1 /*- 2 * SPDX-License-Identifier: BSD-2-Clause-FreeBSD 3 * 4 * Copyright (c) 2007 Attilio Rao <attilio@freebsd.org> 5 * Copyright (c) 2001 Jason Evans <jasone@freebsd.org> 6 * All rights reserved. 7 * 8 * Redistribution and use in source and binary forms, with or without 9 * modification, are permitted provided that the following conditions 10 * are met: 11 * 1. Redistributions of source code must retain the above copyright 12 * notice(s), this list of conditions and the following disclaimer as 13 * the first lines of this file unmodified other than the possible 14 * addition of one or more copyright notices. 15 * 2. Redistributions in binary form must reproduce the above copyright 16 * notice(s), this list of conditions and the following disclaimer in the 17 * documentation and/or other materials provided with the distribution. 18 * 19 * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDER(S) ``AS IS'' AND ANY 20 * EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED 21 * WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE 22 * DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT HOLDER(S) BE LIABLE FOR ANY 23 * DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES 24 * (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR 25 * SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER 26 * CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT 27 * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY 28 * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH 29 * DAMAGE. 30 */ 31 32 /* 33 * Shared/exclusive locks. This implementation attempts to ensure 34 * deterministic lock granting behavior, so that slocks and xlocks are 35 * interleaved. 36 * 37 * Priority propagation will not generally raise the priority of lock holders, 38 * so should not be relied upon in combination with sx locks. 39 */ 40 41 #include "opt_ddb.h" 42 #include "opt_hwpmc_hooks.h" 43 #include "opt_no_adaptive_sx.h" 44 45 #include <sys/cdefs.h> 46 __FBSDID("$FreeBSD$"); 47 48 #include <sys/param.h> 49 #include <sys/systm.h> 50 #include <sys/kdb.h> 51 #include <sys/kernel.h> 52 #include <sys/ktr.h> 53 #include <sys/lock.h> 54 #include <sys/mutex.h> 55 #include <sys/proc.h> 56 #include <sys/sched.h> 57 #include <sys/sleepqueue.h> 58 #include <sys/sx.h> 59 #include <sys/smp.h> 60 #include <sys/sysctl.h> 61 62 #if defined(SMP) && !defined(NO_ADAPTIVE_SX) 63 #include <machine/cpu.h> 64 #endif 65 66 #ifdef DDB 67 #include <ddb/ddb.h> 68 #endif 69 70 #if defined(SMP) && !defined(NO_ADAPTIVE_SX) 71 #define ADAPTIVE_SX 72 #endif 73 74 CTASSERT((SX_NOADAPTIVE & LO_CLASSFLAGS) == SX_NOADAPTIVE); 75 76 #ifdef HWPMC_HOOKS 77 #include <sys/pmckern.h> 78 PMC_SOFT_DECLARE( , , lock, failed); 79 #endif 80 81 /* Handy macros for sleep queues. */ 82 #define SQ_EXCLUSIVE_QUEUE 0 83 #define SQ_SHARED_QUEUE 1 84 85 /* 86 * Variations on DROP_GIANT()/PICKUP_GIANT() for use in this file. We 87 * drop Giant anytime we have to sleep or if we adaptively spin. 88 */ 89 #define GIANT_DECLARE \ 90 int _giantcnt = 0; \ 91 WITNESS_SAVE_DECL(Giant) \ 92 93 #define GIANT_SAVE(work) do { \ 94 if (__predict_false(mtx_owned(&Giant))) { \ 95 work++; \ 96 WITNESS_SAVE(&Giant.lock_object, Giant); \ 97 while (mtx_owned(&Giant)) { \ 98 _giantcnt++; \ 99 mtx_unlock(&Giant); \ 100 } \ 101 } \ 102 } while (0) 103 104 #define GIANT_RESTORE() do { \ 105 if (_giantcnt > 0) { \ 106 mtx_assert(&Giant, MA_NOTOWNED); \ 107 while (_giantcnt--) \ 108 mtx_lock(&Giant); \ 109 WITNESS_RESTORE(&Giant.lock_object, Giant); \ 110 } \ 111 } while (0) 112 113 /* 114 * Returns true if an exclusive lock is recursed. It assumes 115 * curthread currently has an exclusive lock. 116 */ 117 #define sx_recursed(sx) ((sx)->sx_recurse != 0) 118 119 static void assert_sx(const struct lock_object *lock, int what); 120 #ifdef DDB 121 static void db_show_sx(const struct lock_object *lock); 122 #endif 123 static void lock_sx(struct lock_object *lock, uintptr_t how); 124 #ifdef KDTRACE_HOOKS 125 static int owner_sx(const struct lock_object *lock, struct thread **owner); 126 #endif 127 static uintptr_t unlock_sx(struct lock_object *lock); 128 129 struct lock_class lock_class_sx = { 130 .lc_name = "sx", 131 .lc_flags = LC_SLEEPLOCK | LC_SLEEPABLE | LC_RECURSABLE | LC_UPGRADABLE, 132 .lc_assert = assert_sx, 133 #ifdef DDB 134 .lc_ddb_show = db_show_sx, 135 #endif 136 .lc_lock = lock_sx, 137 .lc_unlock = unlock_sx, 138 #ifdef KDTRACE_HOOKS 139 .lc_owner = owner_sx, 140 #endif 141 }; 142 143 #ifndef INVARIANTS 144 #define _sx_assert(sx, what, file, line) 145 #endif 146 147 #ifdef ADAPTIVE_SX 148 static __read_frequently u_int asx_retries; 149 static __read_frequently u_int asx_loops; 150 static SYSCTL_NODE(_debug, OID_AUTO, sx, CTLFLAG_RD, NULL, "sxlock debugging"); 151 SYSCTL_UINT(_debug_sx, OID_AUTO, retries, CTLFLAG_RW, &asx_retries, 0, ""); 152 SYSCTL_UINT(_debug_sx, OID_AUTO, loops, CTLFLAG_RW, &asx_loops, 0, ""); 153 154 static struct lock_delay_config __read_frequently sx_delay; 155 156 SYSCTL_INT(_debug_sx, OID_AUTO, delay_base, CTLFLAG_RW, &sx_delay.base, 157 0, ""); 158 SYSCTL_INT(_debug_sx, OID_AUTO, delay_max, CTLFLAG_RW, &sx_delay.max, 159 0, ""); 160 161 static void 162 sx_lock_delay_init(void *arg __unused) 163 { 164 165 lock_delay_default_init(&sx_delay); 166 asx_retries = 10; 167 asx_loops = max(10000, sx_delay.max); 168 } 169 LOCK_DELAY_SYSINIT(sx_lock_delay_init); 170 #endif 171 172 void 173 assert_sx(const struct lock_object *lock, int what) 174 { 175 176 sx_assert((const struct sx *)lock, what); 177 } 178 179 void 180 lock_sx(struct lock_object *lock, uintptr_t how) 181 { 182 struct sx *sx; 183 184 sx = (struct sx *)lock; 185 if (how) 186 sx_slock(sx); 187 else 188 sx_xlock(sx); 189 } 190 191 uintptr_t 192 unlock_sx(struct lock_object *lock) 193 { 194 struct sx *sx; 195 196 sx = (struct sx *)lock; 197 sx_assert(sx, SA_LOCKED | SA_NOTRECURSED); 198 if (sx_xlocked(sx)) { 199 sx_xunlock(sx); 200 return (0); 201 } else { 202 sx_sunlock(sx); 203 return (1); 204 } 205 } 206 207 #ifdef KDTRACE_HOOKS 208 int 209 owner_sx(const struct lock_object *lock, struct thread **owner) 210 { 211 const struct sx *sx; 212 uintptr_t x; 213 214 sx = (const struct sx *)lock; 215 x = sx->sx_lock; 216 *owner = NULL; 217 return ((x & SX_LOCK_SHARED) != 0 ? (SX_SHARERS(x) != 0) : 218 ((*owner = (struct thread *)SX_OWNER(x)) != NULL)); 219 } 220 #endif 221 222 void 223 sx_sysinit(void *arg) 224 { 225 struct sx_args *sargs = arg; 226 227 sx_init_flags(sargs->sa_sx, sargs->sa_desc, sargs->sa_flags); 228 } 229 230 void 231 sx_init_flags(struct sx *sx, const char *description, int opts) 232 { 233 int flags; 234 235 MPASS((opts & ~(SX_QUIET | SX_RECURSE | SX_NOWITNESS | SX_DUPOK | 236 SX_NOPROFILE | SX_NOADAPTIVE | SX_NEW)) == 0); 237 ASSERT_ATOMIC_LOAD_PTR(sx->sx_lock, 238 ("%s: sx_lock not aligned for %s: %p", __func__, description, 239 &sx->sx_lock)); 240 241 flags = LO_SLEEPABLE | LO_UPGRADABLE; 242 if (opts & SX_DUPOK) 243 flags |= LO_DUPOK; 244 if (opts & SX_NOPROFILE) 245 flags |= LO_NOPROFILE; 246 if (!(opts & SX_NOWITNESS)) 247 flags |= LO_WITNESS; 248 if (opts & SX_RECURSE) 249 flags |= LO_RECURSABLE; 250 if (opts & SX_QUIET) 251 flags |= LO_QUIET; 252 if (opts & SX_NEW) 253 flags |= LO_NEW; 254 255 flags |= opts & SX_NOADAPTIVE; 256 lock_init(&sx->lock_object, &lock_class_sx, description, NULL, flags); 257 sx->sx_lock = SX_LOCK_UNLOCKED; 258 sx->sx_recurse = 0; 259 } 260 261 void 262 sx_destroy(struct sx *sx) 263 { 264 265 KASSERT(sx->sx_lock == SX_LOCK_UNLOCKED, ("sx lock still held")); 266 KASSERT(sx->sx_recurse == 0, ("sx lock still recursed")); 267 sx->sx_lock = SX_LOCK_DESTROYED; 268 lock_destroy(&sx->lock_object); 269 } 270 271 int 272 sx_try_slock_int(struct sx *sx LOCK_FILE_LINE_ARG_DEF) 273 { 274 uintptr_t x; 275 276 if (SCHEDULER_STOPPED()) 277 return (1); 278 279 KASSERT(kdb_active != 0 || !TD_IS_IDLETHREAD(curthread), 280 ("sx_try_slock() by idle thread %p on sx %s @ %s:%d", 281 curthread, sx->lock_object.lo_name, file, line)); 282 283 x = sx->sx_lock; 284 for (;;) { 285 KASSERT(x != SX_LOCK_DESTROYED, 286 ("sx_try_slock() of destroyed sx @ %s:%d", file, line)); 287 if (!(x & SX_LOCK_SHARED)) 288 break; 289 if (atomic_fcmpset_acq_ptr(&sx->sx_lock, &x, x + SX_ONE_SHARER)) { 290 LOCK_LOG_TRY("SLOCK", &sx->lock_object, 0, 1, file, line); 291 WITNESS_LOCK(&sx->lock_object, LOP_TRYLOCK, file, line); 292 LOCKSTAT_PROFILE_OBTAIN_RWLOCK_SUCCESS(sx__acquire, 293 sx, 0, 0, file, line, LOCKSTAT_READER); 294 TD_LOCKS_INC(curthread); 295 curthread->td_sx_slocks++; 296 return (1); 297 } 298 } 299 300 LOCK_LOG_TRY("SLOCK", &sx->lock_object, 0, 0, file, line); 301 return (0); 302 } 303 304 int 305 sx_try_slock_(struct sx *sx, const char *file, int line) 306 { 307 308 return (sx_try_slock_int(sx LOCK_FILE_LINE_ARG)); 309 } 310 311 int 312 _sx_xlock(struct sx *sx, int opts, const char *file, int line) 313 { 314 uintptr_t tid, x; 315 int error = 0; 316 317 KASSERT(kdb_active != 0 || SCHEDULER_STOPPED() || 318 !TD_IS_IDLETHREAD(curthread), 319 ("sx_xlock() by idle thread %p on sx %s @ %s:%d", 320 curthread, sx->lock_object.lo_name, file, line)); 321 KASSERT(sx->sx_lock != SX_LOCK_DESTROYED, 322 ("sx_xlock() of destroyed sx @ %s:%d", file, line)); 323 WITNESS_CHECKORDER(&sx->lock_object, LOP_NEWORDER | LOP_EXCLUSIVE, file, 324 line, NULL); 325 tid = (uintptr_t)curthread; 326 x = SX_LOCK_UNLOCKED; 327 if (!atomic_fcmpset_acq_ptr(&sx->sx_lock, &x, tid)) 328 error = _sx_xlock_hard(sx, x, opts LOCK_FILE_LINE_ARG); 329 else 330 LOCKSTAT_PROFILE_OBTAIN_RWLOCK_SUCCESS(sx__acquire, sx, 331 0, 0, file, line, LOCKSTAT_WRITER); 332 if (!error) { 333 LOCK_LOG_LOCK("XLOCK", &sx->lock_object, 0, sx->sx_recurse, 334 file, line); 335 WITNESS_LOCK(&sx->lock_object, LOP_EXCLUSIVE, file, line); 336 TD_LOCKS_INC(curthread); 337 } 338 339 return (error); 340 } 341 342 int 343 sx_try_xlock_int(struct sx *sx LOCK_FILE_LINE_ARG_DEF) 344 { 345 struct thread *td; 346 uintptr_t tid, x; 347 int rval; 348 bool recursed; 349 350 td = curthread; 351 tid = (uintptr_t)td; 352 if (SCHEDULER_STOPPED_TD(td)) 353 return (1); 354 355 KASSERT(kdb_active != 0 || !TD_IS_IDLETHREAD(td), 356 ("sx_try_xlock() by idle thread %p on sx %s @ %s:%d", 357 curthread, sx->lock_object.lo_name, file, line)); 358 KASSERT(sx->sx_lock != SX_LOCK_DESTROYED, 359 ("sx_try_xlock() of destroyed sx @ %s:%d", file, line)); 360 361 rval = 1; 362 recursed = false; 363 x = SX_LOCK_UNLOCKED; 364 for (;;) { 365 if (atomic_fcmpset_acq_ptr(&sx->sx_lock, &x, tid)) 366 break; 367 if (x == SX_LOCK_UNLOCKED) 368 continue; 369 if (x == tid && (sx->lock_object.lo_flags & LO_RECURSABLE)) { 370 sx->sx_recurse++; 371 atomic_set_ptr(&sx->sx_lock, SX_LOCK_RECURSED); 372 break; 373 } 374 rval = 0; 375 break; 376 } 377 378 LOCK_LOG_TRY("XLOCK", &sx->lock_object, 0, rval, file, line); 379 if (rval) { 380 WITNESS_LOCK(&sx->lock_object, LOP_EXCLUSIVE | LOP_TRYLOCK, 381 file, line); 382 if (!recursed) 383 LOCKSTAT_PROFILE_OBTAIN_RWLOCK_SUCCESS(sx__acquire, 384 sx, 0, 0, file, line, LOCKSTAT_WRITER); 385 TD_LOCKS_INC(curthread); 386 } 387 388 return (rval); 389 } 390 391 int 392 sx_try_xlock_(struct sx *sx, const char *file, int line) 393 { 394 395 return (sx_try_xlock_int(sx LOCK_FILE_LINE_ARG)); 396 } 397 398 void 399 _sx_xunlock(struct sx *sx, const char *file, int line) 400 { 401 402 KASSERT(sx->sx_lock != SX_LOCK_DESTROYED, 403 ("sx_xunlock() of destroyed sx @ %s:%d", file, line)); 404 _sx_assert(sx, SA_XLOCKED, file, line); 405 WITNESS_UNLOCK(&sx->lock_object, LOP_EXCLUSIVE, file, line); 406 LOCK_LOG_LOCK("XUNLOCK", &sx->lock_object, 0, sx->sx_recurse, file, 407 line); 408 #if LOCK_DEBUG > 0 409 _sx_xunlock_hard(sx, (uintptr_t)curthread, file, line); 410 #else 411 __sx_xunlock(sx, curthread, file, line); 412 #endif 413 TD_LOCKS_DEC(curthread); 414 } 415 416 /* 417 * Try to do a non-blocking upgrade from a shared lock to an exclusive lock. 418 * This will only succeed if this thread holds a single shared lock. 419 * Return 1 if if the upgrade succeed, 0 otherwise. 420 */ 421 int 422 sx_try_upgrade_int(struct sx *sx LOCK_FILE_LINE_ARG_DEF) 423 { 424 uintptr_t x; 425 uintptr_t waiters; 426 int success; 427 428 if (SCHEDULER_STOPPED()) 429 return (1); 430 431 KASSERT(sx->sx_lock != SX_LOCK_DESTROYED, 432 ("sx_try_upgrade() of destroyed sx @ %s:%d", file, line)); 433 _sx_assert(sx, SA_SLOCKED, file, line); 434 435 /* 436 * Try to switch from one shared lock to an exclusive lock. We need 437 * to maintain the SX_LOCK_EXCLUSIVE_WAITERS flag if set so that 438 * we will wake up the exclusive waiters when we drop the lock. 439 */ 440 success = 0; 441 x = SX_READ_VALUE(sx); 442 for (;;) { 443 if (SX_SHARERS(x) > 1) 444 break; 445 waiters = (x & SX_LOCK_WAITERS); 446 if (atomic_fcmpset_acq_ptr(&sx->sx_lock, &x, 447 (uintptr_t)curthread | waiters)) { 448 success = 1; 449 break; 450 } 451 } 452 LOCK_LOG_TRY("XUPGRADE", &sx->lock_object, 0, success, file, line); 453 if (success) { 454 curthread->td_sx_slocks--; 455 WITNESS_UPGRADE(&sx->lock_object, LOP_EXCLUSIVE | LOP_TRYLOCK, 456 file, line); 457 LOCKSTAT_RECORD0(sx__upgrade, sx); 458 } 459 return (success); 460 } 461 462 int 463 sx_try_upgrade_(struct sx *sx, const char *file, int line) 464 { 465 466 return (sx_try_upgrade_int(sx LOCK_FILE_LINE_ARG)); 467 } 468 469 /* 470 * Downgrade an unrecursed exclusive lock into a single shared lock. 471 */ 472 void 473 sx_downgrade_int(struct sx *sx LOCK_FILE_LINE_ARG_DEF) 474 { 475 uintptr_t x; 476 int wakeup_swapper; 477 478 if (SCHEDULER_STOPPED()) 479 return; 480 481 KASSERT(sx->sx_lock != SX_LOCK_DESTROYED, 482 ("sx_downgrade() of destroyed sx @ %s:%d", file, line)); 483 _sx_assert(sx, SA_XLOCKED | SA_NOTRECURSED, file, line); 484 #ifndef INVARIANTS 485 if (sx_recursed(sx)) 486 panic("downgrade of a recursed lock"); 487 #endif 488 489 WITNESS_DOWNGRADE(&sx->lock_object, 0, file, line); 490 491 /* 492 * Try to switch from an exclusive lock with no shared waiters 493 * to one sharer with no shared waiters. If there are 494 * exclusive waiters, we don't need to lock the sleep queue so 495 * long as we preserve the flag. We do one quick try and if 496 * that fails we grab the sleepq lock to keep the flags from 497 * changing and do it the slow way. 498 * 499 * We have to lock the sleep queue if there are shared waiters 500 * so we can wake them up. 501 */ 502 x = sx->sx_lock; 503 if (!(x & SX_LOCK_SHARED_WAITERS) && 504 atomic_cmpset_rel_ptr(&sx->sx_lock, x, SX_SHARERS_LOCK(1) | 505 (x & SX_LOCK_EXCLUSIVE_WAITERS))) 506 goto out; 507 508 /* 509 * Lock the sleep queue so we can read the waiters bits 510 * without any races and wakeup any shared waiters. 511 */ 512 sleepq_lock(&sx->lock_object); 513 514 /* 515 * Preserve SX_LOCK_EXCLUSIVE_WAITERS while downgraded to a single 516 * shared lock. If there are any shared waiters, wake them up. 517 */ 518 wakeup_swapper = 0; 519 x = sx->sx_lock; 520 atomic_store_rel_ptr(&sx->sx_lock, SX_SHARERS_LOCK(1) | 521 (x & SX_LOCK_EXCLUSIVE_WAITERS)); 522 if (x & SX_LOCK_SHARED_WAITERS) 523 wakeup_swapper = sleepq_broadcast(&sx->lock_object, SLEEPQ_SX, 524 0, SQ_SHARED_QUEUE); 525 sleepq_release(&sx->lock_object); 526 527 if (wakeup_swapper) 528 kick_proc0(); 529 530 out: 531 curthread->td_sx_slocks++; 532 LOCK_LOG_LOCK("XDOWNGRADE", &sx->lock_object, 0, 0, file, line); 533 LOCKSTAT_RECORD0(sx__downgrade, sx); 534 } 535 536 void 537 sx_downgrade_(struct sx *sx, const char *file, int line) 538 { 539 540 sx_downgrade_int(sx LOCK_FILE_LINE_ARG); 541 } 542 543 #ifdef ADAPTIVE_SX 544 static inline void 545 sx_drop_critical(uintptr_t x, bool *in_critical, int *extra_work) 546 { 547 548 if (x & SX_LOCK_WRITE_SPINNER) 549 return; 550 if (*in_critical) { 551 critical_exit(); 552 *in_critical = false; 553 (*extra_work)--; 554 } 555 } 556 #else 557 #define sx_drop_critical(x, in_critical, extra_work) do { } while(0) 558 #endif 559 560 /* 561 * This function represents the so-called 'hard case' for sx_xlock 562 * operation. All 'easy case' failures are redirected to this. Note 563 * that ideally this would be a static function, but it needs to be 564 * accessible from at least sx.h. 565 */ 566 int 567 _sx_xlock_hard(struct sx *sx, uintptr_t x, int opts LOCK_FILE_LINE_ARG_DEF) 568 { 569 GIANT_DECLARE; 570 uintptr_t tid, setx; 571 #ifdef ADAPTIVE_SX 572 volatile struct thread *owner; 573 u_int i, n, spintries = 0; 574 enum { READERS, WRITER } sleep_reason = READERS; 575 bool adaptive; 576 bool in_critical = false; 577 #endif 578 #ifdef LOCK_PROFILING 579 uint64_t waittime = 0; 580 int contested = 0; 581 #endif 582 int error = 0; 583 #if defined(ADAPTIVE_SX) || defined(KDTRACE_HOOKS) 584 struct lock_delay_arg lda; 585 #endif 586 #ifdef KDTRACE_HOOKS 587 u_int sleep_cnt = 0; 588 int64_t sleep_time = 0; 589 int64_t all_time = 0; 590 #endif 591 #if defined(KDTRACE_HOOKS) || defined(LOCK_PROFILING) 592 uintptr_t state = 0; 593 int doing_lockprof = 0; 594 #endif 595 int extra_work = 0; 596 597 tid = (uintptr_t)curthread; 598 599 #ifdef KDTRACE_HOOKS 600 if (LOCKSTAT_PROFILE_ENABLED(sx__acquire)) { 601 while (x == SX_LOCK_UNLOCKED) { 602 if (atomic_fcmpset_acq_ptr(&sx->sx_lock, &x, tid)) 603 goto out_lockstat; 604 } 605 extra_work = 1; 606 doing_lockprof = 1; 607 all_time -= lockstat_nsecs(&sx->lock_object); 608 state = x; 609 } 610 #endif 611 #ifdef LOCK_PROFILING 612 extra_work = 1; 613 doing_lockprof = 1; 614 state = x; 615 #endif 616 617 if (SCHEDULER_STOPPED()) 618 return (0); 619 620 #if defined(ADAPTIVE_SX) 621 lock_delay_arg_init(&lda, &sx_delay); 622 #elif defined(KDTRACE_HOOKS) 623 lock_delay_arg_init(&lda, NULL); 624 #endif 625 626 if (__predict_false(x == SX_LOCK_UNLOCKED)) 627 x = SX_READ_VALUE(sx); 628 629 /* If we already hold an exclusive lock, then recurse. */ 630 if (__predict_false(lv_sx_owner(x) == (struct thread *)tid)) { 631 KASSERT((sx->lock_object.lo_flags & LO_RECURSABLE) != 0, 632 ("_sx_xlock_hard: recursed on non-recursive sx %s @ %s:%d\n", 633 sx->lock_object.lo_name, file, line)); 634 sx->sx_recurse++; 635 atomic_set_ptr(&sx->sx_lock, SX_LOCK_RECURSED); 636 if (LOCK_LOG_TEST(&sx->lock_object, 0)) 637 CTR2(KTR_LOCK, "%s: %p recursing", __func__, sx); 638 return (0); 639 } 640 641 if (LOCK_LOG_TEST(&sx->lock_object, 0)) 642 CTR5(KTR_LOCK, "%s: %s contested (lock=%p) at %s:%d", __func__, 643 sx->lock_object.lo_name, (void *)sx->sx_lock, file, line); 644 645 #ifdef ADAPTIVE_SX 646 adaptive = ((sx->lock_object.lo_flags & SX_NOADAPTIVE) == 0); 647 #endif 648 649 #ifdef HWPMC_HOOKS 650 PMC_SOFT_CALL( , , lock, failed); 651 #endif 652 lock_profile_obtain_lock_failed(&sx->lock_object, &contested, 653 &waittime); 654 655 #ifndef INVARIANTS 656 GIANT_SAVE(extra_work); 657 #endif 658 659 for (;;) { 660 if (x == SX_LOCK_UNLOCKED) { 661 if (atomic_fcmpset_acq_ptr(&sx->sx_lock, &x, tid)) 662 break; 663 continue; 664 } 665 #ifdef INVARIANTS 666 GIANT_SAVE(extra_work); 667 #endif 668 #ifdef KDTRACE_HOOKS 669 lda.spin_cnt++; 670 #endif 671 #ifdef ADAPTIVE_SX 672 if (__predict_false(!adaptive)) 673 goto sleepq; 674 /* 675 * If the lock is write locked and the owner is 676 * running on another CPU, spin until the owner stops 677 * running or the state of the lock changes. 678 */ 679 if ((x & SX_LOCK_SHARED) == 0) { 680 sx_drop_critical(x, &in_critical, &extra_work); 681 sleep_reason = WRITER; 682 owner = lv_sx_owner(x); 683 if (!TD_IS_RUNNING(owner)) 684 goto sleepq; 685 if (LOCK_LOG_TEST(&sx->lock_object, 0)) 686 CTR3(KTR_LOCK, "%s: spinning on %p held by %p", 687 __func__, sx, owner); 688 KTR_STATE1(KTR_SCHED, "thread", sched_tdname(curthread), 689 "spinning", "lockname:\"%s\"", 690 sx->lock_object.lo_name); 691 do { 692 lock_delay(&lda); 693 x = SX_READ_VALUE(sx); 694 owner = lv_sx_owner(x); 695 } while (owner != NULL && TD_IS_RUNNING(owner)); 696 KTR_STATE0(KTR_SCHED, "thread", sched_tdname(curthread), 697 "running"); 698 continue; 699 } else if (SX_SHARERS(x) > 0) { 700 sleep_reason = READERS; 701 if (spintries == asx_retries) 702 goto sleepq; 703 if (!(x & SX_LOCK_WRITE_SPINNER)) { 704 if (!in_critical) { 705 critical_enter(); 706 in_critical = true; 707 extra_work++; 708 } 709 if (!atomic_fcmpset_ptr(&sx->sx_lock, &x, 710 x | SX_LOCK_WRITE_SPINNER)) { 711 critical_exit(); 712 in_critical = false; 713 extra_work--; 714 continue; 715 } 716 } 717 spintries++; 718 KTR_STATE1(KTR_SCHED, "thread", sched_tdname(curthread), 719 "spinning", "lockname:\"%s\"", 720 sx->lock_object.lo_name); 721 n = SX_SHARERS(x); 722 for (i = 0; i < asx_loops; i += n) { 723 lock_delay_spin(n); 724 x = SX_READ_VALUE(sx); 725 if (!(x & SX_LOCK_WRITE_SPINNER)) 726 break; 727 if (!(x & SX_LOCK_SHARED)) 728 break; 729 n = SX_SHARERS(x); 730 if (n == 0) 731 break; 732 } 733 #ifdef KDTRACE_HOOKS 734 lda.spin_cnt += i; 735 #endif 736 KTR_STATE0(KTR_SCHED, "thread", sched_tdname(curthread), 737 "running"); 738 if (i < asx_loops) 739 continue; 740 } 741 sleepq: 742 #endif 743 sleepq_lock(&sx->lock_object); 744 x = SX_READ_VALUE(sx); 745 retry_sleepq: 746 747 /* 748 * If the lock was released while spinning on the 749 * sleep queue chain lock, try again. 750 */ 751 if (x == SX_LOCK_UNLOCKED) { 752 sleepq_release(&sx->lock_object); 753 sx_drop_critical(x, &in_critical, &extra_work); 754 continue; 755 } 756 757 #ifdef ADAPTIVE_SX 758 /* 759 * The current lock owner might have started executing 760 * on another CPU (or the lock could have changed 761 * owners) while we were waiting on the sleep queue 762 * chain lock. If so, drop the sleep queue lock and try 763 * again. 764 */ 765 if (adaptive) { 766 if (!(x & SX_LOCK_SHARED)) { 767 owner = (struct thread *)SX_OWNER(x); 768 if (TD_IS_RUNNING(owner)) { 769 sleepq_release(&sx->lock_object); 770 sx_drop_critical(x, &in_critical, 771 &extra_work); 772 continue; 773 } 774 } else if (SX_SHARERS(x) > 0 && sleep_reason == WRITER) { 775 sleepq_release(&sx->lock_object); 776 sx_drop_critical(x, &in_critical, &extra_work); 777 continue; 778 } 779 } 780 #endif 781 782 /* 783 * If an exclusive lock was released with both shared 784 * and exclusive waiters and a shared waiter hasn't 785 * woken up and acquired the lock yet, sx_lock will be 786 * set to SX_LOCK_UNLOCKED | SX_LOCK_EXCLUSIVE_WAITERS. 787 * If we see that value, try to acquire it once. Note 788 * that we have to preserve SX_LOCK_EXCLUSIVE_WAITERS 789 * as there are other exclusive waiters still. If we 790 * fail, restart the loop. 791 */ 792 setx = x & (SX_LOCK_WAITERS | SX_LOCK_WRITE_SPINNER); 793 if ((x & ~setx) == SX_LOCK_SHARED) { 794 setx &= ~SX_LOCK_WRITE_SPINNER; 795 if (!atomic_fcmpset_acq_ptr(&sx->sx_lock, &x, tid | setx)) 796 goto retry_sleepq; 797 sleepq_release(&sx->lock_object); 798 CTR2(KTR_LOCK, "%s: %p claimed by new writer", 799 __func__, sx); 800 break; 801 } 802 803 #ifdef ADAPTIVE_SX 804 /* 805 * It is possible we set the SX_LOCK_WRITE_SPINNER bit. 806 * It is an invariant that when the bit is set, there is 807 * a writer ready to grab the lock. Thus clear the bit since 808 * we are going to sleep. 809 */ 810 if (in_critical) { 811 if ((x & SX_LOCK_WRITE_SPINNER) || 812 !((x & SX_LOCK_EXCLUSIVE_WAITERS))) { 813 setx = x & ~SX_LOCK_WRITE_SPINNER; 814 setx |= SX_LOCK_EXCLUSIVE_WAITERS; 815 if (!atomic_fcmpset_ptr(&sx->sx_lock, &x, 816 setx)) { 817 goto retry_sleepq; 818 } 819 } 820 critical_exit(); 821 in_critical = false; 822 } else { 823 #endif 824 /* 825 * Try to set the SX_LOCK_EXCLUSIVE_WAITERS. If we fail, 826 * than loop back and retry. 827 */ 828 if (!(x & SX_LOCK_EXCLUSIVE_WAITERS)) { 829 if (!atomic_fcmpset_ptr(&sx->sx_lock, &x, 830 x | SX_LOCK_EXCLUSIVE_WAITERS)) { 831 goto retry_sleepq; 832 } 833 if (LOCK_LOG_TEST(&sx->lock_object, 0)) 834 CTR2(KTR_LOCK, "%s: %p set excl waiters flag", 835 __func__, sx); 836 } 837 #ifdef ADAPTIVE_SX 838 } 839 #endif 840 841 /* 842 * Since we have been unable to acquire the exclusive 843 * lock and the exclusive waiters flag is set, we have 844 * to sleep. 845 */ 846 if (LOCK_LOG_TEST(&sx->lock_object, 0)) 847 CTR2(KTR_LOCK, "%s: %p blocking on sleep queue", 848 __func__, sx); 849 850 #ifdef KDTRACE_HOOKS 851 sleep_time -= lockstat_nsecs(&sx->lock_object); 852 #endif 853 sleepq_add(&sx->lock_object, NULL, sx->lock_object.lo_name, 854 SLEEPQ_SX | ((opts & SX_INTERRUPTIBLE) ? 855 SLEEPQ_INTERRUPTIBLE : 0), SQ_EXCLUSIVE_QUEUE); 856 if (!(opts & SX_INTERRUPTIBLE)) 857 sleepq_wait(&sx->lock_object, 0); 858 else 859 error = sleepq_wait_sig(&sx->lock_object, 0); 860 #ifdef KDTRACE_HOOKS 861 sleep_time += lockstat_nsecs(&sx->lock_object); 862 sleep_cnt++; 863 #endif 864 if (error) { 865 if (LOCK_LOG_TEST(&sx->lock_object, 0)) 866 CTR2(KTR_LOCK, 867 "%s: interruptible sleep by %p suspended by signal", 868 __func__, sx); 869 break; 870 } 871 if (LOCK_LOG_TEST(&sx->lock_object, 0)) 872 CTR2(KTR_LOCK, "%s: %p resuming from sleep queue", 873 __func__, sx); 874 x = SX_READ_VALUE(sx); 875 } 876 if (__predict_true(!extra_work)) 877 return (error); 878 #ifdef ADAPTIVE_SX 879 if (in_critical) 880 critical_exit(); 881 #endif 882 GIANT_RESTORE(); 883 #if defined(KDTRACE_HOOKS) || defined(LOCK_PROFILING) 884 if (__predict_true(!doing_lockprof)) 885 return (error); 886 #endif 887 #ifdef KDTRACE_HOOKS 888 all_time += lockstat_nsecs(&sx->lock_object); 889 if (sleep_time) 890 LOCKSTAT_RECORD4(sx__block, sx, sleep_time, 891 LOCKSTAT_WRITER, (state & SX_LOCK_SHARED) == 0, 892 (state & SX_LOCK_SHARED) == 0 ? 0 : SX_SHARERS(state)); 893 if (lda.spin_cnt > sleep_cnt) 894 LOCKSTAT_RECORD4(sx__spin, sx, all_time - sleep_time, 895 LOCKSTAT_WRITER, (state & SX_LOCK_SHARED) == 0, 896 (state & SX_LOCK_SHARED) == 0 ? 0 : SX_SHARERS(state)); 897 out_lockstat: 898 #endif 899 if (!error) 900 LOCKSTAT_PROFILE_OBTAIN_RWLOCK_SUCCESS(sx__acquire, sx, 901 contested, waittime, file, line, LOCKSTAT_WRITER); 902 return (error); 903 } 904 905 /* 906 * This function represents the so-called 'hard case' for sx_xunlock 907 * operation. All 'easy case' failures are redirected to this. Note 908 * that ideally this would be a static function, but it needs to be 909 * accessible from at least sx.h. 910 */ 911 void 912 _sx_xunlock_hard(struct sx *sx, uintptr_t x LOCK_FILE_LINE_ARG_DEF) 913 { 914 uintptr_t tid, setx; 915 int queue, wakeup_swapper; 916 917 if (SCHEDULER_STOPPED()) 918 return; 919 920 tid = (uintptr_t)curthread; 921 922 if (__predict_false(x == tid)) 923 x = SX_READ_VALUE(sx); 924 925 MPASS(!(x & SX_LOCK_SHARED)); 926 927 if (__predict_false(x & SX_LOCK_RECURSED)) { 928 /* The lock is recursed, unrecurse one level. */ 929 if ((--sx->sx_recurse) == 0) 930 atomic_clear_ptr(&sx->sx_lock, SX_LOCK_RECURSED); 931 if (LOCK_LOG_TEST(&sx->lock_object, 0)) 932 CTR2(KTR_LOCK, "%s: %p unrecursing", __func__, sx); 933 return; 934 } 935 936 LOCKSTAT_PROFILE_RELEASE_RWLOCK(sx__release, sx, LOCKSTAT_WRITER); 937 if (x == tid && 938 atomic_cmpset_rel_ptr(&sx->sx_lock, tid, SX_LOCK_UNLOCKED)) 939 return; 940 941 if (LOCK_LOG_TEST(&sx->lock_object, 0)) 942 CTR2(KTR_LOCK, "%s: %p contested", __func__, sx); 943 944 sleepq_lock(&sx->lock_object); 945 x = SX_READ_VALUE(sx); 946 MPASS(x & (SX_LOCK_SHARED_WAITERS | SX_LOCK_EXCLUSIVE_WAITERS)); 947 948 /* 949 * The wake up algorithm here is quite simple and probably not 950 * ideal. It gives precedence to shared waiters if they are 951 * present. For this condition, we have to preserve the 952 * state of the exclusive waiters flag. 953 * If interruptible sleeps left the shared queue empty avoid a 954 * starvation for the threads sleeping on the exclusive queue by giving 955 * them precedence and cleaning up the shared waiters bit anyway. 956 */ 957 setx = SX_LOCK_UNLOCKED; 958 queue = SQ_SHARED_QUEUE; 959 if ((x & SX_LOCK_EXCLUSIVE_WAITERS) != 0 && 960 sleepq_sleepcnt(&sx->lock_object, SQ_EXCLUSIVE_QUEUE) != 0) { 961 queue = SQ_EXCLUSIVE_QUEUE; 962 setx |= (x & SX_LOCK_SHARED_WAITERS); 963 } 964 atomic_store_rel_ptr(&sx->sx_lock, setx); 965 966 /* Wake up all the waiters for the specific queue. */ 967 if (LOCK_LOG_TEST(&sx->lock_object, 0)) 968 CTR3(KTR_LOCK, "%s: %p waking up all threads on %s queue", 969 __func__, sx, queue == SQ_SHARED_QUEUE ? "shared" : 970 "exclusive"); 971 972 wakeup_swapper = sleepq_broadcast(&sx->lock_object, SLEEPQ_SX, 0, 973 queue); 974 sleepq_release(&sx->lock_object); 975 if (wakeup_swapper) 976 kick_proc0(); 977 } 978 979 static bool __always_inline 980 __sx_can_read(struct thread *td, uintptr_t x, bool fp) 981 { 982 983 if ((x & (SX_LOCK_SHARED | SX_LOCK_EXCLUSIVE_WAITERS | SX_LOCK_WRITE_SPINNER)) 984 == SX_LOCK_SHARED) 985 return (true); 986 if (!fp && td->td_sx_slocks && (x & SX_LOCK_SHARED)) 987 return (true); 988 return (false); 989 } 990 991 static bool __always_inline 992 __sx_slock_try(struct sx *sx, struct thread *td, uintptr_t *xp, bool fp 993 LOCK_FILE_LINE_ARG_DEF) 994 { 995 996 /* 997 * If no other thread has an exclusive lock then try to bump up 998 * the count of sharers. Since we have to preserve the state 999 * of SX_LOCK_EXCLUSIVE_WAITERS, if we fail to acquire the 1000 * shared lock loop back and retry. 1001 */ 1002 while (__sx_can_read(td, *xp, fp)) { 1003 if (atomic_fcmpset_acq_ptr(&sx->sx_lock, xp, 1004 *xp + SX_ONE_SHARER)) { 1005 if (LOCK_LOG_TEST(&sx->lock_object, 0)) 1006 CTR4(KTR_LOCK, "%s: %p succeed %p -> %p", 1007 __func__, sx, (void *)*xp, 1008 (void *)(*xp + SX_ONE_SHARER)); 1009 td->td_sx_slocks++; 1010 return (true); 1011 } 1012 } 1013 return (false); 1014 } 1015 1016 static int __noinline 1017 _sx_slock_hard(struct sx *sx, int opts, uintptr_t x LOCK_FILE_LINE_ARG_DEF) 1018 { 1019 GIANT_DECLARE; 1020 struct thread *td; 1021 #ifdef ADAPTIVE_SX 1022 volatile struct thread *owner; 1023 u_int i, n, spintries = 0; 1024 bool adaptive; 1025 #endif 1026 #ifdef LOCK_PROFILING 1027 uint64_t waittime = 0; 1028 int contested = 0; 1029 #endif 1030 int error = 0; 1031 #if defined(ADAPTIVE_SX) || defined(KDTRACE_HOOKS) 1032 struct lock_delay_arg lda; 1033 #endif 1034 #ifdef KDTRACE_HOOKS 1035 u_int sleep_cnt = 0; 1036 int64_t sleep_time = 0; 1037 int64_t all_time = 0; 1038 #endif 1039 #if defined(KDTRACE_HOOKS) || defined(LOCK_PROFILING) 1040 uintptr_t state = 0; 1041 #endif 1042 int extra_work = 0; 1043 1044 td = curthread; 1045 1046 #ifdef KDTRACE_HOOKS 1047 if (LOCKSTAT_PROFILE_ENABLED(sx__acquire)) { 1048 if (__sx_slock_try(sx, td, &x, false LOCK_FILE_LINE_ARG)) 1049 goto out_lockstat; 1050 extra_work = 1; 1051 all_time -= lockstat_nsecs(&sx->lock_object); 1052 state = x; 1053 } 1054 #endif 1055 #ifdef LOCK_PROFILING 1056 extra_work = 1; 1057 state = x; 1058 #endif 1059 1060 if (SCHEDULER_STOPPED()) 1061 return (0); 1062 1063 #if defined(ADAPTIVE_SX) 1064 lock_delay_arg_init(&lda, &sx_delay); 1065 #elif defined(KDTRACE_HOOKS) 1066 lock_delay_arg_init(&lda, NULL); 1067 #endif 1068 1069 #ifdef ADAPTIVE_SX 1070 adaptive = ((sx->lock_object.lo_flags & SX_NOADAPTIVE) == 0); 1071 #endif 1072 1073 #ifdef HWPMC_HOOKS 1074 PMC_SOFT_CALL( , , lock, failed); 1075 #endif 1076 lock_profile_obtain_lock_failed(&sx->lock_object, &contested, 1077 &waittime); 1078 1079 #ifndef INVARIANTS 1080 GIANT_SAVE(extra_work); 1081 #endif 1082 1083 /* 1084 * As with rwlocks, we don't make any attempt to try to block 1085 * shared locks once there is an exclusive waiter. 1086 */ 1087 for (;;) { 1088 if (__sx_slock_try(sx, td, &x, false LOCK_FILE_LINE_ARG)) 1089 break; 1090 #ifdef INVARIANTS 1091 GIANT_SAVE(extra_work); 1092 #endif 1093 #ifdef KDTRACE_HOOKS 1094 lda.spin_cnt++; 1095 #endif 1096 1097 #ifdef ADAPTIVE_SX 1098 if (__predict_false(!adaptive)) 1099 goto sleepq; 1100 1101 /* 1102 * If the owner is running on another CPU, spin until 1103 * the owner stops running or the state of the lock 1104 * changes. 1105 */ 1106 if ((x & SX_LOCK_SHARED) == 0) { 1107 owner = lv_sx_owner(x); 1108 if (TD_IS_RUNNING(owner)) { 1109 if (LOCK_LOG_TEST(&sx->lock_object, 0)) 1110 CTR3(KTR_LOCK, 1111 "%s: spinning on %p held by %p", 1112 __func__, sx, owner); 1113 KTR_STATE1(KTR_SCHED, "thread", 1114 sched_tdname(curthread), "spinning", 1115 "lockname:\"%s\"", sx->lock_object.lo_name); 1116 do { 1117 lock_delay(&lda); 1118 x = SX_READ_VALUE(sx); 1119 owner = lv_sx_owner(x); 1120 } while (owner != NULL && TD_IS_RUNNING(owner)); 1121 KTR_STATE0(KTR_SCHED, "thread", 1122 sched_tdname(curthread), "running"); 1123 continue; 1124 } 1125 } else { 1126 if ((x & SX_LOCK_WRITE_SPINNER) && SX_SHARERS(x) == 0) { 1127 MPASS(!__sx_can_read(td, x, false)); 1128 lock_delay_spin(2); 1129 x = SX_READ_VALUE(sx); 1130 continue; 1131 } 1132 if (spintries < asx_retries) { 1133 KTR_STATE1(KTR_SCHED, "thread", sched_tdname(curthread), 1134 "spinning", "lockname:\"%s\"", 1135 sx->lock_object.lo_name); 1136 n = SX_SHARERS(x); 1137 for (i = 0; i < asx_loops; i += n) { 1138 lock_delay_spin(n); 1139 x = SX_READ_VALUE(sx); 1140 if (!(x & SX_LOCK_SHARED)) 1141 break; 1142 n = SX_SHARERS(x); 1143 if (n == 0) 1144 break; 1145 if (__sx_can_read(td, x, false)) 1146 break; 1147 } 1148 #ifdef KDTRACE_HOOKS 1149 lda.spin_cnt += i; 1150 #endif 1151 KTR_STATE0(KTR_SCHED, "thread", sched_tdname(curthread), 1152 "running"); 1153 if (i < asx_loops) 1154 continue; 1155 } 1156 } 1157 sleepq: 1158 #endif 1159 1160 /* 1161 * Some other thread already has an exclusive lock, so 1162 * start the process of blocking. 1163 */ 1164 sleepq_lock(&sx->lock_object); 1165 x = SX_READ_VALUE(sx); 1166 retry_sleepq: 1167 if (((x & SX_LOCK_WRITE_SPINNER) && SX_SHARERS(x) == 0) || 1168 __sx_can_read(td, x, false)) { 1169 sleepq_release(&sx->lock_object); 1170 continue; 1171 } 1172 1173 #ifdef ADAPTIVE_SX 1174 /* 1175 * If the owner is running on another CPU, spin until 1176 * the owner stops running or the state of the lock 1177 * changes. 1178 */ 1179 if (!(x & SX_LOCK_SHARED) && adaptive) { 1180 owner = (struct thread *)SX_OWNER(x); 1181 if (TD_IS_RUNNING(owner)) { 1182 sleepq_release(&sx->lock_object); 1183 x = SX_READ_VALUE(sx); 1184 continue; 1185 } 1186 } 1187 #endif 1188 1189 /* 1190 * Try to set the SX_LOCK_SHARED_WAITERS flag. If we 1191 * fail to set it drop the sleep queue lock and loop 1192 * back. 1193 */ 1194 if (!(x & SX_LOCK_SHARED_WAITERS)) { 1195 if (!atomic_fcmpset_ptr(&sx->sx_lock, &x, 1196 x | SX_LOCK_SHARED_WAITERS)) 1197 goto retry_sleepq; 1198 if (LOCK_LOG_TEST(&sx->lock_object, 0)) 1199 CTR2(KTR_LOCK, "%s: %p set shared waiters flag", 1200 __func__, sx); 1201 } 1202 1203 /* 1204 * Since we have been unable to acquire the shared lock, 1205 * we have to sleep. 1206 */ 1207 if (LOCK_LOG_TEST(&sx->lock_object, 0)) 1208 CTR2(KTR_LOCK, "%s: %p blocking on sleep queue", 1209 __func__, sx); 1210 1211 #ifdef KDTRACE_HOOKS 1212 sleep_time -= lockstat_nsecs(&sx->lock_object); 1213 #endif 1214 sleepq_add(&sx->lock_object, NULL, sx->lock_object.lo_name, 1215 SLEEPQ_SX | ((opts & SX_INTERRUPTIBLE) ? 1216 SLEEPQ_INTERRUPTIBLE : 0), SQ_SHARED_QUEUE); 1217 if (!(opts & SX_INTERRUPTIBLE)) 1218 sleepq_wait(&sx->lock_object, 0); 1219 else 1220 error = sleepq_wait_sig(&sx->lock_object, 0); 1221 #ifdef KDTRACE_HOOKS 1222 sleep_time += lockstat_nsecs(&sx->lock_object); 1223 sleep_cnt++; 1224 #endif 1225 if (error) { 1226 if (LOCK_LOG_TEST(&sx->lock_object, 0)) 1227 CTR2(KTR_LOCK, 1228 "%s: interruptible sleep by %p suspended by signal", 1229 __func__, sx); 1230 break; 1231 } 1232 if (LOCK_LOG_TEST(&sx->lock_object, 0)) 1233 CTR2(KTR_LOCK, "%s: %p resuming from sleep queue", 1234 __func__, sx); 1235 x = SX_READ_VALUE(sx); 1236 } 1237 #if defined(KDTRACE_HOOKS) || defined(LOCK_PROFILING) 1238 if (__predict_true(!extra_work)) 1239 return (error); 1240 #endif 1241 #ifdef KDTRACE_HOOKS 1242 all_time += lockstat_nsecs(&sx->lock_object); 1243 if (sleep_time) 1244 LOCKSTAT_RECORD4(sx__block, sx, sleep_time, 1245 LOCKSTAT_READER, (state & SX_LOCK_SHARED) == 0, 1246 (state & SX_LOCK_SHARED) == 0 ? 0 : SX_SHARERS(state)); 1247 if (lda.spin_cnt > sleep_cnt) 1248 LOCKSTAT_RECORD4(sx__spin, sx, all_time - sleep_time, 1249 LOCKSTAT_READER, (state & SX_LOCK_SHARED) == 0, 1250 (state & SX_LOCK_SHARED) == 0 ? 0 : SX_SHARERS(state)); 1251 out_lockstat: 1252 #endif 1253 if (error == 0) { 1254 LOCKSTAT_PROFILE_OBTAIN_RWLOCK_SUCCESS(sx__acquire, sx, 1255 contested, waittime, file, line, LOCKSTAT_READER); 1256 } 1257 GIANT_RESTORE(); 1258 return (error); 1259 } 1260 1261 int 1262 _sx_slock_int(struct sx *sx, int opts LOCK_FILE_LINE_ARG_DEF) 1263 { 1264 struct thread *td; 1265 uintptr_t x; 1266 int error; 1267 1268 KASSERT(kdb_active != 0 || SCHEDULER_STOPPED() || 1269 !TD_IS_IDLETHREAD(curthread), 1270 ("sx_slock() by idle thread %p on sx %s @ %s:%d", 1271 curthread, sx->lock_object.lo_name, file, line)); 1272 KASSERT(sx->sx_lock != SX_LOCK_DESTROYED, 1273 ("sx_slock() of destroyed sx @ %s:%d", file, line)); 1274 WITNESS_CHECKORDER(&sx->lock_object, LOP_NEWORDER, file, line, NULL); 1275 1276 error = 0; 1277 td = curthread; 1278 x = SX_READ_VALUE(sx); 1279 if (__predict_false(LOCKSTAT_PROFILE_ENABLED(sx__acquire) || 1280 !__sx_slock_try(sx, td, &x, true LOCK_FILE_LINE_ARG))) 1281 error = _sx_slock_hard(sx, opts, x LOCK_FILE_LINE_ARG); 1282 else 1283 lock_profile_obtain_lock_success(&sx->lock_object, 0, 0, 1284 file, line); 1285 if (error == 0) { 1286 LOCK_LOG_LOCK("SLOCK", &sx->lock_object, 0, 0, file, line); 1287 WITNESS_LOCK(&sx->lock_object, 0, file, line); 1288 TD_LOCKS_INC(curthread); 1289 } 1290 return (error); 1291 } 1292 1293 int 1294 _sx_slock(struct sx *sx, int opts, const char *file, int line) 1295 { 1296 1297 return (_sx_slock_int(sx, opts LOCK_FILE_LINE_ARG)); 1298 } 1299 1300 static bool __always_inline 1301 _sx_sunlock_try(struct sx *sx, struct thread *td, uintptr_t *xp) 1302 { 1303 1304 for (;;) { 1305 if (SX_SHARERS(*xp) > 1 || !(*xp & SX_LOCK_WAITERS)) { 1306 if (atomic_fcmpset_rel_ptr(&sx->sx_lock, xp, 1307 *xp - SX_ONE_SHARER)) { 1308 if (LOCK_LOG_TEST(&sx->lock_object, 0)) 1309 CTR4(KTR_LOCK, 1310 "%s: %p succeeded %p -> %p", 1311 __func__, sx, (void *)*xp, 1312 (void *)(*xp - SX_ONE_SHARER)); 1313 td->td_sx_slocks--; 1314 return (true); 1315 } 1316 continue; 1317 } 1318 break; 1319 } 1320 return (false); 1321 } 1322 1323 static void __noinline 1324 _sx_sunlock_hard(struct sx *sx, struct thread *td, uintptr_t x 1325 LOCK_FILE_LINE_ARG_DEF) 1326 { 1327 int wakeup_swapper = 0; 1328 uintptr_t setx, queue; 1329 1330 if (SCHEDULER_STOPPED()) 1331 return; 1332 1333 if (_sx_sunlock_try(sx, td, &x)) 1334 goto out_lockstat; 1335 1336 sleepq_lock(&sx->lock_object); 1337 x = SX_READ_VALUE(sx); 1338 for (;;) { 1339 if (_sx_sunlock_try(sx, td, &x)) 1340 break; 1341 1342 /* 1343 * Wake up semantic here is quite simple: 1344 * Just wake up all the exclusive waiters. 1345 * Note that the state of the lock could have changed, 1346 * so if it fails loop back and retry. 1347 */ 1348 setx = SX_LOCK_UNLOCKED; 1349 queue = SQ_SHARED_QUEUE; 1350 if (x & SX_LOCK_EXCLUSIVE_WAITERS) { 1351 setx |= (x & SX_LOCK_SHARED_WAITERS); 1352 queue = SQ_EXCLUSIVE_QUEUE; 1353 } 1354 setx |= (x & SX_LOCK_WRITE_SPINNER); 1355 if (!atomic_fcmpset_rel_ptr(&sx->sx_lock, &x, setx)) 1356 continue; 1357 if (LOCK_LOG_TEST(&sx->lock_object, 0)) 1358 CTR2(KTR_LOCK, "%s: %p waking up all thread on" 1359 "exclusive queue", __func__, sx); 1360 wakeup_swapper = sleepq_broadcast(&sx->lock_object, SLEEPQ_SX, 1361 0, queue); 1362 td->td_sx_slocks--; 1363 break; 1364 } 1365 sleepq_release(&sx->lock_object); 1366 if (wakeup_swapper) 1367 kick_proc0(); 1368 out_lockstat: 1369 LOCKSTAT_PROFILE_RELEASE_RWLOCK(sx__release, sx, LOCKSTAT_READER); 1370 } 1371 1372 void 1373 _sx_sunlock_int(struct sx *sx LOCK_FILE_LINE_ARG_DEF) 1374 { 1375 struct thread *td; 1376 uintptr_t x; 1377 1378 KASSERT(sx->sx_lock != SX_LOCK_DESTROYED, 1379 ("sx_sunlock() of destroyed sx @ %s:%d", file, line)); 1380 _sx_assert(sx, SA_SLOCKED, file, line); 1381 WITNESS_UNLOCK(&sx->lock_object, 0, file, line); 1382 LOCK_LOG_LOCK("SUNLOCK", &sx->lock_object, 0, 0, file, line); 1383 1384 td = curthread; 1385 x = SX_READ_VALUE(sx); 1386 if (__predict_false(LOCKSTAT_PROFILE_ENABLED(sx__release) || 1387 !_sx_sunlock_try(sx, td, &x))) 1388 _sx_sunlock_hard(sx, td, x LOCK_FILE_LINE_ARG); 1389 else 1390 lock_profile_release_lock(&sx->lock_object); 1391 1392 TD_LOCKS_DEC(curthread); 1393 } 1394 1395 void 1396 _sx_sunlock(struct sx *sx, const char *file, int line) 1397 { 1398 1399 _sx_sunlock_int(sx LOCK_FILE_LINE_ARG); 1400 } 1401 1402 #ifdef INVARIANT_SUPPORT 1403 #ifndef INVARIANTS 1404 #undef _sx_assert 1405 #endif 1406 1407 /* 1408 * In the non-WITNESS case, sx_assert() can only detect that at least 1409 * *some* thread owns an slock, but it cannot guarantee that *this* 1410 * thread owns an slock. 1411 */ 1412 void 1413 _sx_assert(const struct sx *sx, int what, const char *file, int line) 1414 { 1415 #ifndef WITNESS 1416 int slocked = 0; 1417 #endif 1418 1419 if (panicstr != NULL) 1420 return; 1421 switch (what) { 1422 case SA_SLOCKED: 1423 case SA_SLOCKED | SA_NOTRECURSED: 1424 case SA_SLOCKED | SA_RECURSED: 1425 #ifndef WITNESS 1426 slocked = 1; 1427 /* FALLTHROUGH */ 1428 #endif 1429 case SA_LOCKED: 1430 case SA_LOCKED | SA_NOTRECURSED: 1431 case SA_LOCKED | SA_RECURSED: 1432 #ifdef WITNESS 1433 witness_assert(&sx->lock_object, what, file, line); 1434 #else 1435 /* 1436 * If some other thread has an exclusive lock or we 1437 * have one and are asserting a shared lock, fail. 1438 * Also, if no one has a lock at all, fail. 1439 */ 1440 if (sx->sx_lock == SX_LOCK_UNLOCKED || 1441 (!(sx->sx_lock & SX_LOCK_SHARED) && (slocked || 1442 sx_xholder(sx) != curthread))) 1443 panic("Lock %s not %slocked @ %s:%d\n", 1444 sx->lock_object.lo_name, slocked ? "share " : "", 1445 file, line); 1446 1447 if (!(sx->sx_lock & SX_LOCK_SHARED)) { 1448 if (sx_recursed(sx)) { 1449 if (what & SA_NOTRECURSED) 1450 panic("Lock %s recursed @ %s:%d\n", 1451 sx->lock_object.lo_name, file, 1452 line); 1453 } else if (what & SA_RECURSED) 1454 panic("Lock %s not recursed @ %s:%d\n", 1455 sx->lock_object.lo_name, file, line); 1456 } 1457 #endif 1458 break; 1459 case SA_XLOCKED: 1460 case SA_XLOCKED | SA_NOTRECURSED: 1461 case SA_XLOCKED | SA_RECURSED: 1462 if (sx_xholder(sx) != curthread) 1463 panic("Lock %s not exclusively locked @ %s:%d\n", 1464 sx->lock_object.lo_name, file, line); 1465 if (sx_recursed(sx)) { 1466 if (what & SA_NOTRECURSED) 1467 panic("Lock %s recursed @ %s:%d\n", 1468 sx->lock_object.lo_name, file, line); 1469 } else if (what & SA_RECURSED) 1470 panic("Lock %s not recursed @ %s:%d\n", 1471 sx->lock_object.lo_name, file, line); 1472 break; 1473 case SA_UNLOCKED: 1474 #ifdef WITNESS 1475 witness_assert(&sx->lock_object, what, file, line); 1476 #else 1477 /* 1478 * If we hold an exclusve lock fail. We can't 1479 * reliably check to see if we hold a shared lock or 1480 * not. 1481 */ 1482 if (sx_xholder(sx) == curthread) 1483 panic("Lock %s exclusively locked @ %s:%d\n", 1484 sx->lock_object.lo_name, file, line); 1485 #endif 1486 break; 1487 default: 1488 panic("Unknown sx lock assertion: %d @ %s:%d", what, file, 1489 line); 1490 } 1491 } 1492 #endif /* INVARIANT_SUPPORT */ 1493 1494 #ifdef DDB 1495 static void 1496 db_show_sx(const struct lock_object *lock) 1497 { 1498 struct thread *td; 1499 const struct sx *sx; 1500 1501 sx = (const struct sx *)lock; 1502 1503 db_printf(" state: "); 1504 if (sx->sx_lock == SX_LOCK_UNLOCKED) 1505 db_printf("UNLOCKED\n"); 1506 else if (sx->sx_lock == SX_LOCK_DESTROYED) { 1507 db_printf("DESTROYED\n"); 1508 return; 1509 } else if (sx->sx_lock & SX_LOCK_SHARED) 1510 db_printf("SLOCK: %ju\n", (uintmax_t)SX_SHARERS(sx->sx_lock)); 1511 else { 1512 td = sx_xholder(sx); 1513 db_printf("XLOCK: %p (tid %d, pid %d, \"%s\")\n", td, 1514 td->td_tid, td->td_proc->p_pid, td->td_name); 1515 if (sx_recursed(sx)) 1516 db_printf(" recursed: %d\n", sx->sx_recurse); 1517 } 1518 1519 db_printf(" waiters: "); 1520 switch(sx->sx_lock & 1521 (SX_LOCK_SHARED_WAITERS | SX_LOCK_EXCLUSIVE_WAITERS)) { 1522 case SX_LOCK_SHARED_WAITERS: 1523 db_printf("shared\n"); 1524 break; 1525 case SX_LOCK_EXCLUSIVE_WAITERS: 1526 db_printf("exclusive\n"); 1527 break; 1528 case SX_LOCK_SHARED_WAITERS | SX_LOCK_EXCLUSIVE_WAITERS: 1529 db_printf("exclusive and shared\n"); 1530 break; 1531 default: 1532 db_printf("none\n"); 1533 } 1534 } 1535 1536 /* 1537 * Check to see if a thread that is blocked on a sleep queue is actually 1538 * blocked on an sx lock. If so, output some details and return true. 1539 * If the lock has an exclusive owner, return that in *ownerp. 1540 */ 1541 int 1542 sx_chain(struct thread *td, struct thread **ownerp) 1543 { 1544 struct sx *sx; 1545 1546 /* 1547 * Check to see if this thread is blocked on an sx lock. 1548 * First, we check the lock class. If that is ok, then we 1549 * compare the lock name against the wait message. 1550 */ 1551 sx = td->td_wchan; 1552 if (LOCK_CLASS(&sx->lock_object) != &lock_class_sx || 1553 sx->lock_object.lo_name != td->td_wmesg) 1554 return (0); 1555 1556 /* We think we have an sx lock, so output some details. */ 1557 db_printf("blocked on sx \"%s\" ", td->td_wmesg); 1558 *ownerp = sx_xholder(sx); 1559 if (sx->sx_lock & SX_LOCK_SHARED) 1560 db_printf("SLOCK (count %ju)\n", 1561 (uintmax_t)SX_SHARERS(sx->sx_lock)); 1562 else 1563 db_printf("XLOCK\n"); 1564 return (1); 1565 } 1566 #endif 1567