2 * DO NOT ALTER OR REMOVE COPYRIGHT NOTICES OR THIS FILE HEADER.
4 * This code is free software; you can redistribute it and/or modify it
5 * under the terms of the GNU General Public License version 2 only, as
6 * published by the Free Software Foundation. Oracle designates this
7 * particular file as subject to the "Classpath" exception as provided
8 * by Oracle in the LICENSE file that accompanied this code.
10 * This code is distributed in the hope that it will be useful, but WITHOUT
11 * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
12 * FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License
13 * version 2 for more details (a copy is included in the LICENSE file that
14 * accompanied this code).
16 * You should have received a copy of the GNU General Public License version
17 * 2 along with this work; if not, write to the Free Software Foundation,
18 * Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA.
20 * Please contact Oracle, 500 Oracle Parkway, Redwood Shores, CA 94065 USA
21 * or visit www.oracle.com if you need additional information or have any
26 * This file is available under and governed by the GNU General Public
27 * License version 2 only, as published by the Free Software Foundation.
28 * However, the following notice accompanied the original version of this
31 * Written by Doug Lea, Bill Scherer, and Michael Scott with
32 * assistance from members of JCP JSR-166 Expert Group and released to
33 * the public domain, as explained at
34 * http://creativecommons.org/publicdomain/zero/1.0/
37 package java.util.concurrent;
38 import java.util.concurrent.locks.*;
42 * A {@linkplain BlockingQueue blocking queue} in which each insert
43 * operation must wait for a corresponding remove operation by another
44 * thread, and vice versa. A synchronous queue does not have any
45 * internal capacity, not even a capacity of one. You cannot
46 * <tt>peek</tt> at a synchronous queue because an element is only
47 * present when you try to remove it; you cannot insert an element
48 * (using any method) unless another thread is trying to remove it;
49 * you cannot iterate as there is nothing to iterate. The
50 * <em>head</em> of the queue is the element that the first queued
51 * inserting thread is trying to add to the queue; if there is no such
52 * queued thread then no element is available for removal and
53 * <tt>poll()</tt> will return <tt>null</tt>. For purposes of other
54 * <tt>Collection</tt> methods (for example <tt>contains</tt>), a
55 * <tt>SynchronousQueue</tt> acts as an empty collection. This queue
56 * does not permit <tt>null</tt> elements.
58 * <p>Synchronous queues are similar to rendezvous channels used in
59 * CSP and Ada. They are well suited for handoff designs, in which an
60 * object running in one thread must sync up with an object running
61 * in another thread in order to hand it some information, event, or
64 * <p> This class supports an optional fairness policy for ordering
65 * waiting producer and consumer threads. By default, this ordering
66 * is not guaranteed. However, a queue constructed with fairness set
67 * to <tt>true</tt> grants threads access in FIFO order.
69 * <p>This class and its iterator implement all of the
70 * <em>optional</em> methods of the {@link Collection} and {@link
71 * Iterator} interfaces.
73 * <p>This class is a member of the
74 * <a href="{@docRoot}/../technotes/guides/collections/index.html">
75 * Java Collections Framework</a>.
78 * @author Doug Lea and Bill Scherer and Michael Scott
79 * @param <E> the type of elements held in this collection
81 public class SynchronousQueue<E> extends AbstractQueue<E>
82 implements BlockingQueue<E>, java.io.Serializable {
83 private static final long serialVersionUID = -3223113410248163686L;
86 * This class implements extensions of the dual stack and dual
87 * queue algorithms described in "Nonblocking Concurrent Objects
88 * with Condition Synchronization", by W. N. Scherer III and
89 * M. L. Scott. 18th Annual Conf. on Distributed Computing,
91 * http://www.cs.rochester.edu/u/scott/synchronization/pseudocode/duals.html).
92 * The (Lifo) stack is used for non-fair mode, and the (Fifo)
93 * queue for fair mode. The performance of the two is generally
94 * similar. Fifo usually supports higher throughput under
95 * contention but Lifo maintains higher thread locality in common
98 * A dual queue (and similarly stack) is one that at any given
99 * time either holds "data" -- items provided by put operations,
100 * or "requests" -- slots representing take operations, or is
101 * empty. A call to "fulfill" (i.e., a call requesting an item
102 * from a queue holding data or vice versa) dequeues a
103 * complementary node. The most interesting feature of these
104 * queues is that any operation can figure out which mode the
105 * queue is in, and act accordingly without needing locks.
107 * Both the queue and stack extend abstract class Transferer
108 * defining the single method transfer that does a put or a
109 * take. These are unified into a single method because in dual
110 * data structures, the put and take operations are symmetrical,
111 * so nearly all code can be combined. The resulting transfer
112 * methods are on the long side, but are easier to follow than
113 * they would be if broken up into nearly-duplicated parts.
115 * The queue and stack data structures share many conceptual
116 * similarities but very few concrete details. For simplicity,
117 * they are kept distinct so that they can later evolve
120 * The algorithms here differ from the versions in the above paper
121 * in extending them for use in synchronous queues, as well as
122 * dealing with cancellation. The main differences include:
124 * 1. The original algorithms used bit-marked pointers, but
125 * the ones here use mode bits in nodes, leading to a number
126 * of further adaptations.
127 * 2. SynchronousQueues must block threads waiting to become
129 * 3. Support for cancellation via timeout and interrupts,
130 * including cleaning out cancelled nodes/threads
131 * from lists to avoid garbage retention and memory depletion.
133 * Blocking is mainly accomplished using LockSupport park/unpark,
134 * except that nodes that appear to be the next ones to become
135 * fulfilled first spin a bit (on multiprocessors only). On very
136 * busy synchronous queues, spinning can dramatically improve
137 * throughput. And on less busy ones, the amount of spinning is
138 * small enough not to be noticeable.
140 * Cleaning is done in different ways in queues vs stacks. For
141 * queues, we can almost always remove a node immediately in O(1)
142 * time (modulo retries for consistency checks) when it is
143 * cancelled. But if it may be pinned as the current tail, it must
144 * wait until some subsequent cancellation. For stacks, we need a
145 * potentially O(n) traversal to be sure that we can remove the
146 * node, but this can run concurrently with other threads
147 * accessing the stack.
149 * While garbage collection takes care of most node reclamation
150 * issues that otherwise complicate nonblocking algorithms, care
151 * is taken to "forget" references to data, other nodes, and
152 * threads that might be held on to long-term by blocked
153 * threads. In cases where setting to null would otherwise
154 * conflict with main algorithms, this is done by changing a
155 * node's link to now point to the node itself. This doesn't arise
156 * much for Stack nodes (because blocked threads do not hang on to
157 * old head pointers), but references in Queue nodes must be
158 * aggressively forgotten to avoid reachability of everything any
159 * node has ever referred to since arrival.
163 * Shared internal API for dual stacks and queues.
165 abstract static class Transferer {
167 * Performs a put or take.
169 * @param e if non-null, the item to be handed to a consumer;
170 * if null, requests that transfer return an item
171 * offered by producer.
172 * @param timed if this operation should timeout
173 * @param nanos the timeout, in nanoseconds
174 * @return if non-null, the item provided or received; if null,
175 * the operation failed due to timeout or interrupt --
176 * the caller can distinguish which of these occurred
177 * by checking Thread.interrupted.
179 abstract Object transfer(Object e, boolean timed, long nanos);
182 /** The number of CPUs, for spin control */
183 static final int NCPUS = 1;
186 * The number of times to spin before blocking in timed waits.
187 * The value is empirically derived -- it works well across a
188 * variety of processors and OSes. Empirically, the best value
189 * seems not to vary with number of CPUs (beyond 2) so is just
192 static final int maxTimedSpins = (NCPUS < 2) ? 0 : 32;
195 * The number of times to spin before blocking in untimed waits.
196 * This is greater than timed value because untimed waits spin
197 * faster since they don't need to check times on each spin.
199 static final int maxUntimedSpins = maxTimedSpins * 16;
202 * The number of nanoseconds for which it is faster to spin
203 * rather than to use timed park. A rough estimate suffices.
205 static final long spinForTimeoutThreshold = 1000L;
208 static final class TransferStack extends Transferer {
210 * This extends Scherer-Scott dual stack algorithm, differing,
211 * among other ways, by using "covering" nodes rather than
212 * bit-marked pointers: Fulfilling operations push on marker
213 * nodes (with FULFILLING bit set in mode) to reserve a spot
214 * to match a waiting node.
217 /* Modes for SNodes, ORed together in node fields */
218 /** Node represents an unfulfilled consumer */
219 static final int REQUEST = 0;
220 /** Node represents an unfulfilled producer */
221 static final int DATA = 1;
222 /** Node is fulfilling another unfulfilled DATA or REQUEST */
223 static final int FULFILLING = 2;
225 /** Return true if m has fulfilling bit set */
226 static boolean isFulfilling(int m) { return (m & FULFILLING) != 0; }
228 /** Node class for TransferStacks. */
229 static final class SNode {
230 volatile SNode next; // next node in stack
231 volatile SNode match; // the node matched to this
232 volatile Thread waiter; // to control park/unpark
233 Object item; // data; or null for REQUESTs
235 // Note: item and mode fields don't need to be volatile
236 // since they are always written before, and read after,
237 // other volatile/atomic operations.
243 boolean casNext(SNode cmp, SNode val) {
252 * Tries to match node s to this node, if so, waking up thread.
253 * Fulfillers call tryMatch to identify their waiters.
254 * Waiters block until they have been matched.
256 * @param s the node to match
257 * @return true if successfully matched to s
259 boolean tryMatch(SNode s) {
263 if (w != null) { // waiters need at most one unpark
265 LockSupport.unpark(w);
273 * Tries to cancel a wait by matching node to itself.
281 boolean isCancelled() {
282 return match == this;
286 /** The head (top) of the stack */
289 boolean casHead(SNode h, SNode nh) {
298 * Creates or resets fields of a node. Called only from transfer
299 * where the node to push on stack is lazily created and
300 * reused when possible to help reduce intervals between reads
301 * and CASes of head and to avoid surges of garbage when CASes
302 * to push nodes fail due to contention.
304 static SNode snode(SNode s, Object e, SNode next, int mode) {
305 if (s == null) s = new SNode(e);
312 * Puts or takes an item.
314 Object transfer(Object e, boolean timed, long nanos) {
316 * Basic algorithm is to loop trying one of three actions:
318 * 1. If apparently empty or already containing nodes of same
319 * mode, try to push node on stack and wait for a match,
320 * returning it, or null if cancelled.
322 * 2. If apparently containing node of complementary mode,
323 * try to push a fulfilling node on to stack, match
324 * with corresponding waiting node, pop both from
325 * stack, and return matched item. The matching or
326 * unlinking might not actually be necessary because of
327 * other threads performing action 3:
329 * 3. If top of stack already holds another fulfilling node,
330 * help it out by doing its match and/or pop
331 * operations, and then continue. The code for helping
332 * is essentially the same as for fulfilling, except
333 * that it doesn't return the item.
336 SNode s = null; // constructed/reused as needed
337 int mode = (e == null) ? REQUEST : DATA;
341 if (h == null || h.mode == mode) { // empty or same-mode
342 if (timed && nanos <= 0) { // can't wait
343 if (h != null && h.isCancelled())
344 casHead(h, h.next); // pop cancelled node
347 } else if (casHead(h, s = snode(s, e, h, mode))) {
348 SNode m = awaitFulfill(s, timed, nanos);
349 if (m == s) { // wait was cancelled
353 if ((h = head) != null && h.next == s)
354 casHead(h, s.next); // help s's fulfiller
355 return (mode == REQUEST) ? m.item : s.item;
357 } else if (!isFulfilling(h.mode)) { // try to fulfill
358 if (h.isCancelled()) // already cancelled
359 casHead(h, h.next); // pop and retry
360 else if (casHead(h, s=snode(s, e, h, FULFILLING|mode))) {
361 for (;;) { // loop until matched or waiters disappear
362 SNode m = s.next; // m is s's match
363 if (m == null) { // all waiters are gone
364 casHead(s, null); // pop fulfill node
365 s = null; // use new node next time
366 break; // restart main loop
370 casHead(s, mn); // pop both s and m
371 return (mode == REQUEST) ? m.item : s.item;
373 s.casNext(m, mn); // help unlink
376 } else { // help a fulfiller
377 SNode m = h.next; // m is h's match
378 if (m == null) // waiter is gone
379 casHead(h, null); // pop fulfilling node
382 if (m.tryMatch(h)) // help match
383 casHead(h, mn); // pop both h and m
385 h.casNext(m, mn); // help unlink
392 * Spins/blocks until node s is matched by a fulfill operation.
394 * @param s the waiting node
395 * @param timed true if timed wait
396 * @param nanos timeout value
397 * @return matched node, or s if cancelled
399 SNode awaitFulfill(SNode s, boolean timed, long nanos) {
401 * When a node/thread is about to block, it sets its waiter
402 * field and then rechecks state at least one more time
403 * before actually parking, thus covering race vs
404 * fulfiller noticing that waiter is non-null so should be
407 * When invoked by nodes that appear at the point of call
408 * to be at the head of the stack, calls to park are
409 * preceded by spins to avoid blocking when producers and
410 * consumers are arriving very close in time. This can
411 * happen enough to bother only on multiprocessors.
413 * The order of checks for returning out of main loop
414 * reflects fact that interrupts have precedence over
415 * normal returns, which have precedence over
416 * timeouts. (So, on timeout, one last check for match is
417 * done before giving up.) Except that calls from untimed
418 * SynchronousQueue.{poll/offer} don't check interrupts
419 * and don't wait at all, so are trapped in transfer
420 * method rather than calling awaitFulfill.
422 long lastTime = timed ? System.nanoTime() : 0;
423 Thread w = Thread.currentThread();
425 int spins = (shouldSpin(s) ?
426 (timed ? maxTimedSpins : maxUntimedSpins) : 0);
428 if (w.isInterrupted())
434 long now = System.nanoTime();
435 nanos -= now - lastTime;
443 spins = shouldSpin(s) ? (spins-1) : 0;
444 else if (s.waiter == null)
445 s.waiter = w; // establish waiter so can park next iter
447 LockSupport.park(this);
448 else if (nanos > spinForTimeoutThreshold)
449 LockSupport.parkNanos(this, nanos);
454 * Returns true if node s is at head or there is an active
457 boolean shouldSpin(SNode s) {
459 return (h == s || h == null || isFulfilling(h.mode));
463 * Unlinks s from the stack.
465 void clean(SNode s) {
466 s.item = null; // forget item
467 s.waiter = null; // forget thread
470 * At worst we may need to traverse entire stack to unlink
471 * s. If there are multiple concurrent calls to clean, we
472 * might not see s if another thread has already removed
473 * it. But we can stop when we see any node known to
474 * follow s. We use s.next unless it too is cancelled, in
475 * which case we try the node one past. We don't check any
476 * further because we don't want to doubly traverse just to
481 if (past != null && past.isCancelled())
484 // Absorb cancelled nodes at head
486 while ((p = head) != null && p != past && p.isCancelled())
489 // Unsplice embedded nodes
490 while (p != null && p != past) {
492 if (n != null && n.isCancelled())
493 p.casNext(n, n.next);
501 static final class TransferQueue extends Transferer {
503 * This extends Scherer-Scott dual queue algorithm, differing,
504 * among other ways, by using modes within nodes rather than
505 * marked pointers. The algorithm is a little simpler than
506 * that for stacks because fulfillers do not need explicit
507 * nodes, and matching is done by CAS'ing QNode.item field
508 * from non-null to null (for put) or vice versa (for take).
511 /** Node class for TransferQueue. */
512 static final class QNode {
513 volatile QNode next; // next node in queue
514 volatile Object item; // CAS'ed to or from null
515 volatile Thread waiter; // to control park/unpark
516 final boolean isData;
518 QNode(Object item, boolean isData) {
520 this.isData = isData;
523 boolean casNext(QNode cmp, QNode val) {
531 boolean casItem(Object cmp, Object val) {
540 * Tries to cancel by CAS'ing ref to this as item.
542 void tryCancel(Object cmp) {
548 boolean isCancelled() {
553 * Returns true if this node is known to be off the queue
554 * because its next pointer has been forgotten due to
555 * an advanceHead operation.
557 boolean isOffList() {
563 transient volatile QNode head;
565 transient volatile QNode tail;
567 * Reference to a cancelled node that might not yet have been
568 * unlinked from queue because it was the last inserted node
571 transient volatile QNode cleanMe;
574 QNode h = new QNode(null, false); // initialize to dummy node.
580 * Tries to cas nh as new head; if successful, unlink
581 * old head's next node to avoid garbage retention.
583 void advanceHead(QNode h, QNode nh) {
586 h.next = h; // forget old next
591 * Tries to cas nt as new tail.
593 void advanceTail(QNode t, QNode nt) {
600 * Tries to CAS cleanMe slot.
602 boolean casCleanMe(QNode cmp, QNode val) {
603 if (cleanMe == cmp) {
611 * Puts or takes an item.
613 Object transfer(Object e, boolean timed, long nanos) {
614 /* Basic algorithm is to loop trying to take either of
617 * 1. If queue apparently empty or holding same-mode nodes,
618 * try to add node to queue of waiters, wait to be
619 * fulfilled (or cancelled) and return matching item.
621 * 2. If queue apparently contains waiting items, and this
622 * call is of complementary mode, try to fulfill by CAS'ing
623 * item field of waiting node and dequeuing it, and then
624 * returning matching item.
626 * In each case, along the way, check for and try to help
627 * advance head and tail on behalf of other stalled/slow
630 * The loop starts off with a null check guarding against
631 * seeing uninitialized head or tail values. This never
632 * happens in current SynchronousQueue, but could if
633 * callers held non-volatile/final ref to the
634 * transferer. The check is here anyway because it places
635 * null checks at top of loop, which is usually faster
636 * than having them implicitly interspersed.
639 QNode s = null; // constructed/reused as needed
640 boolean isData = (e != null);
645 if (t == null || h == null) // saw uninitialized value
648 if (h == t || t.isData == isData) { // empty or same-mode
650 if (t != tail) // inconsistent read
652 if (tn != null) { // lagging tail
656 if (timed && nanos <= 0) // can't wait
659 s = new QNode(e, isData);
660 if (!t.casNext(null, s)) // failed to link in
663 advanceTail(t, s); // swing tail and wait
664 Object x = awaitFulfill(s, e, timed, nanos);
665 if (x == s) { // wait was cancelled
670 if (!s.isOffList()) { // not already unlinked
671 advanceHead(t, s); // unlink if head
672 if (x != null) // and forget fields
676 return (x != null) ? x : e;
678 } else { // complementary-mode
679 QNode m = h.next; // node to fulfill
680 if (t != tail || m == null || h != head)
681 continue; // inconsistent read
684 if (isData == (x != null) || // m already fulfilled
685 x == m || // m cancelled
686 !m.casItem(x, e)) { // lost CAS
687 advanceHead(h, m); // dequeue and retry
691 advanceHead(h, m); // successfully fulfilled
692 LockSupport.unpark(m.waiter);
693 return (x != null) ? x : e;
699 * Spins/blocks until node s is fulfilled.
701 * @param s the waiting node
702 * @param e the comparison value for checking match
703 * @param timed true if timed wait
704 * @param nanos timeout value
705 * @return matched item, or s if cancelled
707 Object awaitFulfill(QNode s, Object e, boolean timed, long nanos) {
708 /* Same idea as TransferStack.awaitFulfill */
709 long lastTime = timed ? System.nanoTime() : 0;
710 Thread w = Thread.currentThread();
711 int spins = ((head.next == s) ?
712 (timed ? maxTimedSpins : maxUntimedSpins) : 0);
714 if (w.isInterrupted())
720 long now = System.nanoTime();
721 nanos -= now - lastTime;
730 else if (s.waiter == null)
733 LockSupport.park(this);
734 else if (nanos > spinForTimeoutThreshold)
735 LockSupport.parkNanos(this, nanos);
740 * Gets rid of cancelled node s with original predecessor pred.
742 void clean(QNode pred, QNode s) {
743 s.waiter = null; // forget thread
745 * At any given time, exactly one node on list cannot be
746 * deleted -- the last inserted node. To accommodate this,
747 * if we cannot delete s, we save its predecessor as
748 * "cleanMe", deleting the previously saved version
749 * first. At least one of node s or the node previously
750 * saved can always be deleted, so this always terminates.
752 while (pred.next == s) { // Return early if already unlinked
754 QNode hn = h.next; // Absorb cancelled first node as head
755 if (hn != null && hn.isCancelled()) {
759 QNode t = tail; // Ensure consistent read for tail
769 if (s != t) { // If not tail, try to unsplice
771 if (sn == s || pred.casNext(s, sn))
775 if (dp != null) { // Try unlinking previous cancelled node
778 if (d == null || // d is gone or
779 d == dp || // d is off list or
780 !d.isCancelled() || // d not cancelled or
781 (d != t && // d not tail and
782 (dn = d.next) != null && // has successor
783 dn != d && // that is on list
784 dp.casNext(d, dn))) // d unspliced
785 casCleanMe(dp, null);
787 return; // s is already saved node
788 } else if (casCleanMe(null, pred))
789 return; // Postpone cleaning s
795 * The transferer. Set only in constructor, but cannot be declared
796 * as final without further complicating serialization. Since
797 * this is accessed only at most once per public method, there
798 * isn't a noticeable performance penalty for using volatile
799 * instead of final here.
801 private transient volatile Transferer transferer;
804 * Creates a <tt>SynchronousQueue</tt> with nonfair access policy.
806 public SynchronousQueue() {
811 * Creates a <tt>SynchronousQueue</tt> with the specified fairness policy.
813 * @param fair if true, waiting threads contend in FIFO order for
814 * access; otherwise the order is unspecified.
816 public SynchronousQueue(boolean fair) {
817 transferer = fair ? new TransferQueue() : new TransferStack();
821 * Adds the specified element to this queue, waiting if necessary for
822 * another thread to receive it.
824 * @throws InterruptedException {@inheritDoc}
825 * @throws NullPointerException {@inheritDoc}
827 public void put(E o) throws InterruptedException {
828 if (o == null) throw new NullPointerException();
829 if (transferer.transfer(o, false, 0) == null) {
830 Thread.interrupted();
831 throw new InterruptedException();
836 * Inserts the specified element into this queue, waiting if necessary
837 * up to the specified wait time for another thread to receive it.
839 * @return <tt>true</tt> if successful, or <tt>false</tt> if the
840 * specified waiting time elapses before a consumer appears.
841 * @throws InterruptedException {@inheritDoc}
842 * @throws NullPointerException {@inheritDoc}
844 public boolean offer(E o, long timeout, TimeUnit unit)
845 throws InterruptedException {
846 if (o == null) throw new NullPointerException();
847 if (transferer.transfer(o, true, unit.toNanos(timeout)) != null)
849 if (!Thread.interrupted())
851 throw new InterruptedException();
855 * Inserts the specified element into this queue, if another thread is
856 * waiting to receive it.
858 * @param e the element to add
859 * @return <tt>true</tt> if the element was added to this queue, else
861 * @throws NullPointerException if the specified element is null
863 public boolean offer(E e) {
864 if (e == null) throw new NullPointerException();
865 return transferer.transfer(e, true, 0) != null;
869 * Retrieves and removes the head of this queue, waiting if necessary
870 * for another thread to insert it.
872 * @return the head of this queue
873 * @throws InterruptedException {@inheritDoc}
875 public E take() throws InterruptedException {
876 Object e = transferer.transfer(null, false, 0);
879 Thread.interrupted();
880 throw new InterruptedException();
884 * Retrieves and removes the head of this queue, waiting
885 * if necessary up to the specified wait time, for another thread
888 * @return the head of this queue, or <tt>null</tt> if the
889 * specified waiting time elapses before an element is present.
890 * @throws InterruptedException {@inheritDoc}
892 public E poll(long timeout, TimeUnit unit) throws InterruptedException {
893 Object e = transferer.transfer(null, true, unit.toNanos(timeout));
894 if (e != null || !Thread.interrupted())
896 throw new InterruptedException();
900 * Retrieves and removes the head of this queue, if another thread
901 * is currently making an element available.
903 * @return the head of this queue, or <tt>null</tt> if no
904 * element is available.
907 return (E)transferer.transfer(null, true, 0);
911 * Always returns <tt>true</tt>.
912 * A <tt>SynchronousQueue</tt> has no internal capacity.
914 * @return <tt>true</tt>
916 public boolean isEmpty() {
921 * Always returns zero.
922 * A <tt>SynchronousQueue</tt> has no internal capacity.
931 * Always returns zero.
932 * A <tt>SynchronousQueue</tt> has no internal capacity.
936 public int remainingCapacity() {
942 * A <tt>SynchronousQueue</tt> has no internal capacity.
944 public void clear() {
948 * Always returns <tt>false</tt>.
949 * A <tt>SynchronousQueue</tt> has no internal capacity.
951 * @param o the element
952 * @return <tt>false</tt>
954 public boolean contains(Object o) {
959 * Always returns <tt>false</tt>.
960 * A <tt>SynchronousQueue</tt> has no internal capacity.
962 * @param o the element to remove
963 * @return <tt>false</tt>
965 public boolean remove(Object o) {
970 * Returns <tt>false</tt> unless the given collection is empty.
971 * A <tt>SynchronousQueue</tt> has no internal capacity.
973 * @param c the collection
974 * @return <tt>false</tt> unless given collection is empty
976 public boolean containsAll(Collection<?> c) {
981 * Always returns <tt>false</tt>.
982 * A <tt>SynchronousQueue</tt> has no internal capacity.
984 * @param c the collection
985 * @return <tt>false</tt>
987 public boolean removeAll(Collection<?> c) {
992 * Always returns <tt>false</tt>.
993 * A <tt>SynchronousQueue</tt> has no internal capacity.
995 * @param c the collection
996 * @return <tt>false</tt>
998 public boolean retainAll(Collection<?> c) {
1003 * Always returns <tt>null</tt>.
1004 * A <tt>SynchronousQueue</tt> does not return elements
1005 * unless actively waited on.
1007 * @return <tt>null</tt>
1014 * Returns an empty iterator in which <tt>hasNext</tt> always returns
1017 * @return an empty iterator
1019 public Iterator<E> iterator() {
1020 return Collections.emptyIterator();
1024 * Returns a zero-length array.
1025 * @return a zero-length array
1027 public Object[] toArray() {
1028 return new Object[0];
1032 * Sets the zeroeth element of the specified array to <tt>null</tt>
1033 * (if the array has non-zero length) and returns it.
1035 * @param a the array
1036 * @return the specified array
1037 * @throws NullPointerException if the specified array is null
1039 public <T> T[] toArray(T[] a) {
1046 * @throws UnsupportedOperationException {@inheritDoc}
1047 * @throws ClassCastException {@inheritDoc}
1048 * @throws NullPointerException {@inheritDoc}
1049 * @throws IllegalArgumentException {@inheritDoc}
1051 public int drainTo(Collection<? super E> c) {
1053 throw new NullPointerException();
1055 throw new IllegalArgumentException();
1058 while ( (e = poll()) != null) {
1066 * @throws UnsupportedOperationException {@inheritDoc}
1067 * @throws ClassCastException {@inheritDoc}
1068 * @throws NullPointerException {@inheritDoc}
1069 * @throws IllegalArgumentException {@inheritDoc}
1071 public int drainTo(Collection<? super E> c, int maxElements) {
1073 throw new NullPointerException();
1075 throw new IllegalArgumentException();
1078 while (n < maxElements && (e = poll()) != null) {
1086 * To cope with serialization strategy in the 1.5 version of
1087 * SynchronousQueue, we declare some unused classes and fields
1088 * that exist solely to enable serializability across versions.
1089 * These fields are never used, so are initialized only if this
1090 * object is ever serialized or deserialized.
1093 static class WaitQueue implements java.io.Serializable { }
1094 static class LifoWaitQueue extends WaitQueue {
1095 private static final long serialVersionUID = -3633113410248163686L;
1097 static class FifoWaitQueue extends WaitQueue {
1098 private static final long serialVersionUID = -3623113410248163686L;
1100 private ReentrantLock qlock;
1101 private WaitQueue waitingProducers;
1102 private WaitQueue waitingConsumers;
1105 * Save the state to a stream (that is, serialize it).
1107 * @param s the stream
1109 private void writeObject(java.io.ObjectOutputStream s)
1110 throws java.io.IOException {
1111 boolean fair = transferer instanceof TransferQueue;
1113 qlock = new ReentrantLock(true);
1114 waitingProducers = new FifoWaitQueue();
1115 waitingConsumers = new FifoWaitQueue();
1118 qlock = new ReentrantLock();
1119 waitingProducers = new LifoWaitQueue();
1120 waitingConsumers = new LifoWaitQueue();
1122 s.defaultWriteObject();
1125 private void readObject(final java.io.ObjectInputStream s)
1126 throws java.io.IOException, ClassNotFoundException {
1127 s.defaultReadObject();
1128 if (waitingProducers instanceof FifoWaitQueue)
1129 transferer = new TransferQueue();
1131 transferer = new TransferStack();