSECONDS} effectively
- * disables idle threads from ever terminating prior to shut down. By
- * default, the keep-alive policy applies only when there are more
- * than corePoolSizeThreads. But method {@link
- * #allowCoreThreadTimeOut(boolean)} can be used to apply this
- * time-out policy to core threads as well, so long as the
+ * than the keepAliveTime (see {@link #getKeepAliveTime(TimeUnit)}).
+ * This provides a means of reducing resource consumption when the
+ * pool is not being actively used. If the pool becomes more active
+ * later, new threads will be constructed. This parameter can also be
+ * changed dynamically using method {@link #setKeepAliveTime(long,
+ * TimeUnit)}.  Using a value of {@code Long.MAX_VALUE} {@link
+ * TimeUnit#NANOSECONDS} effectively disables idle threads from ever
+ * terminating prior to shut down. By default, the keep-alive policy
+ * applies only when there are more than corePoolSize threads. But
+ * method {@link #allowCoreThreadTimeOut(boolean)} can be used to
+ * apply this time-out policy to core threads as well, so long as the
  * keepAliveTime value is non-zero. </dd>
  *
  * <dt>Queuing</dt>
@@ -165,14 +168,14 @@
  *
  * <dt>Rejected tasks</dt>
  *
- * <dd> New tasks submitted in method {@link #execute} will be
- * <em>rejected</em> when the Executor has been shut down, and also
- * when the Executor uses finite bounds for both maximum threads and
- * work queue capacity, and is saturated.  In either case, the {@code
- * execute} method invokes the {@link
- * RejectedExecutionHandler#rejectedExecution} method of its {@link
- * RejectedExecutionHandler}.  Four predefined handler policies are
- * provided:
+ * <dd>New tasks submitted in method {@link #execute(Runnable)} will be
+ * <em>rejected</em> when the Executor has been shut down, and also when
+ * the Executor uses finite bounds for both maximum threads and work queue
+ * capacity, and is saturated.  In either case, the {@code execute} method
+ * invokes the {@link
+ * RejectedExecutionHandler#rejectedExecution(Runnable, ThreadPoolExecutor)}
+ * method of its {@link RejectedExecutionHandler}.  Four predefined handler
+ * policies are provided:
  *
  * <ol>
  *
@@ -202,30 +205,31 @@
  *
  * <dt>Hook methods</dt>
  *
- * <dd>This class provides {@code protected} overridable {@link
- * #beforeExecute} and {@link #afterExecute} methods that are called
+ * <dd>This class provides {@code protected} overridable
+ * {@link #beforeExecute(Thread, Runnable)} and
+ * {@link #afterExecute(Runnable, Throwable)} methods that are called
  * before and after execution of each task.  These can be used to
  * manipulate the execution environment; for example, reinitializing
- * ThreadLocals, gathering statistics, or adding log
- * entries. Additionally, method {@link #terminated} can be overridden
- * to perform any special processing that needs to be done once the
- * Executor has fully terminated.
+ * ThreadLocals, gathering statistics, or adding log entries.
+ * Additionally, method {@link #terminated} can be overridden to perform
+ * any special processing that needs to be done once the Executor has
+ * fully terminated.
  *
  * <p>If hook or callback methods throw exceptions, internal worker
  * threads may in turn fail and abruptly terminate.</dd>
  *
  * <dt>Queue maintenance</dt>
  *
- * <dd> Method {@link #getQueue} allows access to the work queue for
- * purposes of monitoring and debugging.  Use of this method for any
- * other purpose is strongly discouraged.  Two supplied methods,
- * {@link #remove} and {@link #purge} are available to assist in
- * storage reclamation when large numbers of queued tasks become
- * cancelled.</dd>
+ * <dd>Method {@link #getQueue()} allows access to the work queue
+ * for purposes of monitoring and debugging.  Use of this method for
+ * any other purpose is strongly discouraged.  Two supplied methods,
+ * {@link #remove(Runnable)} and {@link #purge} are available to
+ * assist in storage reclamation when large numbers of queued tasks
+ * become cancelled.</dd>
  *
  * <dt>Finalization</dt>
  *
- * <dd> A pool that is no longer referenced in a program <em>AND</em>
+ * <dd>A pool that is no longer referenced in a program <em>AND</em>
  * has no remaining threads will be {@code shutdown} automatically. If
  * you would like to ensure that unreferenced pools are reclaimed even
  * if users forget to call {@link #shutdown}, then you must arrange
@@ -235,7 +239,7 @@
  *
  * </dl>
  *
- * <p> <b>Extension example</b>. Most extensions of this class
+ * <p><b>Extension example</b>. Most extensions of this class
  * override one or more of the protected hook methods. For example,
  * here is a subclass that adds a simple pause/resume feature:
  *
@@ -304,7 +308,7 @@
      * bookkeeping before terminating. The user-visible pool size is
      * reported as the current size of the workers set.
      *
-     * The runState provides the main lifecyle control, taking on values:
+     * The runState provides the main lifecycle control, taking on values:
      *
      *   RUNNING:  Accept new tasks and process queued tasks
      *   SHUTDOWN: Don't accept new tasks, but process queued tasks
@@ -374,14 +378,14 @@
     }
 
     /**
-     * Attempt to CAS-increment the workerCount field of ctl.
+     * Attempts to CAS-increment the workerCount field of ctl.
      */
     private boolean compareAndIncrementWorkerCount(int expect) {
         return ctl.compareAndSet(expect, expect + 1);
     }
 
     /**
-     * Attempt to CAS-decrement the workerCount field of ctl.
+     * Attempts to CAS-decrement the workerCount field of ctl.
      */
     private boolean compareAndDecrementWorkerCount(int expect) {
         return ctl.compareAndSet(expect, expect - 1);
@@ -461,10 +465,15 @@
      * policy limiting the number of threads.  Even though it is not
      * treated as an error, failure to create threads may result in
      * new tasks being rejected or existing ones remaining stuck in
-     * the queue. On the other hand, no special precautions exist to
-     * handle OutOfMemoryErrors that might be thrown while trying to
-     * create threads, since there is generally no recourse from
-     * within this class.
+     * the queue.
+     *
+     * We go further and preserve pool invariants even in the face of
+     * errors such as OutOfMemoryError, that might be thrown while
+     * trying to create threads.  Such errors are rather common due to
+     * the need to allocate a native stack in Thread.start, and users
+     * will want to perform clean pool shutdown to clean up.  There
+     * will likely be enough memory available for the cleanup code to
+     * complete without encountering yet another OutOfMemoryError.
      */
     private volatile ThreadFactory threadFactory;
 
@@ -538,9 +547,13 @@
      * task execution.  This protects against interrupts that are
      * intended to wake up a worker thread waiting for a task from
      * instead interrupting a task being run.  We implement a simple
-     * non-reentrant mutual exclusion lock rather than use ReentrantLock
-     * because we do not want worker tasks to be able to reacquire the
-     * lock when they invoke pool control methods like setCorePoolSize.
+     * non-reentrant mutual exclusion lock rather than use
+     * ReentrantLock because we do not want worker tasks to be able to
+     * reacquire the lock when they invoke pool control methods like
+     * setCorePoolSize.  Additionally, to suppress interrupts until
+     * the thread actually starts running tasks, we initialize lock
+     * state to a negative value, and clear it upon start (in
+     * runWorker).
      */
     private final class Worker
         extends AbstractQueuedSynchronizer
@@ -564,6 +577,7 @@
          * @param firstTask the first task (null if none)
          */
         Worker(Runnable firstTask) {
+            setState(-1); // inhibit interrupts until runWorker
             this.firstTask = firstTask;
             this.thread = getThreadFactory().newThread(this);
         }
@@ -579,7 +593,7 @@
         // The value 1 represents the locked state.
 
         protected boolean isHeldExclusively() {
-            return getState() == 1;
+            return getState() != 0;
         }
 
         protected boolean tryAcquire(int unused) {
@@ -600,6 +614,16 @@
         public boolean tryLock()  { return tryAcquire(1); }
         public void unlock()      { release(1); }
         public boolean isLocked() { return isHeldExclusively(); }
+
+        void interruptIfStarted() {
+            Thread t;
+            if (getState() >= 0 && (t = thread) != null && !t.isInterrupted()) {
+                try {
+                    t.interrupt();
+                } catch (SecurityException ignore) {
+                }
+            }
+        }
     }
 
     /*
@@ -698,12 +722,8 @@
         final ReentrantLock mainLock = this.mainLock;
         mainLock.lock();
         try {
-            for (Worker w : workers) {
-                try {
-                    w.thread.interrupt();
-                } catch (SecurityException ignore) {
-                }
-            }
+            for (Worker w : workers)
+                w.interruptIfStarted();
         } finally {
             mainLock.unlock();
         }
@@ -760,19 +780,6 @@
 
     private static final boolean ONLY_ONE = true;
 
-    /**
-     * Ensures that unless the pool is stopping, the current thread
-     * does not have its interrupt set. This requires a double-check
-     * of state in case the interrupt was cleared concurrently with a
-     * shutdownNow -- if so, the interrupt is re-enabled.
-     */
-    private void clearInterruptsForTaskRun() {
-        if (runStateLessThan(ctl.get(), STOP) &&
-            Thread.interrupted() &&
-            runStateAtLeast(ctl.get(), STOP))
-            Thread.currentThread().interrupt();
-    }
-
     /*
      * Misc utilities, most of which are also exported to
      * ScheduledThreadPoolExecutor
@@ -813,7 +820,7 @@
      */
     private List<Runnable> drainQueue() {
         BlockingQueue<Runnable> q = workQueue;
-        List<Runnable> taskList = new ArrayList<Runnable>();
+        ArrayList<Runnable> taskList = new ArrayList<Runnable>();
         q.drainTo(taskList);
         if (!q.isEmpty()) {
             for (Runnable r : q.toArray(new Runnable[0])) {
@@ -832,12 +839,13 @@
      * Checks if a new worker can be added with respect to current
      * pool state and the given bound (either core or maximum). If so,
      * the worker count is adjusted accordingly, and, if possible, a
-     * new worker is created and started running firstTask as its
+     * new worker is created and started, running firstTask as its
      * first task. This method returns false if the pool is stopped or
      * eligible to shut down. It also returns false if the thread
-     * factory fails to create a thread when asked, which requires a
-     * backout of workerCount, and a recheck for termination, in case
-     * the existence of this worker was holding up termination.
+     * factory fails to create a thread when asked.  If the thread
+     * creation fails, either due to the thread factory returning
+     * null, or due to an exception (typically OutOfMemoryError in
+     * Thread.start()), we roll back cleanly.
      *
      * @param firstTask the task the new thread should run first (or
      * null if none). Workers are created with an initial first task
@@ -880,46 +888,64 @@
             }
         }
 
-        Worker w = new Worker(firstTask);
-        Thread t = w.thread;
+        boolean workerStarted = false;
+        boolean workerAdded = false;
+        Worker w = null;
+        try {
+            w = new Worker(firstTask);
+            final Thread t = w.thread;
+            if (t != null) {
+                final ReentrantLock mainLock = this.mainLock;
+                mainLock.lock();
+                try {
+                    // Recheck while holding lock.
+                    // Back out on ThreadFactory failure or if
+                    // shut down before lock acquired.
+                    int rs = runStateOf(ctl.get());
 
+                    if (rs < SHUTDOWN ||
+                        (rs == SHUTDOWN && firstTask == null)) {
+                        if (t.isAlive()) // precheck that t is startable
+                            throw new IllegalThreadStateException();
+                        workers.add(w);
+                        int s = workers.size();
+                        if (s > largestPoolSize)
+                            largestPoolSize = s;
+                        workerAdded = true;
+                    }
+                } finally {
+                    mainLock.unlock();
+                }
+                if (workerAdded) {
+                    t.start();
+                    workerStarted = true;
+                }
+            }
+        } finally {
+            if (! workerStarted)
+                addWorkerFailed(w);
+        }
+        return workerStarted;
+    }
+
+    /**
+     * Rolls back the worker thread creation.
+     * - removes worker from workers, if present
+     * - decrements worker count
+     * - rechecks for termination, in case the existence of this
+     *   worker was holding up termination
+     */
+    private void addWorkerFailed(Worker w) {
         final ReentrantLock mainLock = this.mainLock;
         mainLock.lock();
         try {
-            // Recheck while holding lock.
-            // Back out on ThreadFactory failure or if
-            // shut down before lock acquired.
-            int c = ctl.get();
-            int rs = runStateOf(c);
-
-            if (t == null ||
-                (rs >= SHUTDOWN &&
-                 ! (rs == SHUTDOWN &&
-                    firstTask == null))) {
-                decrementWorkerCount();
-                tryTerminate();
-                return false;
-            }
-
-            workers.add(w);
-
-            int s = workers.size();
-            if (s > largestPoolSize)
-                largestPoolSize = s;
+            if (w != null)
+                workers.remove(w);
+            decrementWorkerCount();
+            tryTerminate();
         } finally {
             mainLock.unlock();
         }
-
-        t.start();
-        // It is possible (but unlikely) for a thread to have been
-        // added to workers, but not yet started, during transition to
-        // STOP, which could result in a rare missed interrupt,
-        // because Thread.interrupt is not guaranteed to have any effect
-        // on a non-yet-started Thread (see Thread#interrupt).
-        if (runStateOf(ctl.get()) == STOP && ! t.isInterrupted())
-            t.interrupt();
-
-        return true;
     }
 
     /**
@@ -974,7 +1000,8 @@
      * 4. This worker timed out waiting for a task, and timed-out
      *    workers are subject to termination (that is,
      *    {@code allowCoreThreadTimeOut || workerCount > corePoolSize})
-     *    both before and after the timed wait.
+     *    both before and after the timed wait, and if the queue is
+     *    non-empty, this worker is not the last thread in the pool.
      *
      * @return task, or null if the worker must exit, in which case
      *         workerCount is decremented
@@ -982,7 +1009,6 @@
     private Runnable getTask() {
         boolean timedOut = false; // Did the last poll() time out?
 
-        retry:
         for (;;) {
             int c = ctl.get();
             int rs = runStateOf(c);
@@ -993,20 +1019,16 @@
                 return null;
             }
 
-            boolean timed;      // Are workers subject to culling?
+            int wc = workerCountOf(c);
 
-            for (;;) {
-                int wc = workerCountOf(c);
-                timed = allowCoreThreadTimeOut || wc > corePoolSize;
+            // Are workers subject to culling?
+            boolean timed = allowCoreThreadTimeOut || wc > corePoolSize;
 
-                if (wc <= maximumPoolSize && ! (timedOut && timed))
-                    break;
+            if ((wc > maximumPoolSize || (timed && timedOut))
+                && (wc > 1 || workQueue.isEmpty())) {
                 if (compareAndDecrementWorkerCount(c))
                     return null;
-                c = ctl.get();  // Re-read ctl
-                if (runStateOf(c) != rs)
-                    continue retry;
-                // else CAS failed due to workerCount change; retry inner loop
+                continue;
             }
 
             try {
@@ -1035,9 +1057,9 @@
      * usually leads processWorkerExit to replace this thread.
      *
      * 2. Before running any task, the lock is acquired to prevent
-     * other pool interrupts while the task is executing, and
-     * clearInterruptsForTaskRun called to ensure that unless pool is
-     * stopping, this thread does not have its interrupt set.
+     * other pool interrupts while the task is executing, and then we
+     * ensure that unless pool is stopping, this thread does not have
+     * its interrupt set.
      *
      * 3. Each task run is preceded by a call to beforeExecute, which
      * might throw an exception, in which case we cause thread to die
@@ -1045,12 +1067,12 @@
      * the task.
      *
      * 4. Assuming beforeExecute completes normally, we run the task,
-     * gathering any of its thrown exceptions to send to
-     * afterExecute. We separately handle RuntimeException, Error
-     * (both of which the specs guarantee that we trap) and arbitrary
-     * Throwables.  Because we cannot rethrow Throwables within
-     * Runnable.run, we wrap them within Errors on the way out (to the
-     * thread's UncaughtExceptionHandler).  Any thrown exception also
+     * gathering any of its thrown exceptions to send to afterExecute.
+     * We separately handle RuntimeException, Error (both of which the
+     * specs guarantee that we trap) and arbitrary Throwables.
+     * Because we cannot rethrow Throwables within Runnable.run, we
+     * wrap them within Errors on the way out (to the thread's
+     * UncaughtExceptionHandler).  Any thrown exception also
      * conservatively causes thread to die.
      *
      * 5. After task.run completes, we call afterExecute, which may
@@ -1066,15 +1088,25 @@
      * @param w the worker
      */
     final void runWorker(Worker w) {
+        Thread wt = Thread.currentThread();
         Runnable task = w.firstTask;
         w.firstTask = null;
+        w.unlock(); // allow interrupts
         boolean completedAbruptly = true;
         try {
             while (task != null || (task = getTask()) != null) {
                 w.lock();
-                clearInterruptsForTaskRun();
+                // If pool is stopping, ensure thread is interrupted;
+                // if not, ensure thread is not interrupted.  This
+                // requires a recheck in second case to deal with
+                // shutdownNow race while clearing interrupt
+                if ((runStateAtLeast(ctl.get(), STOP) ||
+                     (Thread.interrupted() &&
+                      runStateAtLeast(ctl.get(), STOP))) &&
+                    !wt.isInterrupted())
+                    wt.interrupt();
                 try {
-                    beforeExecute(w.thread, task);
+                    beforeExecute(wt, task);
                     Throwable thrown = null;
                     try {
                         task.run();
@@ -1374,7 +1406,7 @@
      * ignored or suppressed interruption, causing this executor not
      * to properly terminate.
      *
-     * @return true if terminating but not yet terminated
+     * @return {@code true} if terminating but not yet terminated
      */
     public boolean isTerminating() {
         int c = ctl.get();
@@ -1428,7 +1460,7 @@
      * Returns the thread factory used to create new threads.
      *
      * @return the current thread factory
-     * @see #setThreadFactory
+     * @see #setThreadFactory(ThreadFactory)
      */
     public ThreadFactory getThreadFactory() {
         return threadFactory;
@@ -1451,7 +1483,7 @@
      * Returns the current handler for unexecutable tasks.
      *
      * @return the current handler
-     * @see #setRejectedExecutionHandler
+     * @see #setRejectedExecutionHandler(RejectedExecutionHandler)
      */
     public RejectedExecutionHandler getRejectedExecutionHandler() {
         return handler;
@@ -1623,7 +1655,7 @@
      * @param unit the time unit of the {@code time} argument
      * @throws IllegalArgumentException if {@code time} less than zero or
      *         if {@code time} is zero and {@code allowsCoreThreadTimeOut}
-     * @see #getKeepAliveTime
+     * @see #getKeepAliveTime(TimeUnit)
      */
     public void setKeepAliveTime(long time, TimeUnit unit) {
         if (time < 0)
@@ -1644,7 +1676,7 @@
      *
      * @param unit the desired time unit of the result
      * @return the time limit
-     * @see #setKeepAliveTime
+     * @see #setKeepAliveTime(long, TimeUnit)
      */
     public long getKeepAliveTime(TimeUnit unit) {
         return unit.convert(keepAliveTime, TimeUnit.NANOSECONDS);
@@ -1669,7 +1701,7 @@
      * present, thus causing it not to be run if it has not already
      * started.
      *
-     * <p> This method may be useful as one part of a cancellation
+     * <p>This method may be useful as one part of a cancellation
      * scheme.  It may fail to remove tasks that have been converted
      * into other forms before being placed on the internal queue. For
      * example, a task entered using {@code submit} might be
@@ -1678,7 +1710,7 @@
      * remove those Futures that have been cancelled.
      *
      * @param task the task to remove
-     * @return true if the task was removed
+     * @return {@code true} if the task was removed
      */
     public boolean remove(Runnable task) {
         boolean removed = workQueue.remove(task);
@@ -1973,7 +2005,7 @@
          *
          * @param r the runnable task requested to be executed
          * @param e the executor attempting to execute this task
-         * @throws RejectedExecutionException always.
+         * @throws RejectedExecutionException always
          */
         public void rejectedExecution(Runnable r, ThreadPoolExecutor e) {
             throw new RejectedExecutionException("Task " + r.toString() +
diff --git a/luni/src/main/java/java/util/concurrent/TimeUnit.java b/luni/src/main/java/java/util/concurrent/TimeUnit.java
index 50f6ce0..eb2c495 100644
--- a/luni/src/main/java/java/util/concurrent/TimeUnit.java
+++ b/luni/src/main/java/java/util/concurrent/TimeUnit.java
@@ -7,10 +7,10 @@
 package java.util.concurrent;
 
 /**
- * A <tt>TimeUnit</tt> represents time durations at a given unit of
+ * A {@code TimeUnit} represents time durations at a given unit of
  * granularity and provides utility methods to convert across units,
  * and to perform timing and delay operations in these units.  A
- * <tt>TimeUnit</tt> does not maintain time information, but only
+ * {@code TimeUnit} does not maintain time information, but only
  * helps organize and use time representations that may be maintained
  * separately across various contexts.  A nanosecond is defined as one
  * thousandth of a microsecond, a microsecond as one thousandth of a
@@ -18,7 +18,7 @@
  * as sixty seconds, an hour as sixty minutes, and a day as twenty four
  * hours.
  *
- * <p>A <tt>TimeUnit</tt> is mainly used to inform time-based methods
+ * <p>A {@code TimeUnit} is mainly used to inform time-based methods
  * how a given timing parameter should be interpreted. For example,
  * the following code will timeout in 50 milliseconds if the {@link
  * java.util.concurrent.locks.Lock lock} is not available:
@@ -34,7 +34,7 @@
  *
  * Note however, that there is no guarantee that a particular timeout
  * implementation will be able to notice the passage of time at the
- * same granularity as the given <tt>TimeUnit</tt>.
+ * same granularity as the given {@code TimeUnit}.
  *
  * @since 1.5
  * @author Doug Lea
@@ -148,31 +148,31 @@
      * Convert the given time duration in the given unit to this
      * unit.  Conversions from finer to coarser granularities
      * truncate, so lose precision. For example converting
-     * <tt>999</tt> milliseconds to seconds results in
-     * <tt>0</tt>. Conversions from coarser to finer granularities
+     * {@code 999} milliseconds to seconds results in
+     * {@code 0}. Conversions from coarser to finer granularities
      * with arguments that would numerically overflow saturate to
-     * <tt>Long.MIN_VALUE</tt> if negative or <tt>Long.MAX_VALUE</tt>
+     * {@code Long.MIN_VALUE} if negative or {@code Long.MAX_VALUE}
      * if positive.
      *
      * <p>For example, to convert 10 minutes to milliseconds, use:
-     * <tt>TimeUnit.MILLISECONDS.convert(10L, TimeUnit.MINUTES)</tt>
+     * {@code TimeUnit.MILLISECONDS.convert(10L, TimeUnit.MINUTES)}
      *
-     * @param sourceDuration the time duration in the given <tt>sourceUnit</tt>
-     * @param sourceUnit the unit of the <tt>sourceDuration</tt> argument
+     * @param sourceDuration the time duration in the given {@code sourceUnit}
+     * @param sourceUnit the unit of the {@code sourceDuration} argument
      * @return the converted duration in this unit,
-     * or <tt>Long.MIN_VALUE</tt> if conversion would negatively
-     * overflow, or <tt>Long.MAX_VALUE</tt> if it would positively overflow.
+     * or {@code Long.MIN_VALUE} if conversion would negatively
+     * overflow, or {@code Long.MAX_VALUE} if it would positively overflow.
      */
     public long convert(long sourceDuration, TimeUnit sourceUnit) {
         throw new AbstractMethodError();
     }
 
     /**
-     * Equivalent to <tt>NANOSECONDS.convert(duration, this)</tt>.
+     * Equivalent to {@code NANOSECONDS.convert(duration, this)}.
      * @param duration the duration
      * @return the converted duration,
-     * or <tt>Long.MIN_VALUE</tt> if conversion would negatively
-     * overflow, or <tt>Long.MAX_VALUE</tt> if it would positively overflow.
+     * or {@code Long.MIN_VALUE} if conversion would negatively
+     * overflow, or {@code Long.MAX_VALUE} if it would positively overflow.
      * @see #convert
      */
     public long toNanos(long duration) {
@@ -180,11 +180,11 @@
     }
 
     /**
-     * Equivalent to <tt>MICROSECONDS.convert(duration, this)</tt>.
+     * Equivalent to {@code MICROSECONDS.convert(duration, this)}.
      * @param duration the duration
      * @return the converted duration,
-     * or <tt>Long.MIN_VALUE</tt> if conversion would negatively
-     * overflow, or <tt>Long.MAX_VALUE</tt> if it would positively overflow.
+     * or {@code Long.MIN_VALUE} if conversion would negatively
+     * overflow, or {@code Long.MAX_VALUE} if it would positively overflow.
      * @see #convert
      */
     public long toMicros(long duration) {
@@ -192,11 +192,11 @@
     }
 
     /**
-     * Equivalent to <tt>MILLISECONDS.convert(duration, this)</tt>.
+     * Equivalent to {@code MILLISECONDS.convert(duration, this)}.
      * @param duration the duration
      * @return the converted duration,
-     * or <tt>Long.MIN_VALUE</tt> if conversion would negatively
-     * overflow, or <tt>Long.MAX_VALUE</tt> if it would positively overflow.
+     * or {@code Long.MIN_VALUE} if conversion would negatively
+     * overflow, or {@code Long.MAX_VALUE} if it would positively overflow.
      * @see #convert
      */
     public long toMillis(long duration) {
@@ -204,11 +204,11 @@
     }
 
     /**
-     * Equivalent to <tt>SECONDS.convert(duration, this)</tt>.
+     * Equivalent to {@code SECONDS.convert(duration, this)}.
      * @param duration the duration
      * @return the converted duration,
-     * or <tt>Long.MIN_VALUE</tt> if conversion would negatively
-     * overflow, or <tt>Long.MAX_VALUE</tt> if it would positively overflow.
+     * or {@code Long.MIN_VALUE} if conversion would negatively
+     * overflow, or {@code Long.MAX_VALUE} if it would positively overflow.
      * @see #convert
      */
     public long toSeconds(long duration) {
@@ -216,11 +216,11 @@
     }
 
     /**
-     * Equivalent to <tt>MINUTES.convert(duration, this)</tt>.
+     * Equivalent to {@code MINUTES.convert(duration, this)}.
      * @param duration the duration
      * @return the converted duration,
-     * or <tt>Long.MIN_VALUE</tt> if conversion would negatively
-     * overflow, or <tt>Long.MAX_VALUE</tt> if it would positively overflow.
+     * or {@code Long.MIN_VALUE} if conversion would negatively
+     * overflow, or {@code Long.MAX_VALUE} if it would positively overflow.
      * @see #convert
      * @since 1.6
      */
@@ -229,11 +229,11 @@
     }
 
     /**
-     * Equivalent to <tt>HOURS.convert(duration, this)</tt>.
+     * Equivalent to {@code HOURS.convert(duration, this)}.
      * @param duration the duration
      * @return the converted duration,
-     * or <tt>Long.MIN_VALUE</tt> if conversion would negatively
-     * overflow, or <tt>Long.MAX_VALUE</tt> if it would positively overflow.
+     * or {@code Long.MIN_VALUE} if conversion would negatively
+     * overflow, or {@code Long.MAX_VALUE} if it would positively overflow.
      * @see #convert
      * @since 1.6
      */
@@ -242,7 +242,7 @@
     }
 
     /**
-     * Equivalent to <tt>DAYS.convert(duration, this)</tt>.
+     * Equivalent to {@code DAYS.convert(duration, this)}.
      * @param duration the duration
      * @return the converted duration
      * @see #convert
@@ -265,9 +265,9 @@
      * Performs a timed {@link Object#wait(long, int) Object.wait}
      * using this time unit.
      * This is a convenience method that converts timeout arguments
-     * into the form required by the <tt>Object.wait</tt> method.
+     * into the form required by the {@code Object.wait} method.
      *
-     * <p>For example, you could implement a blocking <tt>poll</tt>
+     * <p>For example, you could implement a blocking {@code poll}
      * method (see {@link BlockingQueue#poll BlockingQueue.poll})
      * using:
      *
@@ -298,7 +298,7 @@
      * Performs a timed {@link Thread#join(long, int) Thread.join}
      * using this time unit.
      * This is a convenience method that converts time arguments into the
-     * form required by the <tt>Thread.join</tt> method.
+     * form required by the {@code Thread.join} method.
      *
      * @param thread the thread to wait for
      * @param timeout the maximum time to wait. If less than
@@ -318,7 +318,7 @@
      * Performs a {@link Thread#sleep(long, int) Thread.sleep} using
      * this time unit.
      * This is a convenience method that converts time arguments into the
-     * form required by the <tt>Thread.sleep</tt> method.
+     * form required by the {@code Thread.sleep} method.
      *
      * @param timeout the minimum time to sleep. If less than
      * or equal to zero, do not sleep at all.
diff --git a/luni/src/main/java/java/util/concurrent/TimeoutException.java b/luni/src/main/java/java/util/concurrent/TimeoutException.java
index 83934f0..1d7e634 100644
--- a/luni/src/main/java/java/util/concurrent/TimeoutException.java
+++ b/luni/src/main/java/java/util/concurrent/TimeoutException.java
@@ -11,7 +11,7 @@
  * operations for which a timeout is specified need a means to
  * indicate that the timeout has occurred. For many such operations it
  * is possible to return a value that indicates timeout; when that is
- * not possible or desirable then <tt>TimeoutException</tt> should be
+ * not possible or desirable then {@code TimeoutException} should be
  * declared and thrown.
  *
  * @since 1.5
@@ -21,13 +21,13 @@
     private static final long serialVersionUID = 1900926677490660714L;
 
     /**
-     * Constructs a <tt>TimeoutException</tt> with no specified detail
+     * Constructs a {@code TimeoutException} with no specified detail
      * message.
      */
     public TimeoutException() {}
 
     /**
-     * Constructs a <tt>TimeoutException</tt> with the specified detail
+     * Constructs a {@code TimeoutException} with the specified detail
      * message.
      *
      * @param message the detail message
diff --git a/luni/src/main/java/java/util/concurrent/atomic/AtomicBoolean.java b/luni/src/main/java/java/util/concurrent/atomic/AtomicBoolean.java
index d531f25..13b12aa 100644
--- a/luni/src/main/java/java/util/concurrent/atomic/AtomicBoolean.java
+++ b/luni/src/main/java/java/util/concurrent/atomic/AtomicBoolean.java
@@ -76,13 +76,13 @@
      * Atomically sets the value to the given updated value
      * if the current value {@code ==} the expected value.
      *
-     * <p>May <a href="package-summary.html#Spurious">fail spuriously</a>
-     * and does not provide ordering guarantees, so is only rarely an
-     * appropriate alternative to {@code compareAndSet}.
+     * <p><a href="package-summary.html#weakCompareAndSet">May fail
+     * spuriously and does not provide ordering guarantees</a>, so is
+     * only rarely an appropriate alternative to {@code compareAndSet}.
      *
      * @param expect the expected value
      * @param update the new value
-     * @return true if successful.
+     * @return true if successful
      */
     public boolean weakCompareAndSet(boolean expect, boolean update) {
         int e = expect ? 1 : 0;
@@ -126,7 +126,7 @@
 
     /**
      * Returns the String representation of the current value.
-     * @return the String representation of the current value.
+     * @return the String representation of the current value
      */
     public String toString() {
         return Boolean.toString(get());
diff --git a/luni/src/main/java/java/util/concurrent/atomic/AtomicInteger.java b/luni/src/main/java/java/util/concurrent/atomic/AtomicInteger.java
index e0a0018..d67b20a 100644
--- a/luni/src/main/java/java/util/concurrent/atomic/AtomicInteger.java
+++ b/luni/src/main/java/java/util/concurrent/atomic/AtomicInteger.java
@@ -110,13 +110,13 @@
      * Atomically sets the value to the given updated value
      * if the current value {@code ==} the expected value.
      *
-     * <p>May <a href="package-summary.html#Spurious">fail spuriously</a>
-     * and does not provide ordering guarantees, so is only rarely an
-     * appropriate alternative to {@code compareAndSet}.
+     * <p><a href="package-summary.html#weakCompareAndSet">May fail
+     * spuriously and does not provide ordering guarantees</a>, so is
+     * only rarely an appropriate alternative to {@code compareAndSet}.
      *
      * @param expect the expected value
      * @param update the new value
-     * @return true if successful.
+     * @return true if successful
      */
     public final boolean weakCompareAndSet(int expect, int update) {
         return unsafe.compareAndSwapInt(this, valueOffset, expect, update);
@@ -210,13 +210,12 @@
 
     /**
      * Returns the String representation of the current value.
-     * @return the String representation of the current value.
+     * @return the String representation of the current value
      */
     public String toString() {
         return Integer.toString(get());
     }
 
-
     /**
      * Returns the value of this {@code AtomicInteger} as an {@code int}.
      */
diff --git a/luni/src/main/java/java/util/concurrent/atomic/AtomicIntegerArray.java b/luni/src/main/java/java/util/concurrent/atomic/AtomicIntegerArray.java
index 804a51e..1f6980d 100644
--- a/luni/src/main/java/java/util/concurrent/atomic/AtomicIntegerArray.java
+++ b/luni/src/main/java/java/util/concurrent/atomic/AtomicIntegerArray.java
@@ -146,14 +146,14 @@
      * Atomically sets the element at position {@code i} to the given
      * updated value if the current value {@code ==} the expected value.
      *
-     * <p>May <a href="package-summary.html#Spurious">fail spuriously</a>
-     * and does not provide ordering guarantees, so is only rarely an
-     * appropriate alternative to {@code compareAndSet}.
+     * <p><a href="package-summary.html#weakCompareAndSet">May fail
+     * spuriously and does not provide ordering guarantees</a>, so is
+     * only rarely an appropriate alternative to {@code compareAndSet}.
      *
      * @param i the index
      * @param expect the expected value
      * @param update the new value
-     * @return true if successful.
+     * @return true if successful
      */
     public final boolean weakCompareAndSet(int i, int expect, int update) {
         return compareAndSet(i, expect, update);
diff --git a/luni/src/main/java/java/util/concurrent/atomic/AtomicIntegerFieldUpdater.java b/luni/src/main/java/java/util/concurrent/atomic/AtomicIntegerFieldUpdater.java
index c7ed158..6067152 100644
--- a/luni/src/main/java/java/util/concurrent/atomic/AtomicIntegerFieldUpdater.java
+++ b/luni/src/main/java/java/util/concurrent/atomic/AtomicIntegerFieldUpdater.java
@@ -8,7 +8,8 @@
 
 import dalvik.system.VMStack; // android-added
 import sun.misc.Unsafe;
-import java.lang.reflect.*;
+import java.lang.reflect.Field;
+import java.lang.reflect.Modifier;
 
 /**
  * A reflection-based utility that enables atomic updates to
@@ -28,7 +29,7 @@
  * @author Doug Lea
  * @param <T> The type of the object holding the updatable field
  */
-public abstract class  AtomicIntegerFieldUpdater<T> {
+public abstract class AtomicIntegerFieldUpdater<T> {
     /**
      * Creates and returns an updater for objects with the given field.
      * The Class argument is needed to check that reflective types and
@@ -40,7 +41,9 @@
      * @throws IllegalArgumentException if the field is not a
      * volatile integer type
      * @throws RuntimeException with a nested reflection-based
-     * exception if the class does not hold field or is the wrong type
+     * exception if the class does not hold field or is the wrong type,
+     * or the field is inaccessible to the caller according to Java language
+     * access control
      */
     public static <U> AtomicIntegerFieldUpdater<U> newUpdater(Class<U> tclass, String fieldName) {
         return new AtomicIntegerFieldUpdaterImpl<U>(tclass, fieldName);
@@ -75,9 +78,9 @@
      * other calls to {@code compareAndSet} and {@code set}, but not
      * necessarily with respect to other changes in the field.
      *
-     * <p>May <a href="package-summary.html#Spurious">fail spuriously</a>
-     * and does not provide ordering guarantees, so is only rarely an
-     * appropriate alternative to {@code compareAndSet}.
+     * <p><a href="package-summary.html#weakCompareAndSet">May fail
+     * spuriously and does not provide ordering guarantees</a>, so is
+     * only rarely an appropriate alternative to {@code compareAndSet}.
      *
      * @param obj An object whose field to conditionally set
      * @param expect the expected value
@@ -108,7 +111,6 @@
      */
     public abstract void lazySet(T obj, int newValue);
 
-
     /**
      * Gets the current value held in the field of the given object managed
      * by this updater.
@@ -242,19 +244,28 @@
         private final Class<?> cclass;
 
         AtomicIntegerFieldUpdaterImpl(Class<T> tclass, String fieldName) {
-            Field field = null;
-            Class<?> caller = null;
-            int modifiers = 0;
+            final Field field;
+            final Class<?> caller;
+            final int modifiers;
             try {
-                field = tclass.getDeclaredField(fieldName);
+                field = tclass.getDeclaredField(fieldName); // android-changed
                 caller = VMStack.getStackClass2(); // android-changed
-                modifiers = field.getModifiers();
 
+                modifiers = field.getModifiers();
                 // BEGIN android-removed
                 // sun.reflect.misc.ReflectUtil.ensureMemberAccess(
                 //     caller, tclass, null, modifiers);
-                // sun.reflect.misc.ReflectUtil.checkPackageAccess(tclass);
+                // ClassLoader cl = tclass.getClassLoader();
+                // ClassLoader ccl = caller.getClassLoader();
+                // if ((ccl != null) && (ccl != cl) &&
+                //     ((cl == null) || !isAncestor(cl, ccl))) {
+                //   sun.reflect.misc.ReflectUtil.checkPackageAccess(tclass);
+                // }
                 // END android-removed
+            // BEGIN android-removed
+            // } catch (PrivilegedActionException pae) {
+            //     throw new RuntimeException(pae.getException());
+            // END android-removed
             } catch (Exception ex) {
                 throw new RuntimeException(ex);
             }
@@ -272,6 +283,23 @@
             offset = unsafe.objectFieldOffset(field);
         }
 
+        // BEGIN android-removed
+        // /**
+        //  * Returns true if the second classloader can be found in the first
+        //  * classloader's delegation chain.
+        //  * Equivalent to the inaccessible: first.isAncestor(second).
+        //  */
+        //  private static boolean isAncestor(ClassLoader first, ClassLoader second) {
+        //     ClassLoader acl = first;
+        //     do {
+        //         acl = acl.getParent();
+        //         if (second == acl) {
+        //             return true;
+        //         }
+        //     } while (acl != null);
+        //     return false;
+        // }
+        // END android-removed
         private void fullCheck(T obj) {
             if (!tclass.isInstance(obj))
                 throw new ClassCastException();
diff --git a/luni/src/main/java/java/util/concurrent/atomic/AtomicLong.java b/luni/src/main/java/java/util/concurrent/atomic/AtomicLong.java
index 5e799f7..278c5b5 100644
--- a/luni/src/main/java/java/util/concurrent/atomic/AtomicLong.java
+++ b/luni/src/main/java/java/util/concurrent/atomic/AtomicLong.java
@@ -124,13 +124,13 @@
      * Atomically sets the value to the given updated value
      * if the current value {@code ==} the expected value.
      *
-     * <p>May <a href="package-summary.html#Spurious">fail spuriously</a>
-     * and does not provide ordering guarantees, so is only rarely an
-     * appropriate alternative to {@code compareAndSet}.
+     * <p><a href="package-summary.html#weakCompareAndSet">May fail
+     * spuriously and does not provide ordering guarantees</a>, so is
+     * only rarely an appropriate alternative to {@code compareAndSet}.
      *
      * @param expect the expected value
      * @param update the new value
-     * @return true if successful.
+     * @return true if successful
      */
     public final boolean weakCompareAndSet(long expect, long update) {
         return unsafe.compareAndSwapLong(this, valueOffset, expect, update);
@@ -224,13 +224,12 @@
 
     /**
      * Returns the String representation of the current value.
-     * @return the String representation of the current value.
+     * @return the String representation of the current value
      */
     public String toString() {
         return Long.toString(get());
     }
 
-
     /**
      * Returns the value of this {@code AtomicLong} as an {@code int}
      * after a narrowing primitive conversion.
diff --git a/luni/src/main/java/java/util/concurrent/atomic/AtomicLongArray.java b/luni/src/main/java/java/util/concurrent/atomic/AtomicLongArray.java
index 22edb3f..2e8c2b7 100644
--- a/luni/src/main/java/java/util/concurrent/atomic/AtomicLongArray.java
+++ b/luni/src/main/java/java/util/concurrent/atomic/AtomicLongArray.java
@@ -106,7 +106,6 @@
         unsafe.putOrderedLong(array, checkedByteOffset(i), newValue);
     }
 
-
     /**
      * Atomically sets the element at position {@code i} to the given value
      * and returns the old value.
@@ -146,14 +145,14 @@
      * Atomically sets the element at position {@code i} to the given
      * updated value if the current value {@code ==} the expected value.
      *
-     * <p>May <a href="package-summary.html#Spurious">fail spuriously</a>
-     * and does not provide ordering guarantees, so is only rarely an
-     * appropriate alternative to {@code compareAndSet}.
+     * <p><a href="package-summary.html#weakCompareAndSet">May fail
+     * spuriously and does not provide ordering guarantees</a>, so is
+     * only rarely an appropriate alternative to {@code compareAndSet}.
      *
      * @param i the index
      * @param expect the expected value
      * @param update the new value
-     * @return true if successful.
+     * @return true if successful
      */
     public final boolean weakCompareAndSet(int i, long expect, long update) {
         return compareAndSet(i, expect, update);
diff --git a/luni/src/main/java/java/util/concurrent/atomic/AtomicLongFieldUpdater.java b/luni/src/main/java/java/util/concurrent/atomic/AtomicLongFieldUpdater.java
index 748ae69..0096a6b 100644
--- a/luni/src/main/java/java/util/concurrent/atomic/AtomicLongFieldUpdater.java
+++ b/luni/src/main/java/java/util/concurrent/atomic/AtomicLongFieldUpdater.java
@@ -8,7 +8,8 @@
 
 import dalvik.system.VMStack; // android-added
 import sun.misc.Unsafe;
-import java.lang.reflect.*;
+import java.lang.reflect.Field;
+import java.lang.reflect.Modifier;
 
 /**
  * A reflection-based utility that enables atomic updates to
@@ -28,19 +29,21 @@
  * @author Doug Lea
  * @param <T> The type of the object holding the updatable field
  */
-public abstract class  AtomicLongFieldUpdater<T> {
+public abstract class AtomicLongFieldUpdater<T> {
     /**
      * Creates and returns an updater for objects with the given field.
      * The Class argument is needed to check that reflective types and
      * generic types match.
      *
      * @param tclass the class of the objects holding the field
-     * @param fieldName the name of the field to be updated.
+     * @param fieldName the name of the field to be updated
      * @return the updater
      * @throws IllegalArgumentException if the field is not a
-     * volatile long type.
+     * volatile long type
      * @throws RuntimeException with a nested reflection-based
-     * exception if the class does not hold field or is the wrong type.
+     * exception if the class does not hold field or is the wrong type,
+     * or the field is inaccessible to the caller according to Java language
+     * access control
      */
     public static <U> AtomicLongFieldUpdater<U> newUpdater(Class<U> tclass, String fieldName) {
         if (AtomicLong.VM_SUPPORTS_LONG_CAS)
@@ -65,9 +68,9 @@
      * @param obj An object whose field to conditionally set
      * @param expect the expected value
      * @param update the new value
-     * @return true if successful.
+     * @return true if successful
      * @throws ClassCastException if {@code obj} is not an instance
-     * of the class possessing the field established in the constructor.
+     * of the class possessing the field established in the constructor
      */
     public abstract boolean compareAndSet(T obj, long expect, long update);
 
@@ -78,16 +81,16 @@
      * other calls to {@code compareAndSet} and {@code set}, but not
      * necessarily with respect to other changes in the field.
      *
-     * <p>May <a href="package-summary.html#Spurious">fail spuriously</a>
-     * and does not provide ordering guarantees, so is only rarely an
-     * appropriate alternative to {@code compareAndSet}.
+     * <p><a href="package-summary.html#weakCompareAndSet">May fail
+     * spuriously and does not provide ordering guarantees</a>, so is
+     * only rarely an appropriate alternative to {@code compareAndSet}.
      *
      * @param obj An object whose field to conditionally set
      * @param expect the expected value
      * @param update the new value
-     * @return true if successful.
+     * @return true if successful
      * @throws ClassCastException if {@code obj} is not an instance
-     * of the class possessing the field established in the constructor.
+     * of the class possessing the field established in the constructor
      */
     public abstract boolean weakCompareAndSet(T obj, long expect, long update);
 
@@ -241,18 +244,27 @@
         private final Class<?> cclass;
 
         CASUpdater(Class<T> tclass, String fieldName) {
-            Field field = null;
-            Class<?> caller = null;
-            int modifiers = 0;
+            final Field field;
+            final Class<?> caller;
+            final int modifiers;
             try {
-                field = tclass.getDeclaredField(fieldName);
+                field = tclass.getDeclaredField(fieldName); // android-changed
                 caller = VMStack.getStackClass2(); // android-changed
                 modifiers = field.getModifiers();
                 // BEGIN android-removed
                 // sun.reflect.misc.ReflectUtil.ensureMemberAccess(
                 //     caller, tclass, null, modifiers);
-                // sun.reflect.misc.ReflectUtil.checkPackageAccess(tclass);
+                // ClassLoader cl = tclass.getClassLoader();
+                // ClassLoader ccl = caller.getClassLoader();
+                // if ((ccl != null) && (ccl != cl) &&
+                //     ((cl == null) || !isAncestor(cl, ccl))) {
+                //   sun.reflect.misc.ReflectUtil.checkPackageAccess(tclass);
+                // }
                 // END android-removed
+            // BEGIN android-removed
+            // } catch (PrivilegedActionException pae) {
+            //    throw new RuntimeException(pae.getException());
+            // END android-removed
             } catch (Exception ex) {
                 throw new RuntimeException(ex);
             }
@@ -330,14 +342,23 @@
             Class<?> caller = null;
             int modifiers = 0;
             try {
-                field = tclass.getDeclaredField(fieldName);
+                field = tclass.getDeclaredField(fieldName); // android-changed
                 caller = VMStack.getStackClass2(); // android-changed
                 modifiers = field.getModifiers();
                 // BEGIN android-removed
                 // sun.reflect.misc.ReflectUtil.ensureMemberAccess(
                 //     caller, tclass, null, modifiers);
-                // sun.reflect.misc.ReflectUtil.checkPackageAccess(tclass);
+                // ClassLoader cl = tclass.getClassLoader();
+                // ClassLoader ccl = caller.getClassLoader();
+                // if ((ccl != null) && (ccl != cl) &&
+                //     ((cl == null) || !isAncestor(cl, ccl))) {
+                //   sun.reflect.misc.ReflectUtil.checkPackageAccess(tclass);
+                // }
                 // END android-removed
+            // BEGIN android-removed
+            // } catch (PrivilegedActionException pae) {
+            //     throw new RuntimeException(pae.getException());
+            // END android-removed
             } catch (Exception ex) {
                 throw new RuntimeException(ex);
             }
@@ -410,4 +431,22 @@
             );
         }
     }
+
+    // BEGIN android-removed
+    // /**
+    //  * Returns true if the second classloader can be found in the first
+    //  * classloader's delegation chain.
+    //  * Equivalent to the inaccessible: first.isAncestor(second).
+    //  */
+    // private static boolean isAncestor(ClassLoader first, ClassLoader second) {
+    //     ClassLoader acl = first;
+    //     do {
+    //         acl = acl.getParent();
+    //         if (second == acl) {
+    //             return true;
+    //        }
+    //     } while (acl != null);
+    //     return false;
+    // }
+    // END android-removed
 }
diff --git a/luni/src/main/java/java/util/concurrent/atomic/AtomicMarkableReference.java b/luni/src/main/java/java/util/concurrent/atomic/AtomicMarkableReference.java
index eaf700c..1257be0 100644
--- a/luni/src/main/java/java/util/concurrent/atomic/AtomicMarkableReference.java
+++ b/luni/src/main/java/java/util/concurrent/atomic/AtomicMarkableReference.java
@@ -83,9 +83,9 @@
      * current reference is {@code ==} to the expected reference
      * and the current mark is equal to the expected mark.
      *
-     * <p>May <a href="package-summary.html#Spurious">fail spuriously</a>
-     * and does not provide ordering guarantees, so is only rarely an
-     * appropriate alternative to {@code compareAndSet}.
+     * <p><a href="package-summary.html#weakCompareAndSet">May fail
+     * spuriously and does not provide ordering guarantees</a>, so is
+     * only rarely an appropriate alternative to {@code compareAndSet}.
      *
      * @param expectedReference the expected value of the reference
      * @param newReference the new value for the reference
diff --git a/luni/src/main/java/java/util/concurrent/atomic/AtomicReference.java b/luni/src/main/java/java/util/concurrent/atomic/AtomicReference.java
index b21e9b6..98b402a 100644
--- a/luni/src/main/java/java/util/concurrent/atomic/AtomicReference.java
+++ b/luni/src/main/java/java/util/concurrent/atomic/AtomicReference.java
@@ -15,7 +15,7 @@
  * @author Doug Lea
  * @param <V> The type of object referred to by this reference
  */
-public class AtomicReference<V>  implements java.io.Serializable {
+public class AtomicReference<V> implements java.io.Serializable {
     private static final long serialVersionUID = -1848883965231344442L;
 
     private static final Unsafe unsafe = Unsafe.getUnsafe();
@@ -89,13 +89,13 @@
      * Atomically sets the value to the given updated value
      * if the current value {@code ==} the expected value.
      *
-     * <p>May <a href="package-summary.html#Spurious">fail spuriously</a>
-     * and does not provide ordering guarantees, so is only rarely an
-     * appropriate alternative to {@code compareAndSet}.
+     * <p><a href="package-summary.html#weakCompareAndSet">May fail
+     * spuriously and does not provide ordering guarantees</a>, so is
+     * only rarely an appropriate alternative to {@code compareAndSet}.
      *
      * @param expect the expected value
      * @param update the new value
-     * @return true if successful.
+     * @return true if successful
      */
     public final boolean weakCompareAndSet(V expect, V update) {
         return unsafe.compareAndSwapObject(this, valueOffset, expect, update);
@@ -117,7 +117,7 @@
 
     /**
      * Returns the String representation of the current value.
-     * @return the String representation of the current value.
+     * @return the String representation of the current value
      */
     public String toString() {
         return String.valueOf(get());
diff --git a/luni/src/main/java/java/util/concurrent/atomic/AtomicReferenceArray.java b/luni/src/main/java/java/util/concurrent/atomic/AtomicReferenceArray.java
index c47728d..052b839 100644
--- a/luni/src/main/java/java/util/concurrent/atomic/AtomicReferenceArray.java
+++ b/luni/src/main/java/java/util/concurrent/atomic/AtomicReferenceArray.java
@@ -29,19 +29,18 @@
     private final Object[] array; // must have exact type Object[]
 
     static {
-        int scale;
         try {
             unsafe = Unsafe.getUnsafe();
             arrayFieldOffset = unsafe.objectFieldOffset
                 (AtomicReferenceArray.class.getDeclaredField("array"));
             base = unsafe.arrayBaseOffset(Object[].class);
-            scale = unsafe.arrayIndexScale(Object[].class);
+            int scale = unsafe.arrayIndexScale(Object[].class);
+            if ((scale & (scale - 1)) != 0)
+                throw new Error("data type scale not a power of two");
+            shift = 31 - Integer.numberOfLeadingZeros(scale);
         } catch (Exception e) {
             throw new Error(e);
         }
-        if ((scale & (scale - 1)) != 0)
-            throw new Error("data type scale not a power of two");
-        shift = 31 - Integer.numberOfLeadingZeros(scale);
     }
 
     private long checkedByteOffset(int i) {
@@ -96,6 +95,7 @@
         return getRaw(checkedByteOffset(i));
     }
 
+    @SuppressWarnings("unchecked")
     private E getRaw(long offset) {
         return (E) unsafe.getObjectVolatile(array, offset);
     }
@@ -121,7 +121,6 @@
         unsafe.putOrderedObject(array, checkedByteOffset(i), newValue);
     }
 
-
     /**
      * Atomically sets the element at position {@code i} to the given
      * value and returns the old value.
@@ -161,14 +160,14 @@
      * Atomically sets the element at position {@code i} to the given
      * updated value if the current value {@code ==} the expected value.
      *
-     * <p>May <a href="package-summary.html#Spurious">fail spuriously</a>
-     * and does not provide ordering guarantees, so is only rarely an
-     * appropriate alternative to {@code compareAndSet}.
+     * <p><a href="package-summary.html#weakCompareAndSet">May fail
+     * spuriously and does not provide ordering guarantees</a>, so is
+     * only rarely an appropriate alternative to {@code compareAndSet}.
      *
      * @param i the index
      * @param expect the expected value
      * @param update the new value
-     * @return true if successful.
+     * @return true if successful
      */
     public final boolean weakCompareAndSet(int i, E expect, E update) {
         return compareAndSet(i, expect, update);
@@ -195,7 +194,6 @@
 
     /**
      * Reconstitutes the instance from a stream (that is, deserializes it).
-     * @param s the stream
      */
     private void readObject(java.io.ObjectInputStream s)
         throws java.io.IOException, ClassNotFoundException,
diff --git a/luni/src/main/java/java/util/concurrent/atomic/AtomicReferenceFieldUpdater.java b/luni/src/main/java/java/util/concurrent/atomic/AtomicReferenceFieldUpdater.java
index d23d766..eb2d73e 100644
--- a/luni/src/main/java/java/util/concurrent/atomic/AtomicReferenceFieldUpdater.java
+++ b/luni/src/main/java/java/util/concurrent/atomic/AtomicReferenceFieldUpdater.java
@@ -7,7 +7,8 @@
 package java.util.concurrent.atomic;
 import dalvik.system.VMStack; // android-added
 import sun.misc.Unsafe;
-import java.lang.reflect.*;
+import java.lang.reflect.Field;
+import java.lang.reflect.Modifier;
 
 /**
  * A reflection-based utility that enables atomic updates to
@@ -45,20 +46,22 @@
  * @param <T> The type of the object holding the updatable field
  * @param <V> The type of the field
  */
-public abstract class AtomicReferenceFieldUpdater<T, V> {
+public abstract class AtomicReferenceFieldUpdater<T,V> {
 
     /**
      * Creates and returns an updater for objects with the given field.
      * The Class arguments are needed to check that reflective types and
      * generic types match.
      *
-     * @param tclass the class of the objects holding the field.
+     * @param tclass the class of the objects holding the field
      * @param vclass the class of the field
-     * @param fieldName the name of the field to be updated.
+     * @param fieldName the name of the field to be updated
      * @return the updater
-     * @throws IllegalArgumentException if the field is not a volatile reference type.
+     * @throws IllegalArgumentException if the field is not a volatile reference type
      * @throws RuntimeException with a nested reflection-based
-     * exception if the class does not hold field or is the wrong type.
+     * exception if the class does not hold field or is the wrong type,
+     * or the field is inaccessible to the caller according to Java language
+     * access control
      */
     public static <U, W> AtomicReferenceFieldUpdater<U,W> newUpdater(Class<U> tclass, Class<W> vclass, String fieldName) {
         return new AtomicReferenceFieldUpdaterImpl<U,W>(tclass,
@@ -82,7 +85,7 @@
      * @param obj An object whose field to conditionally set
      * @param expect the expected value
      * @param update the new value
-     * @return true if successful.
+     * @return true if successful
      */
     public abstract boolean compareAndSet(T obj, V expect, V update);
 
@@ -93,14 +96,14 @@
      * other calls to {@code compareAndSet} and {@code set}, but not
      * necessarily with respect to other changes in the field.
      *
-     * <p>May <a href="package-summary.html#Spurious">fail spuriously</a>
-     * and does not provide ordering guarantees, so is only rarely an
-     * appropriate alternative to {@code compareAndSet}.
+     * <p><a href="package-summary.html#weakCompareAndSet">May fail
+     * spuriously and does not provide ordering guarantees</a>, so is
+     * only rarely an appropriate alternative to {@code compareAndSet}.
      *
      * @param obj An object whose field to conditionally set
      * @param expect the expected value
      * @param update the new value
-     * @return true if successful.
+     * @return true if successful
      */
     public abstract boolean weakCompareAndSet(T obj, V expect, V update);
 
@@ -169,23 +172,32 @@
          * screenings fail.
          */
 
-        AtomicReferenceFieldUpdaterImpl(Class<T> tclass,
+        AtomicReferenceFieldUpdaterImpl(final Class<T> tclass,
                                         Class<V> vclass,
-                                        String fieldName) {
-            Field field = null;
-            Class<?> fieldClass = null;
-            Class<?> caller = null;
-            int modifiers = 0;
+                                        final String fieldName) {
+            final Field field;
+            final Class<?> fieldClass;
+            final Class<?> caller;
+            final int modifiers;
             try {
-                field = tclass.getDeclaredField(fieldName);
+                field = tclass.getDeclaredField(fieldName); // android-changed
                 caller = VMStack.getStackClass2(); // android-changed
                 modifiers = field.getModifiers();
-                // BEGIN android-removed
-                // sun.reflect.misc.ReflectUtil.ensureMemberAccess(
-                //     caller, tclass, null, modifiers);
-                // sun.reflect.misc.ReflectUtil.checkPackageAccess(tclass);
-                // END android-removed
+            // BEGIN android-removed
+            //     sun.reflect.misc.ReflectUtil.ensureMemberAccess(
+            //         caller, tclass, null, modifiers);
+            //     ClassLoader cl = tclass.getClassLoader();
+            //     ClassLoader ccl = caller.getClassLoader();
+            //     if ((ccl != null) && (ccl != cl) &&
+            //         ((cl == null) || !isAncestor(cl, ccl))) {
+            //       sun.reflect.misc.ReflectUtil.checkPackageAccess(tclass);
+            //     }
+            // END android-removed
                 fieldClass = field.getType();
+            // BEGIN android-removed
+            // } catch (PrivilegedActionException pae) {
+            //     throw new RuntimeException(pae.getException());
+            // END android-removed
             } catch (Exception ex) {
                 throw new RuntimeException(ex);
             }
@@ -206,6 +218,25 @@
             offset = unsafe.objectFieldOffset(field);
         }
 
+        // BEGIN android-removed
+        // /**
+        //  * Returns true if the second classloader can be found in the first
+        //  * classloader's delegation chain.
+        //  * Equivalent to the inaccessible: first.isAncestor(second).
+        //  */
+        //
+        // private static boolean isAncestor(ClassLoader first, ClassLoader second) {
+        //     ClassLoader acl = first;
+        //     do {
+        //         acl = acl.getParent();
+        //         if (second == acl) {
+        //             return true;
+        //        }
+        //     } while (acl != null);
+        //     return false;
+        // }
+        // END android-removed
+
         void targetCheck(T obj) {
             if (!tclass.isInstance(obj))
                 throw new ClassCastException();
@@ -254,6 +285,7 @@
             unsafe.putOrderedObject(obj, offset, newValue);
         }
 
+        @SuppressWarnings("unchecked")
         public V get(T obj) {
             if (obj == null || obj.getClass() != tclass || cclass != null)
                 targetCheck(obj);
diff --git a/luni/src/main/java/java/util/concurrent/atomic/AtomicStampedReference.java b/luni/src/main/java/java/util/concurrent/atomic/AtomicStampedReference.java
index a0cb492..b93a6f3 100644
--- a/luni/src/main/java/java/util/concurrent/atomic/AtomicStampedReference.java
+++ b/luni/src/main/java/java/util/concurrent/atomic/AtomicStampedReference.java
@@ -83,9 +83,9 @@
      * current reference is {@code ==} to the expected reference
      * and the current stamp is equal to the expected stamp.
      *
-     * <p>May <a href="package-summary.html#Spurious">fail spuriously</a>
-     * and does not provide ordering guarantees, so is only rarely an
-     * appropriate alternative to {@code compareAndSet}.
+     * <p><a href="package-summary.html#weakCompareAndSet">May fail
+     * spuriously and does not provide ordering guarantees</a>, so is
+     * only rarely an appropriate alternative to {@code compareAndSet}.
      *
      * @param expectedReference the expected value of the reference
      * @param newReference the new value for the reference
@@ -126,7 +126,6 @@
              casPair(current, Pair.of(newReference, newStamp)));
     }
 
-
     /**
      * Unconditionally sets the value of both the reference and stamp.
      *
diff --git a/luni/src/main/java/java/util/concurrent/atomic/package-info.java b/luni/src/main/java/java/util/concurrent/atomic/package-info.java
index efbb413..568d2c6 100644
--- a/luni/src/main/java/java/util/concurrent/atomic/package-info.java
+++ b/luni/src/main/java/java/util/concurrent/atomic/package-info.java
@@ -54,19 +54,20 @@
  *
  * write your utility method as follows:
  *  <pre> {@code
- * boolean getAndTransform(AtomicLong var) {
+ * long getAndTransform(AtomicLong var) {
  *   while (true) {
  *     long current = var.get();
  *     long next = transform(current);
  *     if (var.compareAndSet(current, next))
  *         return current;
+ *         // return next; for transformAndGet
  *   }
  * }}</pre>
  *
  * <p>The memory effects for accesses and updates of atomics generally
  * follow the rules for volatiles, as stated in
- * <a href="http://java.sun.com/docs/books/jls/"> The Java Language
- * Specification, Third Edition (17.4 Memory Model)</a>:
+ * <a href="http://docs.oracle.com/javase/specs/jls/se7/html/jls-17.html#jls-17.4">
+ * The Java Language Specification (17.4 Memory Model)</a>:
  *
  * <ul>
  *
@@ -122,13 +123,12 @@
  * semantics for their array elements, which is not supported for
  * ordinary arrays.
  *
- * <a name="Spurious">
- * <p>The atomic classes also support method {@code weakCompareAndSet},
- * which has limited applicability.  On some platforms, the weak version
- * may be more efficient than {@code compareAndSet} in the normal case,
- * but differs in that any given invocation of the
- * {@code weakCompareAndSet} method may return {@code false}
- * <em>spuriously</em> (that is, for no apparent reason)</a>.  A
+ * <p id="weakCompareAndSet">The atomic classes also support method
+ * {@code weakCompareAndSet}, which has limited applicability.  On some
+ * platforms, the weak version may be more efficient than {@code
+ * compareAndSet} in the normal case, but differs in that any given
+ * invocation of the {@code weakCompareAndSet} method may return {@code
+ * false} <em>spuriously</em> (that is, for no apparent reason).  A
  * {@code false} return means only that the operation may be retried if
  * desired, relying on the guarantee that repeated invocation when the
  * variable holds {@code expectedValue} and no other thread is also
@@ -164,7 +164,7 @@
  *
  * <p>Atomic classes are not general purpose replacements for
  * {@code java.lang.Integer} and related classes.  They do <em>not</em>
- * define methods such as {@code hashCode} and
+ * define methods such as {@code equals}, {@code hashCode} and
  * {@code compareTo}.  (Because atomic variables are expected to be
  * mutated, they are poor choices for hash table keys.)  Additionally,
  * classes are provided only for those types that are commonly useful in
diff --git a/luni/src/main/java/java/util/concurrent/locks/AbstractOwnableSynchronizer.java b/luni/src/main/java/java/util/concurrent/locks/AbstractOwnableSynchronizer.java
index 4bec0cf..fa01824 100644
--- a/luni/src/main/java/java/util/concurrent/locks/AbstractOwnableSynchronizer.java
+++ b/luni/src/main/java/java/util/concurrent/locks/AbstractOwnableSynchronizer.java
@@ -10,7 +10,7 @@
  * A synchronizer that may be exclusively owned by a thread.  This
  * class provides a basis for creating locks and related synchronizers
  * that may entail a notion of ownership.  The
- * <tt>AbstractOwnableSynchronizer</tt> class itself does not manage or
+ * {@code AbstractOwnableSynchronizer} class itself does not manage or
  * use this information. However, subclasses and tools may use
  * appropriately maintained values to help control and monitor access
  * and provide diagnostics.
@@ -36,9 +36,9 @@
 
     /**
      * Sets the thread that currently owns exclusive access. A
-     * <tt>null</tt> argument indicates that no thread owns access.
+     * {@code null} argument indicates that no thread owns access.
      * This method does not otherwise impose any synchronization or
-     * <tt>volatile</tt> field accesses.
+     * {@code volatile} field accesses.
      */
     protected final void setExclusiveOwnerThread(Thread t) {
         exclusiveOwnerThread = t;
@@ -46,9 +46,9 @@
 
     /**
      * Returns the thread last set by
-     * <tt>setExclusiveOwnerThread</tt>, or <tt>null</tt> if never
+     * {@code setExclusiveOwnerThread}, or {@code null} if never
      * set.  This method does not otherwise impose any synchronization
-     * or <tt>volatile</tt> field accesses.
+     * or {@code volatile} field accesses.
      * @return the owner thread
      */
     protected final Thread getExclusiveOwnerThread() {
diff --git a/luni/src/main/java/java/util/concurrent/locks/AbstractQueuedLongSynchronizer.java b/luni/src/main/java/java/util/concurrent/locks/AbstractQueuedLongSynchronizer.java
index 7b36460..4c5e280 100644
--- a/luni/src/main/java/java/util/concurrent/locks/AbstractQueuedLongSynchronizer.java
+++ b/luni/src/main/java/java/util/concurrent/locks/AbstractQueuedLongSynchronizer.java
@@ -5,17 +5,19 @@
  */
 
 package java.util.concurrent.locks;
-import java.util.*;
-import java.util.concurrent.*;
+import java.util.concurrent.TimeUnit;
+import java.util.ArrayList;
+import java.util.Collection;
+import java.util.Date;
 import sun.misc.Unsafe;
 
 /**
  * A version of {@link AbstractQueuedSynchronizer} in
- * which synchronization state is maintained as a <tt>long</tt>.
+ * which synchronization state is maintained as a {@code long}.
  * This class has exactly the same structure, properties, and methods
- * as <tt>AbstractQueuedSynchronizer</tt> with the exception
+ * as {@code AbstractQueuedSynchronizer} with the exception
  * that all state-related parameters and results are defined
- * as <tt>long</tt> rather than <tt>int</tt>. This class
+ * as {@code long} rather than {@code int}. This class
  * may be useful when creating synchronizers such as
  * multilevel locks and barriers that require
  * 64 bits of state.
@@ -40,7 +42,7 @@
     */
 
     /**
-     * Creates a new <tt>AbstractQueuedLongSynchronizer</tt> instance
+     * Creates a new {@code AbstractQueuedLongSynchronizer} instance
      * with initial synchronization state of zero.
      */
     protected AbstractQueuedLongSynchronizer() { }
@@ -73,7 +75,7 @@
      *
      * <p>Insertion into a CLH queue requires only a single atomic
      * operation on "tail", so there is a simple atomic point of
-     * demarcation from unqueued to queued. Similarly, dequeing
+     * demarcation from unqueued to queued. Similarly, dequeuing
      * involves only updating the "head". However, it takes a bit
      * more work for nodes to determine who their successors are,
      * in part to deal with possible cancellation due to timeouts
@@ -180,7 +182,7 @@
 
         /**
          * Link to predecessor node that current node/thread relies on
-         * for checking waitStatus. Assigned during enqueing, and nulled
+         * for checking waitStatus. Assigned during enqueuing, and nulled
          * out (for sake of GC) only upon dequeuing.  Also, upon
          * cancellation of a predecessor, we short-circuit while
          * finding a non-cancelled one, which will always exist
@@ -225,7 +227,7 @@
         Node nextWaiter;
 
         /**
-         * Returns true if node is waiting in shared mode
+         * @return true if node is waiting in shared mode
          */
         final boolean isShared() {
             return nextWaiter == SHARED;
@@ -281,7 +283,7 @@
 
     /**
      * Returns the current value of synchronization state.
-     * This operation has memory semantics of a <tt>volatile</tt> read.
+     * This operation has memory semantics of a {@code volatile} read.
      * @return current state value
      */
     protected final long getState() {
@@ -290,7 +292,7 @@
 
     /**
      * Sets the value of synchronization state.
-     * This operation has memory semantics of a <tt>volatile</tt> write.
+     * This operation has memory semantics of a {@code volatile} write.
      * @param newState the new state value
      */
     protected final void setState(long newState) {
@@ -300,7 +302,7 @@
     /**
      * Atomically sets synchronization state to the given updated
      * value if the current state value equals the expected value.
-     * This operation has memory semantics of a <tt>volatile</tt> read
+     * This operation has memory semantics of a {@code volatile} read
      * and write.
      *
      * @param expect the expected value
@@ -410,7 +412,7 @@
     }
 
     /**
-     * Release action for shared mode -- signal successor and ensure
+     * Release action for shared mode -- signals successor and ensures
      * propagation. (Note: For exclusive mode, release just amounts
      * to calling unparkSuccessor of head if it needs signal.)
      */
@@ -531,7 +533,7 @@
     /**
      * Checks and updates status for a node that failed to acquire.
      * Returns true if thread should block. This is the main signal
-     * control in all acquire loops.  Requires that pred == node.prev
+     * control in all acquire loops.  Requires that pred == node.prev.
      *
      * @param pred node's predecessor holding status
      * @param node the node
@@ -656,8 +658,10 @@
      * @return {@code true} if acquired
      */
     private boolean doAcquireNanos(long arg, long nanosTimeout)
-        throws InterruptedException {
-        long lastTime = System.nanoTime();
+            throws InterruptedException {
+        if (nanosTimeout <= 0L)
+            return false;
+        final long deadline = System.nanoTime() + nanosTimeout;
         final Node node = addWaiter(Node.EXCLUSIVE);
         boolean failed = true;
         try {
@@ -669,14 +673,12 @@
                     failed = false;
                     return true;
                 }
-                if (nanosTimeout <= 0)
+                nanosTimeout = deadline - System.nanoTime();
+                if (nanosTimeout <= 0L)
                     return false;
                 if (shouldParkAfterFailedAcquire(p, node) &&
                     nanosTimeout > spinForTimeoutThreshold)
                     LockSupport.parkNanos(this, nanosTimeout);
-                long now = System.nanoTime();
-                nanosTimeout -= now - lastTime;
-                lastTime = now;
                 if (Thread.interrupted())
                     throw new InterruptedException();
             }
@@ -756,9 +758,10 @@
      * @return {@code true} if acquired
      */
     private boolean doAcquireSharedNanos(long arg, long nanosTimeout)
-        throws InterruptedException {
-
-        long lastTime = System.nanoTime();
+            throws InterruptedException {
+        if (nanosTimeout <= 0L)
+            return false;
+        final long deadline = System.nanoTime() + nanosTimeout;
         final Node node = addWaiter(Node.SHARED);
         boolean failed = true;
         try {
@@ -773,14 +776,12 @@
                         return true;
                     }
                 }
-                if (nanosTimeout <= 0)
+                nanosTimeout = deadline - System.nanoTime();
+                if (nanosTimeout <= 0L)
                     return false;
                 if (shouldParkAfterFailedAcquire(p, node) &&
                     nanosTimeout > spinForTimeoutThreshold)
                     LockSupport.parkNanos(this, nanosTimeout);
-                long now = System.nanoTime();
-                nanosTimeout -= now - lastTime;
-                lastTime = now;
                 if (Thread.interrupted())
                     throw new InterruptedException();
             }
@@ -1036,7 +1037,7 @@
      * thread is queued, possibly repeatedly blocking and unblocking,
      * invoking {@link #tryAcquireShared} until success or the thread
      * is interrupted.
-     * @param arg the acquire argument
+     * @param arg the acquire argument.
      * This value is conveyed to {@link #tryAcquireShared} but is
      * otherwise uninterpreted and can represent anything
      * you like.
@@ -1254,8 +1255,9 @@
      *         current thread, and {@code false} if the current thread
      *         is at the head of the queue or the queue is empty
      * @since 1.7
+     * @hide
      */
-    /*public*/ final boolean hasQueuedPredecessors() { // android-changed
+    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.
@@ -1411,7 +1413,7 @@
      * Returns true if successful.
      * @param node the node
      * @return true if successfully transferred (else the node was
-     * cancelled before signal).
+     * cancelled before signal)
      */
     final boolean transferForSignal(Node node) {
         /*
@@ -1434,10 +1436,10 @@
     }
 
     /**
-     * Transfers node, if necessary, to sync queue after a cancelled
-     * wait. Returns true if thread was cancelled before being
-     * signalled.
-     * @param node its node
+     * Transfers node, if necessary, to sync queue after a cancelled wait.
+     * Returns true if thread was cancelled before being signalled.
+     *
+     * @param node the node
      * @return true if cancelled before the node was signalled
      */
     final boolean transferAfterCancelledWait(Node node) {
@@ -1485,25 +1487,23 @@
      * uses this synchronizer as its lock.
      *
      * @param condition the condition
-     * @return <tt>true</tt> if owned
+     * @return {@code true} if owned
      * @throws NullPointerException if the condition is null
      */
     public final boolean owns(ConditionObject condition) {
-        if (condition == null)
-            throw new NullPointerException();
         return condition.isOwnedBy(this);
     }
 
     /**
      * Queries whether any threads are waiting on the given condition
      * associated with this synchronizer. Note that because timeouts
-     * and interrupts may occur at any time, a <tt>true</tt> return
-     * does not guarantee that a future <tt>signal</tt> will awaken
+     * and interrupts may occur at any time, a {@code true} return
+     * does not guarantee that a future {@code signal} will awaken
      * any threads.  This method is designed primarily for use in
      * monitoring of the system state.
      *
      * @param condition the condition
-     * @return <tt>true</tt> if there are any waiting threads
+     * @return {@code true} if there are any waiting threads
      * @throws IllegalMonitorStateException if exclusive synchronization
      *         is not held
      * @throws IllegalArgumentException if the given condition is
@@ -1570,7 +1570,7 @@
      * and Condition users. Exported versions of this class will in
      * general need to be accompanied by documentation describing
      * condition semantics that rely on those of the associated
-     * <tt>AbstractQueuedLongSynchronizer</tt>.
+     * {@code AbstractQueuedLongSynchronizer}.
      *
      * <p>This class is Serializable, but all fields are transient,
      * so deserialized conditions have no waiters.
@@ -1585,7 +1585,7 @@
         private transient Node lastWaiter;
 
         /**
-         * Creates a new <tt>ConditionObject</tt> instance.
+         * Creates a new {@code ConditionObject} instance.
          */
         public ConditionObject() { }
 
@@ -1711,9 +1711,8 @@
          * Implements uninterruptible condition wait.
          * <ol>
          * <li> Save lock state returned by {@link #getState}.
-         * <li> Invoke {@link #release} with
-         *      saved state as argument, throwing
-         *      IllegalMonitorStateException if it fails.
+         * <li> Invoke {@link #release} with saved state as argument,
+         *      throwing IllegalMonitorStateException if it fails.
          * <li> Block until signalled.
          * <li> Reacquire by invoking specialized version of
          *      {@link #acquire} with saved state as argument.
@@ -1772,9 +1771,8 @@
          * <ol>
          * <li> If current thread is interrupted, throw InterruptedException.
          * <li> Save lock state returned by {@link #getState}.
-         * <li> Invoke {@link #release} with
-         *      saved state as argument, throwing
-         *      IllegalMonitorStateException if it fails.
+         * <li> Invoke {@link #release} with saved state as argument,
+         *      throwing IllegalMonitorStateException if it fails.
          * <li> Block until signalled or interrupted.
          * <li> Reacquire by invoking specialized version of
          *      {@link #acquire} with saved state as argument.
@@ -1805,9 +1803,8 @@
          * <ol>
          * <li> If current thread is interrupted, throw InterruptedException.
          * <li> Save lock state returned by {@link #getState}.
-         * <li> Invoke {@link #release} with
-         *      saved state as argument, throwing
-         *      IllegalMonitorStateException if it fails.
+         * <li> Invoke {@link #release} with saved state as argument,
+         *      throwing IllegalMonitorStateException if it fails.
          * <li> Block until signalled, interrupted, or timed out.
          * <li> Reacquire by invoking specialized version of
          *      {@link #acquire} with saved state as argument.
@@ -1820,20 +1817,18 @@
                 throw new InterruptedException();
             Node node = addConditionWaiter();
             long savedState = fullyRelease(node);
-            long lastTime = System.nanoTime();
+            final long deadline = System.nanoTime() + nanosTimeout;
             int interruptMode = 0;
             while (!isOnSyncQueue(node)) {
                 if (nanosTimeout <= 0L) {
                     transferAfterCancelledWait(node);
                     break;
                 }
-                LockSupport.parkNanos(this, nanosTimeout);
+                if (nanosTimeout >= spinForTimeoutThreshold)
+                    LockSupport.parkNanos(this, nanosTimeout);
                 if ((interruptMode = checkInterruptWhileWaiting(node)) != 0)
                     break;
-
-                long now = System.nanoTime();
-                nanosTimeout -= now - lastTime;
-                lastTime = now;
+                nanosTimeout = deadline - System.nanoTime();
             }
             if (acquireQueued(node, savedState) && interruptMode != THROW_IE)
                 interruptMode = REINTERRUPT;
@@ -1841,7 +1836,7 @@
                 unlinkCancelledWaiters();
             if (interruptMode != 0)
                 reportInterruptAfterWait(interruptMode);
-            return nanosTimeout - (System.nanoTime() - lastTime);
+            return deadline - System.nanoTime();
         }
 
         /**
@@ -1849,9 +1844,8 @@
          * <ol>
          * <li> If current thread is interrupted, throw InterruptedException.
          * <li> Save lock state returned by {@link #getState}.
-         * <li> Invoke {@link #release} with
-         *      saved state as argument, throwing
-         *      IllegalMonitorStateException if it fails.
+         * <li> Invoke {@link #release} with saved state as argument,
+         *      throwing IllegalMonitorStateException if it fails.
          * <li> Block until signalled, interrupted, or timed out.
          * <li> Reacquire by invoking specialized version of
          *      {@link #acquire} with saved state as argument.
@@ -1861,8 +1855,6 @@
          */
         public final boolean awaitUntil(Date deadline)
                 throws InterruptedException {
-            if (deadline == null)
-                throw new NullPointerException();
             long abstime = deadline.getTime();
             if (Thread.interrupted())
                 throw new InterruptedException();
@@ -1893,9 +1885,8 @@
          * <ol>
          * <li> If current thread is interrupted, throw InterruptedException.
          * <li> Save lock state returned by {@link #getState}.
-         * <li> Invoke {@link #release} with
-         *      saved state as argument, throwing
-         *      IllegalMonitorStateException if it fails.
+         * <li> Invoke {@link #release} with saved state as argument,
+         *      throwing IllegalMonitorStateException if it fails.
          * <li> Block until signalled, interrupted, or timed out.
          * <li> Reacquire by invoking specialized version of
          *      {@link #acquire} with saved state as argument.
@@ -1905,14 +1896,12 @@
          */
         public final boolean await(long time, TimeUnit unit)
                 throws InterruptedException {
-            if (unit == null)
-                throw new NullPointerException();
             long nanosTimeout = unit.toNanos(time);
             if (Thread.interrupted())
                 throw new InterruptedException();
             Node node = addConditionWaiter();
             long savedState = fullyRelease(node);
-            long lastTime = System.nanoTime();
+            final long deadline = System.nanoTime() + nanosTimeout;
             boolean timedout = false;
             int interruptMode = 0;
             while (!isOnSyncQueue(node)) {
@@ -1924,9 +1913,7 @@
                     LockSupport.parkNanos(this, nanosTimeout);
                 if ((interruptMode = checkInterruptWhileWaiting(node)) != 0)
                     break;
-                long now = System.nanoTime();
-                nanosTimeout -= now - lastTime;
-                lastTime = now;
+                nanosTimeout = deadline - System.nanoTime();
             }
             if (acquireQueued(node, savedState) && interruptMode != THROW_IE)
                 interruptMode = REINTERRUPT;
diff --git a/luni/src/main/java/java/util/concurrent/locks/AbstractQueuedSynchronizer.java b/luni/src/main/java/java/util/concurrent/locks/AbstractQueuedSynchronizer.java
index 42029f0..0350060 100644
--- a/luni/src/main/java/java/util/concurrent/locks/AbstractQueuedSynchronizer.java
+++ b/luni/src/main/java/java/util/concurrent/locks/AbstractQueuedSynchronizer.java
@@ -5,8 +5,10 @@
  */
 
 package java.util.concurrent.locks;
-import java.util.*;
-import java.util.concurrent.*;
+import java.util.concurrent.TimeUnit;
+import java.util.ArrayList;
+import java.util.Collection;
+import java.util.Date;
 import sun.misc.Unsafe;
 
 // BEGIN android-note
@@ -18,12 +20,12 @@
  * synchronizers (semaphores, events, etc) that rely on
  * first-in-first-out (FIFO) wait queues.  This class is designed to
  * be a useful basis for most kinds of synchronizers that rely on a
- * single atomic <tt>int</tt> value to represent state. Subclasses
+ * single atomic {@code int} value to represent state. Subclasses
  * must define the protected methods that change this state, and which
  * define what that state means in terms of this object being acquired
  * or released.  Given these, the other methods in this class carry
  * out all queuing and blocking mechanics. Subclasses can maintain
- * other state fields, but only the atomically updated <tt>int</tt>
+ * other state fields, but only the atomically updated {@code int}
  * value manipulated using methods {@link #getState}, {@link
  * #setState} and {@link #compareAndSetState} is tracked with respect
  * to synchronization.
@@ -31,7 +33,7 @@
  * <p>Subclasses should be defined as non-public internal helper
  * classes that are used to implement the synchronization properties
  * of their enclosing class.  Class
- * <tt>AbstractQueuedSynchronizer</tt> does not implement any
+ * {@code AbstractQueuedSynchronizer} does not implement any
  * synchronization interface.  Instead it defines methods such as
  * {@link #acquireInterruptibly} that can be invoked as
  * appropriate by concrete locks and related synchronizers to
@@ -58,7 +60,7 @@
  * invoked with the current {@link #getState} value fully releases
  * this object, and {@link #acquire}, given this saved state value,
  * eventually restores this object to its previous acquired state.  No
- * <tt>AbstractQueuedSynchronizer</tt> method otherwise creates such a
+ * {@code AbstractQueuedSynchronizer} method otherwise creates such a
  * condition, so if this constraint cannot be met, do not use it.  The
  * behavior of {@link ConditionObject} depends of course on the
  * semantics of its synchronizer implementation.
@@ -66,13 +68,13 @@
  * <p>This class provides inspection, instrumentation, and monitoring
  * methods for the internal queue, as well as similar methods for
  * condition objects. These can be exported as desired into classes
- * using an <tt>AbstractQueuedSynchronizer</tt> for their
+ * using an {@code AbstractQueuedSynchronizer} for their
  * synchronization mechanics.
  *
  * <p>Serialization of this class stores only the underlying atomic
  * integer maintaining state, so deserialized objects have empty
  * thread queues. Typical subclasses requiring serializability will
- * define a <tt>readObject</tt> method that restores this to a known
+ * define a {@code readObject} method that restores this to a known
  * initial state upon deserialization.
  *
  * <h3>Usage</h3>
@@ -88,14 +90,14 @@
  * <li> {@link #tryAcquireShared}
  * <li> {@link #tryReleaseShared}
  * <li> {@link #isHeldExclusively}
- *</ul>
+ * </ul>
  *
  * Each of these methods by default throws {@link
  * UnsupportedOperationException}.  Implementations of these methods
  * must be internally thread-safe, and should in general be short and
  * not block. Defining these methods is the <em>only</em> supported
  * means of using this class. All other methods are declared
- * <tt>final</tt> because they cannot be independently varied.
+ * {@code final} because they cannot be independently varied.
  *
  * <p>You may also find the inherited methods from {@link
  * AbstractOwnableSynchronizer} useful to keep track of the thread
@@ -121,19 +123,19 @@
  *
  * (Shared mode is similar but may involve cascading signals.)
  *
- * <p><a name="barging">Because checks in acquire are invoked before
+ * <p id="barging">Because checks in acquire are invoked before
  * enqueuing, a newly acquiring thread may <em>barge</em> ahead of
- * others that are blocked and queued. However, you can, if desired,
- * define <tt>tryAcquire</tt> and/or <tt>tryAcquireShared</tt> to
+ * others that are blocked and queued.  However, you can, if desired,
+ * define {@code tryAcquire} and/or {@code tryAcquireShared} to
  * disable barging by internally invoking one or more of the inspection
  * methods. In particular, a strict FIFO lock can define
- * <tt>tryAcquire</tt> to immediately return <tt>false</tt> if {@link
+ * {@code tryAcquire} to immediately return {@code false} if {@link
  * #getFirstQueuedThread} does not return the current thread.  A
  * normally preferable non-strict fair version can immediately return
- * <tt>false</tt> only if {@link #hasQueuedThreads} returns
- * <tt>true</tt> and <tt>getFirstQueuedThread</tt> is not the current
- * thread; or equivalently, that <tt>getFirstQueuedThread</tt> is both
- * non-null and not the current thread.  Further variations are
+ * {@code false} only if {@link #hasQueuedThreads} returns
+ * {@code true} and {@code getFirstQueuedThread} is not the current
+ * thread; or equivalently, that {@code getFirstQueuedThread} is both
+ * non-null and not the current thread. Further variations are
  * possible.
  *
  * <p>Throughput and scalability are generally highest for the
@@ -144,7 +146,7 @@
  * threads, and each recontention has an unbiased chance to succeed
  * against incoming threads.  Also, while acquires do not
  * &quot;spin&quot; in the usual sense, they may perform multiple
- * invocations of <tt>tryAcquire</tt> interspersed with other
+ * invocations of {@code tryAcquire} interspersed with other
  * computations before blocking.  This gives most of the benefits of
  * spins when exclusive synchronization is only briefly held, without
  * most of the liabilities when it isn't. If so desired, you can
@@ -155,7 +157,7 @@
  *
  * <p>This class provides an efficient and scalable basis for
  * synchronization in part by specializing its range of use to
- * synchronizers that can rely on <tt>int</tt> state, acquire, and
+ * synchronizers that can rely on {@code int} state, acquire, and
  * release parameters, and an internal FIFO wait queue. When this does
  * not suffice, you can build synchronizers from a lower level using
  * {@link java.util.concurrent.atomic atomic} classes, your own custom
@@ -177,12 +179,12 @@
  *
  *   // Our internal helper class
  *   private static class Sync extends AbstractQueuedSynchronizer {
- *     // Report whether in locked state
+ *     // Reports whether in locked state
  *     protected boolean isHeldExclusively() {
  *       return getState() == 1;
  *     }
  *
- *     // Acquire the lock if state is zero
+ *     // Acquires the lock if state is zero
  *     public boolean tryAcquire(int acquires) {
  *       assert acquires == 1; // Otherwise unused
  *       if (compareAndSetState(0, 1)) {
@@ -192,7 +194,7 @@
  *       return false;
  *     }
  *
- *     // Release the lock by setting state to zero
+ *     // Releases the lock by setting state to zero
  *     protected boolean tryRelease(int releases) {
  *       assert releases == 1; // Otherwise unused
  *       if (getState() == 0) throw new IllegalMonitorStateException();
@@ -201,10 +203,10 @@
  *       return true;
  *     }
  *
- *     // Provide a Condition
+ *     // Provides a Condition
  *     Condition newCondition() { return new ConditionObject(); }
  *
- *     // Deserialize properly
+ *     // Deserializes properly
  *     private void readObject(ObjectInputStream s)
  *         throws IOException, ClassNotFoundException {
  *       s.defaultReadObject();
@@ -230,9 +232,10 @@
  *   }
  * }}</pre>
  *
- * <p>Here is a latch class that is like a {@link CountDownLatch}
- * except that it only requires a single <tt>signal</tt> to
- * fire. Because a latch is non-exclusive, it uses the <tt>shared</tt>
+ * <p>Here is a latch class that is like a
+ * {@link java.util.concurrent.CountDownLatch CountDownLatch}
+ * except that it only requires a single {@code signal} to
+ * fire. Because a latch is non-exclusive, it uses the {@code shared}
  * acquire and release methods.
  *
  *  <pre> {@code
@@ -269,7 +272,7 @@
     private static final long serialVersionUID = 7373984972572414691L;
 
     /**
-     * Creates a new <tt>AbstractQueuedSynchronizer</tt> instance
+     * Creates a new {@code AbstractQueuedSynchronizer} instance
      * with initial synchronization state of zero.
      */
     protected AbstractQueuedSynchronizer() { }
@@ -302,7 +305,7 @@
      *
      * <p>Insertion into a CLH queue requires only a single atomic
      * operation on "tail", so there is a simple atomic point of
-     * demarcation from unqueued to queued. Similarly, dequeing
+     * demarcation from unqueued to queued. Similarly, dequeuing
      * involves only updating the "head". However, it takes a bit
      * more work for nodes to determine who their successors are,
      * in part to deal with possible cancellation due to timeouts
@@ -409,7 +412,7 @@
 
         /**
          * Link to predecessor node that current node/thread relies on
-         * for checking waitStatus. Assigned during enqueing, and nulled
+         * for checking waitStatus. Assigned during enqueuing, and nulled
          * out (for sake of GC) only upon dequeuing.  Also, upon
          * cancellation of a predecessor, we short-circuit while
          * finding a non-cancelled one, which will always exist
@@ -454,7 +457,7 @@
         Node nextWaiter;
 
         /**
-         * Returns true if node is waiting in shared mode
+         * Returns true if node is waiting in shared mode.
          */
         final boolean isShared() {
             return nextWaiter == SHARED;
@@ -510,7 +513,7 @@
 
     /**
      * Returns the current value of synchronization state.
-     * This operation has memory semantics of a <tt>volatile</tt> read.
+     * This operation has memory semantics of a {@code volatile} read.
      * @return current state value
      */
     protected final int getState() {
@@ -519,7 +522,7 @@
 
     /**
      * Sets the value of synchronization state.
-     * This operation has memory semantics of a <tt>volatile</tt> write.
+     * This operation has memory semantics of a {@code volatile} write.
      * @param newState the new state value
      */
     protected final void setState(int newState) {
@@ -529,7 +532,7 @@
     /**
      * Atomically sets synchronization state to the given updated
      * value if the current state value equals the expected value.
-     * This operation has memory semantics of a <tt>volatile</tt> read
+     * This operation has memory semantics of a {@code volatile} read
      * and write.
      *
      * @param expect the expected value
@@ -639,7 +642,7 @@
     }
 
     /**
-     * Release action for shared mode -- signal successor and ensure
+     * Release action for shared mode -- signals successor and ensures
      * propagation. (Note: For exclusive mode, release just amounts
      * to calling unparkSuccessor of head if it needs signal.)
      */
@@ -760,7 +763,7 @@
     /**
      * Checks and updates status for a node that failed to acquire.
      * Returns true if thread should block. This is the main signal
-     * control in all acquire loops.  Requires that pred == node.prev
+     * control in all acquire loops.  Requires that pred == node.prev.
      *
      * @param pred node's predecessor holding status
      * @param node the node
@@ -885,8 +888,10 @@
      * @return {@code true} if acquired
      */
     private boolean doAcquireNanos(int arg, long nanosTimeout)
-        throws InterruptedException {
-        long lastTime = System.nanoTime();
+            throws InterruptedException {
+        if (nanosTimeout <= 0L)
+            return false;
+        final long deadline = System.nanoTime() + nanosTimeout;
         final Node node = addWaiter(Node.EXCLUSIVE);
         boolean failed = true;
         try {
@@ -898,14 +903,12 @@
                     failed = false;
                     return true;
                 }
-                if (nanosTimeout <= 0)
+                nanosTimeout = deadline - System.nanoTime();
+                if (nanosTimeout <= 0L)
                     return false;
                 if (shouldParkAfterFailedAcquire(p, node) &&
                     nanosTimeout > spinForTimeoutThreshold)
                     LockSupport.parkNanos(this, nanosTimeout);
-                long now = System.nanoTime();
-                nanosTimeout -= now - lastTime;
-                lastTime = now;
                 if (Thread.interrupted())
                     throw new InterruptedException();
             }
@@ -985,9 +988,10 @@
      * @return {@code true} if acquired
      */
     private boolean doAcquireSharedNanos(int arg, long nanosTimeout)
-        throws InterruptedException {
-
-        long lastTime = System.nanoTime();
+            throws InterruptedException {
+        if (nanosTimeout <= 0L)
+            return false;
+        final long deadline = System.nanoTime() + nanosTimeout;
         final Node node = addWaiter(Node.SHARED);
         boolean failed = true;
         try {
@@ -1002,14 +1006,12 @@
                         return true;
                     }
                 }
-                if (nanosTimeout <= 0)
+                nanosTimeout = deadline - System.nanoTime();
+                if (nanosTimeout <= 0L)
                     return false;
                 if (shouldParkAfterFailedAcquire(p, node) &&
                     nanosTimeout > spinForTimeoutThreshold)
                     LockSupport.parkNanos(this, nanosTimeout);
-                long now = System.nanoTime();
-                nanosTimeout -= now - lastTime;
-                lastTime = now;
                 if (Thread.interrupted())
                     throw new InterruptedException();
             }
@@ -1265,7 +1267,7 @@
      * thread is queued, possibly repeatedly blocking and unblocking,
      * invoking {@link #tryAcquireShared} until success or the thread
      * is interrupted.
-     * @param arg the acquire argument
+     * @param arg the acquire argument.
      * This value is conveyed to {@link #tryAcquireShared} but is
      * otherwise uninterpreted and can represent anything
      * you like.
@@ -1641,7 +1643,7 @@
      * Returns true if successful.
      * @param node the node
      * @return true if successfully transferred (else the node was
-     * cancelled before signal).
+     * cancelled before signal)
      */
     final boolean transferForSignal(Node node) {
         /*
@@ -1664,10 +1666,10 @@
     }
 
     /**
-     * Transfers node, if necessary, to sync queue after a cancelled
-     * wait. Returns true if thread was cancelled before being
-     * signalled.
-     * @param node its node
+     * Transfers node, if necessary, to sync queue after a cancelled wait.
+     * Returns true if thread was cancelled before being signalled.
+     *
+     * @param node the node
      * @return true if cancelled before the node was signalled
      */
     final boolean transferAfterCancelledWait(Node node) {
@@ -1715,25 +1717,23 @@
      * uses this synchronizer as its lock.
      *
      * @param condition the condition
-     * @return <tt>true</tt> if owned
+     * @return {@code true} if owned
      * @throws NullPointerException if the condition is null
      */
     public final boolean owns(ConditionObject condition) {
-        if (condition == null)
-            throw new NullPointerException();
         return condition.isOwnedBy(this);
     }
 
     /**
      * Queries whether any threads are waiting on the given condition
      * associated with this synchronizer. Note that because timeouts
-     * and interrupts may occur at any time, a <tt>true</tt> return
-     * does not guarantee that a future <tt>signal</tt> will awaken
+     * and interrupts may occur at any time, a {@code true} return
+     * does not guarantee that a future {@code signal} will awaken
      * any threads.  This method is designed primarily for use in
      * monitoring of the system state.
      *
      * @param condition the condition
-     * @return <tt>true</tt> if there are any waiting threads
+     * @return {@code true} if there are any waiting threads
      * @throws IllegalMonitorStateException if exclusive synchronization
      *         is not held
      * @throws IllegalArgumentException if the given condition is
@@ -1800,7 +1800,7 @@
      * and Condition users. Exported versions of this class will in
      * general need to be accompanied by documentation describing
      * condition semantics that rely on those of the associated
-     * <tt>AbstractQueuedSynchronizer</tt>.
+     * {@code AbstractQueuedSynchronizer}.
      *
      * <p>This class is Serializable, but all fields are transient,
      * so deserialized conditions have no waiters.
@@ -1813,7 +1813,7 @@
         private transient Node lastWaiter;
 
         /**
-         * Creates a new <tt>ConditionObject</tt> instance.
+         * Creates a new {@code ConditionObject} instance.
          */
         public ConditionObject() { }
 
@@ -1939,9 +1939,8 @@
          * Implements uninterruptible condition wait.
          * <ol>
          * <li> Save lock state returned by {@link #getState}.
-         * <li> Invoke {@link #release} with
-         *      saved state as argument, throwing
-         *      IllegalMonitorStateException if it fails.
+         * <li> Invoke {@link #release} with saved state as argument,
+         *      throwing IllegalMonitorStateException if it fails.
          * <li> Block until signalled.
          * <li> Reacquire by invoking specialized version of
          *      {@link #acquire} with saved state as argument.
@@ -2000,9 +1999,8 @@
          * <ol>
          * <li> If current thread is interrupted, throw InterruptedException.
          * <li> Save lock state returned by {@link #getState}.
-         * <li> Invoke {@link #release} with
-         *      saved state as argument, throwing
-         *      IllegalMonitorStateException if it fails.
+         * <li> Invoke {@link #release} with saved state as argument,
+         *      throwing IllegalMonitorStateException if it fails.
          * <li> Block until signalled or interrupted.
          * <li> Reacquire by invoking specialized version of
          *      {@link #acquire} with saved state as argument.
@@ -2033,9 +2031,8 @@
          * <ol>
          * <li> If current thread is interrupted, throw InterruptedException.
          * <li> Save lock state returned by {@link #getState}.
-         * <li> Invoke {@link #release} with
-         *      saved state as argument, throwing
-         *      IllegalMonitorStateException if it fails.
+         * <li> Invoke {@link #release} with saved state as argument,
+         *      throwing IllegalMonitorStateException if it fails.
          * <li> Block until signalled, interrupted, or timed out.
          * <li> Reacquire by invoking specialized version of
          *      {@link #acquire} with saved state as argument.
@@ -2048,20 +2045,18 @@
                 throw new InterruptedException();
             Node node = addConditionWaiter();
             int savedState = fullyRelease(node);
-            long lastTime = System.nanoTime();
+            final long deadline = System.nanoTime() + nanosTimeout;
             int interruptMode = 0;
             while (!isOnSyncQueue(node)) {
                 if (nanosTimeout <= 0L) {
                     transferAfterCancelledWait(node);
                     break;
                 }
-                LockSupport.parkNanos(this, nanosTimeout);
+                if (nanosTimeout >= spinForTimeoutThreshold)
+                    LockSupport.parkNanos(this, nanosTimeout);
                 if ((interruptMode = checkInterruptWhileWaiting(node)) != 0)
                     break;
-
-                long now = System.nanoTime();
-                nanosTimeout -= now - lastTime;
-                lastTime = now;
+                nanosTimeout = deadline - System.nanoTime();
             }
             if (acquireQueued(node, savedState) && interruptMode != THROW_IE)
                 interruptMode = REINTERRUPT;
@@ -2069,7 +2064,7 @@
                 unlinkCancelledWaiters();
             if (interruptMode != 0)
                 reportInterruptAfterWait(interruptMode);
-            return nanosTimeout - (System.nanoTime() - lastTime);
+            return deadline - System.nanoTime();
         }
 
         /**
@@ -2077,9 +2072,8 @@
          * <ol>
          * <li> If current thread is interrupted, throw InterruptedException.
          * <li> Save lock state returned by {@link #getState}.
-         * <li> Invoke {@link #release} with
-         *      saved state as argument, throwing
-         *      IllegalMonitorStateException if it fails.
+         * <li> Invoke {@link #release} with saved state as argument,
+         *      throwing IllegalMonitorStateException if it fails.
          * <li> Block until signalled, interrupted, or timed out.
          * <li> Reacquire by invoking specialized version of
          *      {@link #acquire} with saved state as argument.
@@ -2089,8 +2083,6 @@
          */
         public final boolean awaitUntil(Date deadline)
                 throws InterruptedException {
-            if (deadline == null)
-                throw new NullPointerException();
             long abstime = deadline.getTime();
             if (Thread.interrupted())
                 throw new InterruptedException();
@@ -2121,9 +2113,8 @@
          * <ol>
          * <li> If current thread is interrupted, throw InterruptedException.
          * <li> Save lock state returned by {@link #getState}.
-         * <li> Invoke {@link #release} with
-         *      saved state as argument, throwing
-         *      IllegalMonitorStateException if it fails.
+         * <li> Invoke {@link #release} with saved state as argument,
+         *      throwing IllegalMonitorStateException if it fails.
          * <li> Block until signalled, interrupted, or timed out.
          * <li> Reacquire by invoking specialized version of
          *      {@link #acquire} with saved state as argument.
@@ -2133,14 +2124,12 @@
          */
         public final boolean await(long time, TimeUnit unit)
                 throws InterruptedException {
-            if (unit == null)
-                throw new NullPointerException();
             long nanosTimeout = unit.toNanos(time);
             if (Thread.interrupted())
                 throw new InterruptedException();
             Node node = addConditionWaiter();
             int savedState = fullyRelease(node);
-            long lastTime = System.nanoTime();
+            final long deadline = System.nanoTime() + nanosTimeout;
             boolean timedout = false;
             int interruptMode = 0;
             while (!isOnSyncQueue(node)) {
@@ -2152,9 +2141,7 @@
                     LockSupport.parkNanos(this, nanosTimeout);
                 if ((interruptMode = checkInterruptWhileWaiting(node)) != 0)
                     break;
-                long now = System.nanoTime();
-                nanosTimeout -= now - lastTime;
-                lastTime = now;
+                nanosTimeout = deadline - System.nanoTime();
             }
             if (acquireQueued(node, savedState) && interruptMode != THROW_IE)
                 interruptMode = REINTERRUPT;
diff --git a/luni/src/main/java/java/util/concurrent/locks/Condition.java b/luni/src/main/java/java/util/concurrent/locks/Condition.java
index 7050df9..522e9e2 100644
--- a/luni/src/main/java/java/util/concurrent/locks/Condition.java
+++ b/luni/src/main/java/java/util/concurrent/locks/Condition.java
@@ -5,7 +5,7 @@
  */
 
 package java.util.concurrent.locks;
-import java.util.concurrent.*;
+import java.util.concurrent.TimeUnit;
 import java.util.Date;
 
 /**
@@ -295,7 +295,7 @@
      *   }
      * }}</pre>
      *
-     * <p> Design note: This method requires a nanosecond argument so
+     * <p>Design note: This method requires a nanosecond argument so
      * as to avoid truncation errors in reporting remaining times.
      * Such precision loss would make it difficult for programmers to
      * ensure that total waiting times are not systematically shorter
diff --git a/luni/src/main/java/java/util/concurrent/locks/Lock.java b/luni/src/main/java/java/util/concurrent/locks/Lock.java
index d5c6294..6eeb236 100644
--- a/luni/src/main/java/java/util/concurrent/locks/Lock.java
+++ b/luni/src/main/java/java/util/concurrent/locks/Lock.java
@@ -91,8 +91,9 @@
  *
  * <p>All {@code Lock} implementations <em>must</em> enforce the same
  * memory synchronization semantics as provided by the built-in monitor
- * lock, as described in <a href="http://java.sun.com/docs/books/jls/">
- * The Java Language Specification, Third Edition (17.4 Memory Model)</a>:
+ * lock, as described in
+ * <a href="http://docs.oracle.com/javase/specs/jls/se7/html/jls-17.html#jls-17.4">
+ * The Java Language Specification (17.4 Memory Model)</a>:
  * <ul>
  * <li>A successful {@code lock} operation has the same memory
  * synchronization effects as a successful <em>Lock</em> action.
@@ -106,7 +107,7 @@
  *
  * <h3>Implementation Considerations</h3>
  *
- * <p> The three forms of lock acquisition (interruptible,
+ * <p>The three forms of lock acquisition (interruptible,
  * non-interruptible, and timed) may differ in their performance
  * characteristics, ordering guarantees, or other implementation
  * qualities.  Further, the ability to interrupt the <em>ongoing</em>
@@ -197,7 +198,7 @@
      *
      * @throws InterruptedException if the current thread is
      *         interrupted while acquiring the lock (and interruption
-     *         of lock acquisition is supported).
+     *         of lock acquisition is supported)
      */
     void lockInterruptibly() throws InterruptedException;
 
diff --git a/luni/src/main/java/java/util/concurrent/locks/LockSupport.java b/luni/src/main/java/java/util/concurrent/locks/LockSupport.java
index 422e428..875b2bf 100644
--- a/luni/src/main/java/java/util/concurrent/locks/LockSupport.java
+++ b/luni/src/main/java/java/util/concurrent/locks/LockSupport.java
@@ -7,7 +7,6 @@
 package java.util.concurrent.locks;
 import sun.misc.Unsafe;
 
-
 /**
  * Basic thread blocking primitives for creating locks and other
  * synchronization classes.
@@ -73,14 +72,14 @@
  *     // Block while not first in queue or cannot acquire lock
  *     while (waiters.peek() != current ||
  *            !locked.compareAndSet(false, true)) {
- *        LockSupport.park(this);
- *        if (Thread.interrupted()) // ignore interrupts while waiting
- *          wasInterrupted = true;
+ *       LockSupport.park(this);
+ *       if (Thread.interrupted()) // ignore interrupts while waiting
+ *         wasInterrupted = true;
  *     }
  *
  *     waiters.remove();
  *     if (wasInterrupted)          // reassert interrupt status on exit
- *        current.interrupt();
+ *       current.interrupt();
  *   }
  *
  *   public void unlock() {
@@ -89,7 +88,6 @@
  *   }
  * }}</pre>
  */
-
 public class LockSupport {
     private LockSupport() {} // Cannot be instantiated.
 
diff --git a/luni/src/main/java/java/util/concurrent/locks/ReadWriteLock.java b/luni/src/main/java/java/util/concurrent/locks/ReadWriteLock.java
index bb7b388..8690355 100644
--- a/luni/src/main/java/java/util/concurrent/locks/ReadWriteLock.java
+++ b/luni/src/main/java/java/util/concurrent/locks/ReadWriteLock.java
@@ -7,16 +7,16 @@
 package java.util.concurrent.locks;
 
 /**
- * A <tt>ReadWriteLock</tt> maintains a pair of associated {@link
+ * A {@code ReadWriteLock} maintains a pair of associated {@link
  * Lock locks}, one for read-only operations and one for writing.
  * The {@link #readLock read lock} may be held simultaneously by
  * multiple reader threads, so long as there are no writers.  The
  * {@link #writeLock write lock} is exclusive.
  *
- * <p>All <tt>ReadWriteLock</tt> implementations must guarantee that
- * the memory synchronization effects of <tt>writeLock</tt> operations
+ * <p>All {@code ReadWriteLock} implementations must guarantee that
+ * the memory synchronization effects of {@code writeLock} operations
  * (as specified in the {@link Lock} interface) also hold with respect
- * to the associated <tt>readLock</tt>. That is, a thread successfully
+ * to the associated {@code readLock}. That is, a thread successfully
  * acquiring the read lock will see all updates made upon previous
  * release of the write lock.
  *
@@ -91,14 +91,14 @@
     /**
      * Returns the lock used for reading.
      *
-     * @return the lock used for reading.
+     * @return the lock used for reading
      */
     Lock readLock();
 
     /**
      * Returns the lock used for writing.
      *
-     * @return the lock used for writing.
+     * @return the lock used for writing
      */
     Lock writeLock();
 }
diff --git a/luni/src/main/java/java/util/concurrent/locks/ReentrantLock.java b/luni/src/main/java/java/util/concurrent/locks/ReentrantLock.java
index 07baf41..bde4741 100644
--- a/luni/src/main/java/java/util/concurrent/locks/ReentrantLock.java
+++ b/luni/src/main/java/java/util/concurrent/locks/ReentrantLock.java
@@ -5,8 +5,8 @@
  */
 
 package java.util.concurrent.locks;
-import java.util.*;
-import java.util.concurrent.*;
+import java.util.concurrent.TimeUnit;
+import java.util.Collection;
 
 /**
  * A reentrant mutual exclusion {@link Lock} with the same basic
@@ -35,7 +35,7 @@
  * fair lock may obtain it multiple times in succession while other
  * active threads are not progressing and not currently holding the
  * lock.
- * Also note that the untimed {@link #tryLock() tryLock} method does not
+ * Also note that the untimed {@link #tryLock()} method does not
  * honor the fairness setting. It will succeed if the lock
  * is available even if other threads are waiting.
  *
@@ -59,10 +59,9 @@
  * }}</pre>
  *
  * <p>In addition to implementing the {@link Lock} interface, this
- * class defines methods {@code isLocked} and
- * {@code getLockQueueLength}, as well as some associated
- * {@code protected} access methods that may be useful for
- * instrumentation and monitoring.
+ * class defines a number of {@code public} and {@code protected}
+ * methods for inspecting the state of the lock.  Some of these
+ * methods are only useful for instrumentation and monitoring.
  *
  * <p>Serialization of this class behaves in the same way as built-in
  * locks: a deserialized lock is in the unlocked state, regardless of
@@ -95,9 +94,8 @@
         abstract void lock();
 
         /**
-         * Performs non-fair tryLock.  tryAcquire is
-         * implemented in subclasses, but both need nonfair
-         * try for trylock method.
+         * Performs non-fair tryLock.  tryAcquire is implemented in
+         * subclasses, but both need nonfair try for trylock method.
          */
         final boolean nonfairTryAcquire(int acquires) {
             final Thread current = Thread.currentThread();
@@ -156,8 +154,7 @@
         }
 
         /**
-         * Reconstitutes this lock instance from a stream.
-         * @param s the stream
+         * Reconstitutes the instance from a stream (that is, deserializes it).
          */
         private void readObject(java.io.ObjectInputStream s)
             throws java.io.IOException, ClassNotFoundException {
@@ -325,7 +322,7 @@
      * {@link #tryLock(long, TimeUnit) tryLock(0, TimeUnit.SECONDS) }
      * which is almost equivalent (it also detects interruption).
      *
-     * <p> If the current thread already holds this lock then the hold
+     * <p>If the current thread already holds this lock then the hold
      * count is incremented by one and the method returns {@code true}.
      *
      * <p>If the lock is held by another thread then this method will return
@@ -410,7 +407,6 @@
      *         the lock could be acquired
      * @throws InterruptedException if the current thread is interrupted
      * @throws NullPointerException if the time unit is null
-     *
      */
     public boolean tryLock(long timeout, TimeUnit unit)
             throws InterruptedException {
@@ -605,7 +601,6 @@
         return sync.hasQueuedThreads();
     }
 
-
     /**
      * Queries whether the given thread is waiting to acquire this
      * lock. Note that because cancellations may occur at any time, a
@@ -621,7 +616,6 @@
         return sync.isQueued(thread);
     }
 
-
     /**
      * Returns an estimate of the number of threads waiting to
      * acquire this lock.  The value is only an estimate because the number of
diff --git a/luni/src/main/java/java/util/concurrent/locks/ReentrantReadWriteLock.java b/luni/src/main/java/java/util/concurrent/locks/ReentrantReadWriteLock.java
index 244a4a7..2d3c65d 100644
--- a/luni/src/main/java/java/util/concurrent/locks/ReentrantReadWriteLock.java
+++ b/luni/src/main/java/java/util/concurrent/locks/ReentrantReadWriteLock.java
@@ -5,8 +5,8 @@
  */
 
 package java.util.concurrent.locks;
-import java.util.concurrent.*;
-import java.util.*;
+import java.util.concurrent.TimeUnit;
+import java.util.Collection;
 
 /**
  * An implementation of {@link ReadWriteLock} supporting similar
@@ -16,7 +16,7 @@
  * <ul>
  * <li><b>Acquisition order</b>
  *
- * <p> This class does not impose a reader or writer preference
+ * <p>This class does not impose a reader or writer preference
  * ordering for lock access.  However, it does support an optional
  * <em>fairness</em> policy.
  *
@@ -30,7 +30,7 @@
  * <p>
  *
  * <dt><b><i>Fair mode</i></b>
- * <dd> When constructed as fair, threads contend for entry using an
+ * <dd>When constructed as fair, threads contend for entry using an
  * approximately arrival-order policy. When the currently held lock
  * is released, either the longest-waiting single writer thread will
  * be assigned the write lock, or if there is a group of reader threads
@@ -183,7 +183,6 @@
  *
  * @since 1.5
  * @author Doug Lea
- *
  */
 public class ReentrantReadWriteLock
         implements ReadWriteLock, java.io.Serializable {
@@ -624,10 +623,7 @@
         }
 
         /**
-         * Reconstitutes this lock instance from a stream (that is,
-         * deserializes it).
-         *
-         * @param s the stream
+         * Reconstitutes the instance from a stream (that is, deserializes it).
          */
         private void readObject(java.io.ObjectInputStream s)
             throws java.io.IOException, ClassNotFoundException {
@@ -771,7 +767,7 @@
          *
          * @return {@code true} if the read lock was acquired
          */
-        public  boolean tryLock() {
+        public boolean tryLock() {
             return sync.tryReadLock();
         }
 
@@ -841,7 +837,6 @@
          * @return {@code true} if the read lock was acquired
          * @throws InterruptedException if the current thread is interrupted
          * @throws NullPointerException if the time unit is null
-         *
          */
         public boolean tryLock(long timeout, TimeUnit unit)
                 throws InterruptedException {
@@ -851,10 +846,10 @@
         /**
          * Attempts to release this lock.
          *
-         * <p> If the number of readers is now zero then the lock
+         * <p>If the number of readers is now zero then the lock
          * is made available for write lock attempts.
          */
-        public  void unlock() {
+        public void unlock() {
             sync.releaseShared(1);
         }
 
@@ -993,7 +988,7 @@
          * #tryLock(long, TimeUnit) tryLock(0, TimeUnit.SECONDS) }
          * which is almost equivalent (it also detects interruption).
          *
-         * <p> If the current thread already holds this lock then the
+         * <p>If the current thread already holds this lock then the
          * hold count is incremented by one and the method returns
          * {@code true}.
          *
@@ -1086,7 +1081,6 @@
          *
          * @throws InterruptedException if the current thread is interrupted
          * @throws NullPointerException if the time unit is null
-         *
          */
         public boolean tryLock(long timeout, TimeUnit unit)
                 throws InterruptedException {
@@ -1103,7 +1097,7 @@
          * IllegalMonitorStateException} is thrown.
          *
          * @throws IllegalMonitorStateException if the current thread does not
-         * hold this lock.
+         * hold this lock
          */
         public void unlock() {
             sync.release(1);
@@ -1231,7 +1225,7 @@
      * Queries the number of read locks held for this lock. This
      * method is designed for use in monitoring system state, not for
      * synchronization control.
-     * @return the number of read locks held.
+     * @return the number of read locks held
      */
     public int getReadLockCount() {
         return sync.getReadLockCount();
diff --git a/luni/src/main/java/java/util/concurrent/package-info.java b/luni/src/main/java/java/util/concurrent/package-info.java
index 155d1b8..51a29e8 100644
--- a/luni/src/main/java/java/util/concurrent/package-info.java
+++ b/luni/src/main/java/java/util/concurrent/package-info.java
@@ -5,7 +5,7 @@
  */
 
 // BEGIN android-note
-// omit links to ForkJoinPool, ForkJoinTask, LinkedTransferQueue, PHaser, TransferQueue
+// omit links to ForkJoinPool, ForkJoinTask, LinkedTransferQueue, Phaser, TransferQueue
 // END android-note
 
 /**
@@ -23,7 +23,7 @@
  *
  * {@link java.util.concurrent.Executor} is a simple standardized
  * interface for defining custom thread-like subsystems, including
- * thread pools, asynchronous IO, and lightweight task frameworks.
+ * thread pools, asynchronous I/O, and lightweight task frameworks.
  * Depending on which concrete Executor class is being used, tasks may
  * execute in a newly created thread, an existing task-execution thread,
  * or the thread calling {@link java.util.concurrent.Executor#execute
@@ -174,7 +174,7 @@
  * it may (or may not) reflect any updates since the iterator was
  * created.
  *
- * <h2><a name="MemoryVisibility">Memory Consistency Properties</a></h2>
+ * <h2 id="MemoryVisibility">Memory Consistency Properties</h2>
  *
  * <a href="http://java.sun.com/docs/books/jls/third_edition/html/memory.html">
  * Chapter 17 of the Java Language Specification</a> defines the
@@ -243,8 +243,7 @@
  *   in each thread <i>happen-before</i> those subsequent to the
  *   corresponding {@code exchange()} in another thread.
  *
- *   <li>Actions prior to calling {@code CyclicBarrier.await} and
- *   {@code Phaser.awaitAdvance} (as well as its variants)
+ *   <li>Actions prior to calling {@code CyclicBarrier.await}
  *   <i>happen-before</i> actions performed by the barrier action, and
  *   actions performed by the barrier action <i>happen-before</i> actions
  *   subsequent to a successful return from the corresponding {@code await}