ViewVC Help

View File | Revision Log | Show Annotations | Download File | Root Listing

root/jsr166/jsr166/src/jsr166y/ForkJoinWorkerThread.java

Comparing jsr166/src/jsr166y/ForkJoinWorkerThread.java (file contents):
Revision 1.30 by jsr166, Tue Oct 6 19:02:48 2009 UTC vs.
Revision 1.70 by jsr166, Wed Jul 4 20:13:53 2012 UTC

#	Line 1 | Line 1
1		/*
2		* Written by Doug Lea with assistance from members of JCP JSR-166
3		* Expert Group and released to the public domain, as explained at
4	<	* http://creativecommons.org/licenses/publicdomain
4	>	* http://creativecommons.org/publicdomain/zero/1.0/
5		*/
6
7		package jsr166y;
8
9	–	import java.util.concurrent.*;
10	–
11	–	import java.util.Collection;
12	–
9		/**
10	<	* A thread managed by a {@link ForkJoinPool}.  This class is
11	<	* subclassable solely for the sake of adding functionality -- there
12	<	* are no overridable methods dealing with scheduling or execution.
13	<	* However, you can override initialization and termination methods
14	<	* surrounding the main task processing loop.  If you do create such a
15	<	* subclass, you will also need to supply a custom {@link
16	<	* ForkJoinPool.ForkJoinWorkerThreadFactory} to use it in a {@code
17	<	* ForkJoinPool}.
10	>	* A thread managed by a {@link ForkJoinPool}, which executes
11	>	* {@link ForkJoinTask}s.
12	>	* This class is subclassable solely for the sake of adding
13	>	* functionality -- there are no overridable methods dealing with
14	>	* scheduling or execution.  However, you can override initialization
15	>	* and termination methods surrounding the main task processing loop.
16	>	* If you do create such a subclass, you will also need to supply a
17	>	* custom {@link ForkJoinPool.ForkJoinWorkerThreadFactory} to use it
18	>	* in a {@code ForkJoinPool}.
19		*
20		* @since 1.7
21		* @author Doug Lea
22		*/
23		public class ForkJoinWorkerThread extends Thread {
24		/*
25	<	* Algorithm overview:
26	<	*
27	<	* 1. Work-Stealing: Work-stealing queues are special forms of
31	<	* Deques that support only three of the four possible
32	<	* end-operations -- push, pop, and deq (aka steal), and only do
33	<	* so under the constraints that push and pop are called only from
34	<	* the owning thread, while deq may be called from other threads.
35	<	* (If you are unfamiliar with them, you probably want to read
36	<	* Herlihy and Shavit's book "The Art of Multiprocessor
37	<	* programming", chapter 16 describing these in more detail before
38	<	* proceeding.)  The main work-stealing queue design is roughly
39	<	* similar to "Dynamic Circular Work-Stealing Deque" by David
40	<	* Chase and Yossi Lev, SPAA 2005
41	<	* (http://research.sun.com/scalable/pubs/index.html).  The main
42	<	* difference ultimately stems from gc requirements that we null
43	<	* out taken slots as soon as we can, to maintain as small a
44	<	* footprint as possible even in programs generating huge numbers
45	<	* of tasks. To accomplish this, we shift the CAS arbitrating pop
46	<	* vs deq (steal) from being on the indices ("base" and "sp") to
47	<	* the slots themselves (mainly via method "casSlotNull()"). So,
48	<	* both a successful pop and deq mainly entail CAS'ing a non-null
49	<	* slot to null.  Because we rely on CASes of references, we do
50	<	* not need tag bits on base or sp.  They are simple ints as used
51	<	* in any circular array-based queue (see for example ArrayDeque).
52	<	* Updates to the indices must still be ordered in a way that
53	<	* guarantees that (sp - base) > 0 means the queue is empty, but
54	<	* otherwise may err on the side of possibly making the queue
55	<	* appear nonempty when a push, pop, or deq have not fully
56	<	* committed. Note that this means that the deq operation,
57	<	* considered individually, is not wait-free. One thief cannot
58	<	* successfully continue until another in-progress one (or, if
59	<	* previously empty, a push) completes.  However, in the
60	<	* aggregate, we ensure at least probabilistic
61	<	* non-blockingness. If an attempted steal fails, a thief always
62	<	* chooses a different random victim target to try next. So, in
63	<	* order for one thief to progress, it suffices for any
64	<	* in-progress deq or new push on any empty queue to complete. One
65	<	* reason this works well here is that apparently-nonempty often
66	<	* means soon-to-be-stealable, which gives threads a chance to
67	<	* activate if necessary before stealing (see below).
68	<	*
69	<	* This approach also enables support for "async mode" where local
70	<	* task processing is in FIFO, not LIFO order; simply by using a
71	<	* version of deq rather than pop when locallyFifo is true (as set
72	<	* by the ForkJoinPool).  This allows use in message-passing
73	<	* frameworks in which tasks are never joined.
74	<	*
75	<	* Efficient implementation of this approach currently relies on
76	<	* an uncomfortable amount of "Unsafe" mechanics. To maintain
77	<	* correct orderings, reads and writes of variable base require
78	<	* volatile ordering.  Variable sp does not require volatile write
79	<	* but needs cheaper store-ordering on writes.  Because they are
80	<	* protected by volatile base reads, reads of the queue array and
81	<	* its slots do not need volatile load semantics, but writes (in
82	<	* push) require store order and CASes (in pop and deq) require
83	<	* (volatile) CAS semantics.  (See "Idempotent work stealing" by
84	<	* Michael, Saraswat, and Vechev, PPoPP 2009
85	<	* http://portal.acm.org/citation.cfm?id=1504186 for an algorithm
86	<	* with similar properties, but without support for nulling
87	<	* slots.)  Since these combinations aren't supported using
88	<	* ordinary volatiles, the only way to accomplish these
89	<	* efficiently is to use direct Unsafe calls. (Using external
90	<	* AtomicIntegers and AtomicReferenceArrays for the indices and
91	<	* array is significantly slower because of memory locality and
92	<	* indirection effects.)
93	<	*
94	<	* Further, performance on most platforms is very sensitive to
95	<	* placement and sizing of the (resizable) queue array.  Even
96	<	* though these queues don't usually become all that big, the
97	<	* initial size must be large enough to counteract cache
98	<	* contention effects across multiple queues (especially in the
99	<	* presence of GC cardmarking). Also, to improve thread-locality,
100	<	* queues are currently initialized immediately after the thread
101	<	* gets the initial signal to start processing tasks.  However,
102	<	* all queue-related methods except pushTask are written in a way
103	<	* that allows them to instead be lazily allocated and/or disposed
104	<	* of when empty. All together, these low-level implementation
105	<	* choices produce as much as a factor of 4 performance
106	<	* improvement compared to naive implementations, and enable the
107	<	* processing of billions of tasks per second, sometimes at the
108	<	* expense of ugliness.
109	<	*
110	<	* 2. Run control: The primary run control is based on a global
111	<	* counter (activeCount) held by the pool. It uses an algorithm
112	<	* similar to that in Herlihy and Shavit section 17.6 to cause
113	<	* threads to eventually block when all threads declare they are
114	<	* inactive. For this to work, threads must be declared active
115	<	* when executing tasks, and before stealing a task. They must be
116	<	* inactive before blocking on the Pool Barrier (awaiting a new
117	<	* submission or other Pool event). In between, there is some free
118	<	* play which we take advantage of to avoid contention and rapid
119	<	* flickering of the global activeCount: If inactive, we activate
120	<	* only if a victim queue appears to be nonempty (see above).
121	<	* Similarly, a thread tries to inactivate only after a full scan
122	<	* of other threads.  The net effect is that contention on
123	<	* activeCount is rarely a measurable performance issue. (There
124	<	* are also a few other cases where we scan for work rather than
125	<	* retry/block upon contention.)
126	<	*
127	<	* 3. Selection control. We maintain policy of always choosing to
128	<	* run local tasks rather than stealing, and always trying to
129	<	* steal tasks before trying to run a new submission. All steals
130	<	* are currently performed in randomly-chosen deq-order. It may be
131	<	* worthwhile to bias these with locality / anti-locality
132	<	* information, but doing this well probably requires more
133	<	* lower-level information from JVMs than currently provided.
134	<	*/
135	<
136	<	/**
137	<	* Capacity of work-stealing queue array upon initialization.
138	<	* Must be a power of two. Initial size must be at least 2, but is
139	<	* padded to minimize cache effects.
140	<	*/
141	<	private static final int INITIAL_QUEUE_CAPACITY = 1 << 13;
142	<
143	<	/**
144	<	* Maximum work-stealing queue array size.  Must be less than or
145	<	* equal to 1 << 28 to ensure lack of index wraparound. (This
146	<	* is less than usual bounds, because we need leftshift by 3
147	<	* to be in int range).
148	<	*/
149	<	private static final int MAXIMUM_QUEUE_CAPACITY = 1 << 28;
150	<
151	<	/**
152	<	* The pool this thread works in. Accessed directly by ForkJoinTask.
153	<	*/
154	<	final ForkJoinPool pool;
155	<
156	<	/**
157	<	* The work-stealing queue array. Size must be a power of two.
158	<	* Initialized when thread starts, to improve memory locality.
159	<	*/
160	<	private ForkJoinTask<?>[] queue;
161	<
162	<	/**
163	<	* Index (mod queue.length) of next queue slot to push to or pop
164	<	* from. It is written only by owner thread, via ordered store.
165	<	* Both sp and base are allowed to wrap around on overflow, but
166	<	* (sp - base) still estimates size.
167	<	*/
168	<	private volatile int sp;
169	<
170	<	/**
171	<	* Index (mod queue.length) of least valid queue slot, which is
172	<	* always the next position to steal from if nonempty.
173	<	*/
174	<	private volatile int base;
175	<
176	<	/**
177	<	* Activity status. When true, this worker is considered active.
178	<	* Must be false upon construction. It must be true when executing
179	<	* tasks, and BEFORE stealing a task. It must be false before
180	<	* calling pool.sync.
181	<	*/
182	<	private boolean active;
183	<
184	<	/**
185	<	* Run state of this worker. Supports simple versions of the usual
186	<	* shutdown/shutdownNow control.
187	<	*/
188	<	private volatile int runState;
189	<
190	<	/**
191	<	* Seed for random number generator for choosing steal victims.
192	<	* Uses Marsaglia xorshift. Must be nonzero upon initialization.
193	<	*/
194	<	private int seed;
195	<
196	<	/**
197	<	* Number of steals, transferred to pool when idle
25	>	* ForkJoinWorkerThreads are managed by ForkJoinPools and perform
26	>	* ForkJoinTasks. For explanation, see the internal documentation
27	>	* of class ForkJoinPool.
28		*/
199	–	private int stealCount;
29
30	<	/**
31	<	* Index of this worker in pool array. Set once by pool before
203	<	* running, and accessed directly by pool during cleanup etc.
204	<	*/
205	<	int poolIndex;
206	<
207	<	/**
208	<	* The last barrier event waited for. Accessed in pool callback
209	<	* methods, but only by current thread.
210	<	*/
211	<	long lastEventCount;
212	<
213	<	/**
214	<	* True if use local fifo, not default lifo, for local polling
215	<	*/
216	<	private boolean locallyFifo;
30	>	final ForkJoinPool.WorkQueue workQueue; // Work-stealing mechanics
31	>	final ForkJoinPool pool;                // the pool this thread works in
32
33		/**
34		* Creates a ForkJoinWorkerThread operating in the given pool.
#	Line 222 | Line 37 | public class ForkJoinWorkerThread extend
37		* @throws NullPointerException if pool is null
38		*/
39		protected ForkJoinWorkerThread(ForkJoinPool pool) {
40	<	if (pool == null) throw new NullPointerException();
40	>	super(pool.nextWorkerName());
41	>	setDaemon(true);
42	>	Thread.UncaughtExceptionHandler ueh = pool.ueh;
43	>	if (ueh != null)
44	>	setUncaughtExceptionHandler(ueh);
45		this.pool = pool;
46	<	// Note: poolIndex is set by pool during construction
47	<	// Remaining initialization is deferred to onStart
46	>	pool.registerWorker(this.workQueue = new ForkJoinPool.WorkQueue
47	>	(pool, this, pool.localMode));
48		}
49
231	–	// Public access methods
232	–
50		/**
51		* Returns the pool hosting this thread.
52		*
#	Line 249 | Line 66 | public class ForkJoinWorkerThread extend
66		* @return the index number
67		*/
68		public int getPoolIndex() {
69	<	return poolIndex;
253	<	}
254	<
255	<	/**
256	<	* Establishes local first-in-first-out scheduling mode for forked
257	<	* tasks that are never joined.
258	<	*
259	<	* @param async if true, use locally FIFO scheduling
260	<	*/
261	<	void setAsyncMode(boolean async) {
262	<	locallyFifo = async;
263	<	}
264	<
265	<	// Runstate management
266	<
267	<	// Runstate values. Order matters
268	<	private static final int RUNNING     = 0;
269	<	private static final int SHUTDOWN    = 1;
270	<	private static final int TERMINATING = 2;
271	<	private static final int TERMINATED  = 3;
272	<
273	<	final boolean isShutdown()    { return runState >= SHUTDOWN;  }
274	<	final boolean isTerminating() { return runState >= TERMINATING;  }
275	<	final boolean isTerminated()  { return runState == TERMINATED; }
276	<	final boolean shutdown()      { return transitionRunStateTo(SHUTDOWN); }
277	<	final boolean shutdownNow()   { return transitionRunStateTo(TERMINATING); }
278	<
279	<	/**
280	<	* Transitions to at least the given state.
281	<	*
282	<	* @return {@code true} if not already at least at given state
283	<	*/
284	<	private boolean transitionRunStateTo(int state) {
285	<	for (;;) {
286	<	int s = runState;
287	<	if (s >= state)
288	<	return false;
289	<	if (UNSAFE.compareAndSwapInt(this, runStateOffset, s, state))
290	<	return true;
291	<	}
292	<	}
293	<
294	<	/**
295	<	* Tries to set status to active; fails on contention.
296	<	*/
297	<	private boolean tryActivate() {
298	<	if (!active) {
299	<	if (!pool.tryIncrementActiveCount())
300	<	return false;
301	<	active = true;
302	<	}
303	<	return true;
304	<	}
305	<
306	<	/**
307	<	* Tries to set status to inactive; fails on contention.
308	<	*/
309	<	private boolean tryInactivate() {
310	<	if (active) {
311	<	if (!pool.tryDecrementActiveCount())
312	<	return false;
313	<	active = false;
314	<	}
315	<	return true;
316	<	}
317	<
318	<	/**
319	<	* Computes next value for random victim probe.  Scans don't
320	<	* require a very high quality generator, but also not a crummy
321	<	* one.  Marsaglia xor-shift is cheap and works well.
322	<	*/
323	<	private static int xorShift(int r) {
324	<	r ^= (r << 13);
325	<	r ^= (r >>> 17);
326	<	return r ^ (r << 5);
327	<	}
328	<
329	<	// Lifecycle methods
330	<
331	<	/**
332	<	* This method is required to be public, but should never be
333	<	* called explicitly. It performs the main run loop to execute
334	<	* ForkJoinTasks.
335	<	*/
336	<	public void run() {
337	<	Throwable exception = null;
338	<	try {
339	<	onStart();
340	<	pool.sync(this); // await first pool event
341	<	mainLoop();
342	<	} catch (Throwable ex) {
343	<	exception = ex;
344	<	} finally {
345	<	onTermination(exception);
346	<	}
347	<	}
348	<
349	<	/**
350	<	* Executes tasks until shut down.
351	<	*/
352	<	private void mainLoop() {
353	<	while (!isShutdown()) {
354	<	ForkJoinTask<?> t = pollTask();
355	<	if (t != null || (t = pollSubmission()) != null)
356	<	t.quietlyExec();
357	<	else if (tryInactivate())
358	<	pool.sync(this);
359	<	}
69	>	return workQueue.poolIndex;
70		}
71
72		/**
73		* Initializes internal state after construction but before
74		* processing any tasks. If you override this method, you must
75	<	* invoke super.onStart() at the beginning of the method.
75	>	* invoke {@code super.onStart()} at the beginning of the method.
76		* Initialization requires care: Most fields must have legal
77		* default values, to ensure that attempted accesses from other
78		* threads work correctly even before this thread starts
79		* processing tasks.
80		*/
81		protected void onStart() {
372	–	// Allocate while starting to improve chances of thread-local
373	–	// isolation
374	–	queue = new ForkJoinTask<?>[INITIAL_QUEUE_CAPACITY];
375	–	// Initial value of seed need not be especially random but
376	–	// should differ across workers and must be nonzero
377	–	int p = poolIndex + 1;
378	–	seed = p + (p << 8) + (p << 16) + (p << 24); // spread bits
82		}
83
84		/**
#	Line 387 | Line 90 | public class ForkJoinWorkerThread extend
90		* to an unrecoverable error, or {@code null} if completed normally
91		*/
92		protected void onTermination(Throwable exception) {
390	–	// Execute remaining local tasks unless aborting or terminating
391	–	while (exception == null && pool.isProcessingTasks() && base != sp) {
392	–	try {
393	–	ForkJoinTask<?> t = popTask();
394	–	if (t != null)
395	–	t.quietlyExec();
396	–	} catch (Throwable ex) {
397	–	exception = ex;
398	–	}
399	–	}
400	–	// Cancel other tasks, transition status, notify pool, and
401	–	// propagate exception to uncaught exception handler
402	–	try {
403	–	do {} while (!tryInactivate()); // ensure inactive
404	–	cancelTasks();
405	–	runState = TERMINATED;
406	–	pool.workerTerminated(this);
407	–	} catch (Throwable ex) {        // Shouldn't ever happen
408	–	if (exception == null)      // but if so, at least rethrown
409	–	exception = ex;
410	–	} finally {
411	–	if (exception != null)
412	–	ForkJoinTask.rethrowException(exception);
413	–	}
414	–	}
415	–
416	–	// Intrinsics-based support for queue operations.
417	–
418	–	private static long slotOffset(int i) {
419	–	return ((long) i << qShift) + qBase;
93		}
94
95		/**
96	<	* Adds in store-order the given task at given slot of q to null.
97	<	* Caller must ensure q is non-null and index is in range.
98	<	*/
426	<	private static void setSlot(ForkJoinTask<?>[] q, int i,
427	<	ForkJoinTask<?> t) {
428	<	UNSAFE.putOrderedObject(q, slotOffset(i), t);
429	<	}
430	<
431	<	/**
432	<	* CAS given slot of q to null. Caller must ensure q is non-null
433	<	* and index is in range.
434	<	*/
435	<	private static boolean casSlotNull(ForkJoinTask<?>[] q, int i,
436	<	ForkJoinTask<?> t) {
437	<	return UNSAFE.compareAndSwapObject(q, slotOffset(i), t, null);
438	<	}
439	<
440	<	/**
441	<	* Sets sp in store-order.
442	<	*/
443	<	private void storeSp(int s) {
444	<	UNSAFE.putOrderedInt(this, spOffset, s);
445	<	}
446	<
447	<	// Main queue methods
448	<
449	<	/**
450	<	* Pushes a task. Called only by current thread.
451	<	*
452	<	* @param t the task. Caller must ensure non-null.
453	<	*/
454	<	final void pushTask(ForkJoinTask<?> t) {
455	<	ForkJoinTask<?>[] q = queue;
456	<	int mask = q.length - 1;
457	<	int s = sp;
458	<	setSlot(q, s & mask, t);
459	<	storeSp(++s);
460	<	if ((s -= base) == 1)
461	<	pool.signalWork();
462	<	else if (s >= mask)
463	<	growQueue();
464	<	}
465	<
466	<	/**
467	<	* Tries to take a task from the base of the queue, failing if
468	<	* either empty or contended.
469	<	*
470	<	* @return a task, or null if none or contended
471	<	*/
472	<	final ForkJoinTask<?> deqTask() {
473	<	ForkJoinTask<?> t;
474	<	ForkJoinTask<?>[] q;
475	<	int i;
476	<	int b;
477	<	if (sp != (b = base) &&
478	<	(q = queue) != null && // must read q after b
479	<	(t = q[i = (q.length - 1) & b]) != null &&
480	<	casSlotNull(q, i, t)) {
481	<	base = b + 1;
482	<	return t;
483	<	}
484	<	return null;
485	<	}
486	<
487	<	/**
488	<	* Tries to take a task from the base of own queue, activating if
489	<	* necessary, failing only if empty. Called only by current thread.
490	<	*
491	<	* @return a task, or null if none
492	<	*/
493	<	final ForkJoinTask<?> locallyDeqTask() {
494	<	int b;
495	<	while (sp != (b = base)) {
496	<	if (tryActivate()) {
497	<	ForkJoinTask<?>[] q = queue;
498	<	int i = (q.length - 1) & b;
499	<	ForkJoinTask<?> t = q[i];
500	<	if (t != null && casSlotNull(q, i, t)) {
501	<	base = b + 1;
502	<	return t;
503	<	}
504	<	}
505	<	}
506	<	return null;
507	<	}
508	<
509	<	/**
510	<	* Returns a popped task, or null if empty. Ensures active status
511	<	* if non-null. Called only by current thread.
512	<	*/
513	<	final ForkJoinTask<?> popTask() {
514	<	int s = sp;
515	<	while (s != base) {
516	<	if (tryActivate()) {
517	<	ForkJoinTask<?>[] q = queue;
518	<	int mask = q.length - 1;
519	<	int i = (s - 1) & mask;
520	<	ForkJoinTask<?> t = q[i];
521	<	if (t == null || !casSlotNull(q, i, t))
522	<	break;
523	<	storeSp(s - 1);
524	<	return t;
525	<	}
526	<	}
527	<	return null;
528	<	}
529	<
530	<	/**
531	<	* Specialized version of popTask to pop only if
532	<	* topmost element is the given task. Called only
533	<	* by current thread while active.
534	<	*
535	<	* @param t the task. Caller must ensure non-null.
536	<	*/
537	<	final boolean unpushTask(ForkJoinTask<?> t) {
538	<	ForkJoinTask<?>[] q = queue;
539	<	int mask = q.length - 1;
540	<	int s = sp - 1;
541	<	if (casSlotNull(q, s & mask, t)) {
542	<	storeSp(s);
543	<	return true;
544	<	}
545	<	return false;
546	<	}
547	<
548	<	/**
549	<	* Returns next task or null if empty or contended
550	<	*/
551	<	final ForkJoinTask<?> peekTask() {
552	<	ForkJoinTask<?>[] q = queue;
553	<	if (q == null)
554	<	return null;
555	<	int mask = q.length - 1;
556	<	int i = locallyFifo ? base : (sp - 1);
557	<	return q[i & mask];
558	<	}
559	<
560	<	/**
561	<	* Doubles queue array size. Transfers elements by emulating
562	<	* steals (deqs) from old array and placing, oldest first, into
563	<	* new array.
564	<	*/
565	<	private void growQueue() {
566	<	ForkJoinTask<?>[] oldQ = queue;
567	<	int oldSize = oldQ.length;
568	<	int newSize = oldSize << 1;
569	<	if (newSize > MAXIMUM_QUEUE_CAPACITY)
570	<	throw new RejectedExecutionException("Queue capacity exceeded");
571	<	ForkJoinTask<?>[] newQ = queue = new ForkJoinTask<?>[newSize];
572	<
573	<	int b = base;
574	<	int bf = b + oldSize;
575	<	int oldMask = oldSize - 1;
576	<	int newMask = newSize - 1;
577	<	do {
578	<	int oldIndex = b & oldMask;
579	<	ForkJoinTask<?> t = oldQ[oldIndex];
580	<	if (t != null && !casSlotNull(oldQ, oldIndex, t))
581	<	t = null;
582	<	setSlot(newQ, b & newMask, t);
583	<	} while (++b != bf);
584	<	pool.signalWork();
585	<	}
586	<
587	<	/**
588	<	* Tries to steal a task from another worker. Starts at a random
589	<	* index of workers array, and probes workers until finding one
590	<	* with non-empty queue or finding that all are empty.  It
591	<	* randomly selects the first n probes. If these are empty, it
592	<	* resorts to a full circular traversal, which is necessary to
593	<	* accurately set active status by caller. Also restarts if pool
594	<	* events occurred since last scan, which forces refresh of
595	<	* workers array, in case barrier was associated with resize.
596	<	*
597	<	* This method must be both fast and quiet -- usually avoiding
598	<	* memory accesses that could disrupt cache sharing etc other than
599	<	* those needed to check for and take tasks. This accounts for,
600	<	* among other things, updating random seed in place without
601	<	* storing it until exit.
602	<	*
603	<	* @return a task, or null if none found
604	<	*/
605	<	private ForkJoinTask<?> scan() {
606	<	ForkJoinTask<?> t = null;
607	<	int r = seed;                    // extract once to keep scan quiet
608	<	ForkJoinWorkerThread[] ws;       // refreshed on outer loop
609	<	int mask;                        // must be power 2 minus 1 and > 0
610	<	outer:do {
611	<	if ((ws = pool.workers) != null && (mask = ws.length - 1) > 0) {
612	<	int idx = r;
613	<	int probes = ~mask;      // use random index while negative
614	<	for (;;) {
615	<	r = xorShift(r);     // update random seed
616	<	ForkJoinWorkerThread v = ws[mask & idx];
617	<	if (v == null || v.sp == v.base) {
618	<	if (probes <= mask)
619	<	idx = (probes++ < 0) ? r : (idx + 1);
620	<	else
621	<	break;
622	<	}
623	<	else if (!tryActivate() || (t = v.deqTask()) == null)
624	<	continue outer;  // restart on contention
625	<	else
626	<	break outer;
627	<	}
628	<	}
629	<	} while (pool.hasNewSyncEvent(this)); // retry on pool events
630	<	seed = r;
631	<	return t;
632	<	}
633	<
634	<	/**
635	<	* Gets and removes a local or stolen task.
636	<	*
637	<	* @return a task, if available
638	<	*/
639	<	final ForkJoinTask<?> pollTask() {
640	<	ForkJoinTask<?> t = locallyFifo ? locallyDeqTask() : popTask();
641	<	if (t == null && (t = scan()) != null)
642	<	++stealCount;
643	<	return t;
644	<	}
645	<
646	<	/**
647	<	* Gets a local task.
648	<	*
649	<	* @return a task, if available
650	<	*/
651	<	final ForkJoinTask<?> pollLocalTask() {
652	<	return locallyFifo ? locallyDeqTask() : popTask();
653	<	}
654	<
655	<	/**
656	<	* Returns a pool submission, if one exists, activating first.
657	<	*
658	<	* @return a submission, if available
659	<	*/
660	<	private ForkJoinTask<?> pollSubmission() {
661	<	ForkJoinPool p = pool;
662	<	while (p.hasQueuedSubmissions()) {
663	<	ForkJoinTask<?> t;
664	<	if (tryActivate() && (t = p.pollSubmission()) != null)
665	<	return t;
666	<	}
667	<	return null;
668	<	}
669	<
670	<	// Methods accessed only by Pool
671	<
672	<	/**
673	<	* Removes and cancels all tasks in queue.  Can be called from any
674	<	* thread.
675	<	*/
676	<	final void cancelTasks() {
677	<	ForkJoinTask<?> t;
678	<	while (base != sp && (t = deqTask()) != null)
679	<	t.cancelIgnoringExceptions();
680	<	}
681	<
682	<	/**
683	<	* Drains tasks to given collection c.
684	<	*
685	<	* @return the number of tasks drained
686	<	*/
687	<	final int drainTasksTo(Collection<? super ForkJoinTask<?>> c) {
688	<	int n = 0;
689	<	ForkJoinTask<?> t;
690	<	while (base != sp && (t = deqTask()) != null) {
691	<	c.add(t);
692	<	++n;
693	<	}
694	<	return n;
695	<	}
696	<
697	<	/**
698	<	* Gets and clears steal count for accumulation by pool.  Called
699	<	* only when known to be idle (in pool.sync and termination).
700	<	*/
701	<	final int getAndClearStealCount() {
702	<	int sc = stealCount;
703	<	stealCount = 0;
704	<	return sc;
705	<	}
706	<
707	<	/**
708	<	* Returns {@code true} if at least one worker in the given array
709	<	* appears to have at least one queued task.
710	<	*
711	<	* @param ws array of workers
712	<	*/
713	<	static boolean hasQueuedTasks(ForkJoinWorkerThread[] ws) {
714	<	if (ws != null) {
715	<	int len = ws.length;
716	<	for (int j = 0; j < 2; ++j) { // need two passes for clean sweep
717	<	for (int i = 0; i < len; ++i) {
718	<	ForkJoinWorkerThread w = ws[i];
719	<	if (w != null && w.sp != w.base)
720	<	return true;
721	<	}
722	<	}
723	<	}
724	<	return false;
725	<	}
726	<
727	<	// Support methods for ForkJoinTask
728	<
729	<	/**
730	<	* Returns an estimate of the number of tasks in the queue.
731	<	*/
732	<	final int getQueueSize() {
733	<	// suppress momentarily negative values
734	<	return Math.max(0, sp - base);
735	<	}
736	<
737	<	/**
738	<	* Returns an estimate of the number of tasks, offset by a
739	<	* function of number of idle workers.
740	<	*/
741	<	final int getEstimatedSurplusTaskCount() {
742	<	// The halving approximates weighting idle vs non-idle workers
743	<	return (sp - base) - (pool.getIdleThreadCount() >>> 1);
744	<	}
745	<
746	<	/**
747	<	* Scans, returning early if joinMe done.
748	<	*/
749	<	final ForkJoinTask<?> scanWhileJoining(ForkJoinTask<?> joinMe) {
750	<	ForkJoinTask<?> t = pollTask();
751	<	if (t != null && joinMe.status < 0 && sp == base) {
752	<	pushTask(t); // unsteal if done and this task would be stealable
753	<	t = null;
754	<	}
755	<	return t;
756	<	}
757	<
758	<	/**
759	<	* Runs tasks until {@code pool.isQuiescent()}.
760	<	*/
761	<	final void helpQuiescePool() {
762	<	for (;;) {
763	<	ForkJoinTask<?> t = pollTask();
764	<	if (t != null)
765	<	t.quietlyExec();
766	<	else if (tryInactivate() && pool.isQuiescent())
767	<	break;
768	<	}
769	<	do {} while (!tryActivate()); // re-activate on exit
770	<	}
771	<
772	<	// Unsafe mechanics
773	<
774	<	private static final sun.misc.Unsafe UNSAFE = getUnsafe();
775	<	private static final long spOffset =
776	<	objectFieldOffset("sp", ForkJoinWorkerThread.class);
777	<	private static final long runStateOffset =
778	<	objectFieldOffset("runState", ForkJoinWorkerThread.class);
779	<	private static final long qBase;
780	<	private static final int qShift;
781	<
782	<	static {
783	<	qBase = UNSAFE.arrayBaseOffset(ForkJoinTask[].class);
784	<	int s = UNSAFE.arrayIndexScale(ForkJoinTask[].class);
785	<	if ((s & (s-1)) != 0)
786	<	throw new Error("data type scale not a power of two");
787	<	qShift = 31 - Integer.numberOfLeadingZeros(s);
788	<	}
789	<
790	<	private static long objectFieldOffset(String field, Class<?> klazz) {
791	<	try {
792	<	return UNSAFE.objectFieldOffset(klazz.getDeclaredField(field));
793	<	} catch (NoSuchFieldException e) {
794	<	// Convert Exception to corresponding Error
795	<	NoSuchFieldError error = new NoSuchFieldError(field);
796	<	error.initCause(e);
797	<	throw error;
798	<	}
799	<	}
800	<
801	<	/**
802	<	* Returns a sun.misc.Unsafe.  Suitable for use in a 3rd party package.
803	<	* Replace with a simple call to Unsafe.getUnsafe when integrating
804	<	* into a jdk.
805	<	*
806	<	* @return a sun.misc.Unsafe
96	>	* This method is required to be public, but should never be
97	>	* called explicitly. It performs the main run loop to execute
98	>	* {@link ForkJoinTask}s.
99		*/
100	<	private static sun.misc.Unsafe getUnsafe() {
100	>	public void run() {
101	>	Throwable exception = null;
102		try {
103	<	return sun.misc.Unsafe.getUnsafe();
104	<	} catch (SecurityException se) {
103	>	onStart();
104	>	pool.runWorker(workQueue);
105	>	} catch (Throwable ex) {
106	>	exception = ex;
107	>	} finally {
108		try {
109	<	return java.security.AccessController.doPrivileged
110	<	(new java.security
111	<	.PrivilegedExceptionAction<sun.misc.Unsafe>() {
112	<	public sun.misc.Unsafe run() throws Exception {
113	<	java.lang.reflect.Field f = sun.misc
114	<	.Unsafe.class.getDeclaredField("theUnsafe");
819	<	f.setAccessible(true);
820	<	return (sun.misc.Unsafe) f.get(null);
821	<	}});
822	<	} catch (java.security.PrivilegedActionException e) {
823	<	throw new RuntimeException("Could not initialize intrinsics",
824	<	e.getCause());
109	>	onTermination(exception);
110	>	} catch (Throwable ex) {
111	>	if (exception == null)
112	>	exception = ex;
113	>	} finally {
114	>	pool.deregisterWorker(this, exception);
115		}
116		}
117		}

Diff Legend

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