[ViewVC] Diff of: jsr166/jsr166/src/jsr166y/ForkJoinPool.java

Comparing jsr166/src/jsr166y/ForkJoinPool.java (file contents):
Revision 1.40 by jsr166, Mon Aug 3 00:53:15 2009 UTC vs.
Revision 1.130 by jsr166, Mon Aug 13 18:25:53 2012 UTC

+/*
+ * 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/licenses/publicdomain
->
+ * http://creativecommons.org/publicdomain/zero/1.0/
+ */
+package jsr166y;
-–
-–
+import java.util.concurrent.*;
-–
+import java.util.ArrayList;
+import java.util.Arrays;
+import java.util.Collection;
+import java.util.Collections;
+import java.util.List;
-<
+import java.util.concurrent.locks.Condition;
-<
+import java.util.concurrent.locks.LockSupport;
-<
+import java.util.concurrent.locks.ReentrantLock;
->
+import java.util.Random;
->
+import java.util.concurrent.AbstractExecutorService;
->
+import java.util.concurrent.Callable;
->
+import java.util.concurrent.ExecutorService;
->
+import java.util.concurrent.Future;
->
+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.atomic.AtomicLong;
-+
+import java.util.concurrent.locks.AbstractQueuedSynchronizer;
-+
+import java.util.concurrent.locks.Condition;
+/**
+ * An {@link ExecutorService} for running {@link ForkJoinTask}s.
+ * A {@code ForkJoinPool} provides the entry point for submissions
-<
+ * from non-{@code ForkJoinTask}s, as well as management and
-<
+ * monitoring operations.  Normally a single {@code ForkJoinPool} is
-<
+ * used for a large number of submitted tasks. Otherwise, use would
-<
+ * not usually outweigh the construction and bookkeeping overhead of
-<
+ * creating a large set of threads.
->
+ * from non-{@code ForkJoinTask} clients, as well as management and
->
+ * monitoring operations.
-<
+ * <p>{@code ForkJoinPool}s differ from other kinds of {@link
-<
+ * Executor}s mainly in that they provide <em>work-stealing</em>: all
-<
+ * threads in the pool attempt to find and execute subtasks created by
-<
+ * other active tasks (eventually blocking if none exist). This makes
-<
+ * them efficient when most tasks spawn other subtasks (as do most
-<
+ * {@code ForkJoinTask}s), as well as the mixed execution of some
-<
+ * plain {@code Runnable}- or {@code Callable}- based activities along
-<
+ * with {@code ForkJoinTask}s. When setting {@linkplain #setAsyncMode
-<
+ * async mode}, a {@code ForkJoinPool} may also be appropriate for use
-<
+ * with fine-grained tasks that are never joined. Otherwise, other
-<
+ * {@code ExecutorService} implementations are typically more
-<
+ * appropriate choices.
->
+ * <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 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} may be constructed with a given
-<
+ * parallelism level (target pool size), which it attempts to maintain
-<
+ * by dynamically adding, suspending, or resuming threads, even if
-<
+ * some tasks are waiting to join others. However, no such adjustments
-<
+ * are performed in the face of blocked IO or other unmanaged
-<
+ * synchronization. The nested {@link ManagedBlocker} interface
-<
+ * enables extension of the kinds of synchronization accommodated.
-<
+ * The target parallelism level may also be changed dynamically
-<
+ * ({@link #setParallelism}) and thread construction can be limited
-<
+ * using methods {@link #setMaximumPoolSize} and/or {@link
-<
+ * #setMaintainsParallelism}.
->
+ * <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
->
+ * synchronization accommodated.
+ * <p>In addition to execution and lifecycle control methods, this
+ * class provides status check methods (for example
+ * {@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 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 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>
-+
+ *    <td></td>
-+
+ *    <td ALIGN=CENTER> <b>Call from non-fork/join clients</b></td>
-+
+ *    <td ALIGN=CENTER> <b>Call from within fork/join computations</b></td>
-+
+ *  </tr>
-+
+ *  <tr>
-+
+ *    <td> <b>Arrange async execution</td>
-+
+ *    <td> {@link #execute(ForkJoinTask)}</td>
-+
+ *    <td> {@link ForkJoinTask#fork}</td>
-+
+ *  </tr>
-+
+ *  <tr>
-+
+ *    <td> <b>Await and obtain result</td>
-+
+ *    <td> {@link #invoke(ForkJoinTask)}</td>
-+
+ *    <td> {@link ForkJoinTask#invoke}</td>
-+
+ *  </tr>
-+
+ *  <tr>
-+
+ *    <td> <b>Arrange exec and obtain Future</td>
-+
+ *    <td> {@link #submit(ForkJoinTask)}</td>
-+
+ *    <td> {@link ForkJoinTask#fork} (ForkJoinTasks <em>are</em> Futures)</td>
-+
+ *  </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><b>Implementation notes</b>: This implementation restricts the
+ * maximum number of running threads to 32767. Attempts to create
-<
+ * pools with greater than the maximum result in
->
+ * pools with greater than the maximum number result in
+ * {@code IllegalArgumentException}.
-+
+ * <p>This implementation rejects submitted tasks (that is, by throwing
-+
+ * {@link RejectedExecutionException}) only when the pool is shut down
-+
+ * or internal resources have been exhausted.
-+
+ *
+ * @since 1.7
+ * @author Doug Lea
+ */
+public class ForkJoinPool extends AbstractExecutorService {
+    /*
-<
+     * See the extended comments interspersed below for design,
-<
+     * rationale, and walkthroughs.
->
+     * Implementation Overview
->
+     *
->
+     * 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 loosely 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 never
->
+     * take tasks, and they are multiplexed on to a finite number of
->
+     * shared work queues. However, classes are set up so that future
->
+     * extensions could allow submitters to optionally help perform
->
+     * tasks as well. 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 runState), 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 two volatile variables
->
+     * that are by far most often read (not written) as status and
->
+     * consistency checks.
->
+     *
->
+     * 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.
->
+     *
->
+     * Field "runState" contains 32 bits needed to register and
->
+     * deregister WorkQueues, as well as to enable shutdown. It is
->
+     * only modified under a lock (normally briefly held, but
->
+     * occasionally protecting allocations and resizings) but even
->
+     * when locked remains available to check consistency.
->
+     *
->
+     * Recording WorkQueues.  WorkQueues are recorded in the
->
+     * "workQueues" array that is created upon pool construction 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. Shared (submission) queues
->
+     * are at even indices, worker queues at odd indices. Grouping
->
+     * them together in this way simplifies and speeds up task
->
+     * scanning.
->
+     *
->
+     * 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. 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 (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 fewer than two 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;
->
+     * together these cover the signals needed in cases when more
->
+     * tasks are pushed but untaken, and improve performance compared
->
+     * to having one thread wake up all workers.
->
+     *
->
+     * 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.
->
+     *
->
+     * Shutdown and Termination. A call to shutdownNow atomically sets
->
+     * a runState bit and then (non-atomically) sets each worker's
->
+     * runState 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.
->
+     *
->
+     * 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.
->
+     *
->
+     *   Compensating: Unless there are already enough live threads,
->
+     *      method tryCompensate() may create or re-activate a spare
->
+     *      thread to compensate for blocked joiners until they unblock.
->
+     *
->
+     * A third form (implemented in tryRemoveAndExec and
->
+     * tryPollForAndExec) 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.  A stealHint 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.
->
+     *
->
+     * 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, 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: (1) We only try
->
+     * compensation after attempting enough helping steps (measured
->
+     * via counting and timing) that we have already consumed the
->
+     * estimated cost of creating and activating a new thread.  (2) We
->
+     * allow up to 50% of threads to be blocked before initially
->
+     * adding any others, and unless completely saturated, check that
->
+     * some work is available for a new worker before adding. Also, we
->
+     * create up to only 50% more threads until entering a mode that
->
+     * only adds a thread if all others are possibly blocked.  All
->
+     * together, this means that we might be half as fast to react,
->
+     * and create half as many threads as possible in the ideal case,
->
+     * but present vastly fewer anomalies in all other cases compared
->
+     * to both more aggressive and more conservative alternatives.
->
+     *
->
+     * 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. 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 that help
->
+     * some methods perform reasonably even when interpreted (not
->
+     * compiled).
->
+     *
->
+     * 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
+     */
-<
+    /** Mask for packing and unpacking shorts */
-<
+    private static final int  shortMask = 0xffff;
->
+    // Static utilities
-<
+    /** Max pool size -- must be a power of two minus 1 */
-<
+    private static final int MAX_THREADS =  0x7FFF;
->
+    /**
->
+     * 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.
+         * Returns a new worker thread operating in the given pool.
+         * @param pool the pool this thread works in
-<
+         * @throws NullPointerException if pool is null
->
+         * @throws NullPointerException if the pool is null
+         */
+        public ForkJoinWorkerThread newThread(ForkJoinPool pool);
+     * Default ForkJoinWorkerThreadFactory implementation; creates a
+     * new ForkJoinWorkerThread.
+     */
-<
+    static class  DefaultForkJoinWorkerThreadFactory
->
+    static class DefaultForkJoinWorkerThreadFactory
+        implements ForkJoinWorkerThreadFactory {
+        public ForkJoinWorkerThread newThread(ForkJoinPool pool) {
-<
+            try {
-<
+                return new ForkJoinWorkerThread(pool);
-<
+            } catch (OutOfMemoryError oom)  {
->
+            return new ForkJoinWorkerThread(pool);
->
+        }
->
+    }
->
->
+    /**
->
+     * A simple non-reentrant lock used for exclusion when managing
->
+     * queues and workers. We use a custom lock so that we can readily
->
+     * probe lock state in constructions that check among alternative
->
+     * actions. The lock is normally only very briefly held, and
->
+     * sometimes treated as a spinlock, but other usages block to
->
+     * reduce overall contention in those cases where locked code
->
+     * bodies perform allocation/resizing.
->
+     */
->
+    static final class Mutex extends AbstractQueuedSynchronizer {
->
+        public final boolean tryAcquire(int ignore) {
->
+            return compareAndSetState(0, 1);
->
+        }
->
+        public final boolean tryRelease(int ignore) {
->
+            setState(0);
->
+            return true;
->
+        }
->
+        public final void lock() { acquire(0); }
->
+        public final void unlock() { release(0); }
->
+        public final boolean isHeldExclusively() { return getState() == 1; }
->
+        public final Condition newCondition() { return new ConditionObject(); }
->
+    }
->
->
+    /**
->
+     * 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> {
->
+        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 trySharedPush, 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 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. (Although we can avoid one case
->
+     * of this when locked in trySharedPush.) 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 (via method growArray) 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
->
->
+        volatile long totalSteals; // cumulative number of steals
->
+        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 rescans;               // remaining scans until block
->
+        int nsteals;               // top-level task executions since last idle
->
+        final int mode;            // lifo, fifo, or shared
->
+        int poolIndex;             // index of this queue in pool (or 0)
->
+        int stealHint;             // index of most recent known stealer
->
+        volatile int runState;     // 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
->
+        // Heuristic padding to ameliorate unfortunate memory placements
->
+        Object p00, p01, p02, p03, p04, p05, p06, p07;
->
+        Object p08, p09, p0a, p0b, p0c, p0d, p0e;
->
->
+        WorkQueue(ForkJoinPool pool, ForkJoinWorkerThread owner, int mode) {
->
+            this.mode = mode;
->
+            this.pool = pool;
->
+            this.owner = owner;
->
+            // 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.getObjectVolatile
->
+                      (a, ((m & (s - 1)) << ASHIFT) + ABASE) == null)));
->
+        }
->
->
+        /**
->
+         * Pushes a task. Call only by owner in unshared queues.
->
+         *
->
+         * @param task the task. Caller must ensure non-null.
->
+         * @throw 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
->
+                U.putOrderedObject
->
+                    (a, (((m = a.length - 1) & s) << ASHIFT) + ABASE, task);
->
+                if ((n = (top = s + 1) - base) <= 2) {
->
+                    if ((p = pool) != null)
->
+                        p.signalWork();
->
+                }
->
+                else if (n >= m)
->
+                    growArray(true);
->
+            }
->
+        }
->
->
+        /**
->
+         * Pushes a task if lock is free and array is either big
->
+         * enough or can be resized to be big enough.
->
+         *
->
+         * @param task the task. Caller must ensure non-null.
->
+         * @return true if submitted
->
+         */
->
+        final boolean trySharedPush(ForkJoinTask<?> task) {
->
+            boolean submitted = false;
->
+            if (runState == 0 && U.compareAndSwapInt(this, RUNSTATE, 0, 1)) {
->
+                ForkJoinTask<?>[] a = array;
->
+                int s = top;
->
+                try {
->
+                    if ((a != null && a.length > s + 1 - base) ||
->
+                        (a = growArray(false)) != null) { // must presize
->
+                        int j = (((a.length - 1) & s) << ASHIFT) + ABASE;
->
+                        U.putObject(a, (long)j, task);    // don't need "ordered"
->
+                        top = s + 1;
->
+                        submitted = true;
->
+                    }
->
+                } finally {
->
+                    runState = 0;                         // unlock
->
+                }
->
+            }
->
+            return submitted;
->
+        }
->
->
+        /**
->
+         * Takes next task, if one exists, in LIFO order.  Call only
->
+         * by owner in unshared queues. (We do not have a shared
->
+         * version of this method because it is never needed.)
->
+         */
->
+        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
->
+                }
->
+            }
->
+            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.
-+
+         */
-+
+        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;
-+
+        }
-+
-+
+        /**
-+
+         * Polls the given task only if it is at the current base.
-+
+         */
-+
+        final boolean pollFor(ForkJoinTask<?> task) {
-+
+            ForkJoinTask<?>[] a; int b;
-+
+            if ((b = base) - top < 0 && (a = array) != null) {
-+
+                int j = (((a.length - 1) & b) << ASHIFT) + ABASE;
-+
+                if (U.getObjectVolatile(a, j) == task && base == b &&
-+
+                    U.compareAndSwapObject(a, j, task, null)) {
-+
+                    base = b + 1;
-+
+                    return true;
-+
+                }
-+
+            }
-+
+            return false;
-+
+        }
-+
-+
+        /**
-+
+         * 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.
-+
+         *
-+
+         * @param rejectOnFailure if true, throw exception if capacity
-+
+         * exceeded (relayed ultimately to user); else return null.
-+
+         */
-+
+        final ForkJoinTask<?>[] growArray(boolean rejectOnFailure) {
-+
+            ForkJoinTask<?>[] oldA = array;
-+
+            int size = oldA != null ? oldA.length << 1 : INITIAL_QUEUE_CAPACITY;
-+
+            if (size <= MAXIMUM_QUEUE_CAPACITY) {
-+
+                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;
-+
+            }
-+
+            else if (!rejectOnFailure)
-+
+                return null;
-+
+            else
-+
+                throw new RejectedExecutionException("Queue capacity exceeded");
-+
+        }
-+
-+
+        /**
-+
+         * 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;
-+
+        }
-+
-+
+        // 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) immediately on any CAS
-+
+         * or consistency check failure so caller can retry.
-+
+         *
-+
+         * @return 0 if no progress can be made, else positive
-+
+         * (this unusual convention simplifies use with tryHelpStealer.)
-+
+         */
-+
+        final int tryRemoveAndExec(ForkJoinTask<?> task) {
-+
+            int stat = 1;
-+
+            boolean 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 = 0;
-+
+                        break;
-+
+                    }
-+
+                }
-+
+            }
-+
+            if (removed)
-+
+                task.doExec();
-+
+            return stat;
-+
+        }
-+
-+
+        /**
-+
+         * Executes a top-level task and any local tasks remaining
-+
+         * after execution.
-+
+         */
-+
+        final void runTask(ForkJoinTask<?> t) {
-+
+            if (t != null) {
-+
+                currentSteal = t;
-+
+                t.doExec();
-+
+                if (top != base) {       // process remaining local tasks
-+
+                    if (mode == 0)
-+
+                        popAndExecAll();
-+
+                    else
-+
+                        pollAndExecAll();
-+
+                }
-+
+                ++nsteals;
-+
+                currentSteal = null;
-+
+            }
-+
+        }
-+
-+
+        /**
-+
+         * Executes a non-top-level (stolen) task.
-+
+         */
-+
+        final void runSubtask(ForkJoinTask<?> t) {
-+
+            if (t != null) {
-+
+                ForkJoinTask<?> ps = currentSteal;
-+
+                currentSteal = t;
-+
+                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);
-+
+        }
-+
-+
+        /**
-+
+         * If this owned and is not already interrupted, try to
-+
+         * interrupt and/or unpark, ignoring exceptions.
-+
+         */
-+
+        final void interruptOwner() {
-+
+            Thread wt, p;
-+
+            if ((wt = owner) != null && !wt.isInterrupted()) {
-+
+                try {
-+
+                    wt.interrupt();
-+
+                } catch (SecurityException ignore) {
-+
+                }
-+
+            }
-+
+            if ((p = parker) != null)
-+
+                U.unpark(p);
-+
+        }
-+
-+
+        // Unsafe mechanics
-+
+        private static final sun.misc.Unsafe U;
-+
+        private static final long RUNSTATE;
-+
+        private static final int ABASE;
-+
+        private static final int ASHIFT;
-+
+        static {
-+
+            int s;
-+
+            try {
-+
+                U = getUnsafe();
-+
+                Class<?> k = WorkQueue.class;
-+
+                Class<?> ak = ForkJoinTask[].class;
-+
+                RUNSTATE = U.objectFieldOffset
-+
+                    (k.getDeclaredField("runState"));
-+
+                ABASE = U.arrayBaseOffset(ak);
-+
+                s = U.arrayIndexScale(ak);
-+
+            } 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);
-+
+        }
-+
+    }
-+
+    /**
-+
+     * 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 doSubmit. 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. This is done during
-+
+     * registration for workers, but requires a separate AtomicInteger
-+
+     * for submitters. Seeds are then randomly modified upon
-+
+     * collisions using xorshifts, which requires a non-zero seed.
-+
+     */
-+
+    static final class Submitter {
-+
+        int seed;
-+
+        Submitter() {
-+
+            int s = nextSubmitterSeed.getAndAdd(SEED_INCREMENT);
-+
+            seed = (s == 0) ? 1 : s; // ensure non-zero
-+
+    /** ThreadLocal class for Submitters */
-+
+    static final class ThreadSubmitter extends ThreadLocal<Submitter> {
-+
+        public Submitter initialValue() { return new Submitter(); }
-+
+    }
-+
-+
+    // static fields (initialized in static initializer below)
-+
+    /**
+     * Creates a new ForkJoinWorkerThread. This factory is used unless
+     * overridden in ForkJoinPool constructors.
+     */
+    public static final ForkJoinWorkerThreadFactory
-<
+        defaultForkJoinWorkerThreadFactory =
-<
+        new DefaultForkJoinWorkerThreadFactory();
->
+        defaultForkJoinWorkerThreadFactory;
+    /**
-<
+     * Permission required for callers of methods that may start or
-<
+     * kill threads.
->
+     * Generator for assigning sequence numbers as pool names.
+     */
-<
+    private static final RuntimePermission modifyThreadPermission =
-<
+        new RuntimePermission("modifyThread");
->
+    private static final AtomicInteger poolNumberGenerator;
+    /**
-<
+     * If there is a security manager, makes sure caller has
-<
+     * permission to modify threads.
->
+     * Generator for initial hashes/seeds for submitters. Accessed by
->
+     * Submitter class constructor.
+     */
-<
+    private static void checkPermission() {
-<
+        SecurityManager security = System.getSecurityManager();
-<
+        if (security != null)
-<
+            security.checkPermission(modifyThreadPermission);
-<
+    }
->
+    static final AtomicInteger nextSubmitterSeed;
+    /**
-<
+     * Generator for assigning sequence numbers as pool names.
->
+     * Permission required for callers of methods that may start or
->
+     * kill threads.
+     */
-<
+    private static final AtomicInteger poolNumberGenerator =
-<
+        new AtomicInteger();
->
+    private static final RuntimePermission modifyThreadPermission;
+    /**
-<
+     * Array holding all worker threads in the pool. Initialized upon
-<
+     * first use. Array size must be a power of two.  Updates and
-<
+     * replacements are protected by workerLock, but it is always kept
-<
+     * in a consistent enough state to be randomly accessed without
-<
+     * locking by workers performing work-stealing.
->
+     * 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.
+     */
-<
+    volatile ForkJoinWorkerThread[] workers;
->
+    private static final ThreadSubmitter submitters;
->
->
+    // static constants
+    /**
-<
+     * Lock protecting access to workers.
->
+     * 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 final ReentrantLock workerLock;
->
+    private static final long SHRINK_RATE =
->
+L * 1000L * 1000L * 1000L; // 4 seconds
+    /**
-<
+     * Condition for awaitTermination.
->
+     * The timeout value for attempted shrinkage, includes
->
+     * some slop to cope with system timer imprecision.
+     */
-<
+    private final Condition termination;
->
+    private static final long SHRINK_TIMEOUT = SHRINK_RATE - (SHRINK_RATE / 10);
+    /**
-<
+     * The uncaught exception handler used when any worker
-<
+     * abruptly terminates
->
+     * The maximum stolen->joining link depth allowed in method
->
+     * tryHelpStealer.  Must be a power of two. This value also
->
+     * controls the maximum number of times to try to help join a task
->
+     * without any apparent progress or change in pool state before
->
+     * giving up and blocking (see awaitJoin).  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 Thread.UncaughtExceptionHandler ueh;
->
+    private static final int MAX_HELP = 64;
+    /**
-<
+     * Creation factory for worker threads.
->
+     * Secondary time-based bound (in nanosecs) for helping attempts
->
+     * before trying compensated blocking in awaitJoin. Used in
->
+     * conjunction with MAX_HELP to reduce variance due to different
->
+     * polling rates associated with different helping options. The
->
+     * value should roughly approximate the time required to create
->
+     * and/or activate a worker thread.
+     */
-<
+    private final ForkJoinWorkerThreadFactory factory;
->
+    private static final long COMPENSATION_DELAY = 1L << 18; // ~0.25 millisec
+    /**
-<
+     * Head of stack of threads that were created to maintain
-<
+     * parallelism when other threads blocked, but have since
-<
+     * suspended when the parallelism level rose.
->
+     * Increment for seed generators. See class ThreadLocal for
->
+     * explanation.
+     */
-<
+    private volatile WaitQueueNode spareStack;
->
+    private static final int SEED_INCREMENT = 0x61c88647;
+    /**
-<
+     * Sum of per-thread steal counts, updated only when threads are
-<
+     * idle or terminating.
->
+     * 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 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 =
->
+     * (int)ctl.  The ec field is never accessed alone, but always
->
+     * together with id and st. The offsets of counts by the target
->
+     * 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, 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 runState is an int packed with:
->
+     * SHUTDOWN: true if shutdown is enabled (1 bit)
->
+     * SEQ:  a sequence number updated upon (de)registering workers (30 bits)
->
+     * INIT: set true after workQueues array construction (1 bit)
->
+     *
->
+     * The sequence number enables simple consistency checks:
->
+     * Staleness of read-only operations on the workQueues array can
->
+     * be checked by comparing runState before vs after the reads.
+     */
-<
+    private final AtomicLong stealCount;
->
->
+    // bit positions/shifts for fields
->
+    private static final int  AC_SHIFT   = 48;
->
+    private static final int  TC_SHIFT   = 32;
->
+    private static final int  ST_SHIFT   = 31;
->
+    private static final int  EC_SHIFT   = 16;
->
->
+    // bounds
->
+    private static final int  SMASK      = 0xffff;  // short bits
->
+    private static final int  MAX_CAP    = 0x7fff;  // max #workers - 1
->
+    private static final int  SQMASK     = 0xfffe;  // even short bits
->
+    private static final int  SHORT_SIGN = 1 << 15;
->
+    private static final int  INT_SIGN   = 1 << 31;
->
->
+    // masks
->
+    private static final long STOP_BIT   = 0x0001L << ST_SHIFT;
->
+    private static final long AC_MASK    = ((long)SMASK) << AC_SHIFT;
->
+    private static final long TC_MASK    = ((long)SMASK) << TC_SHIFT;
->
->
+    // units for incrementing and decrementing
->
+    private static final long TC_UNIT    = 1L << TC_SHIFT;
->
+    private static final long AC_UNIT    = 1L << AC_SHIFT;
->
->
+    // masks and units for dealing with u = (int)(ctl >>> 32)
->
+    private static final int  UAC_SHIFT  = AC_SHIFT - 32;
->
+    private static final int  UTC_SHIFT  = TC_SHIFT - 32;
->
+    private static final int  UAC_MASK   = SMASK << UAC_SHIFT;
->
+    private static final int  UTC_MASK   = SMASK << UTC_SHIFT;
->
+    private static final int  UAC_UNIT   = 1 << UAC_SHIFT;
->
+    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 E_SEQ       = 1 << EC_SHIFT;
->
->
+    // runState bits
->
+    private static final int SHUTDOWN    = 1 << 31;
->
->
+    // access mode for WorkQueue
->
+    static final int LIFO_QUEUE          =  0;
->
+    static final int FIFO_QUEUE          =  1;
->
+    static final int SHARED_QUEUE        = -1;
->
->
+    // Instance fields
->
->
+    /*
->
+     * Field layout order in 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.
->
+     */
->
->
+    volatile long ctl;                         // main pool control
->
+    final int parallelism;                     // parallelism level
->
+    final int localMode;                       // per-worker scheduling mode
->
+    final int submitMask;                      // submit queue index bound
->
+    int nextSeed;                              // for initializing worker seeds
->
+    volatile int runState;                     // shutdown status and seq
->
+    WorkQueue[] workQueues;                    // main registry
->
+    final Mutex lock;                          // for registration
->
+    final Condition termination;               // for awaitTermination
->
+    final ForkJoinWorkerThreadFactory factory; // factory for new workers
->
+    final Thread.UncaughtExceptionHandler ueh; // per-worker UEH
->
+    final AtomicLong stealCount;               // collect counts when terminated
->
+    final AtomicInteger nextWorkerNumber;      // to create worker name string
->
+    final String workerNamePrefix;             // to create worker name string
->
->
+    //  Creating, registering, and deregistering workers
->
->
+    /**
->
+     * Tries to create and start a worker
->
+     */
->
+    private void addWorker() {
->
+        Throwable ex = null;
->
+        ForkJoinWorkerThread wt = null;
->
+        try {
->
+            if ((wt = factory.newThread(this)) != null) {
->
+                wt.start();
->
+                return;
->
+            }
->
+        } catch (Throwable e) {
->
+            ex = e;
->
+        }
->
+        deregisterWorker(wt, ex); // adjust counts etc on failure
->
+    }
+    /**
-<
+     * Queue for external submissions.
->
+     * Callback from ForkJoinWorkerThread constructor to assign a
->
+     * public name. This must be separate from registerWorker because
->
+     * it is called during the "super" constructor call in
->
+     * ForkJoinWorkerThread.
+     */
-<
+    private final LinkedTransferQueue<ForkJoinTask<?>> submissionQueue;
->
+    final String nextWorkerName() {
->
+        return workerNamePrefix.concat
->
+            (Integer.toString(nextWorkerNumber.addAndGet(1)));
->
+    }
+    /**
-<
+     * Head of Treiber stack for barrier sync. See below for explanation.
->
+     * Callback from ForkJoinWorkerThread constructor to establish its
->
+     * poolIndex 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's queue
+     */
-<
+    private volatile WaitQueueNode syncStack;
->
->
+    final void registerWorker(WorkQueue w) {
->
+        Mutex lock = this.lock;
->
+        lock.lock();
->
+        try {
->
+            WorkQueue[] ws = workQueues;
->
+            if (w != null && ws != null) {          // skip on shutdown/failure
->
+                int rs, n =  ws.length, m = n - 1;
->
+                int s = nextSeed += SEED_INCREMENT; // rarely-colliding sequence
->
+                w.seed = (s == 0) ? 1 : s;          // ensure non-zero seed
->
+                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) & SQMASK) + 2;
->
+                    while (ws[r = (r + step) & m] != null) {
->
+                        if (++probes >= n) {
->
+                            workQueues = ws = Arrays.copyOf(ws, n <<= 1);
->
+                            m = n - 1;
->
+                            probes = 0;
->
+                        }
->
+                    }
->
+                }
->
+                w.eventCount = w.poolIndex = r;     // establish before recording
->
+                ws[r] = w;                          // also update seq
->
+                runState = ((rs = runState) & SHUTDOWN) | ((rs + 2) & ~SHUTDOWN);
->
+            }
->
+        } finally {
->
+            lock.unlock();
->
+        }
->
+    }
+    /**
-<
+     * The count for event barrier
-<
+     */
-<
+    private volatile long eventCount;
->
+     * Final callback from terminating worker, as well as upon failure
->
+     * to construct or start a worker in addWorker.  Removes record of
->
+     * worker from array, and adjusts counts. If pool is shutting
->
+     * down, tries to complete termination.
->
+     *
->
+     * @param wt the worker thread or null if addWorker failed
->
+     * @param ex the exception causing failure, or null if none
->
+     */
->
+    final void deregisterWorker(ForkJoinWorkerThread wt, Throwable ex) {
->
+        Mutex lock = this.lock;
->
+        WorkQueue w = null;
->
+        if (wt != null && (w = wt.workQueue) != null) {
->
+            w.runState = -1;                // ensure runState is set
->
+            stealCount.getAndAdd(w.totalSteals + w.nsteals);
->
+            int idx = w.poolIndex;
->
+            lock.lock();
->
+            try {                           // remove record from array
->
+                WorkQueue[] ws = workQueues;
->
+                if (ws != null && idx >= 0 && idx < ws.length && ws[idx] == w)
->
+                    ws[idx] = null;
->
+            } finally {
->
+                lock.unlock();
->
+            }
->
+        }
->
->
+        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.cancelAll();                  // cancel remaining tasks
->
+            if (w.array != null)            // suppress signal if never ran
->
+                signalWork();               // wake up or create replacement
->
+            if (ex == null)                 // help clean refs on way out
->
+                ForkJoinTask.helpExpungeStaleExceptions();
->
+        }
->
->
+        if (ex != null)                     // rethrow
->
+            U.throwException(ex);
->
+    }
->
->
->
+    // Submissions
->
->
+    /**
->
+     * Unless shutting down, adds the given task to a submission queue
->
+     * at submitter's current queue index (modulo submission
->
+     * range). If no queue exists at the index, one is created.  If
->
+     * the queue is busy, another index is randomly chosen. The
->
+     * submitMask bounds the effective number of queues to the
->
+     * (nearest power of two for) parallelism level.
->
+     *
->
+     * @param task the task. Caller must ensure non-null.
->
+     */
->
+    private void doSubmit(ForkJoinTask<?> task) {
->
+        Submitter s = submitters.get();
->
+        for (int r = s.seed, m = submitMask;;) {
->
+            WorkQueue[] ws; WorkQueue q;
->
+            int k = r & m & SQMASK;          // use only even indices
->
+            if (runState < 0 || (ws = workQueues) == null || ws.length <= k)
->
+                throw new RejectedExecutionException(); // shutting down
->
+            else if ((q = ws[k]) == null) {  // create new queue
->
+                WorkQueue nq = new WorkQueue(this, null, SHARED_QUEUE);
->
+                Mutex lock = this.lock;      // construct outside lock
->
+                lock.lock();
->
+                try {                        // recheck under lock
->
+                    int rs = runState;       // to update seq
->
+                    if (ws == workQueues && ws[k] == null) {
->
+                        ws[k] = nq;
->
+                        runState = ((rs & SHUTDOWN) | ((rs + 2) & ~SHUTDOWN));
->
+                    }
->
+                } finally {
->
+                    lock.unlock();
->
+                }
->
+            }
->
+            else if (q.trySharedPush(task)) {
->
+                signalWork();
->
+                return;
->
+            }
->
+            else if (m > 1) {                // move to a different index
->
+                r ^= r << 13;                // same xorshift as WorkQueues
->
+                r ^= r >>> 17;
->
+                s.seed = r ^= r << 5;
->
+            }
->
+            else
->
+                Thread.yield();              // yield if no alternatives
->
+        }
->
+    }
->
->
+    // Maintaining ctl counts
+    /**
-<
+     * Pool number, just for assigning useful names to worker threads
->
+     * Increments active count; mainly called upon return from blocking.
+     */
-<
+    private final int poolNumber;
->
+    final void incrementActiveCount() {
->
+        long c;
->
+        do {} while (!U.compareAndSwapLong(this, CTL, c = ctl, c + AC_UNIT));
->
+    }
+    /**
-<
+     * The maximum allowed pool size
->
+     * Tries to activate or create a worker if too few are active.
+     */
-<
+    private volatile int maxPoolSize;
->
+    final void signalWork() {
->
+        long c; int u;
->
+        while ((u = (int)((c = ctl) >>> 32)) < 0) {     // too few active
->
+            WorkQueue[] ws = workQueues; int e, i; WorkQueue w; Thread p;
->
+            if ((e = (int)c) > 0) {                     // at least one waiting
->
+                if (ws != null && (i = e & SMASK) < ws.length &&
->
+                    (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.eventCount = (e + E_SEQ) & E_MASK;
->
+                        if ((p = w.parker) != null)
->
+                            U.unpark(p);                // activate and release
->
+                        break;
->
+                    }
->
+                }
->
+                else
->
+                    break;
->
+            }
->
+            else if (e == 0 && (u & SHORT_SIGN) != 0) { // too few total
->
+                long nc = (long)(((u + UTC_UNIT) & UTC_MASK) |
->
+                                 ((u + UAC_UNIT) & UAC_MASK)) << 32;
->
+                if (U.compareAndSwapLong(this, CTL, c, nc)) {
->
+                    addWorker();
->
+                    break;
->
+                }
->
+            }
->
+            else
->
+                break;
->
+        }
->
+    }
->
->
+    // Scanning for tasks
+    /**
-<
+     * The desired parallelism level, updated only under workerLock.
->
+     * Top-level runloop for workers, called by ForkJoinWorkerThread.run.
+     */
-<
+    private volatile int parallelism;
->
+    final void runWorker(WorkQueue w) {
->
+        w.growArray(false);         // initialize queue array in this thread
->
+        do { w.runTask(scan(w)); } while (w.runState >= 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 a random permutation of
->
+     * queues (starting at a random index and stepping by a random
->
+     * relative prime, checking each at least once).  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.  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 a complete sweep, try to release a waiting worker.  If
->
+     * the scan terminated because the worker is inactivated, then the
->
+     * released worker will often be the calling worker, and it can
->
+     * succeed obtaining a task on the next call. Or maybe it is
->
+     * another worker, but with same net effect. Releasing in other
->
+     * cases as well ensures that we have enough workers running.
->
+     *
->
+     * * 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 check for
->
+     * termination and possibly shrink pool.
->
+     *
->
+     * * If already inactive, and the caller has run a task since the
->
+     * last empty scan, return (to allow rescan) unless others are
->
+     * also inactivated.  Field WorkQueue.rescans counts down on each
->
+     * scan to ensure eventual inactivation and blocking.
->
+     *
->
+     * * If already enqueued and none of the above apply, park
->
+     * awaiting signal,
->
+     *
->
+     * @param w the worker (via its WorkQueue)
->
+     * @return a task or null of none found
->
+     */
->
+    private final ForkJoinTask<?> scan(WorkQueue w) {
->
+        WorkQueue[] ws;                       // first update random seed
->
+        int r = w.seed; r ^= r << 13; r ^= r >>> 17; w.seed = r ^= r << 5;
->
+        int rs = runState, m;                 // volatile read order matters
->
+        if ((ws = workQueues) != null && (m = ws.length - 1) > 0) {
->
+            int ec = w.eventCount;            // ec is negative if inactive
->
+            int step = (r >>> 16) | 1;        // relative prime
->
+            for (int j = (m + 1) << 2; ; r += step) {
->
+                WorkQueue q; ForkJoinTask<?> t; ForkJoinTask<?>[] a; int b;
->
+                if ((q = ws[r & m]) != null && (b = q.base) - q.top < 0 &&
->
+                    (a = q.array) != null) {  // probably nonempty
->
+                    int i = (((a.length - 1) & b) << ASHIFT) + ABASE;
->
+                    t = (ForkJoinTask<?>)U.getObjectVolatile(a, i);
->
+                    if (q.base == b && ec >= 0 && t != null &&
->
+                        U.compareAndSwapObject(a, i, t, null)) {
->
+                        if (q.top - (q.base = b + 1) > 1)
->
+                            signalWork();    // help pushes signal
->
+                        return t;
->
+                    }
->
+                    else if (ec < 0 || j <= m) {
->
+                        rs = 0;               // mark scan as imcomplete
->
+                        break;                // caller can retry after release
->
+                    }
->
+                }
->
+                if (--j < 0)
->
+                    break;
->
+            }
->
->
+            long c = ctl; int e = (int)c, a = (int)(c >> AC_SHIFT), nr, ns;
->
+            if (e < 0)                        // decode ctl on empty scan
->
+                w.runState = -1;              // pool is terminating
->
+            else if (rs == 0 || rs != runState) { // incomplete scan
->
+                WorkQueue v; Thread p;        // try to release a waiter
->
+                if (e > 0 && a < 0 && w.eventCount == ec &&
->
+                    (v = ws[e & m]) != null && v.eventCount == (e | INT_SIGN)) {
->
+                    long nc = ((long)(v.nextWait & E_MASK) |
->
+                               ((c + AC_UNIT) & (AC_MASK|TC_MASK)));
->
+                    if (ctl == c && U.compareAndSwapLong(this, CTL, c, nc)) {
->
+                        v.eventCount = (e + E_SEQ) & E_MASK;
->
+                        if ((p = v.parker) != null)
->
+                            U.unpark(p);
->
+                    }
->
+                }
->
+            }
->
+            else if (ec >= 0) {               // try to enqueue/inactivate
->
+                long nc = (long)ec | ((c - AC_UNIT) & (AC_MASK|TC_MASK));
->
+                w.nextWait = e;
->
+                w.eventCount = ec | INT_SIGN; // mark as inactive
->
+                if (ctl != c || !U.compareAndSwapLong(this, CTL, c, nc))
->
+                    w.eventCount = ec;        // unmark on CAS failure
->
+                else {
->
+                    if ((ns = w.nsteals) != 0) {
->
+                        w.nsteals = 0;        // set rescans if ran task
->
+                        w.rescans = (a > 0) ? 0 : a + parallelism;
->
+                        w.totalSteals += ns;
->
+                    }
->
+                    if (a == 1 - parallelism) // quiescent
->
+                        idleAwaitWork(w, nc, c);
->
+                }
->
+            }
->
+            else if (w.eventCount < 0) {      // already queued
->
+                if ((nr = w.rescans) > 0) {   // continue rescanning
->
+                    int ac = a + parallelism;
->
+                    if (((w.rescans = (ac < nr) ? ac : nr - 1) & 3) == 0)
->
+                        Thread.yield();       // yield before block
->
+                }
->
+                else {
->
+                    Thread.interrupted();     // clear status
->
+                    Thread wt = Thread.currentThread();
->
+                    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);
->
+                }
->
+            }
->
+        }
->
+        return null;
->
+    }
+    /**
-<
+     * True if use local fifo, not default lifo, for local polling
-<
+     */
-<
+    private volatile boolean locallyFifo;
->
+     * 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 SHRINK_RATE
->
+     * nanosecs.  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.eventCount < 0 && !tryTerminate(false, false) &&
->
+            (int)prevCtl != 0 && !hasQueuedSubmissions() && ctl == currentCtl) {
->
+            Thread wt = Thread.currentThread();
->
+            Thread.yield();            // yield before block
->
+            while (ctl == currentCtl) {
->
+                long startTime = System.nanoTime();
->
+                Thread.interrupted();  // timed variant of version in scan()
->
+                U.putObject(wt, PARKBLOCKER, this);
->
+                w.parker = wt;
->
+                if (ctl == currentCtl)
->
+                    U.park(false, SHRINK_RATE);
->
+                w.parker = null;
->
+                U.putObject(wt, PARKBLOCKER, null);
->
+                if (ctl != currentCtl)
->
+                    break;
->
+                if (System.nanoTime() - startTime >= SHRINK_TIMEOUT &&
->
+                    U.compareAndSwapLong(this, CTL, currentCtl, prevCtl)) {
->
+                    w.eventCount = (w.eventCount + E_SEQ) | E_MASK;
->
+                    w.runState = -1;   // shrink
->
+                    break;
->
+                }
->
+            }
->
+        }
->
+    }
+    /**
-<
+     * Holds number of total (i.e., created and not yet terminated)
-<
+     * and running (i.e., not blocked on joins or other managed sync)
-<
+     * threads, packed into one int to ensure consistent snapshot when
-<
+     * making decisions about creating and suspending spare
-<
+     * threads. Updated only by CAS.  Note: CASes in
-<
+     * updateRunningCount and preJoin assume that running active count
-<
+     * is in low word, so need to be modified if this changes.
->
+     * 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 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
->
+     */
->
+    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.stealHint | 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.stealHint = 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;
->
+                            }
->
+                        }
->
+                    }
->
+                }
->
+            }
->
+        }
->
+        return stat;
->
+    }
->
->
+    /**
->
+     * If task is at base of some steal queue, steals and executes it.
->
+     *
->
+     * @param joiner the joining worker
->
+     * @param task the task
+     */
-<
+    private volatile int workerCounts;
->
+    private void tryPollForAndExec(WorkQueue joiner, ForkJoinTask<?> task) {
->
+        WorkQueue[] ws;
->
+        if ((ws = workQueues) != null) {
->
+            for (int j = 1; j < ws.length && task.status >= 0; j += 2) {
->
+                WorkQueue q = ws[j];
->
+                if (q != null && q.pollFor(task)) {
->
+                    joiner.runSubtask(task);
->
+                    break;
->
+                }
->
+            }
->
+        }
->
+    }
-<
+    private static int totalCountOf(int s)           { return s >>> 16;  }
-<
+    private static int runningCountOf(int s)         { return s & shortMask; }
-<
+    private static int workerCountsFor(int t, int r) { return (t << 16) + r; }
->
+    /**
->
+     * 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 either
->
+     * pool would become completely starved or: (at least half
->
+     * starved, and fewer than 50% spares exist, and there is at least
->
+     * one task apparently available). Even though the availability
->
+     * check requires a full scan, it is worthwhile in reducing false
->
+     * alarms.
->
+     *
->
+     * @param task if non-null, a task being waited for
->
+     * @param blocker if non-null, a blocker being waited for
->
+     * @return true if the caller can block, else should recheck and retry
->
+     */
->
+    final boolean tryCompensate(ForkJoinTask<?> task, ManagedBlocker blocker) {
->
+        int pc = parallelism, e;
->
+        long c = ctl;
->
+        WorkQueue[] ws = workQueues;
->
+        if ((e = (int)c) >= 0 && ws != null) {
->
+            int u, a, ac, hc;
->
+            int tc = (short)((u = (int)(c >>> 32)) >>> UTC_SHIFT) + pc;
->
+            boolean replace = false;
->
+            if ((a = u >> UAC_SHIFT) <= 0) {
->
+                if ((ac = a + pc) <= 1)
->
+                    replace = true;
->
+                else if ((e > 0 || (task != null &&
->
+                                    ac <= (hc = pc >>> 1) && tc < pc + hc))) {
->
+                    WorkQueue w;
->
+                    for (int j = 0; j < ws.length; ++j) {
->
+                        if ((w = ws[j]) != null && !w.isEmpty()) {
->
+                            replace = true;
->
+                            break;   // in compensation range and tasks available
->
+                        }
->
+                    }
->
+                }
->
+            }
->
+            if ((task == null || task.status >= 0) && // recheck need to block
->
+                (blocker == null || !blocker.isReleasable()) && ctl == c) {
->
+                if (!replace) {          // no compensation
->
+                    long nc = ((c - AC_UNIT) & AC_MASK) | (c & ~AC_MASK);
->
+                    if (U.compareAndSwapLong(this, CTL, c, nc))
->
+                        return true;
->
+                }
->
+                else if (e != 0) {       // release an idle worker
->
+                    WorkQueue w; Thread p; int i;
->
+                    if ((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 | INT_SIGN) &&
->
+                            U.compareAndSwapLong(this, CTL, c, nc)) {
->
+                            w.eventCount = (e + E_SEQ) & E_MASK;
->
+                            if ((p = w.parker) != null)
->
+                                U.unpark(p);
->
+                            return true;
->
+                        }
->
+                    }
->
+                }
->
+                else if (tc < MAX_CAP) { // create replacement
->
+                    long nc = ((c + TC_UNIT) & TC_MASK) | (c & ~TC_MASK);
->
+                    if (U.compareAndSwapLong(this, CTL, c, nc)) {
->
+                        addWorker();
->
+                        return true;
->
+                    }
->
+                }
->
+            }
->
+        }
->
+        return false;
->
+    }
+    /**
-<
+     * Adds delta (which may be negative) to running count.  This must
-<
+     * be called before (with negative arg) and after (with positive)
-<
+     * any managed synchronization (i.e., mainly, joins).
->
+     * Helps and/or blocks until the given task is done.
-<
+     * @param delta the number to add
->
+     * @param joiner the joining worker
->
+     * @param task the task
->
+     * @return task status on exit
+     */
-<
+    final void updateRunningCount(int delta) {
->
+    final int awaitJoin(WorkQueue joiner, ForkJoinTask<?> task) {
+        int s;
-<
+        do {} while (!casWorkerCounts(s = workerCounts, s + delta));
->
+        if ((s = task.status) >= 0) {
->
+            ForkJoinTask<?> prevJoin = joiner.currentJoin;
->
+            joiner.currentJoin = task;
->
+            long startTime = 0L;
->
+            for (int k = 0;;) {
->
+                if ((s = (joiner.isEmpty() ?           // try to help
->
+                          tryHelpStealer(joiner, task) :
->
+                          joiner.tryRemoveAndExec(task))) == 0 &&
->
+                    (s = task.status) >= 0) {
->
+                    if (k == 0) {
->
+                        startTime = System.nanoTime();
->
+                        tryPollForAndExec(joiner, task); // check uncommon case
->
+                    }
->
+                    else if ((k & (MAX_HELP - 1)) == 0 &&
->
+                             System.nanoTime() - startTime >=
->
+                             COMPENSATION_DELAY &&
->
+                             tryCompensate(task, null)) {
->
+                        if (task.trySetSignal()) {
->
+                            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));
->
+                    }
->
+                }
->
+                if (s < 0 || (s = task.status) < 0) {
->
+                    joiner.currentJoin = prevJoin;
->
+                    break;
->
+                }
->
+                else if ((k++ & (MAX_HELP - 1)) == MAX_HELP >>> 1)
->
+                    Thread.yield();                     // for politeness
->
+            }
->
+        }
->
+        return s;
+    /**
-<
+     * Adds delta (which may be negative) to both total and running
-<
+     * count.  This must be called upon creation and termination of
-<
+     * worker threads.
->
+     * 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 delta the number to add
->
+     * @param joiner the joining worker
->
+     * @param task the task
->
+     * @return task status on exit
+     */
-<
+    private void updateWorkerCount(int delta) {
-<
+        int d = delta + (delta << 16); // add to both lo and hi parts
->
+    final int helpJoinOnce(WorkQueue joiner, ForkJoinTask<?> task) {
+        int s;
-<
+        do {} while (!casWorkerCounts(s = workerCounts, s + d));
->
+        while ((s = task.status) >= 0 &&
->
+               (joiner.isEmpty() ?
->
+                tryHelpStealer(joiner, task) :
->
+                joiner.tryRemoveAndExec(task)) != 0)
->
+            ;
->
+        return s;
+    /**
-<
+     * Lifecycle control. High word contains runState, low word
-<
+     * contains the number of workers that are (probably) executing
-<
+     * tasks. This value is atomically incremented before a worker
-<
+     * gets a task to run, and decremented when worker has no tasks
-<
+     * and cannot find any. These two fields are bundled together to
-<
+     * support correct termination triggering.  Note: activeCount
-<
+     * CAS'es cheat by assuming active count is in low word, so need
-<
+     * to be modified if this changes
-<
+     */
-<
+    private volatile int runControl;
-<
-<
+    // RunState values. Order among values matters
-<
+    private static final int RUNNING     = 0;
-<
+    private static final int SHUTDOWN    = 1;
-<
+    private static final int TERMINATING = 2;
-<
+    private static final int TERMINATED  = 3;
->
+     * 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.
->
+     */
->
+    private WorkQueue findNonEmptyStealQueue(WorkQueue w) {
->
+        // Similar to loop in scan(), but ignoring submissions
->
+        int r = w.seed; r ^= r << 13; r ^= r >>> 17; w.seed = r ^= r << 5;
->
+        int step = (r >>> 16) | 1;
->
+        for (WorkQueue[] ws;;) {
->
+            int rs = runState, m;
->
+            if ((ws = workQueues) == null || (m = ws.length - 1) < 1)
->
+                return null;
->
+            for (int j = (m + 1) << 2; ; r += step) {
->
+                WorkQueue q = ws[((r << 1) | 1) & m];
->
+                if (q != null && !q.isEmpty())
->
+                    return q;
->
+                else if (--j < 0) {
->
+                    if (runState == rs)
->
+                        return null;
->
+                    break;
->
+                }
->
+            }
->
+        }
->
+    }
-–
+    private static int runStateOf(int c)             { return c >>> 16; }
-–
+    private static int activeCountOf(int c)          { return c & shortMask; }
-–
+    private static int runControlFor(int r, int a)   { return (r << 16) + a; }
+    /**
-<
+     * Tries incrementing active count; fails on contention.
-<
+     * Called by workers before/during executing tasks.
-<
+     *
-<
+     * @return true on success
-<
+     */
-<
+    final boolean tryIncrementActiveCount() {
-<
+        int c = runControl;
-<
+        return casRunControl(c, c+1);
->
+     * 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.
->
+     */
->
+    final void helpQuiescePool(WorkQueue w) {
->
+        for (boolean active = true;;) {
->
+            ForkJoinTask<?> localTask; // exhaust local queue
->
+            while ((localTask = w.nextLocalTask()) != null)
->
+                localTask.doExec();
->
+            WorkQueue q = findNonEmptyStealQueue(w);
->
+            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) + parallelism == 0) {
->
+                    do {} while (!U.compareAndSwapLong
->
+                                 (this, CTL, c = ctl, c + AC_UNIT));
->
+                    break;
->
+                }
->
+            }
->
+        }
+    /**
-<
+     * Tries decrementing active count; fails on contention.
-<
+     * Possibly triggers termination on success.
-<
+     * Called by workers when they can't find tasks.
->
+     * Gets and removes a local or stolen task for the given worker.
-<
+     * @return true on success
->
+     * @return a task, if available
+     */
-<
+    final boolean tryDecrementActiveCount() {
-<
+        int c = runControl;
-<
+        int nextc = c - 1;
-<
+        if (!casRunControl(c, nextc))
-<
+            return false;
-<
+        if (canTerminateOnShutdown(nextc))
-<
+            terminateOnShutdown();
-<
+        return true;
-<
+    }
-<
-<
+    /**
-<
+     * Returns {@code true} if argument represents zero active count
-<
+     * and nonzero runstate, which is the triggering condition for
-<
+     * terminating on shutdown.
-<
+     */
-<
+    private static boolean canTerminateOnShutdown(int c) {
-<
+        // i.e. least bit is nonzero runState bit
-<
+        return ((c & -c) >>> 16) != 0;
->
+    final ForkJoinTask<?> nextTaskFor(WorkQueue w) {
->
+        for (ForkJoinTask<?> t;;) {
->
+            WorkQueue q; int b;
->
+            if ((t = w.nextLocalTask()) != null)
->
+                return t;
->
+            if ((q = findNonEmptyStealQueue(w)) == null)
->
+                return null;
->
+            if ((b = q.base) - q.top < 0 && (t = q.pollAt(b)) != null)
->
+                return t;
->
+        }
+    /**
-<
+     * Transition run state to at least the given state. Return true
-<
+     * if not already at least given state.
-<
+     */
-<
+    private boolean transitionRunStateTo(int state) {
-<
+        for (;;) {
-<
+            int c = runControl;
-<
+            if (runStateOf(c) >= state)
-<
+                return false;
-<
+            if (casRunControl(c, runControlFor(state, activeCountOf(c))))
->
+     * Returns the approximate (non-atomic) number of idle threads per
->
+     * active thread to offset steal queue size for method
->
+     * ForkJoinTask.getSurplusQueuedTaskCount().
->
+     */
->
+    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 :
->
+);
->
+    }
->
->
+    //  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) {
->
+        Mutex lock = this.lock;
->
+        for (long c;;) {
->
+            if (((c = ctl) & STOP_BIT) != 0) {      // already terminating
->
+                if ((short)(c >>> TC_SHIFT) == -parallelism) {
->
+                    lock.lock();                    // don't need try/finally
->
+                    termination.signalAll();        // signal when 0 workers
->
+                    lock.unlock();
->
+                }
+                return true;
-+
+            }
-+
+            if (runState >= 0) {                    // not yet enabled
-+
+                if (!enable)
-+
+                    return false;
-+
+                lock.lock();
-+
+                runState |= SHUTDOWN;
-+
+                lock.unlock();
-+
+            }
-+
+            if (!now) {                             // check if idle & no tasks
-+
+                if ((int)(c >> AC_SHIFT) != -parallelism ||
-+
+                    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;
-+
+                        int n = ws.length;
-+
+                        for (int i = 0; i < n; ++i) {
-+
+                            if ((w = ws[i]) != null) {
-+
+                                w.runState = -1;
-+
+                                if (pass > 0) {
-+
+                                    w.cancelAll();
-+
+                                    if (pass > 1)
-+
+                                        w.interruptOwner();
-+
+                                }
-+
+                            }
-+
+                        }
-+
+                        // 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.runState = -1;
-+
+                                if ((p = w.parker) != null)
-+
+                                    U.unpark(p);
-+
+                            }
-+
+                        }
-+
+                    }
-+
+                }
-+
+            }
-<
+    /**
-<
+     * Controls whether to add spares to maintain parallelism
-<
+     */
-<
+    private volatile boolean maintainsParallelism;
->
+    // Exported methods
+    // Constructors
+    /**
-<
+     * Creates a ForkJoinPool with a pool size equal to the number of
-<
+     * processors available on the system, using the default
-<
+     * ForkJoinWorkerThreadFactory.
->
+     * Creates a {@code ForkJoinPool} with parallelism equal to {@link
->
+     * java.lang.Runtime#availableProcessors}, using the {@linkplain
->
+     * #defaultForkJoinWorkerThreadFactory default thread factory},
->
+     * no UncaughtExceptionHandler, and non-async LIFO processing mode.
+     * @throws SecurityException if a security manager exists and
+     *         the caller is not permitted to modify threads
+     */
+    public ForkJoinPool() {
+        this(Runtime.getRuntime().availableProcessors(),
-<
+             defaultForkJoinWorkerThreadFactory);
->
+             defaultForkJoinWorkerThreadFactory, null, false);
+    /**
-<
+     * Creates a ForkJoinPool with the indicated parallelism level
-<
+     * threads and using the default ForkJoinWorkerThreadFactory.
->
+     * Creates a {@code ForkJoinPool} with the indicated parallelism
->
+     * level, the {@linkplain
->
+     * #defaultForkJoinWorkerThreadFactory default thread factory},
->
+     * no UncaughtExceptionHandler, and non-async LIFO processing mode.
-<
+     * @param parallelism the number of worker threads
->
+     * @param parallelism the parallelism level
+     * @throws IllegalArgumentException if parallelism less than or
-<
+     * equal to zero
->
+     *         equal to zero, or greater than implementation limit
+     * @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 ForkJoinPool(int parallelism) {
-<
+        this(parallelism, defaultForkJoinWorkerThreadFactory);
->
+        this(parallelism, defaultForkJoinWorkerThreadFactory, null, false);
+    /**
-<
+     * Creates a ForkJoinPool with parallelism equal to the number of
-<
+     * processors available on the system and using the given
-<
+     * ForkJoinWorkerThreadFactory.
->
+     * Creates a {@code ForkJoinPool} with the given parameters.
-<
+     * @param factory the factory for creating new threads
-<
+     * @throws NullPointerException if factory is null
-<
+     * @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 ForkJoinPool(ForkJoinWorkerThreadFactory factory) {
-<
+        this(Runtime.getRuntime().availableProcessors(), factory);
-<
+    }
-<
-<
+    /**
-<
+     * Creates a ForkJoinPool with the given parallelism and factory.
-<
+     *
-<
+     * @param parallelism the targeted number of worker threads
-<
+     * @param factory the factory for creating new threads
->
+     * @param parallelism the parallelism level. For default value,
->
+     * use {@link java.lang.Runtime#availableProcessors}.
->
+     * @param factory the factory for creating new threads. For default value,
->
+     * use {@link #defaultForkJoinWorkerThreadFactory}.
->
+     * @param handler the handler for internal worker threads that
->
+     * terminate due to unrecoverable errors encountered while executing
->
+     * tasks. For default value, use {@code null}.
->
+     * @param asyncMode if true,
->
+     * establishes local first-in-first-out scheduling mode for forked
->
+     * tasks that are never joined. This mode may be more appropriate
->
+     * than default locally stack-based mode in applications in which
->
+     * worker threads only process event-style asynchronous tasks.
->
+     * For default value, use {@code false}.
+     * @throws IllegalArgumentException if parallelism less than or
-<
+     * equal to zero, or greater than implementation limit
-<
+     * @throws NullPointerException if factory is null
->
+     *         equal to zero, or greater than implementation limit
->
+     * @throws NullPointerException if the factory is null
+     * @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 ForkJoinPool(int parallelism, ForkJoinWorkerThreadFactory factory) {
-<
+        if (parallelism <= 0 || parallelism > MAX_THREADS)
-<
+            throw new IllegalArgumentException();
->
+    public ForkJoinPool(int parallelism,
->
+                        ForkJoinWorkerThreadFactory factory,
->
+                        Thread.UncaughtExceptionHandler handler,
->
+                        boolean asyncMode) {
->
+        checkPermission();
+        if (factory == null)
+            throw new NullPointerException();
-<
+        checkPermission();
-<
+        this.factory = factory;
->
+        if (parallelism <= 0 || parallelism > MAX_CAP)
->
+            throw new IllegalArgumentException();
+        this.parallelism = parallelism;
-<
+        this.maxPoolSize = MAX_THREADS;
-<
+        this.maintainsParallelism = true;
-<
+        this.poolNumber = poolNumberGenerator.incrementAndGet();
-<
+        this.workerLock = new ReentrantLock();
-<
+        this.termination = workerLock.newCondition();
->
+        this.factory = factory;
->
+        this.ueh = handler;
->
+        this.localMode = asyncMode ? FIFO_QUEUE : LIFO_QUEUE;
->
+        long np = (long)(-parallelism); // offset ctl counts
->
+        this.ctl = ((np << AC_SHIFT) & AC_MASK) | ((np << TC_SHIFT) & TC_MASK);
->
+        // Use nearest power 2 for workQueues size. See Hackers Delight sec 3.2.
->
+        int n = parallelism - 1;
->
+        n |= n >>> 1; n |= n >>> 2; n |= n >>> 4; n |= n >>> 8; n |= n >>> 16;
->
+        int size = (n + 1) << 1;        // #slots = 2*#workers
->
+        this.submitMask = size - 1;     // room for max # of submit queues
->
+        this.workQueues = new WorkQueue[size];
->
+        this.termination = (this.lock = new Mutex()).newCondition();
+        this.stealCount = new AtomicLong();
-<
+        this.submissionQueue = new LinkedTransferQueue<ForkJoinTask<?>>();
-<
+        // worker array and workers are lazily constructed
-<
+    }
-<
-<
+    /**
-<
+     * Creates a new worker thread using factory.
-<
+     *
-<
+     * @param index the index to assign worker
-<
+     * @return new worker, or null of factory failed
-<
+     */
-<
+    private ForkJoinWorkerThread createWorker(int index) {
-<
+        Thread.UncaughtExceptionHandler h = ueh;
-<
+        ForkJoinWorkerThread w = factory.newThread(this);
-<
+        if (w != null) {
-<
+            w.poolIndex = index;
-<
+            w.setDaemon(true);
-<
+            w.setAsyncMode(locallyFifo);
-<
+            w.setName("ForkJoinPool-" + poolNumber + "-worker-" + index);
-<
+            if (h != null)
-<
+                w.setUncaughtExceptionHandler(h);
-<
+        }
-<
+        return w;
-<
+    }
-<
-<
+    /**
-<
+     * Returns a good size for worker array given pool size.
-<
+     * Currently requires size to be a power of two.
-<
+     */
-<
+    private static int arraySizeFor(int poolSize) {
-<
+        return (poolSize <= 1) ? 1 :
-<
+            (1 << (32 - Integer.numberOfLeadingZeros(poolSize-1)));
-<
+    }
-<
-<
+    /**
-<
+     * Creates or resizes array if necessary to hold newLength.
-<
+     * Call only under exclusion.
-<
+     *
-<
+     * @return the array
-<
+     */
-<
+    private ForkJoinWorkerThread[] ensureWorkerArrayCapacity(int newLength) {
-<
+        ForkJoinWorkerThread[] ws = workers;
-<
+        if (ws == null)
-<
+            return workers = new ForkJoinWorkerThread[arraySizeFor(newLength)];
-<
+        else if (newLength > ws.length)
-<
+            return workers = Arrays.copyOf(ws, arraySizeFor(newLength));
-<
+        else
-<
+            return ws;
-<
+    }
-<
-<
+    /**
-<
+     * Tries to shrink workers into smaller array after one or more terminate.
-<
+     */
-<
+    private void tryShrinkWorkerArray() {
-<
+        ForkJoinWorkerThread[] ws = workers;
-<
+        if (ws != null) {
-<
+            int len = ws.length;
-<
+            int last = len - 1;
-<
+            while (last >= 0 && ws[last] == null)
-<
+                --last;
-<
+            int newLength = arraySizeFor(last+1);
-<
+            if (newLength < len)
-<
+                workers = Arrays.copyOf(ws, newLength);
-<
+        }
-<
+    }
-<
-<
+    /**
-<
+     * Initializes workers if necessary.
-<
+     */
-<
+    final void ensureWorkerInitialization() {
-<
+        ForkJoinWorkerThread[] ws = workers;
-<
+        if (ws == null) {
-<
+            final ReentrantLock lock = this.workerLock;
-<
+            lock.lock();
-<
+            try {
-<
+                ws = workers;
-<
+                if (ws == null) {
-<
+                    int ps = parallelism;
-<
+                    ws = ensureWorkerArrayCapacity(ps);
-<
+                    for (int i = 0; i < ps; ++i) {
-<
+                        ForkJoinWorkerThread w = createWorker(i);
-<
+                        if (w != null) {
-<
+                            ws[i] = w;
-<
+                            w.start();
-<
+                            updateWorkerCount(1);
-<
+                        }
-<
+                    }
-<
+                }
-<
+            } finally {
-<
+                lock.unlock();
-<
+            }
-<
+        }
-<
+    }
-<
-<
+    /**
-<
+     * Worker creation and startup for threads added via setParallelism.
-<
+     */
-<
+    private void createAndStartAddedWorkers() {
-<
+        resumeAllSpares();  // Allow spares to convert to nonspare
-<
+        int ps = parallelism;
-<
+        ForkJoinWorkerThread[] ws = ensureWorkerArrayCapacity(ps);
-<
+        int len = ws.length;
-<
+        // Sweep through slots, to keep lowest indices most populated
-<
+        int k = 0;
-<
+        while (k < len) {
-<
+            if (ws[k] != null) {
-<
+                ++k;
-<
+                continue;
-<
+            }
-<
+            int s = workerCounts;
-<
+            int tc = totalCountOf(s);
-<
+            int rc = runningCountOf(s);
-<
+            if (rc >= ps || tc >= ps)
-<
+                break;
-<
+            if (casWorkerCounts (s, workerCountsFor(tc+1, rc+1))) {
-<
+                ForkJoinWorkerThread w = createWorker(k);
-<
+                if (w != null) {
-<
+                    ws[k++] = w;
-<
+                    w.start();
-<
+                }
-<
+                else {
-<
+                    updateWorkerCount(-1); // back out on failed creation
-<
+                    break;
-<
+                }
-<
+            }
-<
+        }
->
+        this.nextWorkerNumber = new AtomicInteger();
->
+        int pn = poolNumberGenerator.incrementAndGet();
->
+        StringBuilder sb = new StringBuilder("ForkJoinPool-");
->
+        sb.append(Integer.toString(pn));
->
+        sb.append("-worker-");
->
+        this.workerNamePrefix = sb.toString();
->
+        lock.lock();
->
+        this.runState = 1;              // set init flag
->
+        lock.unlock();
+    // Execution methods
+    /**
-–
+     * Common code for execute, invoke and submit
-–
+     */
-–
+    private <T> void doSubmit(ForkJoinTask<T> task) {
-–
+        if (task == null)
-–
+            throw new NullPointerException();
-–
+        if (isShutdown())
-–
+            throw new RejectedExecutionException();
-–
+        if (workers == null)
-–
+            ensureWorkerInitialization();
-–
+        submissionQueue.offer(task);
-–
+        signalIdleWorkers();
-–
+    }
-–
-–
+    /**
+     * Performs the given task, returning its result upon completion.
-+
+     * If the computation encounters an unchecked Exception or Error,
-+
+     * it is rethrown as the outcome of this invocation.  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 current thread
-+
+     * as well as the thread actually encountering the exception;
-+
+     * minimally only the latter.
+     * @param task the task
+     * @return the task's result
-<
+     * @throws NullPointerException if task is null
-<
+     * @throws RejectedExecutionException if pool is shut down
->
+     * @throws NullPointerException if the task is null
->
+     * @throws RejectedExecutionException if the task cannot be
->
+     *         scheduled for execution
+     */
+    public <T> T invoke(ForkJoinTask<T> task) {
-+
+        if (task == null)
-+
+            throw new NullPointerException();
+        doSubmit(task);
+        return task.join();
+     * Arranges for (asynchronous) execution of the given task.
+     * @param task the task
-<
+     * @throws NullPointerException if task is null
-<
+     * @throws RejectedExecutionException if pool is shut down
->
+     * @throws NullPointerException if the task is null
->
+     * @throws RejectedExecutionException if the task cannot be
->
+     *         scheduled for execution
+     */
+    public void execute(ForkJoinTask<?> task) {
-+
+        if (task == null)
-+
+            throw new NullPointerException();
+        doSubmit(task);
+    // AbstractExecutorService methods
-+
+    /**
-+
+     * @throws NullPointerException if the task is null
-+
+     * @throws RejectedExecutionException if the task cannot be
-+
+     *         scheduled for execution
-+
+     */
+    public void execute(Runnable task) {
-+
+        if (task == null)
-+
+            throw new NullPointerException();
+        ForkJoinTask<?> job;
+        if (task instanceof ForkJoinTask<?>) // avoid re-wrap
+            job = (ForkJoinTask<?>) task;
+        else
-<
+            job = ForkJoinTask.adapt(task, null);
->
+            job = new ForkJoinTask.AdaptedRunnableAction(task);
+        doSubmit(job);
-+
+    /**
-+
+     * Submits a ForkJoinTask for execution.
-+
+     *
-+
+     * @param task the task to submit
-+
+     * @return the task
-+
+     * @throws NullPointerException if the task is null
-+
+     * @throws RejectedExecutionException if the task cannot be
-+
+     *         scheduled for execution
-+
+     */
-+
+    public <T> ForkJoinTask<T> submit(ForkJoinTask<T> task) {
-+
+        if (task == null)
-+
+            throw new NullPointerException();
-+
+        doSubmit(task);
-+
+        return task;
-+
+    }
-+
-+
+    /**
-+
+     * @throws NullPointerException if the task is null
-+
+     * @throws RejectedExecutionException if the task cannot be
-+
+     *         scheduled for execution
-+
+     */
+    public <T> ForkJoinTask<T> submit(Callable<T> task) {
-<
+        ForkJoinTask<T> job = ForkJoinTask.adapt(task);
->
+        ForkJoinTask<T> job = new ForkJoinTask.AdaptedCallable<T>(task);
+        doSubmit(job);
+        return job;
-+
+    /**
-+
+     * @throws NullPointerException if the task is null
-+
+     * @throws RejectedExecutionException if the task cannot be
-+
+     *         scheduled for execution
-+
+     */
+    public <T> ForkJoinTask<T> submit(Runnable task, T result) {
-<
+        ForkJoinTask<T> job = ForkJoinTask.adapt(task, result);
->
+        ForkJoinTask<T> job = new ForkJoinTask.AdaptedRunnable<T>(task, result);
+        doSubmit(job);
+        return job;
-+
+    /**
-+
+     * @throws NullPointerException if the task is null
-+
+     * @throws RejectedExecutionException if the task cannot be
-+
+     *         scheduled for execution
-+
+     */
+    public ForkJoinTask<?> submit(Runnable task) {
-+
+        if (task == null)
-+
+            throw new NullPointerException();
+        ForkJoinTask<?> job;
+        if (task instanceof ForkJoinTask<?>) // avoid re-wrap
+            job = (ForkJoinTask<?>) task;
+        else
-<
+            job = ForkJoinTask.adapt(task, null);
->
+            job = new ForkJoinTask.AdaptedRunnableAction(task);
+        doSubmit(job);
+        return job;
+    /**
-<
+     * Submits a ForkJoinTask for execution.
-<
+     *
-<
+     * @param task the task to submit
-<
+     * @return the task
-<
+     * @throws RejectedExecutionException if the task cannot be
-<
+     *         scheduled for execution
-<
+     * @throws NullPointerException if the task is null
->
+     * @throws NullPointerException       {@inheritDoc}
->
+     * @throws RejectedExecutionException {@inheritDoc}
+     */
-–
+    public <T> ForkJoinTask<T> submit(ForkJoinTask<T> task) {
-–
+        doSubmit(task);
-–
+        return task;
-–
+    }
-–
-–
+    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.
->
+        List<ForkJoinTask<T>> fs = new ArrayList<ForkJoinTask<T>>(tasks.size());
->
+        // Workaround needed because method wasn't declared with
->
+        // wildcards in return type but should have been.
+        @SuppressWarnings({"unchecked", "rawtypes"})
-<
+        List<Future<T>> futures = (List<Future<T>>) (List) forkJoinTasks;
-<
+        return futures;
-<
+    }
->
+            List<Future<T>> futures = (List<Future<T>>) (List) fs;
-<
+    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);
->
+                doSubmit(f);
->
+                fs.add(f);
->
+            }
->
+            for (ForkJoinTask<T> f : fs)
->
+                f.quietlyJoin();
->
+            done = true;
->
+            return futures;
->
+        } finally {
->
+            if (!done)
->
+                for (ForkJoinTask<T> f : fs)
->
+                    f.cancel(false);
-–
+        private static final long serialVersionUID = -7914297376763021607L;
-–
+    // Configuration and status settings and queries
-–
+    /**
+     * Returns the factory used for constructing new workers.
+     * @return the handler, or {@code null} if none
+     */
+    public Thread.UncaughtExceptionHandler getUncaughtExceptionHandler() {
-<
+        Thread.UncaughtExceptionHandler h;
-<
+        final ReentrantLock lock = this.workerLock;
-<
+        lock.lock();
-<
+        try {
-<
+            h = ueh;
-<
+        } finally {
-<
+            lock.unlock();
-<
+        }
-<
+        return h;
->
+        return ueh;
+    /**
-<
+     * Sets the handler for internal worker threads that terminate due
-<
+     * to unrecoverable errors encountered while executing tasks.
-<
+     * Unless set, the current default or ThreadGroup handler is used
-<
+     * as handler.
->
+     * Returns the targeted parallelism level of this pool.
-<
+     * @param h the new handler
-<
+     * @return the old handler, or {@code null} if none
-<
+     * @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 Thread.UncaughtExceptionHandler
-<
+        setUncaughtExceptionHandler(Thread.UncaughtExceptionHandler h) {
-<
+        checkPermission();
-<
+        Thread.UncaughtExceptionHandler old = null;
-<
+        final ReentrantLock lock = this.workerLock;
-<
+        lock.lock();
-<
+        try {
-<
+            old = ueh;
-<
+            ueh = h;
-<
+            ForkJoinWorkerThread[] ws = workers;
-<
+            if (ws != null) {
-<
+                for (int i = 0; i < ws.length; ++i) {
-<
+                    ForkJoinWorkerThread w = ws[i];
-<
+                    if (w != null)
-<
+                        w.setUncaughtExceptionHandler(h);
-<
+                }
-<
+            }
-<
+        } finally {
-<
+            lock.unlock();
-<
+        }
-<
+        return old;
-<
+    }
-<
-<
-<
+    /**
-<
+     * Sets the target parallelism level of this pool.
-<
+     *
-<
+     * @param parallelism the target parallelism
-<
+     * @throws IllegalArgumentException if parallelism less than or
-<
+     * equal to zero or greater than maximum size bounds
-<
+     * @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 setParallelism(int parallelism) {
-<
+        checkPermission();
-<
+        if (parallelism <= 0 || parallelism > maxPoolSize)
-<
+            throw new IllegalArgumentException();
-<
+        final ReentrantLock lock = this.workerLock;
-<
+        lock.lock();
-<
+        try {
-<
+            if (!isTerminating()) {
-<
+                int p = this.parallelism;
-<
+                this.parallelism = parallelism;
-<
+                if (parallelism > p)
-<
+                    createAndStartAddedWorkers();
-<
+                else
-<
+                    trimSpares();
-<
+            }
-<
+        } finally {
-<
+            lock.unlock();
-<
+        }
-<
+        signalIdleWorkers();
-<
+    }
-<
-<
+    /**
-<
+     * Returns the targeted number of worker threads in this pool.
-<
+     *
-<
+     * @return the targeted number of worker threads in this pool
->
+     * @return the targeted parallelism level of this pool
+     */
+    public int getParallelism() {
+        return parallelism;
+    /**
+     * Returns the number of worker threads that have started but not
-<
+     * yet terminated.  This result returned by this method may differ
->
+     * yet terminated.  The result returned by this method may differ
+     * from {@link #getParallelism} when threads are created to
+     * maintain parallelism when others are cooperatively blocked.
+     * @return the number of worker threads
+     */
+    public int getPoolSize() {
-<
+        return totalCountOf(workerCounts);
-<
+    }
-<
-<
+    /**
-<
+     * Returns the maximum number of threads allowed to exist in the
-<
+     * pool, even if there are insufficient unblocked running threads.
-<
+     *
-<
+     * @return the maximum
-<
+     */
-<
+    public int getMaximumPoolSize() {
-<
+        return maxPoolSize;
-<
+    }
-<
-<
+    /**
-<
+     * Sets the maximum number of threads allowed to exist in the
-<
+     * pool, even if there are insufficient unblocked running threads.
-<
+     * Setting this value has no effect on current pool size. It
-<
+     * controls construction of new threads.
-<
+     *
-<
+     * @throws IllegalArgumentException if negative or greater than
-<
+     * internal implementation limit
-<
+     */
-<
+    public void setMaximumPoolSize(int newMax) {
-<
+        if (newMax < 0 || newMax > MAX_THREADS)
-<
+            throw new IllegalArgumentException();
-<
+        maxPoolSize = newMax;
-<
+    }
-<
-<
-<
+    /**
-<
+     * Returns {@code true} if this pool dynamically maintains its
-<
+     * target parallelism level. If false, new threads are added only
-<
+     * to avoid possible starvation.  This setting is by default true.
-<
+     *
-<
+     * @return {@code true} if maintains parallelism
-<
+     */
-<
+    public boolean getMaintainsParallelism() {
-<
+        return maintainsParallelism;
-<
+    }
-<
-<
+    /**
-<
+     * Sets whether this pool dynamically maintains its target
-<
+     * parallelism level. If false, new threads are added only to
-<
+     * avoid possible starvation.
-<
+     *
-<
+     * @param enable {@code true} to maintain parallelism
-<
+     */
-<
+    public void setMaintainsParallelism(boolean enable) {
-<
+        maintainsParallelism = enable;
-<
+    }
-<
-<
+    /**
-<
+     * Establishes local first-in-first-out scheduling mode for forked
-<
+     * tasks that are never joined. This mode may be more appropriate
-<
+     * than default locally stack-based mode in applications in which
-<
+     * worker threads only process asynchronous tasks.  This method is
-<
+     * designed to be invoked only when the pool is quiescent, and
-<
+     * typically only before any tasks are submitted. The effects of
-<
+     * invocations at other times may be unpredictable.
-<
+     *
-<
+     * @param async if {@code true}, use locally FIFO scheduling
-<
+     * @return the previous mode
-<
+     * @see #getAsyncMode
-<
+     */
-<
+    public boolean setAsyncMode(boolean async) {
-<
+        boolean oldMode = locallyFifo;
-<
+        locallyFifo = async;
-<
+        ForkJoinWorkerThread[] ws = workers;
-<
+        if (ws != null) {
-<
+            for (int i = 0; i < ws.length; ++i) {
-<
+                ForkJoinWorkerThread t = ws[i];
-<
+                if (t != null)
-<
+                    t.setAsyncMode(async);
-<
+            }
-<
+        }
-<
+        return oldMode;
->
+        return parallelism + (short)(ctl >>> TC_SHIFT);
+    /**
+     * scheduling mode for forked tasks that are never joined.
+     * @return {@code true} if this pool uses async mode
-–
+     * @see #setAsyncMode
+     */
+    public boolean getAsyncMode() {
-<
+        return locallyFifo;
->
+        return localMode != 0;
+    /**
+     * Returns an estimate of the number of worker threads that are
+     * not blocked waiting to join tasks or for other managed
-<
+     * synchronization.
->
+     * synchronization. This method may overestimate the
->
+     * number of running threads.
+     * @return the number of worker threads
+     */
+    public int getRunningThreadCount() {
-<
+        return runningCountOf(workerCounts);
->
+        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;
+    /**
+     * @return the number of active threads
+     */
+    public int getActiveThreadCount() {
-<
+        return activeCountOf(runControl);
-<
+    }
-<
-<
+    /**
-<
+     * Returns an estimate of the number of threads that are currently
-<
+     * idle waiting for tasks. This method may underestimate the
-<
+     * number of idle threads.
-<
+     *
-<
+     * @return the number of idle threads
-<
+     */
-<
+    final int getIdleThreadCount() {
-<
+        int c = runningCountOf(workerCounts) - activeCountOf(runControl);
-<
+        return (c <= 0) ? 0 : c;
->
+        int r = parallelism + (int)(ctl >> AC_SHIFT);
->
+        return (r <= 0) ? 0 : r; // suppress momentarily negative values
+    /**
+     * @return {@code true} if all threads are currently idle
+     */
+    public boolean isQuiescent() {
-<
+        return activeCountOf(runControl) == 0;
->
+        return (int)(ctl >> AC_SHIFT) + parallelism == 0;
+    /**
+     * @return the number of steals
+     */
+    public long getStealCount() {
-<
+        return stealCount.get();
-<
+    }
-<
-<
+    /**
-<
+     * Accumulates steal count from a worker.
-<
+     * Call only when worker known to be idle.
-<
+     */
-<
+    private void updateStealCount(ForkJoinWorkerThread w) {
-<
+        int sc = w.getAndClearStealCount();
-<
+        if (sc != 0)
-<
+            stealCount.addAndGet(sc);
->
+        long count = stealCount.get();
->
+        WorkQueue[] ws; WorkQueue w;
->
+        if ((ws = workQueues) != null) {
->
+            for (int i = 1; i < ws.length; i += 2) {
->
+                if ((w = ws[i]) != null)
->
+                    count += w.totalSteals;
->
+            }
->
+        }
->
+        return count;
+    /**
+     */
+    public long getQueuedTaskCount() {
+        long count = 0;
-<
+        ForkJoinWorkerThread[] ws = workers;
-<
+        if (ws != null) {
-<
+            for (int i = 0; i < ws.length; ++i) {
-<
+                ForkJoinWorkerThread t = ws[i];
-<
+                if (t != null)
-<
+                    count += t.getQueueSize();
->
+        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;
+    /**
+     * Returns an estimate of the number of tasks submitted to this
-<
+     * pool that have not yet begun executing.  This method takes time
-<
+     * proportional to the number of submissions.
->
+     * pool that have not yet begun executing.  This method may take
->
+     * time proportional to the number of submissions.
+     * @return the number of queued submissions
+     */
+    public int getQueuedSubmissionCount() {
-<
+        return submissionQueue.size();
->
+        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;
+    /**
+     * @return {@code true} if there are any queued submissions
+     */
+    public boolean hasQueuedSubmissions() {
-<
+        return !submissionQueue.isEmpty();
->
+        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;
+    /**
+     * @return the next submission, or {@code null} if none
+     */
+    protected ForkJoinTask<?> pollSubmission() {
-<
+        return submissionQueue.poll();
->
+        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;
+    /**
+     * Removes all available unexecuted submitted and forked tasks
+     * from scheduling queues and adds them to the given collection,
+     * without altering their execution status. These may include
-<
+     * artificially generated or wrapped tasks. This method is designed
-<
+     * to be invoked only when the pool is known to be
->
+     * artificially generated or wrapped tasks. This method is
->
+     * designed to be invoked only when the pool is known to be
+     * quiescent. Invocations at other times may not remove all
+     * tasks. A failure encountered while attempting to add elements
+     * to collection {@code c} may result in elements being in
+     * @return the number of elements transferred
+     */
+    protected int drainTasksTo(Collection<? super ForkJoinTask<?>> c) {
-<
+        int n = submissionQueue.drainTo(c);
-<
+        ForkJoinWorkerThread[] ws = workers;
-<
+        if (ws != null) {
->
+        int count = 0;
->
+        WorkQueue[] ws; WorkQueue w; ForkJoinTask<?> t;
->
+        if ((ws = workQueues) != null) {
+            for (int i = 0; i < ws.length; ++i) {
-<
+                ForkJoinWorkerThread w = ws[i];
-<
+                if (w != null)
-<
+                    n += w.drainTasksTo(c);
->
+                if ((w = ws[i]) != null) {
->
+                    while ((t = w.poll()) != null) {
->
+                        c.add(t);
->
+                        ++count;
->
+                    }
->
+                }
-<
+        return n;
->
+        return count;
+    /**
+     * @return a string identifying this pool, as well as its state
+     */
+    public String toString() {
-<
+        int ps = parallelism;
-<
+        int wc = workerCounts;
-<
+        int rc = runControl;
-<
+        long st = getStealCount();
-<
+        long qt = getQueuedTaskCount();
-<
+        long qs = getQueuedSubmissionCount();
->
+        // Use a single pass through workQueues to collect counts
->
+        long qt = 0L, qs = 0L; int rc = 0;
->
+        long st = stealCount.get();
->
+        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.totalSteals;
->
+                        if (w.isApparentlyUnblocked())
->
+                            ++rc;
->
+                    }
->
+                }
->
+            }
->
+        }
->
+        int pc = parallelism;
->
+        int tc = pc + (short)(c >>> TC_SHIFT);
->
+        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 = runState < 0 ? "Shutting down" : "Running";
+        return super.toString() +
-<
+            "[" + runStateToString(runStateOf(rc)) +
-<
+            ", parallelism = " + ps +
-<
+            ", size = " + totalCountOf(wc) +
-<
+            ", active = " + activeCountOf(rc) +
-<
+            ", running = " + runningCountOf(wc) +
->
+            "[" + level +
->
+            ", parallelism = " + pc +
->
+            ", size = " + tc +
->
+            ", active = " + ac +
->
+            ", running = " + rc +
+            ", steals = " + st +
+            ", tasks = " + qt +
+            ", submissions = " + qs +
+            "]";
-–
+    private static String runStateToString(int rs) {
-–
+        switch(rs) {
-–
+        case RUNNING: return "Running";
-–
+        case SHUTDOWN: return "Shutting down";
-–
+        case TERMINATING: return "Terminating";
-–
+        case TERMINATED: return "Terminated";
-–
+        default: throw new Error("Unknown run state");
-–
+        }
-–
+    }
-–
-–
+    // lifecycle control
-–
+    /**
+     * Initiates an orderly shutdown in which previously submitted
+     * tasks are executed, but no new tasks will be accepted.
+     */
+    public void shutdown() {
+        checkPermission();
-<
+        transitionRunStateTo(SHUTDOWN);
-<
+        if (canTerminateOnShutdown(runControl)) {
-<
+            if (workers == null) { // shutting down before workers created
-<
+                final ReentrantLock lock = this.workerLock;
-<
+                lock.lock();
-<
+                try {
-<
+                    if (workers == null) {
-<
+                        terminate();
-<
+                        transitionRunStateTo(TERMINATED);
-<
+                        termination.signalAll();
-<
+                    }
-<
+                } finally {
-<
+                    lock.unlock();
-<
+                }
-<
+            }
-<
+            terminateOnShutdown();
-<
+        }
->
+        tryTerminate(false, true);
+    /**
-<
+     * Attempts to stop all actively executing tasks, and cancels all
-<
+     * waiting 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. Unlike some other executors,
-<
+     * this method cancels rather than collects non-executed tasks
-<
+     * upon termination, so always returns an empty list. However, you
-<
+     * can use method {@link #drainTasksTo} before invoking this
-<
+     * method to transfer unexecuted tasks to another collection.
->
+     * 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).
+     * @return an empty list
+     * @throws SecurityException if a security manager exists and
+     */
+    public List<Runnable> shutdownNow() {
+        checkPermission();
-<
+        terminate();
->
+        tryTerminate(true, true);
+        return Collections.emptyList();
+     * @return {@code true} if all tasks have completed following shut down
+     */
+    public boolean isTerminated() {
-<
+        return runStateOf(runControl) == TERMINATED;
->
+        long c = ctl;
->
+        return ((c & STOP_BIT) != 0L &&
->
+                (short)(c >>> TC_SHIFT) == -parallelism);
+    /**
+     * Returns {@code true} if the process of termination has
-<
+     * commenced but possibly not yet completed.
->
+     * 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,
->
+     * 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
->
+     * they do, they must abort them on interrupt.)
-<
+     * @return {@code true} if terminating
->
+     * @return {@code true} if terminating but not yet terminated
+     */
+    public boolean isTerminating() {
-<
+        return runStateOf(runControl) >= TERMINATING;
->
+        long c = ctl;
->
+        return ((c & STOP_BIT) != 0L &&
->
+                (short)(c >>> TC_SHIFT) != -parallelism);
+    /**
+     * @return {@code true} if this pool has been shut down
+     */
+    public boolean isShutdown() {
-<
+        return runStateOf(runControl) >= SHUTDOWN;
->
+        return runState < 0;
+    /**
+    public boolean awaitTermination(long timeout, TimeUnit unit)
+        throws InterruptedException {
+        long nanos = unit.toNanos(timeout);
-<
+        final ReentrantLock lock = this.workerLock;
->
+        final Mutex lock = this.lock;
+        lock.lock();
+        try {
+            for (;;) {
-–
+    // Shutdown and termination support
-–
-–
+    /**
-–
+     * Callback from terminating worker. Nulls out the corresponding
-–
+     * workers slot, and if terminating, tries to terminate; else
-–
+     * tries to shrink workers array.
-–
+     *
-–
+     * @param w the worker
-–
+     */
-–
+    final void workerTerminated(ForkJoinWorkerThread w) {
-–
+        updateStealCount(w);
-–
+        updateWorkerCount(-1);
-–
+        final ReentrantLock lock = this.workerLock;
-–
+        lock.lock();
-–
+        try {
-–
+            ForkJoinWorkerThread[] ws = workers;
-–
+            if (ws != null) {
-–
+                int idx = w.poolIndex;
-–
+                if (idx >= 0 && idx < ws.length && ws[idx] == w)
-–
+                    ws[idx] = null;
-–
+                if (totalCountOf(workerCounts) == 0) {
-–
+                    terminate(); // no-op if already terminating
-–
+                    transitionRunStateTo(TERMINATED);
-–
+                    termination.signalAll();
-–
+                }
-–
+                else if (!isTerminating()) {
-–
+                    tryShrinkWorkerArray();
-–
+                    tryResumeSpare(true); // allow replacement
-–
+                }
-–
+            }
-–
+        } finally {
-–
+            lock.unlock();
-–
+        }
-–
+        signalIdleWorkers();
-–
+    }
-–
-–
+    /**
-–
+     * Initiates termination.
-–
+     */
-–
+    private void terminate() {
-–
+        if (transitionRunStateTo(TERMINATING)) {
-–
+            stopAllWorkers();
-–
+            resumeAllSpares();
-–
+            signalIdleWorkers();
-–
+            cancelQueuedSubmissions();
-–
+            cancelQueuedWorkerTasks();
-–
+            interruptUnterminatedWorkers();
-–
+            signalIdleWorkers(); // resignal after interrupt
-–
+        }
-–
+    }
-–
-–
+    /**
-–
+     * Possibly terminates when on shutdown state.
-–
+     */
-–
+    private void terminateOnShutdown() {
-–
+        if (!hasQueuedSubmissions() && canTerminateOnShutdown(runControl))
-–
+            terminate();
-–
+    }
-–
-–
+    /**
-–
+     * Clears out and cancels submissions.
-–
+     */
-–
+    private void cancelQueuedSubmissions() {
-–
+        ForkJoinTask<?> task;
-–
+        while ((task = pollSubmission()) != null)
-–
+            task.cancel(false);
-–
+    }
-–
-–
+    /**
-–
+     * Cleans out worker queues.
-–
+     */
-–
+    private void cancelQueuedWorkerTasks() {
-–
+        final ReentrantLock lock = this.workerLock;
-–
+        lock.lock();
-–
+        try {
-–
+            ForkJoinWorkerThread[] ws = workers;
-–
+            if (ws != null) {
-–
+                for (int i = 0; i < ws.length; ++i) {
-–
+                    ForkJoinWorkerThread t = ws[i];
-–
+                    if (t != null)
-–
+                        t.cancelTasks();
-–
+                }
-–
+            }
-–
+        } finally {
-–
+            lock.unlock();
-–
+        }
-–
+    }
-–
-–
+    /**
-–
+     * Sets each worker's status to terminating. Requires lock to avoid
-–
+     * conflicts with add/remove.
-–
+     */
-–
+    private void stopAllWorkers() {
-–
+        final ReentrantLock lock = this.workerLock;
-–
+        lock.lock();
-–
+        try {
-–
+            ForkJoinWorkerThread[] ws = workers;
-–
+            if (ws != null) {
-–
+                for (int i = 0; i < ws.length; ++i) {
-–
+                    ForkJoinWorkerThread t = ws[i];
-–
+                    if (t != null)
-–
+                        t.shutdownNow();
-–
+                }
-–
+            }
-–
+        } finally {
-–
+            lock.unlock();
-–
+        }
-–
+    }
-–
-–
+    /**
-–
+     * Interrupts all unterminated workers.  This is not required for
-–
+     * sake of internal control, but may help unstick user code during
-–
+     * shutdown.
-–
+     */
-–
+    private void interruptUnterminatedWorkers() {
-–
+        final ReentrantLock lock = this.workerLock;
-–
+        lock.lock();
-–
+        try {
-–
+            ForkJoinWorkerThread[] ws = workers;
-–
+            if (ws != null) {
-–
+                for (int i = 0; i < ws.length; ++i) {
-–
+                    ForkJoinWorkerThread t = ws[i];
-–
+                    if (t != null && !t.isTerminated()) {
-–
+                        try {
-–
+                            t.interrupt();
-–
+                        } catch (SecurityException ignore) {
-–
+                        }
-–
+                    }
-–
+                }
-–
+            }
-–
+        } finally {
-–
+            lock.unlock();
-–
+        }
-–
+    }
-–
-–
-–
+    /*
-–
+     * Nodes for event barrier to manage idle threads.  Queue nodes
-–
+     * are basic Treiber stack nodes, also used for spare stack.
-–
+     *
-–
+     * The event barrier has an event count and a wait queue (actually
-–
+     * a Treiber stack).  Workers are enabled to look for work when
-–
+     * the eventCount is incremented. If they fail to find work, they
-–
+     * may wait for next count. Upon release, threads help others wake
-–
+     * up.
-–
+     *
-–
+     * Synchronization events occur only in enough contexts to
-–
+     * maintain overall liveness:
-–
+     *
-–
+     *   - Submission of a new task to the pool
-–
+     *   - Resizes or other changes to the workers array
-–
+     *   - pool termination
-–
+     *   - A worker pushing a task on an empty queue
-–
+     *
-–
+     * The case of pushing a task occurs often enough, and is heavy
-–
+     * enough compared to simple stack pushes, to require special
-–
+     * handling: Method signalWork returns without advancing count if
-–
+     * the queue appears to be empty.  This would ordinarily result in
-–
+     * races causing some queued waiters not to be woken up. To avoid
-–
+     * this, the first worker enqueued in method sync (see
-–
+     * syncIsReleasable) rescans for tasks after being enqueued, and
-–
+     * helps signal if any are found. This works well because the
-–
+     * worker has nothing better to do, and so might as well help
-–
+     * alleviate the overhead and contention on the threads actually
-–
+     * doing work.  Also, since event counts increments on task
-–
+     * availability exist to maintain liveness (rather than to force
-–
+     * refreshes etc), it is OK for callers to exit early if
-–
+     * contending with another signaller.
-–
+     */
-–
+    static final class WaitQueueNode {
-–
+        WaitQueueNode next; // only written before enqueued
-–
+        volatile ForkJoinWorkerThread thread; // nulled to cancel wait
-–
+        final long count; // unused for spare stack
-–
-–
+        WaitQueueNode(long c, ForkJoinWorkerThread w) {
-–
+            count = c;
-–
+            thread = w;
-–
+        }
-–
-–
+        /**
-–
+         * Wakes up waiter, returning false if known to already
-–
+         */
-–
+        boolean signal() {
-–
+            ForkJoinWorkerThread t = thread;
-–
+            if (t == null)
-–
+                return false;
-–
+            thread = null;
-–
+            LockSupport.unpark(t);
-–
+            return true;
-–
+        }
-–
-–
+        /**
-–
+         * Awaits release on sync.
-–
+         */
-–
+        void awaitSyncRelease(ForkJoinPool p) {
-–
+            while (thread != null && !p.syncIsReleasable(this))
-–
+                LockSupport.park(this);
-–
+        }
-–
-–
+        /**
-–
+         * Awaits resumption as spare.
-–
+         */
-–
+        void awaitSpareRelease() {
-–
+            while (thread != null) {
-–
+                if (!Thread.interrupted())
-–
+                    LockSupport.park(this);
-–
+            }
-–
+        }
-–
+    }
-–
-–
+    /**
-–
+     * Ensures that no thread is waiting for count to advance from the
-–
+     * current value of eventCount read on entry to this method, by
-–
+     * releasing waiting threads if necessary.
-–
+     *
-–
+     * @return the count
-–
+     */
-–
+    final long ensureSync() {
-–
+        long c = eventCount;
-–
+        WaitQueueNode q;
-–
+        while ((q = syncStack) != null && q.count < c) {
-–
+            if (casBarrierStack(q, null)) {
-–
+                do {
-–
+                    q.signal();
-–
+                } while ((q = q.next) != null);
-–
+                break;
-–
+            }
-–
+        }
-–
+        return c;
-–
+    }
-–
-–
+    /**
-–
+     * Increments event count and releases waiting threads.
-–
+     */
-–
+    private void signalIdleWorkers() {
-–
+        long c;
-–
+        do {} while (!casEventCount(c = eventCount, c+1));
-–
+        ensureSync();
-–
+    }
-–
-–
+    /**
-–
+     * Signals threads waiting to poll a task. Because method sync
-–
+     * rechecks availability, it is OK to only proceed if queue
-–
+     * appears to be non-empty, and OK to skip under contention to
-–
+     * increment count (since some other thread succeeded).
-–
+     */
-–
+    final void signalWork() {
-–
+        long c;
-–
+        WaitQueueNode q;
-–
+        if (syncStack != null &&
-–
+            casEventCount(c = eventCount, c+1) &&
-–
+            (((q = syncStack) != null && q.count <= c) &&
-–
+             (!casBarrierStack(q, q.next) || !q.signal())))
-–
+            ensureSync();
-–
+    }
-–
-–
+    /**
-–
+     * Waits until event count advances from last value held by
-–
+     * caller, or if excess threads, caller is resumed as spare, or
-–
+     * caller or pool is terminating. Updates caller's event on exit.
-–
+     *
-–
+     * @param w the calling worker thread
-–
+     */
-–
+    final void sync(ForkJoinWorkerThread w) {
-–
+        updateStealCount(w); // Transfer w's count while it is idle
-–
-–
+        while (!w.isShutdown() && !isTerminating() && !suspendIfSpare(w)) {
-–
+            long prev = w.lastEventCount;
-–
+            WaitQueueNode node = null;
-–
+            WaitQueueNode h;
-–
+            while (eventCount == prev &&
-–
+                   ((h = syncStack) == null || h.count == prev)) {
-–
+                if (node == null)
-–
+                    node = new WaitQueueNode(prev, w);
-–
+                if (casBarrierStack(node.next = h, node)) {
-–
+                    node.awaitSyncRelease(this);
-–
+                    break;
-–
+                }
-–
+            }
-–
+            long ec = ensureSync();
-–
+            if (ec != prev) {
-–
+                w.lastEventCount = ec;
-–
+                break;
-–
+            }
-–
+        }
-–
+    }
-–
-–
+    /**
-–
+     * Returns {@code true} if worker waiting on sync can proceed:
-–
+     *  - on signal (thread == null)
-–
+     *  - on event count advance (winning race to notify vs signaller)
-–
+     *  - on interrupt
-–
+     *  - if the first queued node, we find work available
-–
+     * If node was not signalled and event count not advanced on exit,
-–
+     * then we also help advance event count.
-–
+     *
-–
+     * @return {@code true} if node can be released
-–
+     */
-–
+    final boolean syncIsReleasable(WaitQueueNode node) {
-–
+        long prev = node.count;
-–
+        if (!Thread.interrupted() && node.thread != null &&
-–
+            (node.next != null ||
-–
+             !ForkJoinWorkerThread.hasQueuedTasks(workers)) &&
-–
+            eventCount == prev)
-–
+            return false;
-–
+        if (node.thread != null) {
-–
+            node.thread = null;
-–
+            long ec = eventCount;
-–
+            if (prev <= ec) // help signal
-–
+                casEventCount(ec, ec+1);
-–
+        }
-–
+        return true;
-–
+    }
-–
-–
+    /**
-–
+     * Returns {@code true} if a new sync event occurred since last
-–
+     * call to sync or this method, if so, updating caller's count.
-–
+     */
-–
+    final boolean hasNewSyncEvent(ForkJoinWorkerThread w) {
-–
+        long lc = w.lastEventCount;
-–
+        long ec = ensureSync();
-–
+        if (ec == lc)
-–
+            return false;
-–
+        w.lastEventCount = ec;
-–
+        return true;
-–
+    }
-–
-–
+    //  Parallelism maintenance
-–
-–
+    /**
-–
+     * Decrements running count; if too low, adds spare.
-–
+     *
-–
+     * Conceptually, all we need to do here is add or resume a
-–
+     * spare thread when one is about to block (and remove or
-–
+     * suspend it later when unblocked -- see suspendIfSpare).
-–
+     * However, implementing this idea requires coping with
-–
+     * several problems: we have imperfect information about the
-–
+     * states of threads. Some count updates can and usually do
-–
+     * lag run state changes, despite arrangements to keep them
-–
+     * accurate (for example, when possible, updating counts
-–
+     * before signalling or resuming), especially when running on
-–
+     * dynamic JVMs that don't optimize the infrequent paths that
-–
+     * update counts. Generating too many threads can make these
-–
+     * problems become worse, because excess threads are more
-–
+     * likely to be context-switched with others, slowing them all
-–
+     * down, especially if there is no work available, so all are
-–
+     * busy scanning or idling.  Also, excess spare threads can
-–
+     * only be suspended or removed when they are idle, not
-–
+     * immediately when they aren't needed. So adding threads will
-–
+     * raise parallelism level for longer than necessary.  Also,
-–
+     * FJ applications often encounter highly transient peaks when
-–
+     * many threads are blocked joining, but for less time than it
-–
+     * takes to create or resume spares.
-–
+     *
-–
+     * @param joinMe if non-null, return early if done
-–
+     * @param maintainParallelism if true, try to stay within
-–
+     * target counts, else create only to avoid starvation
-–
+     * @return true if joinMe known to be done
-–
+     */
-–
+    final boolean preJoin(ForkJoinTask<?> joinMe,
-–
+                          boolean maintainParallelism) {
-–
+        maintainParallelism &= maintainsParallelism; // overrride
-–
+        boolean dec = false;  // true when running count decremented
-–
+        while (spareStack == null || !tryResumeSpare(dec)) {
-–
+            int counts = workerCounts;
-–
+            if (dec || (dec = casWorkerCounts(counts, --counts))) {
-–
+                // CAS cheat
-–
+                if (!needSpare(counts, maintainParallelism))
-–
+                    break;
-–
+                if (joinMe.status < 0)
-–
+                    return true;
-–
+                if (tryAddSpare(counts))
-–
+                    break;
-–
+            }
-–
+        }
-–
+        return false;
-–
+    }
-–
-–
+    /**
-–
+     * Same idea as preJoin
-–
+     */
-–
+    final boolean preBlock(ManagedBlocker blocker,
-–
+                           boolean maintainParallelism) {
-–
+        maintainParallelism &= maintainsParallelism;
-–
+        boolean dec = false;
-–
+        while (spareStack == null || !tryResumeSpare(dec)) {
-–
+            int counts = workerCounts;
-–
+            if (dec || (dec = casWorkerCounts(counts, --counts))) {
-–
+                if (!needSpare(counts, maintainParallelism))
-–
+                    break;
-–
+                if (blocker.isReleasable())
-–
+                    return true;
-–
+                if (tryAddSpare(counts))
-–
+                    break;
-–
+            }
-–
+        }
-–
+        return false;
-–
+    }
-–
-–
+    /**
-–
+     * Returns {@code true} if a spare thread appears to be needed.
-–
+     * If maintaining parallelism, returns true when the deficit in
-–
+     * running threads is more than the surplus of total threads, and
-–
+     * there is apparently some work to do.  This self-limiting rule
-–
+     * means that the more threads that have already been added, the
-–
+     * less parallelism we will tolerate before adding another.
-–
+     *
-–
+     * @param counts current worker counts
-–
+     * @param maintainParallelism try to maintain parallelism
-–
+     */
-–
+    private boolean needSpare(int counts, boolean maintainParallelism) {
-–
+        int ps = parallelism;
-–
+        int rc = runningCountOf(counts);
-–
+        int tc = totalCountOf(counts);
-–
+        int runningDeficit = ps - rc;
-–
+        int totalSurplus = tc - ps;
-–
+        return (tc < maxPoolSize &&
-–
+                (rc == 0 || totalSurplus < 0 ||
-–
+                 (maintainParallelism &&
-–
+                  runningDeficit > totalSurplus &&
-–
+                  ForkJoinWorkerThread.hasQueuedTasks(workers))));
-–
+    }
-–
-–
+    /**
-–
+     * Adds a spare worker if lock available and no more than the
-–
+     * expected numbers of threads exist.
-–
+     *
-–
+     * @return true if successful
-–
+     */
-–
+    private boolean tryAddSpare(int expectedCounts) {
-–
+        final ReentrantLock lock = this.workerLock;
-–
+        int expectedRunning = runningCountOf(expectedCounts);
-–
+        int expectedTotal = totalCountOf(expectedCounts);
-–
+        boolean success = false;
-–
+        boolean locked = false;
-–
+        // confirm counts while locking; CAS after obtaining lock
-–
+        try {
-–
+            for (;;) {
-–
+                int s = workerCounts;
-–
+                int tc = totalCountOf(s);
-–
+                int rc = runningCountOf(s);
-–
+                if (rc > expectedRunning || tc > expectedTotal)
-–
+                    break;
-–
+                if (!locked && !(locked = lock.tryLock()))
-–
+                    break;
-–
+                if (casWorkerCounts(s, workerCountsFor(tc+1, rc+1))) {
-–
+                    createAndStartSpare(tc);
-–
+                    success = true;
-–
+                    break;
-–
+                }
-–
+            }
-–
+        } finally {
-–
+            if (locked)
-–
+                lock.unlock();
-–
+        }
-–
+        return success;
-–
+    }
-–
-–
+    /**
-–
+     * Adds the kth spare worker. On entry, pool counts are already
-–
+     * adjusted to reflect addition.
-–
+     */
-–
+    private void createAndStartSpare(int k) {
-–
+        ForkJoinWorkerThread w = null;
-–
+        ForkJoinWorkerThread[] ws = ensureWorkerArrayCapacity(k + 1);
-–
+        int len = ws.length;
-–
+        // Probably, we can place at slot k. If not, find empty slot
-–
+        if (k < len && ws[k] != null) {
-–
+            for (k = 0; k < len && ws[k] != null; ++k)
-–
+                ;
-–
+        }
-–
+        if (k < len && !isTerminating() && (w = createWorker(k)) != null) {
-–
+            ws[k] = w;
-–
+            w.start();
-–
+        }
-–
+        else
-–
+            updateWorkerCount(-1); // adjust on failure
-–
+        signalIdleWorkers();
-–
+    }
-–
-–
+    /**
-–
+     * Suspends calling thread w if there are excess threads.  Called
-–
+     * only from sync.  Spares are enqueued in a Treiber stack using
-–
+     * the same WaitQueueNodes as barriers.  They are resumed mainly
-–
+     * in preJoin, but are also woken on pool events that require all
-–
+     * threads to check run state.
-–
+     *
-–
+     * @param w the caller
-–
+     */
-–
+    private boolean suspendIfSpare(ForkJoinWorkerThread w) {
-–
+        WaitQueueNode node = null;
-–
+        int s;
-–
+        while (parallelism < runningCountOf(s = workerCounts)) {
-–
+            if (node == null)
-–
+                node = new WaitQueueNode(0, w);
-–
+            if (casWorkerCounts(s, s-1)) { // representation-dependent
-–
+                // push onto stack
-–
+                do {} while (!casSpareStack(node.next = spareStack, node));
-–
+                // block until released by resumeSpare
-–
+                node.awaitSpareRelease();
-–
+                return true;
-–
+            }
-–
+        }
-–
+        return false;
-–
+    }
-–
-–
+    /**
-–
+     * Tries to pop and resume a spare thread.
-–
+     *
-–
+     * @param updateCount if true, increment running count on success
-–
+     * @return true if successful
-–
+     */
-–
+    private boolean tryResumeSpare(boolean updateCount) {
-–
+        WaitQueueNode q;
-–
+        while ((q = spareStack) != null) {
-–
+            if (casSpareStack(q, q.next)) {
-–
+                if (updateCount)
-–
+                    updateRunningCount(1);
-–
+                q.signal();
-–
+                return true;
-–
+            }
-–
+        }
-–
+        return false;
-–
+    }
-–
-–
+    /**
-–
+     * Pops and resumes all spare threads. Same idea as ensureSync.
-–
+     *
-–
+     * @return true if any spares released
-–
+     */
-–
+    private boolean resumeAllSpares() {
-–
+        WaitQueueNode q;
-–
+        while ( (q = spareStack) != null) {
-–
+            if (casSpareStack(q, null)) {
-–
+                do {
-–
+                    updateRunningCount(1);
-–
+                    q.signal();
-–
+                } while ((q = q.next) != null);
-–
+                return true;
-–
+            }
-–
+        }
-–
+        return false;
-–
+    }
-–
-–
+    /**
-–
+     * Pops and shuts down excessive spare threads. Call only while
-–
+     * holding lock. This is not guaranteed to eliminate all excess
-–
+     * threads, only those suspended as spares, which are the ones
-–
+     * unlikely to be needed in the future.
-–
+     */
-–
+    private void trimSpares() {
-–
+        int surplus = totalCountOf(workerCounts) - parallelism;
-–
+        WaitQueueNode q;
-–
+        while (surplus > 0 && (q = spareStack) != null) {
-–
+            if (casSpareStack(q, null)) {
-–
+                do {
-–
+                    updateRunningCount(1);
-–
+                    ForkJoinWorkerThread w = q.thread;
-–
+                    if (w != null && surplus > 0 &&
-–
+                        runningCountOf(workerCounts) > 0 && w.shutdown())
-–
+                        --surplus;
-–
+                    q.signal();
-–
+                } while ((q = q.next) != null);
-–
+            }
-–
+        }
-–
+    }
-–
+    /**
+     * Interface for extending managed parallelism for tasks running
+     * in {@link ForkJoinPool}s.
-<
+     * <p>A {@code ManagedBlocker} provides two methods.
-<
+     * Method {@code isReleasable} must return {@code true} if
-<
+     * blocking is not necessary. Method {@code block} blocks the
-<
+     * current thread if necessary (perhaps internally invoking
-<
+     * {@code isReleasable} before actually blocking).
->
+     * <p>A {@code ManagedBlocker} provides two methods.  Method
->
+     * {@code isReleasable} must return {@code true} if blocking is
->
+     * not necessary. Method {@code block} blocks the current thread
->
+     * if necessary (perhaps internally invoking {@code isReleasable}
->
+     * before actually blocking). These actions are performed by any
->
+     * thread invoking {@link ForkJoinPool#managedBlock}.  The
->
+     * unusual methods in this API accommodate synchronizers that may,
->
+     * but don't usually, block for long periods. Similarly, they
->
+     * allow more efficient internal handling of cases in which
->
+     * additional workers may be, but usually are not, needed to
->
+     * ensure sufficient parallelism.  Toward this end,
->
+     * implementations of method {@code isReleasable} must be amenable
->
+     * to repeated invocation.
+     * <p>For example, here is a ManagedBlocker based on a
+     * ReentrantLock:
+     *     return hasLock || (hasLock = lock.tryLock());
+     *   }
+     * }}</pre>
-+
+     *
-+
+     * <p>Here is a class that possibly blocks waiting for an
-+
+     * item on a given queue:
-+
+     *  <pre> {@code
-+
+     * class QueueTaker<E> implements ManagedBlocker {
-+
+     *   final BlockingQueue<E> queue;
-+
+     *   volatile E item = null;
-+
+     *   QueueTaker(BlockingQueue<E> q) { this.queue = q; }
-+
+     *   public boolean block() throws InterruptedException {
-+
+     *     if (item == null)
-+
+     *       item = queue.take();
-+
+     *     return true;
-+
+     *   }
-+
+     *   public boolean isReleasable() {
-+
+     *     return item != null || (item = queue.poll()) != null;
-+
+     *   }
-+
+     *   public E getItem() { // call after pool.managedBlock completes
-+
+     *     return item;
-+
+     *   }
-+
+     * }}</pre>
+     */
+    public static interface ManagedBlocker {
+        /**
+     * Blocks in accord with the given blocker.  If the current thread
+     * is a {@link ForkJoinWorkerThread}, this method possibly
+     * arranges for a spare thread to be activated if necessary to
-<
+     * ensure parallelism while the current thread is blocked.
-<
+     *
-<
+     * <p>If {@code maintainParallelism} is {@code true} and the pool
-<
+     * supports it ({@link #getMaintainsParallelism}), this method
-<
+     * attempts to maintain the pool's nominal parallelism. Otherwise
-<
+     * it activates a thread only if necessary to avoid complete
-<
+     * starvation. This option may be preferable when blockages use
-<
+     * timeouts, or are almost always brief.
->
+     * ensure sufficient parallelism while the current thread is blocked.
+     * <p>If the caller is not a {@link ForkJoinTask}, this method is
+     * behaviorally equivalent to
+     * first be expanded to ensure parallelism, and later adjusted.
+     * @param blocker the blocker
-–
+     * @param maintainParallelism if {@code true} and supported by
-–
+     * this pool, attempt to maintain the pool's nominal parallelism;
-–
+     * otherwise activate a thread only if necessary to avoid
-–
+     * complete starvation.
+     * @throws InterruptedException if blocker.block did so
+     */
-<
+    public static void managedBlock(ManagedBlocker blocker,
-<
+                                    boolean maintainParallelism)
->
+    public static void managedBlock(ManagedBlocker blocker)
+        throws InterruptedException {
+        Thread t = Thread.currentThread();
-<
+        ForkJoinPool pool = ((t instanceof ForkJoinWorkerThread) ?
-<
+                             ((ForkJoinWorkerThread) t).pool : null);
-<
+        if (!blocker.isReleasable()) {
-<
+            try {
-<
+                if (pool == null ||
-<
+                    !pool.preBlock(blocker, maintainParallelism))
-<
+                    awaitBlocker(blocker);
-<
+            } finally {
-<
+                if (pool != null)
-<
+                    pool.updateRunningCount(1);
->
+        ForkJoinPool p = ((t instanceof ForkJoinWorkerThread) ?
->
+                          ((ForkJoinWorkerThread)t).pool : null);
->
+        while (!blocker.isReleasable()) {
->
+            if (p == null || p.tryCompensate(null, blocker)) {
->
+                try {
->
+                    do {} while (!blocker.isReleasable() && !blocker.block());
->
+                } finally {
->
+                    if (p != null)
->
+                        p.incrementActiveCount();
->
+                }
->
+                break;
-–
+    private static void awaitBlocker(ManagedBlocker blocker)
-–
+        throws InterruptedException {
-–
+        do {} while (!blocker.isReleasable() && !blocker.block());
-–
+    }
-–
+    // AbstractExecutorService overrides.  These rely on undocumented
+    // fact that ForkJoinTask.adapt returns ForkJoinTasks that also
+    // 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 = getUnsafe();
-<
+    private static final long eventCountOffset =
-<
+        objectFieldOffset("eventCount", ForkJoinPool.class);
-<
+    private static final long workerCountsOffset =
-<
+        objectFieldOffset("workerCounts", ForkJoinPool.class);
-<
+    private static final long runControlOffset =
-<
+        objectFieldOffset("runControl", ForkJoinPool.class);
-<
+    private static final long syncStackOffset =
-<
+        objectFieldOffset("syncStack",ForkJoinPool.class);
-<
+    private static final long spareStackOffset =
-<
+        objectFieldOffset("spareStack", ForkJoinPool.class);
-<
-<
+    private boolean casEventCount(long cmp, long val) {
-<
+        return UNSAFE.compareAndSwapLong(this, eventCountOffset, cmp, val);
-<
+    }
-<
+    private boolean casWorkerCounts(int cmp, int val) {
-<
+        return UNSAFE.compareAndSwapInt(this, workerCountsOffset, cmp, val);
-<
+    }
-<
+    private boolean casRunControl(int cmp, int val) {
-<
+        return UNSAFE.compareAndSwapInt(this, runControlOffset, cmp, val);
-<
+    }
-<
+    private boolean casSpareStack(WaitQueueNode cmp, WaitQueueNode val) {
-<
+        return UNSAFE.compareAndSwapObject(this, spareStackOffset, cmp, val);
-<
+    }
-<
+    private boolean casBarrierStack(WaitQueueNode cmp, WaitQueueNode val) {
-<
+        return UNSAFE.compareAndSwapObject(this, syncStackOffset, cmp, val);
-<
+    }
-<
-<
+    private static long objectFieldOffset(String field, Class<?> klazz) {
->
+    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;
->
->
+    static {
->
+        poolNumberGenerator = new AtomicInteger();
->
+        nextSubmitterSeed = new AtomicInteger(0x55555555);
->
+        modifyThreadPermission = new RuntimePermission("modifyThread");
->
+        defaultForkJoinWorkerThreadFactory =
->
+            new DefaultForkJoinWorkerThreadFactory();
->
+        submitters = new ThreadSubmitter();
->
+        int s;
+        try {
-<
+            return UNSAFE.objectFieldOffset(klazz.getDeclaredField(field));
-<
+        } catch (NoSuchFieldException e) {
-<
+            // Convert Exception to corresponding Error
-<
+            NoSuchFieldError error = new NoSuchFieldError(field);
-<
+            error.initCause(e);
-<
+            throw error;
-<
+        }
->
+            U = getUnsafe();
->
+            Class<?> k = ForkJoinPool.class;
->
+            Class<?> ak = ForkJoinTask[].class;
->
+            CTL = U.objectFieldOffset
->
+                (k.getDeclaredField("ctl"));
->
+            Class<?> tk = Thread.class;
->
+            PARKBLOCKER = U.objectFieldOffset
->
+                (tk.getDeclaredField("parkBlocker"));
->
+            ABASE = U.arrayBaseOffset(ak);
->
+            s = U.arrayIndexScale(ak);
->
+        } 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 Legend

-–
+Removed lines
-+
+Added lines
-<
+Changed lines
->
+Changed lines

Comparing jsr166/src/jsr166y/ForkJoinPool.java (file contents): Revision 1.40 by jsr166, Mon Aug 3 00:53:15 2009 UTC vs. Revision 1.130 by jsr166, Mon Aug 13 18:25:53 2012 UTC

Diff Legend

Comparing jsr166/src/jsr166y/ForkJoinPool.java (file contents):
Revision 1.40 by jsr166, Mon Aug 3 00:53:15 2009 UTC vs.
Revision 1.130 by jsr166, Mon Aug 13 18:25:53 2012 UTC