| 161 |
|
* targets. Even when using very small slack values, this |
| 162 |
|
* approach works well for dual queues because it allows all |
| 163 |
|
* operations up to the point of matching or appending an item |
| 164 |
< |
* (hence potentially releasing another thread) to be read-only, |
| 165 |
< |
* thus not introducing any further contention. As described |
| 166 |
< |
* below, we implement this by performing slack maintenance |
| 167 |
< |
* retries only after these points. |
| 164 |
> |
* (hence potentially allowing progress by another thread) to be |
| 165 |
> |
* read-only, thus not introducing any further contention. As |
| 166 |
> |
* described below, we implement this by performing slack |
| 167 |
> |
* maintenance retries only after these points. |
| 168 |
|
* |
| 169 |
|
* As an accompaniment to such techniques, traversal overhead can |
| 170 |
|
* be further reduced without increasing contention of head |
| 171 |
< |
* pointer updates. During traversals, threads may sometimes |
| 172 |
< |
* shortcut the "next" link path from the current "head" node to |
| 173 |
< |
* be closer to the currently known first unmatched node. Again, |
| 174 |
< |
* this may be triggered with using thresholds or randomization. |
| 171 |
> |
* pointer updates: Threads may sometimes shortcut the "next" link |
| 172 |
> |
* path from the current "head" node to be closer to the currently |
| 173 |
> |
* known first unmatched node, and similarly for tail. Again, this |
| 174 |
> |
* may be triggered with using thresholds or randomization. |
| 175 |
|
* |
| 176 |
|
* These ideas must be further extended to avoid unbounded amounts |
| 177 |
|
* of costly-to-reclaim garbage caused by the sequential "next" |
| 199 |
|
* mechanics because an update may leave head at a detached node. |
| 200 |
|
* And while direct writes are possible for tail updates, they |
| 201 |
|
* increase the risk of long retraversals, and hence long garbage |
| 202 |
< |
* chains which can be much more costly than is worthwhile |
| 202 |
> |
* chains, which can be much more costly than is worthwhile |
| 203 |
|
* considering that the cost difference of performing a CAS vs |
| 204 |
|
* write is smaller when they are not triggered on each operation |
| 205 |
|
* (especially considering that writes and CASes equally require |
| 206 |
|
* additional GC bookkeeping ("write barriers") that are sometimes |
| 207 |
|
* more costly than the writes themselves because of contention). |
| 208 |
|
* |
| 209 |
– |
* Removal of interior nodes (due to timed out or interrupted |
| 210 |
– |
* waits, or calls to remove or Iterator.remove) uses a scheme |
| 211 |
– |
* roughly similar to that in Scherer, Lea, and Scott's |
| 212 |
– |
* SynchronousQueue. Given a predecessor, we can unsplice any node |
| 213 |
– |
* except the (actual) tail of the queue. To avoid build-up of |
| 214 |
– |
* cancelled trailing nodes, upon a request to remove a trailing |
| 215 |
– |
* node, it is placed in field "cleanMe" to be unspliced upon the |
| 216 |
– |
* next call to unsplice any other node. Situations needing such |
| 217 |
– |
* mechanics are not common but do occur in practice; for example |
| 218 |
– |
* when an unbounded series of short timed calls to poll |
| 219 |
– |
* repeatedly time out but never otherwise fall off the list |
| 220 |
– |
* because of an untimed call to take at the front of the |
| 221 |
– |
* queue. (Note that maintaining field cleanMe does not otherwise |
| 222 |
– |
* much impact garbage retention even if never cleared by some |
| 223 |
– |
* other call because the held node will eventually either |
| 224 |
– |
* directly or indirectly lead to a self-link once off the list.) |
| 225 |
– |
* |
| 209 |
|
* *** Overview of implementation *** |
| 210 |
|
* |
| 211 |
< |
* We use a threshold-based approach to updates, with a target |
| 212 |
< |
* slack of two. The slack value is hard-wired: a path greater |
| 211 |
> |
* We use a threshold-based approach to updates, with a slack |
| 212 |
> |
* threshold of two -- that is, we update head/tail when the |
| 213 |
> |
* current pointer appears to be two or more steps away from the |
| 214 |
> |
* first/last node. The slack value is hard-wired: a path greater |
| 215 |
|
* than one is naturally implemented by checking equality of |
| 216 |
|
* traversal pointers except when the list has only one element, |
| 217 |
< |
* in which case we keep target slack at one. Avoiding tracking |
| 217 |
> |
* in which case we keep slack threshold at one. Avoiding tracking |
| 218 |
|
* explicit counts across method calls slightly simplifies an |
| 219 |
|
* already-messy implementation. Using randomization would |
| 220 |
|
* probably work better if there were a low-quality dirt-cheap |
| 221 |
|
* per-thread one available, but even ThreadLocalRandom is too |
| 222 |
|
* heavy for these purposes. |
| 223 |
|
* |
| 224 |
< |
* With such a small target slack value, it is rarely worthwhile |
| 225 |
< |
* to augment this with path short-circuiting; i.e., unsplicing |
| 226 |
< |
* nodes between head and the first unmatched node, or similarly |
| 227 |
< |
* for tail, rather than advancing head or tail proper. However, |
| 243 |
< |
* it is used (in awaitMatch) immediately before a waiting thread |
| 244 |
< |
* starts to block, as a final bit of helping at a point when |
| 245 |
< |
* contention with others is extremely unlikely (since if other |
| 246 |
< |
* threads that could release it are operating, then the current |
| 247 |
< |
* thread wouldn't be blocking). |
| 224 |
> |
* With such a small slack threshold value, it is not worthwhile |
| 225 |
> |
* to augment this with path short-circuiting (i.e., unsplicing |
| 226 |
> |
* interior nodes) except in the case of cancellation/removal (see |
| 227 |
> |
* below). |
| 228 |
|
* |
| 229 |
|
* We allow both the head and tail fields to be null before any |
| 230 |
|
* nodes are enqueued; initializing upon first append. This |
| 240 |
|
* of offer, put, poll, take, or transfer (each possibly with |
| 241 |
|
* timeout). The relative complexity of using one monolithic |
| 242 |
|
* method outweighs the code bulk and maintenance problems of |
| 243 |
< |
* using nine separate methods. |
| 243 |
> |
* using separate methods for each case. |
| 244 |
|
* |
| 245 |
|
* Operation consists of up to three phases. The first is |
| 246 |
|
* implemented within method xfer, the second in tryAppend, and |
| 265 |
|
* |
| 266 |
|
* 2. Try to append a new node (method tryAppend) |
| 267 |
|
* |
| 268 |
< |
* Starting at current tail pointer, try to append a new node |
| 269 |
< |
* to the list (or if head was null, establish the first |
| 270 |
< |
* node). Nodes can be appended only if their predecessors are |
| 271 |
< |
* either already matched or are of the same mode. If we detect |
| 272 |
< |
* otherwise, then a new node with opposite mode must have been |
| 273 |
< |
* appended during traversal, so must restart at phase 1. The |
| 274 |
< |
* traversal and update steps are otherwise similar to phase 1: |
| 275 |
< |
* Retrying upon CAS misses and checking for staleness. In |
| 276 |
< |
* particular, if a self-link is encountered, then we can |
| 277 |
< |
* safely jump to a node on the list by continuing the |
| 278 |
< |
* traversal at current head. |
| 268 |
> |
* Starting at current tail pointer, find the actual last node |
| 269 |
> |
* and try to append a new node (or if head was null, establish |
| 270 |
> |
* the first node). Nodes can be appended only if their |
| 271 |
> |
* predecessors are either already matched or are of the same |
| 272 |
> |
* mode. If we detect otherwise, then a new node with opposite |
| 273 |
> |
* mode must have been appended during traversal, so we must |
| 274 |
> |
* restart at phase 1. The traversal and update steps are |
| 275 |
> |
* otherwise similar to phase 1: Retrying upon CAS misses and |
| 276 |
> |
* checking for staleness. In particular, if a self-link is |
| 277 |
> |
* encountered, then we can safely jump to a node on the list |
| 278 |
> |
* by continuing the traversal at current head. |
| 279 |
|
* |
| 280 |
|
* On successful append, if the call was ASYNC, return. |
| 281 |
|
* |
| 282 |
|
* 3. Await match or cancellation (method awaitMatch) |
| 283 |
|
* |
| 284 |
|
* Wait for another thread to match node; instead cancelling if |
| 285 |
< |
* current thread was interrupted or the wait timed out. On |
| 285 |
> |
* the current thread was interrupted or the wait timed out. On |
| 286 |
|
* multiprocessors, we use front-of-queue spinning: If a node |
| 287 |
|
* appears to be the first unmatched node in the queue, it |
| 288 |
|
* spins a bit before blocking. In either case, before blocking |
| 297 |
|
* to decide to occasionally perform a Thread.yield. While |
| 298 |
|
* yield has underdefined specs, we assume that might it help, |
| 299 |
|
* and will not hurt in limiting impact of spinning on busy |
| 300 |
< |
* systems. We also use much smaller (1/4) spins for nodes |
| 301 |
< |
* that are not known to be front but whose predecessors have |
| 302 |
< |
* not blocked -- these "chained" spins avoid artifacts of |
| 300 |
> |
* systems. We also use smaller (1/2) spins for nodes that are |
| 301 |
> |
* not known to be front but whose predecessors have not |
| 302 |
> |
* blocked -- these "chained" spins avoid artifacts of |
| 303 |
|
* front-of-queue rules which otherwise lead to alternating |
| 304 |
|
* nodes spinning vs blocking. Further, front threads that |
| 305 |
|
* represent phase changes (from data to request node or vice |
| 306 |
|
* versa) compared to their predecessors receive additional |
| 307 |
< |
* spins, reflecting the longer code path lengths necessary to |
| 308 |
< |
* release them under contention. |
| 307 |
> |
* chained spins, reflecting longer paths typically required to |
| 308 |
> |
* unblock threads during phase changes. |
| 309 |
> |
* |
| 310 |
> |
* |
| 311 |
> |
* ** Unlinking removed interior nodes ** |
| 312 |
> |
* |
| 313 |
> |
* In addition to minimizing garbage retention via self-linking |
| 314 |
> |
* described above, we also unlink removed interior nodes. These |
| 315 |
> |
* may arise due to timed out or interrupted waits, or calls to |
| 316 |
> |
* remove(x) or Iterator.remove. Normally, given a node that was |
| 317 |
> |
* at one time known to be the predecessor of some node s that is |
| 318 |
> |
* to be removed, we can unsplice s by CASing the next field of |
| 319 |
> |
* its predecessor if it still points to s (otherwise s must |
| 320 |
> |
* already have been removed or is now offlist). But there are two |
| 321 |
> |
* situations in which we cannot guarantee to make node s |
| 322 |
> |
* unreachable in this way: (1) If s is the trailing node of list |
| 323 |
> |
* (i.e., with null next), then it is pinned as the target node |
| 324 |
> |
* for appends, so can only be removed later when other nodes are |
| 325 |
> |
* appended. (2) We cannot necessarily unlink s given a |
| 326 |
> |
* predecessor node that is matched (including the case of being |
| 327 |
> |
* cancelled): the predecessor may already be unspliced, in which |
| 328 |
> |
* case some previous reachable node may still point to s. |
| 329 |
> |
* (For further explanation see Herlihy & Shavit "The Art of |
| 330 |
> |
* Multiprocessor Programming" chapter 9). Although, in both |
| 331 |
> |
* cases, we can rule out the need for further action if either s |
| 332 |
> |
* or its predecessor are (or can be made to be) at, or fall off |
| 333 |
> |
* from, the head of list. |
| 334 |
> |
* |
| 335 |
> |
* Without taking these into account, it would be possible for an |
| 336 |
> |
* unbounded number of supposedly removed nodes to remain |
| 337 |
> |
* reachable. Situations leading to such buildup are uncommon but |
| 338 |
> |
* can occur in practice; for example when a series of short timed |
| 339 |
> |
* calls to poll repeatedly time out but never otherwise fall off |
| 340 |
> |
* the list because of an untimed call to take at the front of the |
| 341 |
> |
* queue. |
| 342 |
> |
* |
| 343 |
> |
* When these cases arise, rather than always retraversing the |
| 344 |
> |
* entire list to find an actual predecessor to unlink (which |
| 345 |
> |
* won't help for case (1) anyway), we record a conservative |
| 346 |
> |
* estimate of possible unsplice failures (in "sweepVotes"). We |
| 347 |
> |
* trigger a full sweep when the estimate exceeds a threshold |
| 348 |
> |
* indicating the maximum number of estimated removal failures to |
| 349 |
> |
* tolerate before sweeping through, unlinking cancelled nodes |
| 350 |
> |
* that were not unlinked upon initial removal. We perform sweeps |
| 351 |
> |
* by the thread hitting threshold (rather than background threads |
| 352 |
> |
* or by spreading work to other threads) because in the main |
| 353 |
> |
* contexts in which removal occurs, the caller is already |
| 354 |
> |
* timed-out, cancelled, or performing a potentially O(n) |
| 355 |
> |
* operation (i.e., remove(x)), none of which are time-critical |
| 356 |
> |
* enough to warrant the overhead that alternatives would impose |
| 357 |
> |
* on other threads. |
| 358 |
> |
* |
| 359 |
> |
* Because the sweepVotes estimate is conservative, and because |
| 360 |
> |
* nodes become unlinked "naturally" as they fall off the head of |
| 361 |
> |
* the queue, and because we allow votes to accumulate even while |
| 362 |
> |
* sweeps are in progress, there are typically significantly fewer |
| 363 |
> |
* such nodes than estimated. Choice of a threshold value |
| 364 |
> |
* balances the likelihood of wasted effort and contention, versus |
| 365 |
> |
* providing a worst-case bound on retention of interior nodes in |
| 366 |
> |
* quiescent queues. The value defined below was chosen |
| 367 |
> |
* empirically to balance these under various timeout scenarios. |
| 368 |
> |
* |
| 369 |
> |
* Note that we cannot self-link unlinked interior nodes during |
| 370 |
> |
* sweeps. However, the associated garbage chains terminate when |
| 371 |
> |
* some successor ultimately falls off the head of the list and is |
| 372 |
> |
* self-linked. |
| 373 |
|
*/ |
| 374 |
|
|
| 375 |
|
/** True if on multiprocessor */ |
| 377 |
|
Runtime.getRuntime().availableProcessors() > 1; |
| 378 |
|
|
| 379 |
|
/** |
| 380 |
< |
* The number of times to spin (with on average one randomly |
| 381 |
< |
* interspersed call to Thread.yield) on multiprocessor before |
| 382 |
< |
* blocking when a node is apparently the first waiter in the |
| 383 |
< |
* queue. See above for explanation. Must be a power of two. The |
| 384 |
< |
* value is empirically derived -- it works pretty well across a |
| 385 |
< |
* variety of processors, numbers of CPUs, and OSes. |
| 380 |
> |
* The number of times to spin (with randomly interspersed calls |
| 381 |
> |
* to Thread.yield) on multiprocessor before blocking when a node |
| 382 |
> |
* is apparently the first waiter in the queue. See above for |
| 383 |
> |
* explanation. Must be a power of two. The value is empirically |
| 384 |
> |
* derived -- it works pretty well across a variety of processors, |
| 385 |
> |
* numbers of CPUs, and OSes. |
| 386 |
|
*/ |
| 387 |
|
private static final int FRONT_SPINS = 1 << 7; |
| 388 |
|
|
| 389 |
|
/** |
| 390 |
|
* The number of times to spin before blocking when a node is |
| 391 |
< |
* preceded by another node that is apparently spinning. |
| 391 |
> |
* preceded by another node that is apparently spinning. Also |
| 392 |
> |
* serves as an increment to FRONT_SPINS on phase changes, and as |
| 393 |
> |
* base average frequency for yielding during spins. Must be a |
| 394 |
> |
* power of two. |
| 395 |
|
*/ |
| 396 |
< |
private static final int CHAINED_SPINS = FRONT_SPINS >>> 2; |
| 396 |
> |
private static final int CHAINED_SPINS = FRONT_SPINS >>> 1; |
| 397 |
> |
|
| 398 |
> |
/** |
| 399 |
> |
* The maximum number of estimated removal failures (sweepVotes) |
| 400 |
> |
* to tolerate before sweeping through the queue unlinking |
| 401 |
> |
* cancelled nodes that were not unlinked upon initial |
| 402 |
> |
* removal. See above for explanation. The value must be at least |
| 403 |
> |
* two to avoid useless sweeps when removing trailing nodes. |
| 404 |
> |
*/ |
| 405 |
> |
static final int SWEEP_THRESHOLD = 32; |
| 406 |
|
|
| 407 |
|
/** |
| 408 |
|
* Queue nodes. Uses Object, not E, for items to allow forgetting |
| 409 |
|
* them after use. Relies heavily on Unsafe mechanics to minimize |
| 410 |
< |
* unnecessary ordering constraints: Writes that intrinsically |
| 411 |
< |
* precede or follow CASes use simple relaxed forms. Other |
| 356 |
< |
* cleanups use releasing/lazy writes. |
| 410 |
> |
* unnecessary ordering constraints: Writes that are intrinsically |
| 411 |
> |
* ordered wrt other accesses or CASes use simple relaxed forms. |
| 412 |
|
*/ |
| 413 |
|
static final class Node { |
| 414 |
|
final boolean isData; // false if this is a request node |
| 422 |
|
} |
| 423 |
|
|
| 424 |
|
final boolean casItem(Object cmp, Object val) { |
| 425 |
+ |
assert cmp == null || cmp.getClass() != Node.class; |
| 426 |
|
return UNSAFE.compareAndSwapObject(this, itemOffset, cmp, val); |
| 427 |
|
} |
| 428 |
|
|
| 444 |
|
} |
| 445 |
|
|
| 446 |
|
/** |
| 447 |
< |
* Sets item to self (using a releasing/lazy write) and waiter |
| 448 |
< |
* to null, to avoid garbage retention after extracting or |
| 449 |
< |
* cancelling. |
| 447 |
> |
* Sets item to self and waiter to null, to avoid garbage |
| 448 |
> |
* retention after matching or cancelling. Uses relaxed writes |
| 449 |
> |
* bacause order is already constrained in the only calling |
| 450 |
> |
* contexts: item is forgotten only after volatile/atomic |
| 451 |
> |
* mechanics that extract items. Similarly, clearing waiter |
| 452 |
> |
* follows either CAS or return from park (if ever parked; |
| 453 |
> |
* else we don't care). |
| 454 |
|
*/ |
| 455 |
|
final void forgetContents() { |
| 456 |
< |
UNSAFE.putOrderedObject(this, itemOffset, this); |
| 457 |
< |
UNSAFE.putOrderedObject(this, waiterOffset, null); |
| 456 |
> |
UNSAFE.putObject(this, itemOffset, this); |
| 457 |
> |
UNSAFE.putObject(this, waiterOffset, null); |
| 458 |
|
} |
| 459 |
|
|
| 460 |
|
/** |
| 463 |
|
*/ |
| 464 |
|
final boolean isMatched() { |
| 465 |
|
Object x = item; |
| 466 |
< |
return x == this || (x != null) != isData; |
| 466 |
> |
return (x == this) || ((x == null) == isData); |
| 467 |
> |
} |
| 468 |
> |
|
| 469 |
> |
/** |
| 470 |
> |
* Returns true if this is an unmatched request node. |
| 471 |
> |
*/ |
| 472 |
> |
final boolean isUnmatchedRequest() { |
| 473 |
> |
return !isData && item == null; |
| 474 |
|
} |
| 475 |
|
|
| 476 |
|
/** |
| 488 |
|
* Tries to artificially match a data node -- used by remove. |
| 489 |
|
*/ |
| 490 |
|
final boolean tryMatchData() { |
| 491 |
+ |
assert isData; |
| 492 |
|
Object x = item; |
| 493 |
|
if (x != null && x != this && casItem(x, null)) { |
| 494 |
|
LockSupport.unpark(waiter); |
| 510 |
|
} |
| 511 |
|
|
| 512 |
|
/** head of the queue; null until first enqueue */ |
| 513 |
< |
private transient volatile Node head; |
| 446 |
< |
|
| 447 |
< |
/** predecessor of dangling unspliceable node */ |
| 448 |
< |
private transient volatile Node cleanMe; // decl here to reduce contention |
| 513 |
> |
transient volatile Node head; |
| 514 |
|
|
| 515 |
|
/** tail of the queue; null until first append */ |
| 516 |
|
private transient volatile Node tail; |
| 517 |
|
|
| 518 |
+ |
/** The number of apparent failures to unsplice removed nodes */ |
| 519 |
+ |
private transient volatile int sweepVotes; |
| 520 |
+ |
|
| 521 |
|
// CAS methods for fields |
| 522 |
|
private boolean casTail(Node cmp, Node val) { |
| 523 |
|
return UNSAFE.compareAndSwapObject(this, tailOffset, cmp, val); |
| 527 |
|
return UNSAFE.compareAndSwapObject(this, headOffset, cmp, val); |
| 528 |
|
} |
| 529 |
|
|
| 530 |
< |
private boolean casCleanMe(Node cmp, Node val) { |
| 531 |
< |
return UNSAFE.compareAndSwapObject(this, cleanMeOffset, cmp, val); |
| 530 |
> |
private boolean casSweepVotes(int cmp, int val) { |
| 531 |
> |
return UNSAFE.compareAndSwapInt(this, sweepVotesOffset, cmp, val); |
| 532 |
|
} |
| 533 |
|
|
| 534 |
|
/* |
| 535 |
< |
* Possible values for "how" argument in xfer method. Beware that |
| 468 |
< |
* the order of assigned numerical values matters. |
| 535 |
> |
* Possible values for "how" argument in xfer method. |
| 536 |
|
*/ |
| 537 |
< |
private static final int NOW = 0; // for untimed poll, tryTransfer |
| 538 |
< |
private static final int ASYNC = 1; // for offer, put, add |
| 539 |
< |
private static final int SYNC = 2; // for transfer, take |
| 540 |
< |
private static final int TIMEOUT = 3; // for timed poll, tryTransfer |
| 537 |
> |
private static final int NOW = 0; // for untimed poll, tryTransfer |
| 538 |
> |
private static final int ASYNC = 1; // for offer, put, add |
| 539 |
> |
private static final int SYNC = 2; // for transfer, take |
| 540 |
> |
private static final int TIMED = 3; // for timed poll, tryTransfer |
| 541 |
> |
|
| 542 |
> |
@SuppressWarnings("unchecked") |
| 543 |
> |
static <E> E cast(Object item) { |
| 544 |
> |
assert item == null || item.getClass() != Node.class; |
| 545 |
> |
return (E) item; |
| 546 |
> |
} |
| 547 |
|
|
| 548 |
|
/** |
| 549 |
|
* Implements all queuing methods. See above for explanation. |
| 550 |
|
* |
| 551 |
|
* @param e the item or null for take |
| 552 |
|
* @param haveData true if this is a put, else a take |
| 553 |
< |
* @param how NOW, ASYNC, SYNC, or TIMEOUT |
| 554 |
< |
* @param nanos timeout in nanosecs, used only if mode is TIMEOUT |
| 553 |
> |
* @param how NOW, ASYNC, SYNC, or TIMED |
| 554 |
> |
* @param nanos timeout in nanosecs, used only if mode is TIMED |
| 555 |
|
* @return an item if matched, else e |
| 556 |
|
* @throws NullPointerException if haveData mode but e is null |
| 557 |
|
*/ |
| 558 |
< |
private Object xfer(Object e, boolean haveData, int how, long nanos) { |
| 558 |
> |
private E xfer(E e, boolean haveData, int how, long nanos) { |
| 559 |
|
if (haveData && (e == null)) |
| 560 |
|
throw new NullPointerException(); |
| 561 |
|
Node s = null; // the node to append, if needed |
| 569 |
|
if (isData == haveData) // can't match |
| 570 |
|
break; |
| 571 |
|
if (p.casItem(item, e)) { // match |
| 572 |
< |
Thread w = p.waiter; |
| 573 |
< |
while (p != h) { // update head |
| 574 |
< |
Node n = p.next; // by 2 unless singleton |
| 502 |
< |
if (n != null) |
| 503 |
< |
p = n; |
| 504 |
< |
if (head == h && casHead(h, p)) { |
| 572 |
> |
for (Node q = p; q != h;) { |
| 573 |
> |
Node n = q.next; // update by 2 unless singleton |
| 574 |
> |
if (head == h && casHead(h, n == null? q : n)) { |
| 575 |
|
h.forgetNext(); |
| 576 |
|
break; |
| 577 |
|
} // advance and retry |
| 578 |
|
if ((h = head) == null || |
| 579 |
< |
(p = h.next) == null || !p.isMatched()) |
| 579 |
> |
(q = h.next) == null || !q.isMatched()) |
| 580 |
|
break; // unless slack < 2 |
| 581 |
|
} |
| 582 |
< |
LockSupport.unpark(w); |
| 583 |
< |
return item; |
| 582 |
> |
LockSupport.unpark(p.waiter); |
| 583 |
> |
return this.<E>cast(item); |
| 584 |
|
} |
| 585 |
|
} |
| 586 |
|
Node n = p.next; |
| 587 |
|
p = (p != n) ? n : (h = head); // Use head if p offlist |
| 588 |
|
} |
| 589 |
|
|
| 590 |
< |
if (how >= ASYNC) { // No matches available |
| 590 |
> |
if (how != NOW) { // No matches available |
| 591 |
|
if (s == null) |
| 592 |
|
s = new Node(e, haveData); |
| 593 |
|
Node pred = tryAppend(s, haveData); |
| 594 |
|
if (pred == null) |
| 595 |
|
continue retry; // lost race vs opposite mode |
| 596 |
< |
if (how >= SYNC) |
| 597 |
< |
return awaitMatch(pred, s, e, how, nanos); |
| 596 |
> |
if (how != ASYNC) |
| 597 |
> |
return awaitMatch(s, pred, e, (how == TIMED), nanos); |
| 598 |
|
} |
| 599 |
|
return e; // not waiting |
| 600 |
|
} |
| 610 |
|
* predecessor |
| 611 |
|
*/ |
| 612 |
|
private Node tryAppend(Node s, boolean haveData) { |
| 613 |
< |
for (Node t = tail, p = t;;) { // move p to last node and append |
| 613 |
> |
for (Node t = tail, p = t;;) { // move p to last node and append |
| 614 |
|
Node n, u; // temps for reads of next & tail |
| 615 |
|
if (p == null && (p = head) == null) { |
| 616 |
|
if (casHead(null, s)) |
| 638 |
|
/** |
| 639 |
|
* Spins/yields/blocks until node s is matched or caller gives up. |
| 640 |
|
* |
| 571 |
– |
* @param pred the predecessor of s, or s or null if none |
| 641 |
|
* @param s the waiting node |
| 642 |
+ |
* @param pred the predecessor of s, or s itself if it has no |
| 643 |
+ |
* predecessor, or null if unknown (the null case does not occur |
| 644 |
+ |
* in any current calls but may in possible future extensions) |
| 645 |
|
* @param e the comparison value for checking match |
| 646 |
< |
* @param how either SYNC or TIMEOUT |
| 647 |
< |
* @param nanos timeout value |
| 646 |
> |
* @param timed if true, wait only until timeout elapses |
| 647 |
> |
* @param nanos timeout in nanosecs, used only if timed is true |
| 648 |
|
* @return matched item, or e if unmatched on interrupt or timeout |
| 649 |
|
*/ |
| 650 |
< |
private Object awaitMatch(Node pred, Node s, Object e, |
| 651 |
< |
int how, long nanos) { |
| 580 |
< |
long lastTime = (how == TIMEOUT) ? System.nanoTime() : 0L; |
| 650 |
> |
private E awaitMatch(Node s, Node pred, E e, boolean timed, long nanos) { |
| 651 |
> |
long lastTime = timed ? System.nanoTime() : 0L; |
| 652 |
|
Thread w = Thread.currentThread(); |
| 653 |
|
int spins = -1; // initialized after first item and cancel checks |
| 654 |
|
ThreadLocalRandom randomYields = null; // bound if needed |
| 656 |
|
for (;;) { |
| 657 |
|
Object item = s.item; |
| 658 |
|
if (item != e) { // matched |
| 659 |
+ |
assert item != s; |
| 660 |
|
s.forgetContents(); // avoid garbage |
| 661 |
< |
return item; |
| 661 |
> |
return this.<E>cast(item); |
| 662 |
|
} |
| 663 |
< |
if ((w.isInterrupted() || (how == TIMEOUT && nanos <= 0)) && |
| 664 |
< |
s.casItem(e, s)) { // cancel |
| 663 |
> |
if ((w.isInterrupted() || (timed && nanos <= 0)) && |
| 664 |
> |
s.casItem(e, s)) { // cancel |
| 665 |
|
unsplice(pred, s); |
| 666 |
|
return e; |
| 667 |
|
} |
| 670 |
|
if ((spins = spinsFor(pred, s.isData)) > 0) |
| 671 |
|
randomYields = ThreadLocalRandom.current(); |
| 672 |
|
} |
| 673 |
< |
else if (spins > 0) { // spin, occasionally yield |
| 602 |
< |
if (randomYields.nextInt(FRONT_SPINS) == 0) |
| 603 |
< |
Thread.yield(); |
| 673 |
> |
else if (spins > 0) { // spin |
| 674 |
|
--spins; |
| 675 |
+ |
if (randomYields.nextInt(CHAINED_SPINS) == 0) |
| 676 |
+ |
Thread.yield(); // occasionally yield |
| 677 |
|
} |
| 678 |
|
else if (s.waiter == null) { |
| 679 |
< |
shortenHeadPath(); // reduce slack before blocking |
| 608 |
< |
s.waiter = w; // request unpark |
| 679 |
> |
s.waiter = w; // request unpark then recheck |
| 680 |
|
} |
| 681 |
< |
else if (how == TIMEOUT) { |
| 681 |
> |
else if (timed) { |
| 682 |
|
long now = System.nanoTime(); |
| 683 |
|
if ((nanos -= now - lastTime) > 0) |
| 684 |
|
LockSupport.parkNanos(this, nanos); |
| 686 |
|
} |
| 687 |
|
else { |
| 688 |
|
LockSupport.park(this); |
| 618 |
– |
spins = -1; // spin if front upon wakeup |
| 689 |
|
} |
| 690 |
|
} |
| 691 |
|
} |
| 696 |
|
*/ |
| 697 |
|
private static int spinsFor(Node pred, boolean haveData) { |
| 698 |
|
if (MP && pred != null) { |
| 699 |
< |
boolean predData = pred.isData; |
| 700 |
< |
if (predData != haveData) // front and phase change |
| 701 |
< |
return FRONT_SPINS + (FRONT_SPINS >>> 1); |
| 632 |
< |
if (predData != (pred.item != null)) // probably at front |
| 699 |
> |
if (pred.isData != haveData) // phase change |
| 700 |
> |
return FRONT_SPINS + CHAINED_SPINS; |
| 701 |
> |
if (pred.isMatched()) // probably at front |
| 702 |
|
return FRONT_SPINS; |
| 703 |
|
if (pred.waiter == null) // pred apparently spinning |
| 704 |
|
return CHAINED_SPINS; |
| 706 |
|
return 0; |
| 707 |
|
} |
| 708 |
|
|
| 709 |
+ |
/* -------------- Traversal methods -------------- */ |
| 710 |
+ |
|
| 711 |
|
/** |
| 712 |
< |
* Tries (once) to unsplice nodes between head and first unmatched |
| 713 |
< |
* or trailing node; failing on contention. |
| 714 |
< |
*/ |
| 715 |
< |
private void shortenHeadPath() { |
| 716 |
< |
Node h, hn, p, q; |
| 717 |
< |
if ((p = h = head) != null && h.isMatched() && |
| 718 |
< |
(q = hn = h.next) != null) { |
| 648 |
< |
Node n; |
| 649 |
< |
while ((n = q.next) != q) { |
| 650 |
< |
if (n == null || !q.isMatched()) { |
| 651 |
< |
if (hn != q && h.next == hn) |
| 652 |
< |
h.casNext(hn, q); |
| 653 |
< |
break; |
| 654 |
< |
} |
| 655 |
< |
p = q; |
| 656 |
< |
q = n; |
| 657 |
< |
} |
| 658 |
< |
} |
| 712 |
> |
* Returns the successor of p, or the head node if p.next has been |
| 713 |
> |
* linked to self, which will only be true if traversing with a |
| 714 |
> |
* stale pointer that is now off the list. |
| 715 |
> |
*/ |
| 716 |
> |
final Node succ(Node p) { |
| 717 |
> |
Node next = p.next; |
| 718 |
> |
return (p == next) ? head : next; |
| 719 |
|
} |
| 720 |
|
|
| 661 |
– |
/* -------------- Traversal methods -------------- */ |
| 662 |
– |
|
| 721 |
|
/** |
| 722 |
|
* Returns the first unmatched node of the given mode, or null if |
| 723 |
|
* none. Used by methods isEmpty, hasWaitingConsumer. |
| 724 |
|
*/ |
| 725 |
< |
private Node firstOfMode(boolean data) { |
| 726 |
< |
for (Node p = head; p != null; ) { |
| 725 |
> |
private Node firstOfMode(boolean isData) { |
| 726 |
> |
for (Node p = head; p != null; p = succ(p)) { |
| 727 |
|
if (!p.isMatched()) |
| 728 |
< |
return (p.isData == data) ? p : null; |
| 671 |
< |
Node n = p.next; |
| 672 |
< |
p = (n != p) ? n : head; |
| 728 |
> |
return (p.isData == isData) ? p : null; |
| 729 |
|
} |
| 730 |
|
return null; |
| 731 |
|
} |
| 732 |
|
|
| 733 |
|
/** |
| 734 |
|
* Returns the item in the first unmatched node with isData; or |
| 735 |
< |
* null if none. Used by peek. |
| 735 |
> |
* null if none. Used by peek. |
| 736 |
|
*/ |
| 737 |
< |
private Object firstDataItem() { |
| 738 |
< |
for (Node p = head; p != null; ) { |
| 683 |
< |
boolean isData = p.isData; |
| 737 |
> |
private E firstDataItem() { |
| 738 |
> |
for (Node p = head; p != null; p = succ(p)) { |
| 739 |
|
Object item = p.item; |
| 740 |
< |
if (item != p && (item != null) == isData) |
| 741 |
< |
return isData ? item : null; |
| 742 |
< |
Node n = p.next; |
| 743 |
< |
p = (n != p) ? n : head; |
| 740 |
> |
if (p.isData) { |
| 741 |
> |
if (item != null && item != p) |
| 742 |
> |
return this.<E>cast(item); |
| 743 |
> |
} |
| 744 |
> |
else if (item == null) |
| 745 |
> |
return null; |
| 746 |
|
} |
| 747 |
|
return null; |
| 748 |
|
} |
| 773 |
|
|
| 774 |
|
final class Itr implements Iterator<E> { |
| 775 |
|
private Node nextNode; // next node to return item for |
| 776 |
< |
private Object nextItem; // the corresponding item |
| 776 |
> |
private E nextItem; // the corresponding item |
| 777 |
|
private Node lastRet; // last returned node, to support remove |
| 778 |
+ |
private Node lastPred; // predecessor to unlink lastRet |
| 779 |
|
|
| 780 |
|
/** |
| 781 |
|
* Moves to next node after prev, or first node if prev null. |
| 782 |
|
*/ |
| 783 |
|
private void advance(Node prev) { |
| 784 |
+ |
lastPred = lastRet; |
| 785 |
|
lastRet = prev; |
| 786 |
< |
Node p; |
| 787 |
< |
if (prev == null || (p = prev.next) == prev) |
| 729 |
< |
p = head; |
| 730 |
< |
while (p != null) { |
| 786 |
> |
for (Node p = (prev == null) ? head : succ(prev); |
| 787 |
> |
p != null; p = succ(p)) { |
| 788 |
|
Object item = p.item; |
| 789 |
|
if (p.isData) { |
| 790 |
|
if (item != null && item != p) { |
| 791 |
< |
nextItem = item; |
| 791 |
> |
nextItem = LinkedTransferQueue.this.<E>cast(item); |
| 792 |
|
nextNode = p; |
| 793 |
|
return; |
| 794 |
|
} |
| 795 |
|
} |
| 796 |
|
else if (item == null) |
| 797 |
|
break; |
| 741 |
– |
Node n = p.next; |
| 742 |
– |
p = (n != p) ? n : head; |
| 798 |
|
} |
| 799 |
|
nextNode = null; |
| 800 |
|
} |
| 810 |
|
public final E next() { |
| 811 |
|
Node p = nextNode; |
| 812 |
|
if (p == null) throw new NoSuchElementException(); |
| 813 |
< |
Object e = nextItem; |
| 813 |
> |
E e = nextItem; |
| 814 |
|
advance(p); |
| 815 |
< |
return (E) e; |
| 815 |
> |
return e; |
| 816 |
|
} |
| 817 |
|
|
| 818 |
|
public final void remove() { |
| 819 |
|
Node p = lastRet; |
| 820 |
|
if (p == null) throw new IllegalStateException(); |
| 821 |
< |
lastRet = null; |
| 822 |
< |
findAndRemoveNode(p); |
| 821 |
> |
if (p.tryMatchData()) |
| 822 |
> |
unsplice(lastPred, p); |
| 823 |
|
} |
| 824 |
|
} |
| 825 |
|
|
| 829 |
|
* Unsplices (now or later) the given deleted/cancelled node with |
| 830 |
|
* the given predecessor. |
| 831 |
|
* |
| 832 |
< |
* @param pred predecessor of node to be unspliced |
| 832 |
> |
* @param pred a node that was at one time known to be the |
| 833 |
> |
* predecessor of s, or null or s itself if s is/was at head |
| 834 |
|
* @param s the node to be unspliced |
| 835 |
|
*/ |
| 836 |
< |
private void unsplice(Node pred, Node s) { |
| 837 |
< |
s.forgetContents(); // clear unneeded fields |
| 836 |
> |
final void unsplice(Node pred, Node s) { |
| 837 |
> |
s.forgetContents(); // forget unneeded fields |
| 838 |
|
/* |
| 839 |
< |
* At any given time, exactly one node on list cannot be |
| 840 |
< |
* unlinked -- the last inserted node. To accommodate this, if |
| 841 |
< |
* we cannot unlink s, we save its predecessor as "cleanMe", |
| 842 |
< |
* processing the previously saved version first. Because only |
| 843 |
< |
* one node in the list can have a null next, at least one of |
| 788 |
< |
* node s or the node previously saved can always be |
| 789 |
< |
* processed, so this always terminates. |
| 839 |
> |
* See above for rationale. Briefly: if pred still points to |
| 840 |
> |
* s, try to unlink s. If s cannot be unlinked, because it is |
| 841 |
> |
* trailing node or pred might be unlinked, and neither pred |
| 842 |
> |
* nor s are head or offlist, add to sweepVotes, and if enough |
| 843 |
> |
* votes have accumulated, sweep. |
| 844 |
|
*/ |
| 845 |
< |
if (pred != null && pred != s) { |
| 846 |
< |
while (pred.next == s) { |
| 847 |
< |
Node oldpred = (cleanMe == null) ? null : reclean(); |
| 848 |
< |
Node n = s.next; |
| 849 |
< |
if (n != null) { |
| 850 |
< |
if (n != s) |
| 851 |
< |
pred.casNext(s, n); |
| 852 |
< |
break; |
| 845 |
> |
if (pred != null && pred != s && pred.next == s) { |
| 846 |
> |
Node n = s.next; |
| 847 |
> |
if (n == null || |
| 848 |
> |
(n != s && pred.casNext(s, n) && pred.isMatched())) { |
| 849 |
> |
for (;;) { // check if at, or could be, head |
| 850 |
> |
Node h = head; |
| 851 |
> |
if (h == pred || h == s || h == null) |
| 852 |
> |
return; // at head or list empty |
| 853 |
> |
if (!h.isMatched()) |
| 854 |
> |
break; |
| 855 |
> |
Node hn = h.next; |
| 856 |
> |
if (hn == null) |
| 857 |
> |
return; // now empty |
| 858 |
> |
if (hn != h && casHead(h, hn)) |
| 859 |
> |
h.forgetNext(); // advance head |
| 860 |
> |
} |
| 861 |
> |
if (pred.next != pred && s.next != s) { // recheck if offlist |
| 862 |
> |
for (;;) { // sweep now if enough votes |
| 863 |
> |
int v = sweepVotes; |
| 864 |
> |
if (v < SWEEP_THRESHOLD) { |
| 865 |
> |
if (casSweepVotes(v, v + 1)) |
| 866 |
> |
break; |
| 867 |
> |
} |
| 868 |
> |
else if (casSweepVotes(v, 0)) { |
| 869 |
> |
sweep(); |
| 870 |
> |
break; |
| 871 |
> |
} |
| 872 |
> |
} |
| 873 |
|
} |
| 800 |
– |
if (oldpred == pred || // Already saved |
| 801 |
– |
(oldpred == null && casCleanMe(null, pred))) |
| 802 |
– |
break; // Postpone cleaning |
| 874 |
|
} |
| 875 |
|
} |
| 876 |
|
} |
| 877 |
|
|
| 878 |
|
/** |
| 879 |
< |
* Tries to unsplice the deleted/cancelled node held in cleanMe |
| 809 |
< |
* that was previously uncleanable because it was at tail. |
| 810 |
< |
* |
| 811 |
< |
* @return current cleanMe node (or null) |
| 879 |
> |
* Unlinks matched nodes encountered in a traversal from head. |
| 880 |
|
*/ |
| 881 |
< |
private Node reclean() { |
| 882 |
< |
/* |
| 883 |
< |
* cleanMe is, or at one time was, predecessor of a cancelled |
| 884 |
< |
* node s that was the tail so could not be unspliced. If it |
| 885 |
< |
* is no longer the tail, try to unsplice if necessary and |
| 886 |
< |
* make cleanMe slot available. This differs from similar |
| 887 |
< |
* code in unsplice() because we must check that pred still |
| 820 |
< |
* points to a matched node that can be unspliced -- if not, |
| 821 |
< |
* we can (must) clear cleanMe without unsplicing. This can |
| 822 |
< |
* loop only due to contention. |
| 823 |
< |
*/ |
| 824 |
< |
Node pred; |
| 825 |
< |
while ((pred = cleanMe) != null) { |
| 826 |
< |
Node s = pred.next; |
| 827 |
< |
Node n; |
| 828 |
< |
if (s == null || s == pred || !s.isMatched()) |
| 829 |
< |
casCleanMe(pred, null); // already gone |
| 830 |
< |
else if ((n = s.next) != null) { |
| 831 |
< |
if (n != s) |
| 832 |
< |
pred.casNext(s, n); |
| 833 |
< |
casCleanMe(pred, null); |
| 834 |
< |
} |
| 835 |
< |
else |
| 881 |
> |
private void sweep() { |
| 882 |
> |
for (Node p = head, s, n; p != null && (s = p.next) != null; ) { |
| 883 |
> |
if (p == s) // stale |
| 884 |
> |
p = head; |
| 885 |
> |
else if (!s.isMatched()) |
| 886 |
> |
p = s; |
| 887 |
> |
else if ((n = s.next) == null) // trailing node is pinned |
| 888 |
|
break; |
| 889 |
< |
} |
| 890 |
< |
return pred; |
| 839 |
< |
} |
| 840 |
< |
|
| 841 |
< |
/** |
| 842 |
< |
* Main implementation of Iterator.remove(). Find |
| 843 |
< |
* and unsplice the given node. |
| 844 |
< |
*/ |
| 845 |
< |
final void findAndRemoveNode(Node s) { |
| 846 |
< |
if (s.tryMatchData()) { |
| 847 |
< |
Node pred = null; |
| 848 |
< |
Node p = head; |
| 849 |
< |
while (p != null) { |
| 850 |
< |
if (p == s) { |
| 851 |
< |
unsplice(pred, p); |
| 852 |
< |
break; |
| 853 |
< |
} |
| 854 |
< |
if (!p.isData && !p.isMatched()) |
| 855 |
< |
break; |
| 856 |
< |
pred = p; |
| 857 |
< |
if ((p = p.next) == pred) { // stale |
| 858 |
< |
pred = null; |
| 859 |
< |
p = head; |
| 860 |
< |
} |
| 861 |
< |
} |
| 889 |
> |
else |
| 890 |
> |
p.casNext(s, n); |
| 891 |
|
} |
| 892 |
|
} |
| 893 |
|
|
| 896 |
|
*/ |
| 897 |
|
private boolean findAndRemove(Object e) { |
| 898 |
|
if (e != null) { |
| 899 |
< |
Node pred = null; |
| 871 |
< |
Node p = head; |
| 872 |
< |
while (p != null) { |
| 899 |
> |
for (Node pred = null, p = head; p != null; ) { |
| 900 |
|
Object item = p.item; |
| 901 |
|
if (p.isData) { |
| 902 |
|
if (item != null && item != p && e.equals(item) && |
| 908 |
|
else if (item == null) |
| 909 |
|
break; |
| 910 |
|
pred = p; |
| 911 |
< |
if ((p = p.next) == pred) { |
| 911 |
> |
if ((p = p.next) == pred) { // stale |
| 912 |
|
pred = null; |
| 913 |
|
p = head; |
| 914 |
|
} |
| 1036 |
|
*/ |
| 1037 |
|
public boolean tryTransfer(E e, long timeout, TimeUnit unit) |
| 1038 |
|
throws InterruptedException { |
| 1039 |
< |
if (xfer(e, true, TIMEOUT, unit.toNanos(timeout)) == null) |
| 1039 |
> |
if (xfer(e, true, TIMED, unit.toNanos(timeout)) == null) |
| 1040 |
|
return true; |
| 1041 |
|
if (!Thread.interrupted()) |
| 1042 |
|
return false; |
| 1044 |
|
} |
| 1045 |
|
|
| 1046 |
|
public E take() throws InterruptedException { |
| 1047 |
< |
Object e = xfer(null, false, SYNC, 0); |
| 1047 |
> |
E e = xfer(null, false, SYNC, 0); |
| 1048 |
|
if (e != null) |
| 1049 |
< |
return (E)e; |
| 1049 |
> |
return e; |
| 1050 |
|
Thread.interrupted(); |
| 1051 |
|
throw new InterruptedException(); |
| 1052 |
|
} |
| 1053 |
|
|
| 1054 |
|
public E poll(long timeout, TimeUnit unit) throws InterruptedException { |
| 1055 |
< |
Object e = xfer(null, false, TIMEOUT, unit.toNanos(timeout)); |
| 1055 |
> |
E e = xfer(null, false, TIMED, unit.toNanos(timeout)); |
| 1056 |
|
if (e != null || !Thread.interrupted()) |
| 1057 |
< |
return (E)e; |
| 1057 |
> |
return e; |
| 1058 |
|
throw new InterruptedException(); |
| 1059 |
|
} |
| 1060 |
|
|
| 1061 |
|
public E poll() { |
| 1062 |
< |
return (E)xfer(null, false, NOW, 0); |
| 1062 |
> |
return xfer(null, false, NOW, 0); |
| 1063 |
|
} |
| 1064 |
|
|
| 1065 |
|
/** |
| 1116 |
|
} |
| 1117 |
|
|
| 1118 |
|
public E peek() { |
| 1119 |
< |
return (E) firstDataItem(); |
| 1119 |
> |
return firstDataItem(); |
| 1120 |
|
} |
| 1121 |
|
|
| 1122 |
|
/** |
| 1212 |
|
} |
| 1213 |
|
} |
| 1214 |
|
|
| 1188 |
– |
|
| 1215 |
|
// Unsafe mechanics |
| 1216 |
|
|
| 1217 |
|
private static final sun.misc.Unsafe UNSAFE = getUnsafe(); |
| 1219 |
|
objectFieldOffset(UNSAFE, "head", LinkedTransferQueue.class); |
| 1220 |
|
private static final long tailOffset = |
| 1221 |
|
objectFieldOffset(UNSAFE, "tail", LinkedTransferQueue.class); |
| 1222 |
< |
private static final long cleanMeOffset = |
| 1223 |
< |
objectFieldOffset(UNSAFE, "cleanMe", LinkedTransferQueue.class); |
| 1222 |
> |
private static final long sweepVotesOffset = |
| 1223 |
> |
objectFieldOffset(UNSAFE, "sweepVotes", LinkedTransferQueue.class); |
| 1224 |
|
|
| 1225 |
|
static long objectFieldOffset(sun.misc.Unsafe UNSAFE, |
| 1226 |
|
String field, Class<?> klazz) { |
| 1234 |
|
} |
| 1235 |
|
} |
| 1236 |
|
|
| 1237 |
< |
private static sun.misc.Unsafe getUnsafe() { |
| 1237 |
> |
/** |
| 1238 |
> |
* Returns a sun.misc.Unsafe. Suitable for use in a 3rd party package. |
| 1239 |
> |
* Replace with a simple call to Unsafe.getUnsafe when integrating |
| 1240 |
> |
* into a jdk. |
| 1241 |
> |
* |
| 1242 |
> |
* @return a sun.misc.Unsafe |
| 1243 |
> |
*/ |
| 1244 |
> |
static sun.misc.Unsafe getUnsafe() { |
| 1245 |
|
try { |
| 1246 |
|
return sun.misc.Unsafe.getUnsafe(); |
| 1247 |
|
} catch (SecurityException se) { |
