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

Comparing jsr166/src/jsr166y/ForkJoinPool.java (file contents):
Revision 1.54 by dl, Sun Apr 18 12:51:18 2010 UTC vs.
Revision 1.64 by dl, Tue Aug 17 18:30:32 2010 UTC

#	Line 21 \| Line 21 \| import java.util.concurrent.CountDownLat
21		/**
22		* An {@link ExecutorService} for running {@link ForkJoinTask}s.
23		* A {@code ForkJoinPool} provides the entry point for submissions
24	<	* from non-{@code ForkJoinTask}s, as well as management and
24	>	* from non-{@code ForkJoinTask} clients, as well as management and
25		* monitoring operations.
26		*
27		* <p>A {@code ForkJoinPool} differs from other kinds of {@link
#	Line 30 \| Line 30 \| import java.util.concurrent.CountDownLat
30		* execute subtasks created by other active tasks (eventually blocking
31		* waiting for work if none exist). This enables efficient processing
32		* when most tasks spawn other subtasks (as do most {@code
33	<	* ForkJoinTask}s). A {@code ForkJoinPool} may also be used for mixed
34	<	* execution of some plain {@code Runnable}- or {@code Callable}-
35	<	* based activities along with {@code ForkJoinTask}s. When setting
36	<	* {@linkplain #setAsyncMode async mode}, a {@code ForkJoinPool} may
37	<	* also be appropriate for use with fine-grained tasks of any form
38	<	* that are never joined. Otherwise, other {@code ExecutorService}
39	<	* implementations are typically more appropriate choices.
33	>	* ForkJoinTask}s). When setting <em>asyncMode</em> to true in
34	>	* constructors, {@code ForkJoinPool}s may also be appropriate for use
35	>	* with event-style tasks that are never joined.
36		*
37		* <p>A {@code ForkJoinPool} is constructed with a given target
38		* parallelism level; by default, equal to the number of available
39	<	* processors. Unless configured otherwise via {@link
40	<	* #setMaintainsParallelism}, the pool attempts to maintain this
41	<	* number of active (or available) threads by dynamically adding,
42	<	* suspending, or resuming internal worker threads, even if some tasks
43	<	* are stalled waiting to join others. However, no such adjustments
44	<	* are performed in the face of blocked IO or other unmanaged
45	<	* synchronization. The nested {@link ManagedBlocker} interface
50	<	* enables extension of the kinds of synchronization accommodated.
51	<	* The target parallelism level may also be changed dynamically
52	<	* ({@link #setParallelism}). The total number of threads may be
53	<	* limited using method {@link #setMaximumPoolSize}, in which case it
54	<	* may become possible for the activities of a pool to stall due to
55	<	* the lack of available threads to process new tasks. When the pool
56	<	* is executing tasks, these and other configuration setting methods
57	<	* may only gradually affect actual pool sizes. It is normally best
58	<	* practice to invoke these methods only when the pool is known to be
59	<	* quiescent.
39	>	* processors. The pool attempts to maintain enough active (or
40	>	* available) threads by dynamically adding, suspending, or resuming
41	>	* internal worker threads, even if some tasks are stalled waiting to
42	>	* join others. However, no such adjustments are guaranteed in the
43	>	* face of blocked IO or other unmanaged synchronization. The nested
44	>	* {@link ManagedBlocker} interface enables extension of the kinds of
45	>	* synchronization accommodated.
46		*
47		* <p>In addition to execution and lifecycle control methods, this
48		* class provides status check methods (for example
#	Line 65 \| Line 51 \| import java.util.concurrent.CountDownLat
51		* {@link #toString} returns indications of pool state in a
52		* convenient form for informal monitoring.
53		*
54	+	* <p> As is the case with other ExecutorServices, there are three
55	+	* main task execution methods summarized in the following
56	+	* table. These are designed to be used by clients not already engaged
57	+	* in fork/join computations in the current pool. The main forms of
58	+	* these methods accept instances of {@code ForkJoinTask}, but
59	+	* overloaded forms also allow mixed execution of plain {@code
60	+	* Runnable}- or {@code Callable}- based activities as well. However,
61	+	* tasks that are already executing in a pool should normally
62	+	* <em>NOT</em> use these pool execution methods, but instead use the
63	+	* within-computation forms listed in the table.
64	+	*
65	+	* <table BORDER CELLPADDING=3 CELLSPACING=1>
66	+	* <tr>
67	+	* <td></td>
68	+	* <td ALIGN=CENTER> <b>Call from non-fork/join clients</b></td>
69	+	* <td ALIGN=CENTER> <b>Call from within fork/join computations</b></td>
70	+	* </tr>
71	+	* <tr>
72	+	* <td> <b>Arange async execution</td>
73	+	* <td> {@link #execute(ForkJoinTask)}</td>
74	+	* <td> {@link ForkJoinTask#fork}</td>
75	+	* </tr>
76	+	* <tr>
77	+	* <td> <b>Await and obtain result</td>
78	+	* <td> {@link #invoke(ForkJoinTask)}</td>
79	+	* <td> {@link ForkJoinTask#invoke}</td>
80	+	* </tr>
81	+	* <tr>
82	+	* <td> <b>Arrange exec and obtain Future</td>
83	+	* <td> {@link #submit(ForkJoinTask)}</td>
84	+	* <td> {@link ForkJoinTask#fork} (ForkJoinTasks <em>are</em> Futures)</td>
85	+	* </tr>
86	+	* </table>
87	+	*
88		* <p><b>Sample Usage.</b> Normally a single {@code ForkJoinPool} is
89		* used for all parallel task execution in a program or subsystem.
90		* Otherwise, use would not usually outweigh the construction and
#	Line 89 \| Line 109 \| import java.util.concurrent.CountDownLat
109		* {@code IllegalArgumentException}.
110		*
111		* <p>This implementation rejects submitted tasks (that is, by throwing
112	<	* {@link RejectedExecutionException}) only when the pool is shut down.
112	>	* {@link RejectedExecutionException}) only when the pool is shut down
113	>	* or internal resources have been exhausted.
114		*
115		* @since 1.7
116		* @author Doug Lea
#	Line 116 \| Line 137 \| public class ForkJoinPool extends Abstra
137		* of tasks profit from cache affinities, but others are harmed by
138		* cache pollution effects.)
139		*
140	+	* Beyond work-stealing support and essential bookkeeping, the
141	+	* main responsibility of this framework is to take actions when
142	+	* one worker is waiting to join a task stolen (or always held by)
143	+	* another. Becauae we are multiplexing many tasks on to a pool
144	+	* of workers, we can't just let them block (as in Thread.join).
145	+	* We also cannot just reassign the joiner's run-time stack with
146	+	* another and replace it later, which would be a form of
147	+	* "continuation", that even if possible is not necessarily a good
148	+	* idea. Given that the creation costs of most threads on most
149	+	* systems mainly surrounds setting up runtime stacks, thread
150	+	* creation and switching is usually not much more expensive than
151	+	* stack creation and switching, and is more flexible). Instead we
152	+	* combine two tactics:
153	+	*
154	+	* Helping: Arranging for the joiner to execute some task that it
155	+	* would be running if the steal had not occurred. Method
156	+	* ForkJoinWorkerThread.helpJoinTask tracks joining->stealing
157	+	* links to try to find such a task.
158	+	*
159	+	* Compensating: Unless there are already enough live threads,
160	+	* method helpMaintainParallelism() may create or or
161	+	* re-activate a spare thread to compensate for blocked
162	+	* joiners until they unblock.
163	+	*
164	+	* Because the determining existence of conservatively safe
165	+	* helping targets, the availability of already-created spares,
166	+	* and the apparent need to create new spares are all racy and
167	+	* require heuristic guidance, we rely on multiple retries of
168	+	* each. Further, because it is impossible to keep exactly the
169	+	* target (parallelism) number of threads running at any given
170	+	* time, we allow compensation during joins to fail, and enlist
171	+	* all other threads to help out whenever they are not otherwise
172	+	* occupied (i.e., mainly in method preStep).
173	+	*
174	+	* The ManagedBlocker extension API can't use helping so relies
175	+	* only on compensation in method awaitBlocker.
176	+	*
177		* The main throughput advantages of work-stealing stem from
178		* decentralized control -- workers mostly steal tasks from each
179		* other. We do not want to negate this by creating bottlenecks
180	<	* implementing the management responsibilities of this class. So
181	<	* we use a collection of techniques that avoid, reduce, or cope
182	<	* well with contention. These entail several instances of
183	<	* bit-packing into CASable fields to maintain only the minimally
184	<	* required atomicity. To enable such packing, we restrict maximum
185	<	* parallelism to (1<<15)-1 (enabling twice this to fit into a 16
186	<	* bit field), which is far in excess of normal operating range.
187	<	* Even though updates to some of these bookkeeping fields do
188	<	* sometimes contend with each other, they don't normally
189	<	* cache-contend with updates to others enough to warrant memory
190	<	* padding or isolation. So they are all held as fields of
191	<	* ForkJoinPool objects. The main capabilities are as follows:
180	>	* implementing other management responsibilities. So we use a
181	>	* collection of techniques that avoid, reduce, or cope well with
182	>	* contention. These entail several instances of bit-packing into
183	>	* CASable fields to maintain only the minimally required
184	>	* atomicity. To enable such packing, we restrict maximum
185	>	* parallelism to (1<<15)-1 (enabling twice this (to accommodate
186	>	* unbalanced increments and decrements) to fit into a 16 bit
187	>	* field, which is far in excess of normal operating range. Even
188	>	* though updates to some of these bookkeeping fields do sometimes
189	>	* contend with each other, they don't normally cache-contend with
190	>	* updates to others enough to warrant memory padding or
191	>	* isolation. So they are all held as fields of ForkJoinPool
192	>	* objects. The main capabilities are as follows:
193		*
194		* 1. Creating and removing workers. Workers are recorded in the
195		* "workers" array. This is an array as opposed to some other data
#	Line 140 \| Line 199 \| public class ForkJoinPool extends Abstra
199		* (workerLock) but the array is otherwise concurrently readable,
200		* and accessed directly by workers. To simplify index-based
201		* operations, the array size is always a power of two, and all
202	<	* readers must tolerate null slots. Currently, all but the first
203	<	* worker thread creation is on-demand, triggered by task
204	<	* submissions, replacement of terminated workers, and/or
205	<	* compensation for blocked workers. However, all other support
206	<	* code is set up to work with other policies.
202	>	* readers must tolerate null slots. Currently, all worker thread
203	>	* creation is on-demand, triggered by task submissions,
204	>	* replacement of terminated workers, and/or compensation for
205	>	* blocked workers. However, all other support code is set up to
206	>	* work with other policies.
207	>	*
208	>	* To ensure that we do not hold on to worker references that
209	>	* would prevent GC, ALL accesses to workers are via indices into
210	>	* the workers array (which is one source of some of the unusual
211	>	* code constructions here). In essence, the workers array serves
212	>	* as a WeakReference mechanism. Thus for example the event queue
213	>	* stores worker indices, not worker references. Access to the
214	>	* workers in associated methods (for example releaseEventWaiters)
215	>	* must both index-check and null-check the IDs. All such accesses
216	>	* ignore bad IDs by returning out early from what they are doing,
217	>	* since this can only be associated with shutdown, in which case
218	>	* it is OK to give up. On termination, we just clobber these
219	>	* data structures without trying to use them.
220		*
221		* 2. Bookkeeping for dynamically adding and removing workers. We
222	<	* maintain a given level of parallelism (or, if
223	<	* maintainsParallelism is false, at least avoid starvation). When
152	<	* some workers are known to be blocked (on joins or via
222	>	* aim to approximately maintain the given level of parallelism.
223	>	* When some workers are known to be blocked (on joins or via
224		* ManagedBlocker), we may create or resume others to take their
225		* place until they unblock (see below). Implementing this
226		* requires counts of the number of "running" threads (i.e., those
227		* that are neither blocked nor artifically suspended) as well as
228		* the total number. These two values are packed into one field,
229		* "workerCounts" because we need accurate snapshots when deciding
230	<	* to create, resume or suspend. To support these decisions,
231	<	* updates must be prospective (not retrospective). For example,
232	<	* the running count is decremented before blocking by a thread
233	<	* about to block, but incremented by the thread about to unblock
163	<	* it. (In a few cases, these prospective updates may need to be
164	<	* rolled back, for example when deciding to create a new worker
165	<	* but the thread factory fails or returns null. In these cases,
166	<	* we are no worse off wrt other decisions than we would be
167	<	* otherwise.) Updates to the workerCounts field sometimes
168	<	* transiently encounter a fair amount of contention when join
169	<	* dependencies are such that many threads block or unblock at
170	<	* about the same time. We alleviate this by sometimes bundling
171	<	* updates (for example blocking one thread on join and resuming a
172	<	* spare cancel each other out), and in most other cases
173	<	* performing an alternative action (like releasing waiters and
174	<	* finding spares; see below) as a more productive form of
175	<	* backoff.
230	>	* to create, resume or suspend. Note however that the
231	>	* correspondance of these counts to reality is not guaranteed. In
232	>	* particular updates for unblocked threads may lag until they
233	>	* actually wake up.
234		*
235		* 3. Maintaining global run state. The run state of the pool
236		* consists of a runLevel (SHUTDOWN, TERMINATING, etc) similar to
#	Line 201 \| Line 259 \| public class ForkJoinPool extends Abstra
259		* workers that previously could not find a task to now find one:
260		* Submission of a new task to the pool, or another worker pushing
261		* a task onto a previously empty queue. (We also use this
262	<	* mechanism for termination and reconfiguration actions that
262	>	* mechanism for configuration and termination actions that
263		* require wakeups of idle workers). Each worker maintains its
264		* last known event count, and blocks when a scan for work did not
265		* find a task AND its lastEventCount matches the current
#	Line 212 \| Line 270 \| public class ForkJoinPool extends Abstra
270		* a record (field nextEventWaiter) for the next waiting worker.
271		* In addition to allowing simpler decisions about need for
272		* wakeup, the event count bits in eventWaiters serve the role of
273	<	* tags to avoid ABA errors in Treiber stacks. To reduce delays
274	<	* in task diffusion, workers not otherwise occupied may invoke
275	<	* method releaseWaiters, that removes and signals (unparks)
276	<	* workers not waiting on current count. To minimize task
277	<	* production stalls associate with signalling, any worker pushing
278	<	* a task on an empty queue invokes the weaker method signalWork,
279	<	* that only releases idle workers until it detects interference
222	<	* by other threads trying to release, and lets them take
223	<	* over. The net effect is a tree-like diffusion of signals, where
224	<	* released threads and possibly others) help with unparks. To
225	<	* further reduce contention effects a bit, failed CASes to
226	<	* increment field eventCount are tolerated without retries.
273	>	* tags to avoid ABA errors in Treiber stacks. Upon any wakeup,
274	>	* released threads also try to release others (but give up upon
275	>	* contention to reduce useless flailing). The net effect is a
276	>	* tree-like diffusion of signals, where released threads (and
277	>	* possibly others) help with unparks. To further reduce
278	>	* contention effects a bit, failed CASes to increment field
279	>	* eventCount are tolerated without retries in signalWork.
280		* Conceptually they are merged into the same event, which is OK
281		* when their only purpose is to enable workers to scan for work.
282		*
283		* 5. Managing suspension of extra workers. When a worker is about
284		* to block waiting for a join (or via ManagedBlockers), we may
285		* create a new thread to maintain parallelism level, or at least
286	<	* avoid starvation (see below). Usually, extra threads are needed
287	<	* for only very short periods, yet join dependencies are such
288	<	* that we sometimes need them in bursts. Rather than create new
289	<	* threads each time this happens, we suspend no-longer-needed
290	<	* extra ones as "spares". For most purposes, we don't distinguish
291	<	* "extra" spare threads from normal "core" threads: On each call
292	<	* to preStep (the only point at which we can do this) a worker
286	>	* avoid starvation. Usually, extra threads are needed for only
287	>	* very short periods, yet join dependencies are such that we
288	>	* sometimes need them in bursts. Rather than create new threads
289	>	* each time this happens, we suspend no-longer-needed extra ones
290	>	* as "spares". For most purposes, we don't distinguish "extra"
291	>	* spare threads from normal "core" threads: On each call to
292	>	* preStep (the only point at which we can do this) a worker
293		* checks to see if there are now too many running workers, and if
294	<	* so, suspends itself. Methods preJoin and doBlock look for
294	>	* so, suspends itself. Method helpMaintainParallelism looks for
295		* suspended threads to resume before considering creating a new
296	<	* replacement. We don't need a special data structure to maintain
297	<	* spares; simply scanning the workers array looking for
298	<	* worker.isSuspended() is fine because the calling thread is
246	<	* otherwise not doing anything useful anyway; we are at least as
247	<	* happy if after locating a spare, the caller doesn't actually
248	<	* block because the join is ready before we try to adjust and
249	<	* compensate. Note that this is intrinsically racy. One thread
296	>	* replacement. The spares themselves are encoded on another
297	>	* variant of a Treiber Stack, headed at field "spareWaiters".
298	>	* Note that the use of spares is intrinsically racy. One thread
299		* may become a spare at about the same time as another is
300		* needlessly being created. We counteract this and related slop
301		* in part by requiring resumed spares to immediately recheck (in
302	<	* preStep) to see whether they they should re-suspend. The only
303	<	* effective difference between "extra" and "core" threads is that
304	<	* we allow the "extra" ones to time out and die if they are not
305	<	* resumed within a keep-alive interval of a few seconds. This is
306	<	* implemented mainly within ForkJoinWorkerThread, but requires
307	<	* some coordination (isTrimmed() -- meaning killed while
308	<	* suspended) to correctly maintain pool counts.
309	<	*
310	<	* 6. Deciding when to create new workers. The main dynamic
311	<	* control in this class is deciding when to create extra threads,
312	<	* in methods preJoin and doBlock. We always need to create one
313	<	* when the number of running threads becomes zero. But because
314	<	* blocked joins are typically dependent, we don't necessarily
315	<	* need or want one-to-one replacement. Using a one-to-one
316	<	* compensation rule often leads to enough useless overhead
317	<	* creating, suspending, resuming, and/or killing threads to
318	<	* signficantly degrade throughput. We use a rule reflecting the
319	<	* idea that, the more spare threads you already have, the more
320	<	* evidence you need to create another one; where "evidence" is
321	<	* expressed as the current deficit -- target minus running
322	<	* threads. To reduce flickering and drift around target values,
323	<	* the relation is quadratic: adding a spare if (dcdc)>=(scpc)
324	<	* (where dc is deficit, sc is number of spare threads and pc is
325	<	* target parallelism.) This effectively reduces churn at the
326	<	* price of systematically undershooting target parallelism when
327	<	* many threads are blocked. However, biasing toward undeshooting
328	<	* partially compensates for the above mechanics to suspend extra
329	<	* threads, that normally lead to overshoot because we can only
330	<	* suspend workers in-between top-level actions. It also better
331	<	* copes with the fact that some of the methods in this class tend
332	<	* to never become compiled (but are interpreted), so some
333	<	* components of the entire set of controls might execute many
302	>	* preStep) to see whether they they should re-suspend.
303	>	*
304	>	* 6. Killing off unneeded workers. The Spare and Event queues use
305	>	* similar mechanisms to shed unused workers: The oldest (first)
306	>	* waiter uses a timed rather than hard wait. When this wait times
307	>	* out without a normal wakeup, it tries to shutdown any one (for
308	>	* convenience the newest) other waiter via tryShutdownSpare or
309	>	* tryShutdownWaiter, respectively. The wakeup rates for spares
310	>	* are much shorter than for waiters. Together, they will
311	>	* eventually reduce the number of worker threads to a minimum of
312	>	* one after a long enough period without use.
313	>	*
314	>	* 7. Deciding when to create new workers. The main dynamic
315	>	* control in this class is deciding when to create extra threads
316	>	* in method helpMaintainParallelism. We would like to keep
317	>	* exactly #parallelism threads running, which is an impossble
318	>	* task. We always need to create one when the number of running
319	>	* threads would become zero and all workers are busy. Beyond
320	>	* this, we must rely on heuristics that work well in the the
321	>	* presence of transients phenomena such as GC stalls, dynamic
322	>	* compilation, and wake-up lags. These transients are extremely
323	>	* common -- we are normally trying to fully saturate the CPUs on
324	>	* a machine, so almost any activity other than running tasks
325	>	* impedes accuracy. Our main defense is to allow some slack in
326	>	* creation thresholds, using rules that reflect the fact that the
327	>	* more threads we have running, the more likely that we are
328	>	* underestimating the number running threads. (We also include
329	>	* some heuristic use of Thread.yield when all workers appear to
330	>	* be busy, to improve likelihood of counts settling.) The rules
331	>	* also better cope with the fact that some of the methods in this
332	>	* class tend to never become compiled (but are interpreted), so
333	>	* some components of the entire set of controls might execute 100
334		* times faster than others. And similarly for cases where the
335		* apparent lack of work is just due to GC stalls and other
336		* transient system activity.
337		*
289	–	* 7. Maintaining other configuration parameters and monitoring
290	–	* statistics. Updates to fields controlling parallelism level,
291	–	* max size, etc can only meaningfully take effect for individual
292	–	* threads upon their next top-level actions; i.e., between
293	–	* stealing/running tasks/submission, which are separated by calls
294	–	* to preStep. Memory ordering for these (assumed infrequent)
295	–	* reconfiguration calls is ensured by using reads and writes to
296	–	* volatile field workerCounts (that must be read in preStep anyway)
297	–	* as "fences" -- user-level reads are preceded by reads of
298	–	* workCounts, and writes are followed by no-op CAS to
299	–	* workerCounts. The values reported by other management and
300	–	* monitoring methods are either computed on demand, or are kept
301	–	* in fields that are only updated when threads are otherwise
302	–	* idle.
303	–	*
338		* Beware that there is a lot of representation-level coupling
339		* among classes ForkJoinPool, ForkJoinWorkerThread, and
340		* ForkJoinTask. For example, direct access to "workers" array by
#	Line 312 \| Line 346 \| public class ForkJoinPool extends Abstra
346		*
347		* Style notes: There are lots of inline assignments (of form
348		* "while ((local = field) != 0)") which are usually the simplest
349	<	* way to ensure read orderings. Also several occurrences of the
350	<	* unusual "do {} while(!cas...)" which is the simplest way to
351	<	* force an update of a CAS'ed variable. There are also a few
352	<	* other coding oddities that help some methods perform reasonably
353	<	* even when interpreted (not compiled).
349	>	* way to ensure the required read orderings (which are sometimes
350	>	* critical). Also several occurrences of the unusual "do {}
351	>	* while(!cas...)" which is the simplest way to force an update of
352	>	* a CAS'ed variable. There are also other coding oddities that
353	>	* help some methods perform reasonably even when interpreted (not
354	>	* compiled), at the expense of some messy constructions that
355	>	* reduce byte code counts.
356		*
357		* The order of declarations in this file is: (1) statics (2)
358		* fields (along with constants used when unpacking some of them)
#	Line 345 \| Line 381 \| public class ForkJoinPool extends Abstra
381		* Default ForkJoinWorkerThreadFactory implementation; creates a
382		* new ForkJoinWorkerThread.
383		*/
384	<	static class DefaultForkJoinWorkerThreadFactory
384	>	static class DefaultForkJoinWorkerThreadFactory
385		implements ForkJoinWorkerThreadFactory {
386		public ForkJoinWorkerThread newThread(ForkJoinPool pool) {
387		return new ForkJoinWorkerThread(pool);
#	Line 384 \| Line 420 \| public class ForkJoinPool extends Abstra
420		new AtomicInteger();
421
422		/**
423	<	* Absolute bound for parallelism level. Twice this number must
424	<	* fit into a 16bit field to enable word-packing for some counts.
423	>	* The wakeup interval (in nanoseconds) for the oldest worker
424	>	* worker waiting for an event invokes tryShutdownWaiter to shrink
425	>	* the number of workers. The exact value does not matter too
426	>	* much, but should be long enough to slowly release resources
427	>	* during long periods without use without disrupting normal use.
428		*/
429	<	private static final int MAX_THREADS = 0x7fff;
429	>	private static final long SHRINK_RATE_NANOS =
430	>	60L * 1000L * 1000L * 1000L; // one minute
431	>
432	>	/**
433	>	* Absolute bound for parallelism level. Twice this number plus
434	>	* one (i.e., 0xfff) must fit into a 16bit field to enable
435	>	* word-packing for some counts and indices.
436	>	*/
437	>	private static final int MAX_WORKERS = 0x7fff;
438
439		/**
440		* Array holding all worker threads in the pool. Array size must
#	Line 413 \| Line 460 \| public class ForkJoinPool extends Abstra
460		/**
461		* Latch released upon termination.
462		*/
463	<	private final CountDownLatch terminationLatch;
463	>	private final Phaser termination;
464
465		/**
466		* Creation factory for worker threads.
#	Line 429 \| Line 476 \| public class ForkJoinPool extends Abstra
476		/**
477		* Encoded record of top of treiber stack of threads waiting for
478		* events. The top 32 bits contain the count being waited for. The
479	<	* bottom word contains one plus the pool index of waiting worker
480	<	* thread.
479	>	* bottom 16 bits contains one plus the pool index of waiting
480	>	* worker thread. (Bits 16-31 are unused.)
481		*/
482		private volatile long eventWaiters;
483
484		private static final int EVENT_COUNT_SHIFT = 32;
485	<	private static final long WAITER_INDEX_MASK = (1L << EVENT_COUNT_SHIFT)-1L;
485	>	private static final long WAITER_ID_MASK = (1L << 16) - 1L;
486
487		/**
488		* A counter for events that may wake up worker threads:
489		* - Submission of a new task to the pool
490		* - A worker pushing a task on an empty queue
491	<	* - termination and reconfiguration
491	>	* - termination
492		*/
493		private volatile int eventCount;
494
495		/**
496	+	* Encoded record of top of treiber stack of spare threads waiting
497	+	* for resumption. The top 16 bits contain an arbitrary count to
498	+	* avoid ABA effects. The bottom 16bits contains one plus the pool
499	+	* index of waiting worker thread.
500	+	*/
501	+	private volatile int spareWaiters;
502	+
503	+	private static final int SPARE_COUNT_SHIFT = 16;
504	+	private static final int SPARE_ID_MASK = (1 << 16) - 1;
505	+
506	+	/**
507		* Lifecycle control. The low word contains the number of workers
508		* that are (probably) executing tasks. This value is atomically
509		* incremented before a worker gets a task to run, and decremented
#	Line 456 \| Line 514 \| public class ForkJoinPool extends Abstra
514		* These are bundled together to ensure consistent read for
515		* termination checks (i.e., that runLevel is at least SHUTDOWN
516		* and active threads is zero).
517	+	*
518	+	* Notes: Most direct CASes are dependent on these bitfield
519	+	* positions. Also, this field is non-private to enable direct
520	+	* performance-sensitive CASes in ForkJoinWorkerThread.
521		*/
522	<	private volatile int runState;
522	>	volatile int runState;
523
524		// Note: The order among run level values matters.
525		private static final int RUNLEVEL_SHIFT = 16;
#	Line 465 \| Line 527 \| public class ForkJoinPool extends Abstra
527		private static final int TERMINATING = 1 << (RUNLEVEL_SHIFT + 1);
528		private static final int TERMINATED = 1 << (RUNLEVEL_SHIFT + 2);
529		private static final int ACTIVE_COUNT_MASK = (1 << RUNLEVEL_SHIFT) - 1;
468	–	private static final int ONE_ACTIVE = 1; // active update delta
530
531		/**
532		* Holds number of total (i.e., created and not yet terminated)
#	Line 474 \| Line 535 \| public class ForkJoinPool extends Abstra
535		* making decisions about creating and suspending spare
536		* threads. Updated only by CAS. Note that adding a new worker
537		* requires incrementing both counts, since workers start off in
538	<	* running state. This field is also used for memory-fencing
478	<	* configuration parameters.
538	>	* running state.
539		*/
540		private volatile int workerCounts;
541
#	Line 484 \| Line 544 \| public class ForkJoinPool extends Abstra
544		private static final int ONE_RUNNING = 1;
545		private static final int ONE_TOTAL = 1 << TOTAL_COUNT_SHIFT;
546
487	–	/*
488	–	* Fields parallelism. maxPoolSize, locallyFifo,
489	–	* maintainsParallelism, and ueh are non-volatile, but external
490	–	* reads/writes use workerCount fences to ensure visability.
491	–	*/
492	–
547		/**
548		* The target parallelism level.
549	+	* Accessed directly by ForkJoinWorkerThreads.
550		*/
551	<	private int parallelism;
497	<
498	<	/**
499	<	* The maximum allowed pool size.
500	<	*/
501	<	private int maxPoolSize;
551	>	final int parallelism;
552
553		/**
554		* True if use local fifo, not default lifo, for local polling
555	<	* Replicated by ForkJoinWorkerThreads
555	>	* Read by, and replicated by ForkJoinWorkerThreads
556		*/
557	<	private boolean locallyFifo;
557	>	final boolean locallyFifo;
558
559		/**
560	<	* Controls whether to add spares to maintain parallelism
560	>	* The uncaught exception handler used when any worker abruptly
561	>	* terminates.
562		*/
563	<	private boolean maintainsParallelism;
513	<
514	<	/**
515	<	* The uncaught exception handler used when any worker
516	<	* abruptly terminates
517	<	*/
518	<	private Thread.UncaughtExceptionHandler ueh;
563	>	private final Thread.UncaughtExceptionHandler ueh;
564
565		/**
566		* Pool number, just for assigning useful names to worker threads
567		*/
568		private final int poolNumber;
569
570	<	// utilities for updating fields
570	>
571	>	// Utilities for CASing fields. Note that most of these
572	>	// are usually manually inlined by callers
573
574		/**
575	<	* Adds delta to running count. Used mainly by ForkJoinTask.
529	<	*
530	<	* @param delta the number to add
575	>	* Increments running count part of workerCounts
576		*/
577	<	final void updateRunningCount(int delta) {
578	<	int wc;
577	>	final void incrementRunningCount() {
578	>	int c;
579		do {} while (!UNSAFE.compareAndSwapInt(this, workerCountsOffset,
580	<	wc = workerCounts,
581	<	wc + delta));
580	>	c = workerCounts,
581	>	c + ONE_RUNNING));
582		}
583
584		/**
585	<	* Write fence for user modifications of pool parameters
541	<	* (parallelism. etc). Note that it doesn't matter if CAS fails.
585	>	* Tries to decrement running count unless already zero
586		*/
587	<	private void workerCountWriteFence() {
588	<	int wc;
589	<	UNSAFE.compareAndSwapInt(this, workerCountsOffset,
590	<	wc = workerCounts, wc);
587	>	final boolean tryDecrementRunningCount() {
588	>	int wc = workerCounts;
589	>	if ((wc & RUNNING_COUNT_MASK) == 0)
590	>	return false;
591	>	return UNSAFE.compareAndSwapInt(this, workerCountsOffset,
592	>	wc, wc - ONE_RUNNING);
593		}
594
595		/**
596	<	* Read fence for external reads of pool parameters
597	<	* (parallelism. maxPoolSize, etc).
596	>	* Forces decrement of encoded workerCounts, awaiting nonzero if
597	>	* (rarely) necessary when other count updates lag.
598	>	*
599	>	* @param dr -- either zero or ONE_RUNNING
600	>	* @param dt == either zero or ONE_TOTAL
601		*/
602	<	private void workerCountReadFence() {
603	<	int ignore = workerCounts;
602	>	private void decrementWorkerCounts(int dr, int dt) {
603	>	for (;;) {
604	>	int wc = workerCounts;
605	>	if ((wc & RUNNING_COUNT_MASK) - dr < 0 \|\|
606	>	(wc >>> TOTAL_COUNT_SHIFT) - dt < 0) {
607	>	if ((runState & TERMINATED) != 0)
608	>	return; // lagging termination on a backout
609	>	Thread.yield();
610	>	}
611	>	if (UNSAFE.compareAndSwapInt(this, workerCountsOffset,
612	>	wc, wc - (dr + dt)))
613	>	return;
614	>	}
615	>	}
616	>
617	>	/**
618	>	* Increments event count
619	>	*/
620	>	private void advanceEventCount() {
621	>	int c;
622	>	do {} while(!UNSAFE.compareAndSwapInt(this, eventCountOffset,
623	>	c = eventCount, c+1));
624		}
625
626		/**
#	Line 563 \| Line 632 \| public class ForkJoinPool extends Abstra
632		final boolean tryIncrementActiveCount() {
633		int c;
634		return UNSAFE.compareAndSwapInt(this, runStateOffset,
635	<	c = runState, c + ONE_ACTIVE);
635	>	c = runState, c + 1);
636		}
637
638		/**
#	Line 573 \| Line 642 \| public class ForkJoinPool extends Abstra
642		final boolean tryDecrementActiveCount() {
643		int c;
644		return UNSAFE.compareAndSwapInt(this, runStateOffset,
645	<	c = runState, c - ONE_ACTIVE);
645	>	c = runState, c - 1);
646		}
647
648		/**
#	Line 602 \| Line 671 \| public class ForkJoinPool extends Abstra
671		lock.lock();
672		try {
673		ForkJoinWorkerThread[] ws = workers;
674	<	int len = ws.length;
675	<	if (k < 0 \|\| k >= len \|\| ws[k] != null) {
676	<	for (k = 0; k < len && ws[k] != null; ++k)
674	>	int n = ws.length;
675	>	if (k < 0 \|\| k >= n \|\| ws[k] != null) {
676	>	for (k = 0; k < n && ws[k] != null; ++k)
677		;
678	<	if (k == len)
679	<	ws = Arrays.copyOf(ws, len << 1);
678	>	if (k == n)
679	>	ws = Arrays.copyOf(ws, n << 1);
680		}
681		ws[k] = w;
682		workers = ws; // volatile array write ensures slot visibility
#	Line 641 \| Line 710 \| public class ForkJoinPool extends Abstra
710		* are already updated to accommodate the worker, so adjusts on
711		* failure.
712		*
713	<	* @return new worker or null if creation failed
713	>	* @return the worker, or null on failure
714		*/
715		private ForkJoinWorkerThread addWorker() {
716		ForkJoinWorkerThread w = null;
#	Line 649 \| Line 718 \| public class ForkJoinPool extends Abstra
718		w = factory.newThread(this);
719		} finally { // Adjust on either null or exceptional factory return
720		if (w == null) {
721	<	onWorkerCreationFailure();
722	<	return null;
721	>	decrementWorkerCounts(ONE_RUNNING, ONE_TOTAL);
722	>	tryTerminate(false); // in case of failure during shutdown
723		}
724		}
725	<	w.start(recordWorker(w), locallyFifo, ueh);
726	<	return w;
727	<	}
659	<
660	<	/**
661	<	* Adjusts counts upon failure to create worker
662	<	*/
663	<	private void onWorkerCreationFailure() {
664	<	int c;
665	<	do {} while (!UNSAFE.compareAndSwapInt(this, workerCountsOffset,
666	<	c = workerCounts,
667	<	c - (ONE_RUNNING\|ONE_TOTAL)));
668	<	tryTerminate(false); // in case of failure during shutdown
669	<	}
670	<
671	<	/**
672	<	* Create enough total workers to establish target parallelism,
673	<	* giving up if terminating or addWorker fails
674	<	*/
675	<	private void ensureEnoughTotalWorkers() {
676	<	int wc;
677	<	while (runState < TERMINATING &&
678	<	((wc = workerCounts) >>> TOTAL_COUNT_SHIFT) < parallelism) {
679	<	if ((UNSAFE.compareAndSwapInt(this, workerCountsOffset,
680	<	wc, wc + (ONE_RUNNING\|ONE_TOTAL)) &&
681	<	addWorker() == null))
682	<	break;
725	>	if (w != null) {
726	>	w.start(recordWorker(w), ueh);
727	>	advanceEventCount();
728		}
729	+	return w;
730		}
731
732		/**
733		* Final callback from terminating worker. Removes record of
734		* worker from array, and adjusts counts. If pool is shutting
735	<	* down, tries to complete terminatation, else possibly replaces
690	<	* the worker.
735	>	* down, tries to complete terminatation.
736		*
737		* @param w the worker
738		*/
739		final void workerTerminated(ForkJoinWorkerThread w) {
695	–	if (w.active) { // force inactive
696	–	w.active = false;
697	–	do {} while (!tryDecrementActiveCount());
698	–	}
740		forgetWorker(w);
741	<
742	<	// decrement total count, and if was running, running count
743	<	int unit = w.isTrimmed()? ONE_TOTAL : (ONE_RUNNING\|ONE_TOTAL);
744	<	int wc;
704	<	do {} while (!UNSAFE.compareAndSwapInt(this, workerCountsOffset,
705	<	wc = workerCounts, wc - unit));
706	<
707	<	accumulateStealCount(w); // collect final count
708	<	if (!tryTerminate(false))
709	<	ensureEnoughTotalWorkers();
741	>	decrementWorkerCounts(w.isTrimmed()? 0 : ONE_RUNNING, ONE_TOTAL);
742	>	while (w.stealCount != 0) // collect final count
743	>	tryAccumulateStealCount(w);
744	>	tryTerminate(false);
745		}
746
747		// Waiting for and signalling events
748
749		/**
750	<	* Ensures eventCount on exit is different (mod 2^32) than on
751	<	* entry. CAS failures are OK -- any change in count suffices.
752	<	*/
753	<	private void advanceEventCount() {
754	<	int c;
755	<	UNSAFE.compareAndSwapInt(this, eventCountOffset, c = eventCount, c+1);
750	>	* Releases workers blocked on a count not equal to current count.
751	>	* Normally called after precheck that eventWaiters isn't zero to
752	>	* avoid wasted array checks. Gives up upon a change in count or
753	>	* contention, letting other workers take over.
754	>	*/
755	>	private void releaseEventWaiters() {
756	>	ForkJoinWorkerThread[] ws = workers;
757	>	int n = ws.length;
758	>	long h = eventWaiters;
759	>	int ec = eventCount;
760	>	ForkJoinWorkerThread w; int id;
761	>	while ((int)(h >>> EVENT_COUNT_SHIFT) != ec &&
762	>	(id = ((int)(h & WAITER_ID_MASK)) - 1) >= 0 &&
763	>	id < n && (w = ws[id]) != null &&
764	>	UNSAFE.compareAndSwapLong(this, eventWaitersOffset,
765	>	h, h = w.nextWaiter)) {
766	>	LockSupport.unpark(w);
767	>	if (eventWaiters != h \|\| eventCount != ec)
768	>	break;
769	>	}
770		}
771
772		/**
773	<	* Releases workers blocked on a count not equal to current count.
773	>	* Tries to advance eventCount and releases waiters. Called only
774	>	* from workers.
775		*/
776	<	final void releaseWaiters() {
777	<	long top;
778	<	int id;
779	<	while ((id = (int)((top = eventWaiters) & WAITER_INDEX_MASK)) > 0 &&
780	<	(int)(top >>> EVENT_COUNT_SHIFT) != eventCount) {
731	<	ForkJoinWorkerThread[] ws = workers;
732	<	ForkJoinWorkerThread w;
733	<	if (ws.length >= id && (w = ws[id - 1]) != null &&
734	<	UNSAFE.compareAndSwapLong(this, eventWaitersOffset,
735	<	top, w.nextWaiter))
736	<	LockSupport.unpark(w);
737	<	}
776	>	final void signalWork() {
777	>	int c; // try to increment event count -- CAS failure OK
778	>	UNSAFE.compareAndSwapInt(this, eventCountOffset, c = eventCount, c+1);
779	>	if (eventWaiters != 0L)
780	>	releaseEventWaiters();
781		}
782
783		/**
784	<	* Advances eventCount and releases waiters until interference by
785	<	* other releasing threads is detected.
784	>	* Adds the given worker to event queue and blocks until
785	>	* terminating or event count advances from the workers
786	>	* lastEventCount value
787	>	*
788	>	* @param w the calling worker thread
789		*/
790	<	final void signalWork() {
791	<	int ec;
792	<	UNSAFE.compareAndSwapInt(this, eventCountOffset, ec=eventCount, ec+1);
793	<	outer:for (;;) {
794	<	long top = eventWaiters;
795	<	ec = eventCount;
796	<	for (;;) {
797	<	ForkJoinWorkerThread[] ws; ForkJoinWorkerThread w;
798	<	int id = (int)(top & WAITER_INDEX_MASK);
799	<	if (id <= 0 \|\| (int)(top >>> EVENT_COUNT_SHIFT) == ec)
800	<	return;
801	<	if ((ws = workers).length < id \|\| (w = ws[id - 1]) == null \|\|
756	<	!UNSAFE.compareAndSwapLong(this, eventWaitersOffset,
757	<	top, top = w.nextWaiter))
758	<	continue outer; // possibly stale; reread
759	<	LockSupport.unpark(w);
760	<	if (top != eventWaiters) // let someone else take over
761	<	return;
790	>	private void eventSync(ForkJoinWorkerThread w) {
791	>	int ec = w.lastEventCount;
792	>	long nh = (((long)ec) << EVENT_COUNT_SHIFT) \| ((long)(w.poolIndex+1));
793	>	long h;
794	>	while ((runState < SHUTDOWN \|\| !tryTerminate(false)) &&
795	>	(((int)((h = eventWaiters) & WAITER_ID_MASK)) == 0 \|\|
796	>	(int)(h >>> EVENT_COUNT_SHIFT) == ec) &&
797	>	eventCount == ec) {
798	>	if (UNSAFE.compareAndSwapLong(this, eventWaitersOffset,
799	>	w.nextWaiter = h, nh)) {
800	>	awaitEvent(w, ec);
801	>	break;
802		}
803		}
804		}
805
806		/**
807	<	* If worker is inactive, blocks until terminating or event count
808	<	* advances from last value held by worker; in any case helps
809	<	* release others.
807	>	* Blocks the given worker (that has already been entered as an
808	>	* event waiter) until terminating or event count advances from
809	>	* the given value. The oldest (first) waiter uses a timed wait to
810	>	* occasionally one-by-one shrink the number of workers (to a
811	>	* minumum of one) if the pool has not been used for extended
812	>	* periods.
813		*
814		* @param w the calling worker thread
815	+	* @param ec the count
816		*/
817	<	private void eventSync(ForkJoinWorkerThread w) {
818	<	if (!w.active) {
819	<	int prev = w.lastEventCount;
820	<	long nextTop = (((long)prev << EVENT_COUNT_SHIFT) \|
821	<	((long)(w.poolIndex + 1)));
822	<	long top;
823	<	while ((runState < SHUTDOWN \|\| !tryTerminate(false)) &&
824	<	(((int)(top = eventWaiters) & WAITER_INDEX_MASK) == 0 \|\|
825	<	(int)(top >>> EVENT_COUNT_SHIFT) == prev) &&
782	<	eventCount == prev) {
783	<	if (UNSAFE.compareAndSwapLong(this, eventWaitersOffset,
784	<	w.nextWaiter = top, nextTop)) {
785	<	accumulateStealCount(w); // transfer steals while idle
786	<	Thread.interrupted(); // clear/ignore interrupt
787	<	while (eventCount == prev)
788	<	w.doPark();
817	>	private void awaitEvent(ForkJoinWorkerThread w, int ec) {
818	>	while (eventCount == ec) {
819	>	if (tryAccumulateStealCount(w)) { // transfer while idle
820	>	boolean untimed = (w.nextWaiter != 0L \|\|
821	>	(workerCounts & RUNNING_COUNT_MASK) <= 1);
822	>	long startTime = untimed? 0 : System.nanoTime();
823	>	Thread.interrupted(); // clear/ignore interrupt
824	>	if (eventCount != ec \|\| !w.isRunning() \|\|
825	>	runState >= TERMINATING) // recheck after clear
826		break;
827	+	if (untimed)
828	+	LockSupport.park(w);
829	+	else {
830	+	LockSupport.parkNanos(w, SHRINK_RATE_NANOS);
831	+	if (eventCount != ec \|\| !w.isRunning() \|\|
832	+	runState >= TERMINATING)
833	+	break;
834	+	if (System.nanoTime() - startTime >= SHRINK_RATE_NANOS)
835	+	tryShutdownWaiter(ec);
836		}
837		}
792	–	w.lastEventCount = eventCount;
838		}
794	–	releaseWaiters();
839		}
840
841		/**
842	<	* Callback from workers invoked upon each top-level action (i.e.,
843	<	* stealing a task or taking a submission and running
844	<	* it). Performs one or both of the following:
845	<	*
846	<	* * If the worker cannot find work, updates its active status to
847	<	* inactive and updates activeCount unless there is contention, in
848	<	* which case it may try again (either in this or a subsequent
849	<	* call). Additionally, awaits the next task event and/or helps
850	<	* wake up other releasable waiters.
851	<	*
852	<	* * If there are too many running threads, suspends this worker
853	<	* (first forcing inactivation if necessary). If it is not
854	<	* resumed before a keepAlive elapses, the worker may be "trimmed"
855	<	* -- killed while suspended within suspendAsSpare. Otherwise,
856	<	* upon resume it rechecks to make sure that it is still needed.
842	>	* Callback from the oldest waiter in awaitEvent waking up after a
843	>	* period of non-use. Tries (once) to shutdown an event waiter (or
844	>	* a spare, if one exists). Note that we don't need CAS or locks
845	>	* here because the method is called only from one thread
846	>	* occasionally waking (and even misfires are OK). Note that
847	>	* until the shutdown worker fully terminates, workerCounts
848	>	* will overestimate total count, which is tolerable.
849	>	*
850	>	* @param ec the event count waited on by caller (to abort
851	>	* attempt if count has since changed).
852	>	*/
853	>	private void tryShutdownWaiter(int ec) {
854	>	if (spareWaiters != 0) { // prefer killing spares
855	>	tryShutdownSpare();
856	>	return;
857	>	}
858	>	ForkJoinWorkerThread[] ws = workers;
859	>	int n = ws.length;
860	>	long h = eventWaiters;
861	>	ForkJoinWorkerThread w; int id; long nh;
862	>	if (runState == 0 &&
863	>	submissionQueue.isEmpty() &&
864	>	eventCount == ec &&
865	>	(id = ((int)(h & WAITER_ID_MASK)) - 1) >= 0 &&
866	>	id < n && (w = ws[id]) != null &&
867	>	(nh = w.nextWaiter) != 0L && // keep at least one worker
868	>	UNSAFE.compareAndSwapLong(this, eventWaitersOffset, h, nh)) {
869	>	w.shutdown();
870	>	LockSupport.unpark(w);
871	>	}
872	>	releaseEventWaiters();
873	>	}
874	>
875	>	// Maintaining spares
876	>
877	>	/**
878	>	* Pushes worker onto the spare stack
879	>	*/
880	>	final void pushSpare(ForkJoinWorkerThread w) {
881	>	int ns = (++w.spareCount << SPARE_COUNT_SHIFT) \| (w.poolIndex + 1);
882	>	do {} while (!UNSAFE.compareAndSwapInt(this, spareWaitersOffset,
883	>	w.nextSpare = spareWaiters,ns));
884	>	}
885	>
886	>	/**
887	>	* Callback from oldest spare occasionally waking up. Tries
888	>	* (once) to shutdown a spare. Same idea as tryShutdownWaiter.
889	>	*/
890	>	final void tryShutdownSpare() {
891	>	int sw, id;
892	>	ForkJoinWorkerThread w;
893	>	ForkJoinWorkerThread[] ws;
894	>	if ((id = ((sw = spareWaiters) & SPARE_ID_MASK) - 1) >= 0 &&
895	>	id < (ws = workers).length && (w = ws[id]) != null &&
896	>	(workerCounts & RUNNING_COUNT_MASK) >= parallelism &&
897	>	UNSAFE.compareAndSwapInt(this, spareWaitersOffset,
898	>	sw, w.nextSpare)) {
899	>	w.shutdown();
900	>	LockSupport.unpark(w);
901	>	advanceEventCount();
902	>	}
903	>	}
904	>
905	>	/**
906	>	* Tries (once) to resume a spare if worker counts match
907	>	* the given count.
908		*
909	<	* @param w the worker
815	<	* @param worked false if the worker scanned for work but didn't
816	<	* find any (in which case it may block waiting for work).
909	>	* @param wc workerCounts value on invocation of this method
910		*/
911	<	final void preStep(ForkJoinWorkerThread w, boolean worked) {
912	<	boolean active = w.active;
913	<	boolean inactivate = !worked & active;
914	<	for (;;) {
915	<	if (inactivate) {
916	<	int c = runState;
917	<	if (UNSAFE.compareAndSwapInt(this, runStateOffset,
918	<	c, c - ONE_ACTIVE))
919	<	inactivate = active = w.active = false;
920	<	}
921	<	int wc = workerCounts;
922	<	if ((wc & RUNNING_COUNT_MASK) <= parallelism) {
923	<	if (!worked)
924	<	eventSync(w);
925	<	return;
911	>	private void tryResumeSpare(int wc) {
912	>	ForkJoinWorkerThread[] ws = workers;
913	>	int n = ws.length;
914	>	int sw, id, rs; ForkJoinWorkerThread w;
915	>	if ((id = ((sw = spareWaiters) & SPARE_ID_MASK) - 1) >= 0 &&
916	>	id < n && (w = ws[id]) != null &&
917	>	(rs = runState) < TERMINATING &&
918	>	eventWaiters == 0L && workerCounts == wc) {
919	>	// In case all workers busy, heuristically back off to let settle
920	>	Thread.yield();
921	>	if (eventWaiters == 0L && runState == rs && // recheck
922	>	workerCounts == wc && spareWaiters == sw &&
923	>	UNSAFE.compareAndSwapInt(this, spareWaitersOffset,
924	>	sw, w.nextSpare)) {
925	>	int c; // increment running count before resume
926	>	do {} while(!UNSAFE.compareAndSwapInt
927	>	(this, workerCountsOffset,
928	>	c = workerCounts, c + ONE_RUNNING));
929	>	if (w.tryUnsuspend())
930	>	LockSupport.unpark(w);
931	>	else // back out if w was shutdown
932	>	decrementWorkerCounts(ONE_RUNNING, 0);
933		}
834	–	if (!(inactivate \|= active) && // must inactivate to suspend
835	–	UNSAFE.compareAndSwapInt(this, workerCountsOffset,
836	–	wc, wc - ONE_RUNNING) &&
837	–	!w.suspendAsSpare()) // false if trimmed
838	–	return;
934		}
935		}
936
937	+	// adding workers on demand
938	+
939		/**
940	<	* Adjusts counts and creates or resumes compensating threads for
941	<	* a worker about to block on task joinMe, returning early if
845	<	* joinMe becomes ready. First tries resuming an existing spare
846	<	* (which usually also avoids any count adjustment), but must then
847	<	* decrement running count to determine whether a new thread is
848	<	* needed. See above for fuller explanation.
940	>	* Adds one or more workers if needed to establish target parallelism.
941	>	* Retries upon contention.
942		*/
943	<	final void preJoin(ForkJoinTask<?> joinMe) {
944	<	boolean dec = false; // true when running count decremented
945	<	for (;;) {
946	<	releaseWaiters(); // help other threads progress
947	<
948	<	if (joinMe.status < 0) // surround spare search with done checks
949	<	return;
950	<	ForkJoinWorkerThread spare = null;
858	<	for (ForkJoinWorkerThread w : workers) {
859	<	if (w != null && w.isSuspended()) {
860	<	spare = w;
943	>	private void addWorkerIfBelowTarget() {
944	>	int pc = parallelism;
945	>	int wc;
946	>	while (((wc = workerCounts) >>> TOTAL_COUNT_SHIFT) < pc &&
947	>	runState < TERMINATING) {
948	>	if (UNSAFE.compareAndSwapInt(this, workerCountsOffset, wc,
949	>	wc + (ONE_RUNNING\|ONE_TOTAL))) {
950	>	if (addWorker() == null)
951		break;
862	–	}
863	–	}
864	–	if (joinMe.status < 0)
865	–	return;
866	–
867	–	if (spare != null && spare.tryUnsuspend()) {
868	–	if (dec \|\| joinMe.requestSignal() < 0) {
869	–	int c;
870	–	do {} while (!UNSAFE.compareAndSwapInt(this,
871	–	workerCountsOffset,
872	–	c = workerCounts,
873	–	c + ONE_RUNNING));
874	–	} // else no net count change
875	–	LockSupport.unpark(spare);
876	–	return;
952		}
953	+	}
954	+	}
955
956	<	int wc = workerCounts; // decrement running count
957	<	if (!dec && (wc & RUNNING_COUNT_MASK) != 0 &&
958	<	(dec = UNSAFE.compareAndSwapInt(this, workerCountsOffset,
959	<	wc, wc -= ONE_RUNNING)) &&
960	<	joinMe.requestSignal() < 0) { // cannot block
961	<	int c; // back out
962	<	do {} while (!UNSAFE.compareAndSwapInt(this,
963	<	workerCountsOffset,
964	<	c = workerCounts,
965	<	c + ONE_RUNNING));
966	<	return;
967	<	}
956	>	/**
957	>	* Tries (once) to add a new worker if all existing workers are
958	>	* busy, and there are either no running workers or the deficit is
959	>	* at least twice the surplus.
960	>	*
961	>	* @param wc workerCounts value on invocation of this method
962	>	*/
963	>	private void tryAddWorkerIfBusy(int wc) {
964	>	int tc, rc, rs;
965	>	int pc = parallelism;
966	>	if ((tc = wc >>> TOTAL_COUNT_SHIFT) < MAX_WORKERS &&
967	>	((rc = wc & RUNNING_COUNT_MASK) == 0 \|\|
968	>	rc < pc - ((tc - pc) << 1)) &&
969	>	(rs = runState) < TERMINATING &&
970	>	(rs & ACTIVE_COUNT_MASK) == tc) {
971	>	// Since all workers busy, heuristically back off to let settle
972	>	Thread.yield();
973	>	if (eventWaiters == 0L && spareWaiters == 0 && // recheck
974	>	runState == rs && workerCounts == wc &&
975	>	UNSAFE.compareAndSwapInt(this, workerCountsOffset, wc,
976	>	wc + (ONE_RUNNING\|ONE_TOTAL)))
977	>	addWorker();
978	>	}
979	>	}
980
981	<	if (dec) {
982	<	int tc = wc >>> TOTAL_COUNT_SHIFT;
983	<	int pc = parallelism;
984	<	int dc = pc - (wc & RUNNING_COUNT_MASK); // deficit count
985	<	if ((dc < pc && (dc <= 0 \|\| (dc * dc < (tc - pc) * pc) \|\|
986	<	!maintainsParallelism)) \|\|
987	<	tc >= maxPoolSize) // cannot add
988	<	return;
989	<	if (spare == null &&
990	<	UNSAFE.compareAndSwapInt(this, workerCountsOffset, wc,
991	<	wc + (ONE_RUNNING\|ONE_TOTAL))) {
992	<	addWorker();
993	<	return;
994	<	}
995	<	}
981	>	/**
982	>	* Does at most one of:
983	>	*
984	>	* 1. Help wake up existing workers waiting for work via
985	>	* releaseEventWaiters. (If any exist, then it doesn't
986	>	* matter right now if under target parallelism level.)
987	>	*
988	>	* 2. If a spare exists, try (once) to resume it via tryResumeSpare.
989	>	*
990	>	* 3. If there are not enough total workers, add some
991	>	* via addWorkerIfBelowTarget;
992	>	*
993	>	* 4. Try (once) to add a new worker if all existing workers
994	>	* are busy, via tryAddWorkerIfBusy
995	>	*/
996	>	private void helpMaintainParallelism() {
997	>	long h; int pc, wc;
998	>	if (((int)((h = eventWaiters) & WAITER_ID_MASK)) != 0) {
999	>	if ((int)(h >>> EVENT_COUNT_SHIFT) != eventCount)
1000	>	releaseEventWaiters(); // avoid useless call
1001	>	}
1002	>	else if ((pc = parallelism) >
1003	>	((wc = workerCounts) & RUNNING_COUNT_MASK)) {
1004	>	if (spareWaiters != 0)
1005	>	tryResumeSpare(wc);
1006	>	else if ((wc >>> TOTAL_COUNT_SHIFT) < pc)
1007	>	addWorkerIfBelowTarget();
1008	>	else
1009	>	tryAddWorkerIfBusy(wc);
1010		}
1011		}
1012
1013		/**
1014	<	* Same idea as preJoin but with too many differing details to
1015	<	* integrate: There are no task-based signal counts, and only one
1016	<	* way to do the actual blocking. So for simplicity it is directly
1017	<	* incorporated into this method.
1014	>	* Callback from workers invoked upon each top-level action (i.e.,
1015	>	* stealing a task or taking a submission and running it).
1016	>	* Performs one or more of the following:
1017	>	*
1018	>	* 1. If the worker is active, try to set its active status to
1019	>	* inactive and update activeCount. On contention, we may try
1020	>	* again on this or subsequent call.
1021	>	*
1022	>	* 2. Release any existing event waiters that are now relesable
1023	>	*
1024	>	* 3. If there are too many running threads, suspend this worker
1025	>	* (first forcing inactive if necessary). If it is not
1026	>	* needed, it may be killed while suspended via
1027	>	* tryShutdownSpare. Otherwise, upon resume it rechecks to make
1028	>	* sure that it is still needed.
1029	>	*
1030	>	* 4. If more than 1 miss, await the next task event via
1031	>	* eventSync (first forcing inactivation if necessary), upon
1032	>	* which worker may also be killed, via tryShutdownWaiter.
1033	>	*
1034	>	* 5. Help reactivate other workers via helpMaintainParallelism
1035	>	*
1036	>	* @param w the worker
1037	>	* @param misses the number of scans by caller failing to find work
1038	>	* (saturating at 2 to avoid wraparound)
1039		*/
1040	<	final void doBlock(ManagedBlocker blocker, boolean maintainPar)
1041	<	throws InterruptedException {
1042	<	maintainPar &= maintainsParallelism; // override
919	<	boolean dec = false;
920	<	boolean done = false;
1040	>	final void preStep(ForkJoinWorkerThread w, int misses) {
1041	>	boolean active = w.active;
1042	>	int pc = parallelism;
1043		for (;;) {
1044	<	releaseWaiters();
1045	<	if (done = blocker.isReleasable())
1046	<	break;
1047	<	ForkJoinWorkerThread spare = null;
1048	<	for (ForkJoinWorkerThread w : workers) {
1049	<	if (w != null && w.isSuspended()) {
1050	<	spare = w;
1051	<	break;
1052	<	}
1053	<	}
1054	<	if (done = blocker.isReleasable())
1055	<	break;
1056	<	if (spare != null && spare.tryUnsuspend()) {
1057	<	if (dec) {
1058	<	int c;
1059	<	do {} while (!UNSAFE.compareAndSwapInt(this,
1060	<	workerCountsOffset,
1061	<	c = workerCounts,
940	<	c + ONE_RUNNING));
1044	>	int rs, wc, rc, ec; long h;
1045	>	if (active && UNSAFE.compareAndSwapInt(this, runStateOffset,
1046	>	rs = runState, rs - 1))
1047	>	active = w.active = false;
1048	>	if (((int)((h = eventWaiters) & WAITER_ID_MASK)) != 0 &&
1049	>	(int)(h >>> EVENT_COUNT_SHIFT) != eventCount) {
1050	>	releaseEventWaiters();
1051	>	if (misses > 1)
1052	>	continue; // clear before sync below
1053	>	}
1054	>	if ((rc = ((wc = workerCounts) & RUNNING_COUNT_MASK)) > pc) {
1055	>	if (!active && // must inactivate to suspend
1056	>	workerCounts == wc && // try to suspend as spare
1057	>	UNSAFE.compareAndSwapInt(this, workerCountsOffset,
1058	>	wc, wc - ONE_RUNNING)) {
1059	>	w.suspendAsSpare();
1060	>	if (!w.isRunning())
1061	>	break; // was killed while spare
1062		}
1063	<	LockSupport.unpark(spare);
943	<	break;
1063	>	continue;
1064		}
1065	<	int wc = workerCounts;
1066	<	if (!dec && (wc & RUNNING_COUNT_MASK) != 0)
1067	<	dec = UNSAFE.compareAndSwapInt(this, workerCountsOffset,
1068	<	wc, wc -= ONE_RUNNING);
1069	<	if (dec) {
1070	<	int tc = wc >>> TOTAL_COUNT_SHIFT;
1071	<	int pc = parallelism;
1072	<	int dc = pc - (wc & RUNNING_COUNT_MASK);
1073	<	if ((dc < pc && (dc <= 0 \|\| (dc * dc < (tc - pc) * pc) \|\|
1074	<	!maintainPar)) \|\|
955	<	tc >= maxPoolSize)
956	<	break;
957	<	if (spare == null &&
958	<	UNSAFE.compareAndSwapInt(this, workerCountsOffset, wc,
959	<	wc + (ONE_RUNNING\|ONE_TOTAL))){
960	<	addWorker();
961	<	break;
1065	>	if (misses > 0) {
1066	>	if ((ec = eventCount) == w.lastEventCount && misses > 1) {
1067	>	if (!active) { // must inactivate to sync
1068	>	eventSync(w);
1069	>	if (w.isRunning())
1070	>	misses = 1; // don't re-sync
1071	>	else
1072	>	break; // was killed while waiting
1073	>	}
1074	>	continue;
1075		}
1076	+	w.lastEventCount = ec;
1077		}
1078	<	}
1079	<
1080	<	try {
967	<	if (!done)
968	<	do {} while (!blocker.isReleasable() && !blocker.block());
969	<	} finally {
970	<	if (dec) {
971	<	int c;
972	<	do {} while (!UNSAFE.compareAndSwapInt(this,
973	<	workerCountsOffset,
974	<	c = workerCounts,
975	<	c + ONE_RUNNING));
976	<	}
1078	>	if (rc < pc)
1079	>	helpMaintainParallelism();
1080	>	break;
1081		}
1082		}
1083
1084		/**
1085	<	* Unless there are not enough other running threads, adjusts
1086	<	* counts for a a worker in performing helpJoin that cannot find
1087	<	* any work, so that this worker can now block.
1085	>	* Helps and/or blocks awaiting join of the given task.
1086	>	* Alternates between helpJoinTask() and helpMaintainParallelism()
1087	>	* as many times as there is a deficit in running count (or longer
1088	>	* if running count would become zero), then blocks if task still
1089	>	* not done.
1090		*
1091	<	* @return true if worker may block
1091	>	* @param joinMe the task to join
1092		*/
1093	<	final boolean preBlockHelpingJoin(ForkJoinTask<?> joinMe) {
1093	>	final void awaitJoin(ForkJoinTask<?> joinMe, ForkJoinWorkerThread worker) {
1094	>	int threshold = parallelism; // descend blocking thresholds
1095		while (joinMe.status >= 0) {
1096	<	releaseWaiters(); // help other threads progress
1097	<
1098	<	// if a spare exists, resume it to maintain parallelism level
992	<	if ((workerCounts & RUNNING_COUNT_MASK) <= parallelism) {
993	<	ForkJoinWorkerThread spare = null;
994	<	for (ForkJoinWorkerThread w : workers) {
995	<	if (w != null && w.isSuspended()) {
996	<	spare = w;
997	<	break;
998	<	}
999	<	}
1000	<	if (joinMe.status < 0)
1001	<	break;
1002	<	if (spare != null) {
1003	<	if (spare.tryUnsuspend()) {
1004	<	boolean canBlock = true;
1005	<	if (joinMe.requestSignal() < 0) {
1006	<	canBlock = false; // already done
1007	<	int c;
1008	<	do {} while (!UNSAFE.compareAndSwapInt
1009	<	(this, workerCountsOffset,
1010	<	c = workerCounts, c + ONE_RUNNING));
1011	<	}
1012	<	LockSupport.unpark(spare);
1013	<	return canBlock;
1014	<	}
1015	<	continue; // recheck -- another spare may exist
1016	<	}
1017	<	}
1018	<
1019	<	int wc = workerCounts; // reread to shorten CAS window
1020	<	int rc = wc & RUNNING_COUNT_MASK;
1021	<	if (rc <= 2) // keep this and at most one other thread alive
1096	>	boolean block; int wc;
1097	>	worker.helpJoinTask(joinMe);
1098	>	if (joinMe.status < 0)
1099		break;
1100	<
1101	<	if (UNSAFE.compareAndSwapInt(this, workerCountsOffset,
1102	<	wc, wc - ONE_RUNNING)) {
1103	<	if (joinMe.requestSignal() >= 0)
1104	<	return true;
1105	<	int c; // back out
1100	>	if (((wc = workerCounts) & RUNNING_COUNT_MASK) <= threshold) {
1101	>	if (threshold > 0)
1102	>	--threshold;
1103	>	else
1104	>	advanceEventCount(); // force release
1105	>	block = false;
1106	>	}
1107	>	else
1108	>	block = UNSAFE.compareAndSwapInt(this, workerCountsOffset,
1109	>	wc, wc - ONE_RUNNING);
1110	>	helpMaintainParallelism();
1111	>	if (block) {
1112	>	int c;
1113	>	joinMe.internalAwaitDone();
1114		do {} while (!UNSAFE.compareAndSwapInt
1115		(this, workerCountsOffset,
1116		c = workerCounts, c + ONE_RUNNING));
1117		break;
1118		}
1119		}
1120	<	return false;
1120	>	}
1121	>
1122	>	/**
1123	>	* Same idea as awaitJoin, but no helping
1124	>	*/
1125	>	final void awaitBlocker(ManagedBlocker blocker)
1126	>	throws InterruptedException {
1127	>	int threshold = parallelism;
1128	>	while (!blocker.isReleasable()) {
1129	>	boolean block; int wc;
1130	>	if (((wc = workerCounts) & RUNNING_COUNT_MASK) <= threshold) {
1131	>	if (threshold > 0)
1132	>	--threshold;
1133	>	else
1134	>	advanceEventCount();
1135	>	block = false;
1136	>	}
1137	>	else
1138	>	block = UNSAFE.compareAndSwapInt(this, workerCountsOffset,
1139	>	wc, wc - ONE_RUNNING);
1140	>	helpMaintainParallelism();
1141	>	if (block) {
1142	>	try {
1143	>	do {} while (!blocker.isReleasable() && !blocker.block());
1144	>	} finally {
1145	>	int c;
1146	>	do {} while (!UNSAFE.compareAndSwapInt
1147	>	(this, workerCountsOffset,
1148	>	c = workerCounts, c + ONE_RUNNING));
1149	>	}
1150	>	break;
1151	>	}
1152	>	}
1153		}
1154
1155		/**
#	Line 1056 \| Line 1173 \| public class ForkJoinPool extends Abstra
1173		// Finish now if all threads terminated; else in some subsequent call
1174		if ((workerCounts >>> TOTAL_COUNT_SHIFT) == 0) {
1175		advanceRunLevel(TERMINATED);
1176	<	terminationLatch.countDown();
1176	>	termination.arrive();
1177		}
1178		return true;
1179		}
1180
1181		/**
1182		* Actions on transition to TERMINATING
1183	+	*
1184	+	* Runs up to four passes through workers: (0) shutting down each
1185	+	* (without waking up if parked) to quickly spread notifications
1186	+	* without unnecessary bouncing around event queues etc (1) wake
1187	+	* up and help cancel tasks (2) interrupt (3) mop up races with
1188	+	* interrupted workers
1189		*/
1190		private void startTerminating() {
1191	<	// Clear out and cancel submissions, ignoring exceptions
1191	>	cancelSubmissions();
1192	>	for (int passes = 0; passes < 4 && workerCounts != 0; ++passes) {
1193	>	advanceEventCount();
1194	>	eventWaiters = 0L; // clobber lists
1195	>	spareWaiters = 0;
1196	>	ForkJoinWorkerThread[] ws = workers;
1197	>	int n = ws.length;
1198	>	for (int i = 0; i < n; ++i) {
1199	>	ForkJoinWorkerThread w = ws[i];
1200	>	if (w != null) {
1201	>	w.shutdown();
1202	>	if (passes > 0 && !w.isTerminated()) {
1203	>	w.cancelTasks();
1204	>	LockSupport.unpark(w);
1205	>	if (passes > 1) {
1206	>	try {
1207	>	w.interrupt();
1208	>	} catch (SecurityException ignore) {
1209	>	}
1210	>	}
1211	>	}
1212	>	}
1213	>	}
1214	>	}
1215	>	}
1216	>
1217	>	/**
1218	>	* Clear out and cancel submissions, ignoring exceptions
1219	>	*/
1220	>	private void cancelSubmissions() {
1221		ForkJoinTask<?> task;
1222		while ((task = submissionQueue.poll()) != null) {
1223		try {
#	Line 1073 \| Line 1225 \| public class ForkJoinPool extends Abstra
1225		} catch (Throwable ignore) {
1226		}
1227		}
1076	–	// Propagate run level
1077	–	for (ForkJoinWorkerThread w : workers) {
1078	–	if (w != null)
1079	–	w.shutdown(); // also resumes suspended workers
1080	–	}
1081	–	// Ensure no straggling local tasks
1082	–	for (ForkJoinWorkerThread w : workers) {
1083	–	if (w != null)
1084	–	w.cancelTasks();
1085	–	}
1086	–	// Wake up idle workers
1087	–	advanceEventCount();
1088	–	releaseWaiters();
1089	–	// Unstick pending joins
1090	–	for (ForkJoinWorkerThread w : workers) {
1091	–	if (w != null && !w.isTerminated()) {
1092	–	try {
1093	–	w.interrupt();
1094	–	} catch (SecurityException ignore) {
1095	–	}
1096	–	}
1097	–	}
1228		}
1229
1230		// misc support for ForkJoinWorkerThread
#	Line 1107 \| Line 1237 \| public class ForkJoinPool extends Abstra
1237		}
1238
1239		/**
1240	<	* Accumulates steal count from a worker, clearing
1241	<	* the worker's value
1240	>	* Tries to accumulates steal count from a worker, clearing
1241	>	* the worker's value.
1242	>	*
1243	>	* @return true if worker steal count now zero
1244		*/
1245	<	final void accumulateStealCount(ForkJoinWorkerThread w) {
1245	>	final boolean tryAccumulateStealCount(ForkJoinWorkerThread w) {
1246		int sc = w.stealCount;
1247	<	if (sc != 0) {
1248	<	long c;
1249	<	w.stealCount = 0;
1250	<	do {} while (!UNSAFE.compareAndSwapLong(this, stealCountOffset,
1251	<	c = stealCount, c + sc));
1247	>	long c = stealCount;
1248	>	// CAS even if zero, for fence effects
1249	>	if (UNSAFE.compareAndSwapLong(this, stealCountOffset, c, c + sc)) {
1250	>	if (sc != 0)
1251	>	w.stealCount = 0;
1252	>	return true;
1253		}
1254	+	return sc == 0;
1255		}
1256
1257		/**
#	Line 1125 \| Line 1259 \| public class ForkJoinPool extends Abstra
1259		* active thread.
1260		*/
1261		final int idlePerActive() {
1262	+	int pc = parallelism; // use parallelism, not rc
1263		int ac = runState; // no mask -- artifically boosts during shutdown
1129	–	int pc = parallelism; // use targeted parallelism, not rc
1264		// Use exact results for small values, saturate past 4
1265		return pc <= ac? 0 : pc >>> 1 <= ac? 1 : pc >>> 2 <= ac? 3 : pc >>> 3;
1266		}
#	Line 1137 \| Line 1271 \| public class ForkJoinPool extends Abstra
1271
1272		/**
1273		* Creates a {@code ForkJoinPool} with parallelism equal to {@link
1274	<	* java.lang.Runtime#availableProcessors}, and using the {@linkplain
1275	<	* #defaultForkJoinWorkerThreadFactory default thread factory}.
1274	>	* java.lang.Runtime#availableProcessors}, using the {@linkplain
1275	>	* #defaultForkJoinWorkerThreadFactory default thread factory},
1276	>	* no UncaughtExceptionHandler, and non-async LIFO processing mode.
1277		*
1278		* @throws SecurityException if a security manager exists and
1279		* the caller is not permitted to modify threads
#	Line 1147 \| Line 1282 \| public class ForkJoinPool extends Abstra
1282		*/
1283		public ForkJoinPool() {
1284		this(Runtime.getRuntime().availableProcessors(),
1285	<	defaultForkJoinWorkerThreadFactory);
1285	>	defaultForkJoinWorkerThreadFactory, null, false);
1286		}
1287
1288		/**
1289		* Creates a {@code ForkJoinPool} with the indicated parallelism
1290	<	* level and using the {@linkplain
1291	<	* #defaultForkJoinWorkerThreadFactory default thread factory}.
1290	>	* level, the {@linkplain
1291	>	* #defaultForkJoinWorkerThreadFactory default thread factory},
1292	>	* no UncaughtExceptionHandler, and non-async LIFO processing mode.
1293		*
1294		* @param parallelism the parallelism level
1295		* @throws IllegalArgumentException if parallelism less than or
#	Line 1164 \| Line 1300 \| public class ForkJoinPool extends Abstra
1300		* java.lang.RuntimePermission}{@code ("modifyThread")}
1301		*/
1302		public ForkJoinPool(int parallelism) {
1303	<	this(parallelism, defaultForkJoinWorkerThreadFactory);
1303	>	this(parallelism, defaultForkJoinWorkerThreadFactory, null, false);
1304		}
1305
1306		/**
1307	<	* Creates a {@code ForkJoinPool} with parallelism equal to {@link
1172	<	* java.lang.Runtime#availableProcessors}, and using the given
1173	<	* thread factory.
1307	>	* Creates a {@code ForkJoinPool} with the given parameters.
1308		*
1309	<	* @param factory the factory for creating new threads
1310	<	* @throws NullPointerException if the factory is null
1311	<	* @throws SecurityException if a security manager exists and
1312	<	* the caller is not permitted to modify threads
1313	<	* because it does not hold {@link
1314	<	* java.lang.RuntimePermission}{@code ("modifyThread")}
1315	<	*/
1316	<	public ForkJoinPool(ForkJoinWorkerThreadFactory factory) {
1317	<	this(Runtime.getRuntime().availableProcessors(), factory);
1318	<	}
1319	<
1320	<	/**
1321	<	* Creates a {@code ForkJoinPool} with the given parallelism and
1188	<	* thread factory.
1189	<	*
1190	<	* @param parallelism the parallelism level
1191	<	* @param factory the factory for creating new threads
1309	>	* @param parallelism the parallelism level. For default value,
1310	>	* use {@link java.lang.Runtime#availableProcessors}.
1311	>	* @param factory the factory for creating new threads. For default value,
1312	>	* use {@link #defaultForkJoinWorkerThreadFactory}.
1313	>	* @param handler the handler for internal worker threads that
1314	>	* terminate due to unrecoverable errors encountered while executing
1315	>	* tasks. For default value, use <code>null</code>.
1316	>	* @param asyncMode if true,
1317	>	* establishes local first-in-first-out scheduling mode for forked
1318	>	* tasks that are never joined. This mode may be more appropriate
1319	>	* than default locally stack-based mode in applications in which
1320	>	* worker threads only process event-style asynchronous tasks.
1321	>	* For default value, use <code>false</code>.
1322		* @throws IllegalArgumentException if parallelism less than or
1323		* equal to zero, or greater than implementation limit
1324		* @throws NullPointerException if the factory is null
#	Line 1197 \| Line 1327 \| public class ForkJoinPool extends Abstra
1327		* because it does not hold {@link
1328		* java.lang.RuntimePermission}{@code ("modifyThread")}
1329		*/
1330	<	public ForkJoinPool(int parallelism, ForkJoinWorkerThreadFactory factory) {
1330	>	public ForkJoinPool(int parallelism,
1331	>	ForkJoinWorkerThreadFactory factory,
1332	>	Thread.UncaughtExceptionHandler handler,
1333	>	boolean asyncMode) {
1334		checkPermission();
1335		if (factory == null)
1336		throw new NullPointerException();
1337	<	if (parallelism <= 0 \|\| parallelism > MAX_THREADS)
1337	>	if (parallelism <= 0 \|\| parallelism > MAX_WORKERS)
1338		throw new IllegalArgumentException();
1206	–	this.poolNumber = poolNumberGenerator.incrementAndGet();
1207	–	int arraySize = initialArraySizeFor(parallelism);
1339		this.parallelism = parallelism;
1340		this.factory = factory;
1341	<	this.maxPoolSize = MAX_THREADS;
1342	<	this.maintainsParallelism = true;
1341	>	this.ueh = handler;
1342	>	this.locallyFifo = asyncMode;
1343	>	int arraySize = initialArraySizeFor(parallelism);
1344		this.workers = new ForkJoinWorkerThread[arraySize];
1345		this.submissionQueue = new LinkedTransferQueue<ForkJoinTask<?>>();
1346		this.workerLock = new ReentrantLock();
1347	<	this.terminationLatch = new CountDownLatch(1);
1348	<	// Start first worker; remaining workers added upon first submission
1217	<	workerCounts = ONE_RUNNING \| ONE_TOTAL;
1218	<	addWorker();
1347	>	this.termination = new Phaser(1);
1348	>	this.poolNumber = poolNumberGenerator.incrementAndGet();
1349		}
1350
1351		/**
#	Line 1223 \| Line 1353 \| public class ForkJoinPool extends Abstra
1353		* @param pc the initial parallelism level
1354		*/
1355		private static int initialArraySizeFor(int pc) {
1356	<	// See Hackers Delight, sec 3.2. We know MAX_THREADS < (1 >>> 16)
1357	<	int size = pc < MAX_THREADS ? pc + 1 : MAX_THREADS;
1356	>	// See Hackers Delight, sec 3.2. We know MAX_WORKERS < (1 >>> 16)
1357	>	int size = pc < MAX_WORKERS ? pc + 1 : MAX_WORKERS;
1358		size \|= size >>> 1;
1359		size \|= size >>> 2;
1360		size \|= size >>> 4;
#	Line 1244 \| Line 1374 \| public class ForkJoinPool extends Abstra
1374		throw new RejectedExecutionException();
1375		submissionQueue.offer(task);
1376		advanceEventCount();
1377	<	releaseWaiters();
1377	>	if (eventWaiters != 0L)
1378	>	releaseEventWaiters();
1379		if ((workerCounts >>> TOTAL_COUNT_SHIFT) < parallelism)
1380	<	ensureEnoughTotalWorkers();
1380	>	addWorkerIfBelowTarget();
1381		}
1382
1383		/**
#	Line 1292 \| Line 1423 \| public class ForkJoinPool extends Abstra
1423		}
1424
1425		/**
1426	+	* Submits a ForkJoinTask for execution.
1427	+	*
1428	+	* @param task the task to submit
1429	+	* @return the task
1430	+	* @throws NullPointerException if the task is null
1431	+	* @throws RejectedExecutionException if the task cannot be
1432	+	* scheduled for execution
1433	+	*/
1434	+	public <T> ForkJoinTask<T> submit(ForkJoinTask<T> task) {
1435	+	doSubmit(task);
1436	+	return task;
1437	+	}
1438	+
1439	+	/**
1440		* @throws NullPointerException if the task is null
1441		* @throws RejectedExecutionException if the task cannot be
1442		* scheduled for execution
#	Line 1329 \| Line 1474 \| public class ForkJoinPool extends Abstra
1474		}
1475
1476		/**
1332	–	* Submits a ForkJoinTask for execution.
1333	–	*
1334	–	* @param task the task to submit
1335	–	* @return the task
1336	–	* @throws NullPointerException if the task is null
1337	–	* @throws RejectedExecutionException if the task cannot be
1338	–	* scheduled for execution
1339	–	*/
1340	–	public <T> ForkJoinTask<T> submit(ForkJoinTask<T> task) {
1341	–	doSubmit(task);
1342	–	return task;
1343	–	}
1344	–
1345	–	/**
1477		* @throws NullPointerException {@inheritDoc}
1478		* @throws RejectedExecutionException {@inheritDoc}
1479		*/
#	Line 1384 \| Line 1515 \| public class ForkJoinPool extends Abstra
1515		* @return the handler, or {@code null} if none
1516		*/
1517		public Thread.UncaughtExceptionHandler getUncaughtExceptionHandler() {
1387	–	workerCountReadFence();
1518		return ueh;
1519		}
1520
1521		/**
1392	–	* Sets the handler for internal worker threads that terminate due
1393	–	* to unrecoverable errors encountered while executing tasks.
1394	–	* Unless set, the current default or ThreadGroup handler is used
1395	–	* as handler.
1396	–	*
1397	–	* @param h the new handler
1398	–	* @return the old handler, or {@code null} if none
1399	–	* @throws SecurityException if a security manager exists and
1400	–	* the caller is not permitted to modify threads
1401	–	* because it does not hold {@link
1402	–	* java.lang.RuntimePermission}{@code ("modifyThread")}
1403	–	*/
1404	–	public Thread.UncaughtExceptionHandler
1405	–	setUncaughtExceptionHandler(Thread.UncaughtExceptionHandler h) {
1406	–	checkPermission();
1407	–	workerCountReadFence();
1408	–	Thread.UncaughtExceptionHandler old = ueh;
1409	–	if (h != old) {
1410	–	ueh = h;
1411	–	workerCountWriteFence();
1412	–	for (ForkJoinWorkerThread w : workers) {
1413	–	if (w != null)
1414	–	w.setUncaughtExceptionHandler(h);
1415	–	}
1416	–	}
1417	–	return old;
1418	–	}
1419	–
1420	–	/**
1421	–	* Sets the target parallelism level of this pool.
1422	–	*
1423	–	* @param parallelism the target parallelism
1424	–	* @throws IllegalArgumentException if parallelism less than or
1425	–	* equal to zero or greater than maximum size bounds
1426	–	* @throws SecurityException if a security manager exists and
1427	–	* the caller is not permitted to modify threads
1428	–	* because it does not hold {@link
1429	–	* java.lang.RuntimePermission}{@code ("modifyThread")}
1430	–	*/
1431	–	public void setParallelism(int parallelism) {
1432	–	checkPermission();
1433	–	if (parallelism <= 0 \|\| parallelism > maxPoolSize)
1434	–	throw new IllegalArgumentException();
1435	–	workerCountReadFence();
1436	–	int pc = this.parallelism;
1437	–	if (pc != parallelism) {
1438	–	this.parallelism = parallelism;
1439	–	workerCountWriteFence();
1440	–	// Release spares. If too many, some will die after re-suspend
1441	–	for (ForkJoinWorkerThread w : workers) {
1442	–	if (w != null && w.tryUnsuspend()) {
1443	–	updateRunningCount(1);
1444	–	LockSupport.unpark(w);
1445	–	}
1446	–	}
1447	–	ensureEnoughTotalWorkers();
1448	–	advanceEventCount();
1449	–	releaseWaiters(); // force config recheck by existing workers
1450	–	}
1451	–	}
1452	–
1453	–	/**
1522		* Returns the targeted parallelism level of this pool.
1523		*
1524		* @return the targeted parallelism level of this pool
1525		*/
1526		public int getParallelism() {
1459	–	// workerCountReadFence(); // inlined below
1460	–	int ignore = workerCounts;
1527		return parallelism;
1528		}
1529
#	Line 1474 \| Line 1540 \| public class ForkJoinPool extends Abstra
1540		}
1541
1542		/**
1477	–	* Returns the maximum number of threads allowed to exist in the
1478	–	* pool. Unless set using {@link #setMaximumPoolSize}, the
1479	–	* maximum is an implementation-defined value designed only to
1480	–	* prevent runaway growth.
1481	–	*
1482	–	* @return the maximum
1483	–	*/
1484	–	public int getMaximumPoolSize() {
1485	–	workerCountReadFence();
1486	–	return maxPoolSize;
1487	–	}
1488	–
1489	–	/**
1490	–	* Sets the maximum number of threads allowed to exist in the
1491	–	* pool. The given value should normally be greater than or equal
1492	–	* to the {@link #getParallelism parallelism} level. Setting this
1493	–	* value has no effect on current pool size. It controls
1494	–	* construction of new threads. The use of this method may cause
1495	–	* tasks that intrinsically require extra threads for dependent
1496	–	* computations to indefinitely stall. If you are instead trying
1497	–	* to minimize internal thread creation, consider setting {link
1498	–	* #setMaintainsParallelism} as false.
1499	–	*
1500	–	* @throws IllegalArgumentException if negative or greater than
1501	–	* internal implementation limit
1502	–	*/
1503	–	public void setMaximumPoolSize(int newMax) {
1504	–	if (newMax < 0 \|\| newMax > MAX_THREADS)
1505	–	throw new IllegalArgumentException();
1506	–	maxPoolSize = newMax;
1507	–	workerCountWriteFence();
1508	–	}
1509	–
1510	–	/**
1511	–	* Returns {@code true} if this pool dynamically maintains its
1512	–	* target parallelism level. If false, new threads are added only
1513	–	* to avoid possible starvation. This setting is by default true.
1514	–	*
1515	–	* @return {@code true} if maintains parallelism
1516	–	*/
1517	–	public boolean getMaintainsParallelism() {
1518	–	workerCountReadFence();
1519	–	return maintainsParallelism;
1520	–	}
1521	–
1522	–	/**
1523	–	* Sets whether this pool dynamically maintains its target
1524	–	* parallelism level. If false, new threads are added only to
1525	–	* avoid possible starvation.
1526	–	*
1527	–	* @param enable {@code true} to maintain parallelism
1528	–	*/
1529	–	public void setMaintainsParallelism(boolean enable) {
1530	–	maintainsParallelism = enable;
1531	–	workerCountWriteFence();
1532	–	}
1533	–
1534	–	/**
1535	–	* Establishes local first-in-first-out scheduling mode for forked
1536	–	* tasks that are never joined. This mode may be more appropriate
1537	–	* than default locally stack-based mode in applications in which
1538	–	* worker threads only process asynchronous tasks. This method is
1539	–	* designed to be invoked only when the pool is quiescent, and
1540	–	* typically only before any tasks are submitted. The effects of
1541	–	* invocations at other times may be unpredictable.
1542	–	*
1543	–	* @param async if {@code true}, use locally FIFO scheduling
1544	–	* @return the previous mode
1545	–	* @see #getAsyncMode
1546	–	*/
1547	–	public boolean setAsyncMode(boolean async) {
1548	–	workerCountReadFence();
1549	–	boolean oldMode = locallyFifo;
1550	–	if (oldMode != async) {
1551	–	locallyFifo = async;
1552	–	workerCountWriteFence();
1553	–	for (ForkJoinWorkerThread w : workers) {
1554	–	if (w != null)
1555	–	w.setAsyncMode(async);
1556	–	}
1557	–	}
1558	–	return oldMode;
1559	–	}
1560	–
1561	–	/**
1543		* Returns {@code true} if this pool uses local first-in-first-out
1544		* scheduling mode for forked tasks that are never joined.
1545		*
1546		* @return {@code true} if this pool uses async mode
1566	–	* @see #setAsyncMode
1547		*/
1548		public boolean getAsyncMode() {
1569	–	workerCountReadFence();
1549		return locallyFifo;
1550		}
1551
#	Line 1635 \| Line 1614 \| public class ForkJoinPool extends Abstra
1614		*/
1615		public long getQueuedTaskCount() {
1616		long count = 0;
1617	<	for (ForkJoinWorkerThread w : workers) {
1617	>	ForkJoinWorkerThread[] ws = workers;
1618	>	int n = ws.length;
1619	>	for (int i = 0; i < n; ++i) {
1620	>	ForkJoinWorkerThread w = ws[i];
1621		if (w != null)
1622		count += w.getQueueSize();
1623		}
#	Line 1692 \| Line 1674 \| public class ForkJoinPool extends Abstra
1674		* @return the number of elements transferred
1675		*/
1676		protected int drainTasksTo(Collection<? super ForkJoinTask<?>> c) {
1677	<	int n = submissionQueue.drainTo(c);
1678	<	for (ForkJoinWorkerThread w : workers) {
1677	>	int count = submissionQueue.drainTo(c);
1678	>	ForkJoinWorkerThread[] ws = workers;
1679	>	int n = ws.length;
1680	>	for (int i = 0; i < n; ++i) {
1681	>	ForkJoinWorkerThread w = ws[i];
1682		if (w != null)
1683	<	n += w.drainTasksTo(c);
1683	>	count += w.drainTasksTo(c);
1684		}
1685	<	return n;
1685	>	return count;
1686		}
1687
1688		/**
#	Line 1821 \| Line 1806 \| public class ForkJoinPool extends Abstra
1806		*/
1807		public boolean awaitTermination(long timeout, TimeUnit unit)
1808		throws InterruptedException {
1809	<	return terminationLatch.await(timeout, unit);
1809	>	try {
1810	>	return termination.awaitAdvanceInterruptibly(0, timeout, unit) > 0;
1811	>	} catch(TimeoutException ex) {
1812	>	return false;
1813	>	}
1814		}
1815
1816		/**
1817		* Interface for extending managed parallelism for tasks running
1818		* in {@link ForkJoinPool}s.
1819		*
1820	<	* <p>A {@code ManagedBlocker} provides two methods.
1821	<	* Method {@code isReleasable} must return {@code true} if
1822	<	* blocking is not necessary. Method {@code block} blocks the
1823	<	* current thread if necessary (perhaps internally invoking
1824	<	* {@code isReleasable} before actually blocking).
1820	>	* <p>A {@code ManagedBlocker} provides two methods. Method
1821	>	* {@code isReleasable} must return {@code true} if blocking is
1822	>	* not necessary. Method {@code block} blocks the current thread
1823	>	* if necessary (perhaps internally invoking {@code isReleasable}
1824	>	* before actually blocking). The unusual methods in this API
1825	>	* accommodate synchronizers that may, but don't usually, block
1826	>	* for long periods. Similarly, they allow more efficient internal
1827	>	* handling of cases in which additional workers may be, but
1828	>	* usually are not, needed to ensure sufficient parallelism.
1829	>	* Toward this end, implementations of method {@code isReleasable}
1830	>	* must be amenable to repeated invocation.
1831		*
1832		* <p>For example, here is a ManagedBlocker based on a
1833		* ReentrantLock:
#	Line 1850 \| Line 1845 \| public class ForkJoinPool extends Abstra
1845		* return hasLock \|\| (hasLock = lock.tryLock());
1846		* }
1847		* }}</pre>
1848	+	*
1849	+	* <p>Here is a class that possibly blocks waiting for an
1850	+	* item on a given queue:
1851	+	* <pre> {@code
1852	+	* class QueueTaker<E> implements ManagedBlocker {
1853	+	* final BlockingQueue<E> queue;
1854	+	* volatile E item = null;
1855	+	* QueueTaker(BlockingQueue<E> q) { this.queue = q; }
1856	+	* public boolean block() throws InterruptedException {
1857	+	* if (item == null)
1858	+	* item = queue.take
1859	+	* return true;
1860	+	* }
1861	+	* public boolean isReleasable() {
1862	+	* return item != null \|\| (item = queue.poll) != null;
1863	+	* }
1864	+	* public E getItem() { // call after pool.managedBlock completes
1865	+	* return item;
1866	+	* }
1867	+	* }}</pre>
1868		*/
1869		public static interface ManagedBlocker {
1870		/**
#	Line 1873 \| Line 1888 \| public class ForkJoinPool extends Abstra
1888		* Blocks in accord with the given blocker. If the current thread
1889		* is a {@link ForkJoinWorkerThread}, this method possibly
1890		* arranges for a spare thread to be activated if necessary to
1891	<	* ensure parallelism while the current thread is blocked.
1877	<	*
1878	<	* <p>If {@code maintainParallelism} is {@code true} and the pool
1879	<	* supports it ({@link #getMaintainsParallelism}), this method
1880	<	* attempts to maintain the pool's nominal parallelism. Otherwise
1881	<	* it activates a thread only if necessary to avoid complete
1882	<	* starvation. This option may be preferable when blockages use
1883	<	* timeouts, or are almost always brief.
1891	>	* ensure sufficient parallelism while the current thread is blocked.
1892		*
1893		* <p>If the caller is not a {@link ForkJoinTask}, this method is
1894		* behaviorally equivalent to
#	Line 1894 \| Line 1902 \| public class ForkJoinPool extends Abstra
1902		* first be expanded to ensure parallelism, and later adjusted.
1903		*
1904		* @param blocker the blocker
1897	–	* @param maintainParallelism if {@code true} and supported by
1898	–	* this pool, attempt to maintain the pool's nominal parallelism;
1899	–	* otherwise activate a thread only if necessary to avoid
1900	–	* complete starvation.
1905		* @throws InterruptedException if blocker.block did so
1906		*/
1907	<	public static void managedBlock(ManagedBlocker blocker,
1904	<	boolean maintainParallelism)
1907	>	public static void managedBlock(ManagedBlocker blocker)
1908		throws InterruptedException {
1909		Thread t = Thread.currentThread();
1910	<	if (t instanceof ForkJoinWorkerThread)
1911	<	((ForkJoinWorkerThread) t).pool.
1912	<	doBlock(blocker, maintainParallelism);
1913	<	else
1914	<	awaitBlocker(blocker);
1915	<	}
1916	<
1914	<	/**
1915	<	* Performs Non-FJ blocking
1916	<	*/
1917	<	private static void awaitBlocker(ManagedBlocker blocker)
1918	<	throws InterruptedException {
1919	<	do {} while (!blocker.isReleasable() && !blocker.block());
1910	>	if (t instanceof ForkJoinWorkerThread) {
1911	>	ForkJoinWorkerThread w = (ForkJoinWorkerThread) t;
1912	>	w.pool.awaitBlocker(blocker);
1913	>	}
1914	>	else {
1915	>	do {} while (!blocker.isReleasable() && !blocker.block());
1916	>	}
1917		}
1918
1919		// AbstractExecutorService overrides. These rely on undocumented
#	Line 1944 \| Line 1941 \| public class ForkJoinPool extends Abstra
1941		objectFieldOffset("eventWaiters",ForkJoinPool.class);
1942		private static final long stealCountOffset =
1943		objectFieldOffset("stealCount",ForkJoinPool.class);
1944	<
1944	>	private static final long spareWaitersOffset =
1945	>	objectFieldOffset("spareWaiters",ForkJoinPool.class);
1946
1947		private static long objectFieldOffset(String field, Class<?> klazz) {
1948		try {

Diff Legend

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

Comparing jsr166/src/jsr166y/ForkJoinPool.java (file contents): Revision 1.54 by dl, Sun Apr 18 12:51:18 2010 UTC vs. Revision 1.64 by dl, Tue Aug 17 18:30:32 2010 UTC

Diff Legend

Comparing jsr166/src/jsr166y/ForkJoinPool.java (file contents):
Revision 1.54 by dl, Sun Apr 18 12:51:18 2010 UTC vs.
Revision 1.64 by dl, Tue Aug 17 18:30:32 2010 UTC