1 /* SPDX-License-Identifier: GPL-2.0+ */ 2 /* 3 * Read-Copy Update mechanism for mutual exclusion (tree-based version) 4 * Internal non-public definitions that provide either classic 5 * or preemptible semantics. 6 * 7 * Copyright Red Hat, 2009 8 * Copyright IBM Corporation, 2009 9 * Copyright SUSE, 2021 10 * 11 * Author: Ingo Molnar <mingo@elte.hu> 12 * Paul E. McKenney <paulmck@linux.ibm.com> 13 * Frederic Weisbecker <frederic@kernel.org> 14 */ 15 16 #ifdef CONFIG_RCU_NOCB_CPU 17 static cpumask_var_t rcu_nocb_mask; /* CPUs to have callbacks offloaded. */ 18 static bool __read_mostly rcu_nocb_poll; /* Offload kthread are to poll. */ 19 static inline int rcu_lockdep_is_held_nocb(struct rcu_data *rdp) 20 { 21 return lockdep_is_held(&rdp->nocb_lock); 22 } 23 24 static inline bool rcu_current_is_nocb_kthread(struct rcu_data *rdp) 25 { 26 /* Race on early boot between thread creation and assignment */ 27 if (!rdp->nocb_cb_kthread || !rdp->nocb_gp_kthread) 28 return true; 29 30 if (current == rdp->nocb_cb_kthread || current == rdp->nocb_gp_kthread) 31 if (in_task()) 32 return true; 33 return false; 34 } 35 36 /* 37 * Offload callback processing from the boot-time-specified set of CPUs 38 * specified by rcu_nocb_mask. For the CPUs in the set, there are kthreads 39 * created that pull the callbacks from the corresponding CPU, wait for 40 * a grace period to elapse, and invoke the callbacks. These kthreads 41 * are organized into GP kthreads, which manage incoming callbacks, wait for 42 * grace periods, and awaken CB kthreads, and the CB kthreads, which only 43 * invoke callbacks. Each GP kthread invokes its own CBs. The no-CBs CPUs 44 * do a wake_up() on their GP kthread when they insert a callback into any 45 * empty list, unless the rcu_nocb_poll boot parameter has been specified, 46 * in which case each kthread actively polls its CPU. (Which isn't so great 47 * for energy efficiency, but which does reduce RCU's overhead on that CPU.) 48 * 49 * This is intended to be used in conjunction with Frederic Weisbecker's 50 * adaptive-idle work, which would seriously reduce OS jitter on CPUs 51 * running CPU-bound user-mode computations. 52 * 53 * Offloading of callbacks can also be used as an energy-efficiency 54 * measure because CPUs with no RCU callbacks queued are more aggressive 55 * about entering dyntick-idle mode. 56 */ 57 58 59 /* 60 * Parse the boot-time rcu_nocb_mask CPU list from the kernel parameters. 61 * If the list is invalid, a warning is emitted and all CPUs are offloaded. 62 */ 63 static int __init rcu_nocb_setup(char *str) 64 { 65 alloc_bootmem_cpumask_var(&rcu_nocb_mask); 66 if (*str == '=') { 67 if (cpulist_parse(++str, rcu_nocb_mask)) { 68 pr_warn("rcu_nocbs= bad CPU range, all CPUs set\n"); 69 cpumask_setall(rcu_nocb_mask); 70 } 71 } 72 rcu_state.nocb_is_setup = true; 73 return 1; 74 } 75 __setup("rcu_nocbs", rcu_nocb_setup); 76 77 static int __init parse_rcu_nocb_poll(char *arg) 78 { 79 rcu_nocb_poll = true; 80 return 0; 81 } 82 early_param("rcu_nocb_poll", parse_rcu_nocb_poll); 83 84 /* 85 * Don't bother bypassing ->cblist if the call_rcu() rate is low. 86 * After all, the main point of bypassing is to avoid lock contention 87 * on ->nocb_lock, which only can happen at high call_rcu() rates. 88 */ 89 static int nocb_nobypass_lim_per_jiffy = 16 * 1000 / HZ; 90 module_param(nocb_nobypass_lim_per_jiffy, int, 0); 91 92 /* 93 * Acquire the specified rcu_data structure's ->nocb_bypass_lock. If the 94 * lock isn't immediately available, increment ->nocb_lock_contended to 95 * flag the contention. 96 */ 97 static void rcu_nocb_bypass_lock(struct rcu_data *rdp) 98 __acquires(&rdp->nocb_bypass_lock) 99 { 100 lockdep_assert_irqs_disabled(); 101 if (raw_spin_trylock(&rdp->nocb_bypass_lock)) 102 return; 103 atomic_inc(&rdp->nocb_lock_contended); 104 WARN_ON_ONCE(smp_processor_id() != rdp->cpu); 105 smp_mb__after_atomic(); /* atomic_inc() before lock. */ 106 raw_spin_lock(&rdp->nocb_bypass_lock); 107 smp_mb__before_atomic(); /* atomic_dec() after lock. */ 108 atomic_dec(&rdp->nocb_lock_contended); 109 } 110 111 /* 112 * Spinwait until the specified rcu_data structure's ->nocb_lock is 113 * not contended. Please note that this is extremely special-purpose, 114 * relying on the fact that at most two kthreads and one CPU contend for 115 * this lock, and also that the two kthreads are guaranteed to have frequent 116 * grace-period-duration time intervals between successive acquisitions 117 * of the lock. This allows us to use an extremely simple throttling 118 * mechanism, and further to apply it only to the CPU doing floods of 119 * call_rcu() invocations. Don't try this at home! 120 */ 121 static void rcu_nocb_wait_contended(struct rcu_data *rdp) 122 { 123 WARN_ON_ONCE(smp_processor_id() != rdp->cpu); 124 while (WARN_ON_ONCE(atomic_read(&rdp->nocb_lock_contended))) 125 cpu_relax(); 126 } 127 128 /* 129 * Conditionally acquire the specified rcu_data structure's 130 * ->nocb_bypass_lock. 131 */ 132 static bool rcu_nocb_bypass_trylock(struct rcu_data *rdp) 133 { 134 lockdep_assert_irqs_disabled(); 135 return raw_spin_trylock(&rdp->nocb_bypass_lock); 136 } 137 138 /* 139 * Release the specified rcu_data structure's ->nocb_bypass_lock. 140 */ 141 static void rcu_nocb_bypass_unlock(struct rcu_data *rdp) 142 __releases(&rdp->nocb_bypass_lock) 143 { 144 lockdep_assert_irqs_disabled(); 145 raw_spin_unlock(&rdp->nocb_bypass_lock); 146 } 147 148 /* 149 * Acquire the specified rcu_data structure's ->nocb_lock, but only 150 * if it corresponds to a no-CBs CPU. 151 */ 152 static void rcu_nocb_lock(struct rcu_data *rdp) 153 { 154 lockdep_assert_irqs_disabled(); 155 if (!rcu_rdp_is_offloaded(rdp)) 156 return; 157 raw_spin_lock(&rdp->nocb_lock); 158 } 159 160 /* 161 * Release the specified rcu_data structure's ->nocb_lock, but only 162 * if it corresponds to a no-CBs CPU. 163 */ 164 static void rcu_nocb_unlock(struct rcu_data *rdp) 165 { 166 if (rcu_rdp_is_offloaded(rdp)) { 167 lockdep_assert_irqs_disabled(); 168 raw_spin_unlock(&rdp->nocb_lock); 169 } 170 } 171 172 /* 173 * Release the specified rcu_data structure's ->nocb_lock and restore 174 * interrupts, but only if it corresponds to a no-CBs CPU. 175 */ 176 static void rcu_nocb_unlock_irqrestore(struct rcu_data *rdp, 177 unsigned long flags) 178 { 179 if (rcu_rdp_is_offloaded(rdp)) { 180 lockdep_assert_irqs_disabled(); 181 raw_spin_unlock_irqrestore(&rdp->nocb_lock, flags); 182 } else { 183 local_irq_restore(flags); 184 } 185 } 186 187 /* Lockdep check that ->cblist may be safely accessed. */ 188 static void rcu_lockdep_assert_cblist_protected(struct rcu_data *rdp) 189 { 190 lockdep_assert_irqs_disabled(); 191 if (rcu_rdp_is_offloaded(rdp)) 192 lockdep_assert_held(&rdp->nocb_lock); 193 } 194 195 /* 196 * Wake up any no-CBs CPUs' kthreads that were waiting on the just-ended 197 * grace period. 198 */ 199 static void rcu_nocb_gp_cleanup(struct swait_queue_head *sq) 200 { 201 swake_up_all(sq); 202 } 203 204 static struct swait_queue_head *rcu_nocb_gp_get(struct rcu_node *rnp) 205 { 206 return &rnp->nocb_gp_wq[rcu_seq_ctr(rnp->gp_seq) & 0x1]; 207 } 208 209 static void rcu_init_one_nocb(struct rcu_node *rnp) 210 { 211 init_swait_queue_head(&rnp->nocb_gp_wq[0]); 212 init_swait_queue_head(&rnp->nocb_gp_wq[1]); 213 } 214 215 static bool __wake_nocb_gp(struct rcu_data *rdp_gp, 216 struct rcu_data *rdp, 217 bool force, unsigned long flags) 218 __releases(rdp_gp->nocb_gp_lock) 219 { 220 bool needwake = false; 221 222 if (!READ_ONCE(rdp_gp->nocb_gp_kthread)) { 223 raw_spin_unlock_irqrestore(&rdp_gp->nocb_gp_lock, flags); 224 trace_rcu_nocb_wake(rcu_state.name, rdp->cpu, 225 TPS("AlreadyAwake")); 226 return false; 227 } 228 229 if (rdp_gp->nocb_defer_wakeup > RCU_NOCB_WAKE_NOT) { 230 WRITE_ONCE(rdp_gp->nocb_defer_wakeup, RCU_NOCB_WAKE_NOT); 231 del_timer(&rdp_gp->nocb_timer); 232 } 233 234 if (force || READ_ONCE(rdp_gp->nocb_gp_sleep)) { 235 WRITE_ONCE(rdp_gp->nocb_gp_sleep, false); 236 needwake = true; 237 } 238 raw_spin_unlock_irqrestore(&rdp_gp->nocb_gp_lock, flags); 239 if (needwake) { 240 trace_rcu_nocb_wake(rcu_state.name, rdp->cpu, TPS("DoWake")); 241 wake_up_process(rdp_gp->nocb_gp_kthread); 242 } 243 244 return needwake; 245 } 246 247 /* 248 * Kick the GP kthread for this NOCB group. 249 */ 250 static bool wake_nocb_gp(struct rcu_data *rdp, bool force) 251 { 252 unsigned long flags; 253 struct rcu_data *rdp_gp = rdp->nocb_gp_rdp; 254 255 raw_spin_lock_irqsave(&rdp_gp->nocb_gp_lock, flags); 256 return __wake_nocb_gp(rdp_gp, rdp, force, flags); 257 } 258 259 /* 260 * Arrange to wake the GP kthread for this NOCB group at some future 261 * time when it is safe to do so. 262 */ 263 static void wake_nocb_gp_defer(struct rcu_data *rdp, int waketype, 264 const char *reason) 265 { 266 unsigned long flags; 267 struct rcu_data *rdp_gp = rdp->nocb_gp_rdp; 268 269 raw_spin_lock_irqsave(&rdp_gp->nocb_gp_lock, flags); 270 271 /* 272 * Bypass wakeup overrides previous deferments. In case 273 * of callback storm, no need to wake up too early. 274 */ 275 if (waketype == RCU_NOCB_WAKE_BYPASS) { 276 mod_timer(&rdp_gp->nocb_timer, jiffies + 2); 277 WRITE_ONCE(rdp_gp->nocb_defer_wakeup, waketype); 278 } else { 279 if (rdp_gp->nocb_defer_wakeup < RCU_NOCB_WAKE) 280 mod_timer(&rdp_gp->nocb_timer, jiffies + 1); 281 if (rdp_gp->nocb_defer_wakeup < waketype) 282 WRITE_ONCE(rdp_gp->nocb_defer_wakeup, waketype); 283 } 284 285 raw_spin_unlock_irqrestore(&rdp_gp->nocb_gp_lock, flags); 286 287 trace_rcu_nocb_wake(rcu_state.name, rdp->cpu, reason); 288 } 289 290 /* 291 * Flush the ->nocb_bypass queue into ->cblist, enqueuing rhp if non-NULL. 292 * However, if there is a callback to be enqueued and if ->nocb_bypass 293 * proves to be initially empty, just return false because the no-CB GP 294 * kthread may need to be awakened in this case. 295 * 296 * Note that this function always returns true if rhp is NULL. 297 */ 298 static bool rcu_nocb_do_flush_bypass(struct rcu_data *rdp, struct rcu_head *rhp, 299 unsigned long j) 300 { 301 struct rcu_cblist rcl; 302 303 WARN_ON_ONCE(!rcu_rdp_is_offloaded(rdp)); 304 rcu_lockdep_assert_cblist_protected(rdp); 305 lockdep_assert_held(&rdp->nocb_bypass_lock); 306 if (rhp && !rcu_cblist_n_cbs(&rdp->nocb_bypass)) { 307 raw_spin_unlock(&rdp->nocb_bypass_lock); 308 return false; 309 } 310 /* Note: ->cblist.len already accounts for ->nocb_bypass contents. */ 311 if (rhp) 312 rcu_segcblist_inc_len(&rdp->cblist); /* Must precede enqueue. */ 313 rcu_cblist_flush_enqueue(&rcl, &rdp->nocb_bypass, rhp); 314 rcu_segcblist_insert_pend_cbs(&rdp->cblist, &rcl); 315 WRITE_ONCE(rdp->nocb_bypass_first, j); 316 rcu_nocb_bypass_unlock(rdp); 317 return true; 318 } 319 320 /* 321 * Flush the ->nocb_bypass queue into ->cblist, enqueuing rhp if non-NULL. 322 * However, if there is a callback to be enqueued and if ->nocb_bypass 323 * proves to be initially empty, just return false because the no-CB GP 324 * kthread may need to be awakened in this case. 325 * 326 * Note that this function always returns true if rhp is NULL. 327 */ 328 static bool rcu_nocb_flush_bypass(struct rcu_data *rdp, struct rcu_head *rhp, 329 unsigned long j) 330 { 331 if (!rcu_rdp_is_offloaded(rdp)) 332 return true; 333 rcu_lockdep_assert_cblist_protected(rdp); 334 rcu_nocb_bypass_lock(rdp); 335 return rcu_nocb_do_flush_bypass(rdp, rhp, j); 336 } 337 338 /* 339 * If the ->nocb_bypass_lock is immediately available, flush the 340 * ->nocb_bypass queue into ->cblist. 341 */ 342 static void rcu_nocb_try_flush_bypass(struct rcu_data *rdp, unsigned long j) 343 { 344 rcu_lockdep_assert_cblist_protected(rdp); 345 if (!rcu_rdp_is_offloaded(rdp) || 346 !rcu_nocb_bypass_trylock(rdp)) 347 return; 348 WARN_ON_ONCE(!rcu_nocb_do_flush_bypass(rdp, NULL, j)); 349 } 350 351 /* 352 * See whether it is appropriate to use the ->nocb_bypass list in order 353 * to control contention on ->nocb_lock. A limited number of direct 354 * enqueues are permitted into ->cblist per jiffy. If ->nocb_bypass 355 * is non-empty, further callbacks must be placed into ->nocb_bypass, 356 * otherwise rcu_barrier() breaks. Use rcu_nocb_flush_bypass() to switch 357 * back to direct use of ->cblist. However, ->nocb_bypass should not be 358 * used if ->cblist is empty, because otherwise callbacks can be stranded 359 * on ->nocb_bypass because we cannot count on the current CPU ever again 360 * invoking call_rcu(). The general rule is that if ->nocb_bypass is 361 * non-empty, the corresponding no-CBs grace-period kthread must not be 362 * in an indefinite sleep state. 363 * 364 * Finally, it is not permitted to use the bypass during early boot, 365 * as doing so would confuse the auto-initialization code. Besides 366 * which, there is no point in worrying about lock contention while 367 * there is only one CPU in operation. 368 */ 369 static bool rcu_nocb_try_bypass(struct rcu_data *rdp, struct rcu_head *rhp, 370 bool *was_alldone, unsigned long flags) 371 { 372 unsigned long c; 373 unsigned long cur_gp_seq; 374 unsigned long j = jiffies; 375 long ncbs = rcu_cblist_n_cbs(&rdp->nocb_bypass); 376 377 lockdep_assert_irqs_disabled(); 378 379 // Pure softirq/rcuc based processing: no bypassing, no 380 // locking. 381 if (!rcu_rdp_is_offloaded(rdp)) { 382 *was_alldone = !rcu_segcblist_pend_cbs(&rdp->cblist); 383 return false; 384 } 385 386 // In the process of (de-)offloading: no bypassing, but 387 // locking. 388 if (!rcu_segcblist_completely_offloaded(&rdp->cblist)) { 389 rcu_nocb_lock(rdp); 390 *was_alldone = !rcu_segcblist_pend_cbs(&rdp->cblist); 391 return false; /* Not offloaded, no bypassing. */ 392 } 393 394 // Don't use ->nocb_bypass during early boot. 395 if (rcu_scheduler_active != RCU_SCHEDULER_RUNNING) { 396 rcu_nocb_lock(rdp); 397 WARN_ON_ONCE(rcu_cblist_n_cbs(&rdp->nocb_bypass)); 398 *was_alldone = !rcu_segcblist_pend_cbs(&rdp->cblist); 399 return false; 400 } 401 402 // If we have advanced to a new jiffy, reset counts to allow 403 // moving back from ->nocb_bypass to ->cblist. 404 if (j == rdp->nocb_nobypass_last) { 405 c = rdp->nocb_nobypass_count + 1; 406 } else { 407 WRITE_ONCE(rdp->nocb_nobypass_last, j); 408 c = rdp->nocb_nobypass_count - nocb_nobypass_lim_per_jiffy; 409 if (ULONG_CMP_LT(rdp->nocb_nobypass_count, 410 nocb_nobypass_lim_per_jiffy)) 411 c = 0; 412 else if (c > nocb_nobypass_lim_per_jiffy) 413 c = nocb_nobypass_lim_per_jiffy; 414 } 415 WRITE_ONCE(rdp->nocb_nobypass_count, c); 416 417 // If there hasn't yet been all that many ->cblist enqueues 418 // this jiffy, tell the caller to enqueue onto ->cblist. But flush 419 // ->nocb_bypass first. 420 if (rdp->nocb_nobypass_count < nocb_nobypass_lim_per_jiffy) { 421 rcu_nocb_lock(rdp); 422 *was_alldone = !rcu_segcblist_pend_cbs(&rdp->cblist); 423 if (*was_alldone) 424 trace_rcu_nocb_wake(rcu_state.name, rdp->cpu, 425 TPS("FirstQ")); 426 WARN_ON_ONCE(!rcu_nocb_flush_bypass(rdp, NULL, j)); 427 WARN_ON_ONCE(rcu_cblist_n_cbs(&rdp->nocb_bypass)); 428 return false; // Caller must enqueue the callback. 429 } 430 431 // If ->nocb_bypass has been used too long or is too full, 432 // flush ->nocb_bypass to ->cblist. 433 if ((ncbs && j != READ_ONCE(rdp->nocb_bypass_first)) || 434 ncbs >= qhimark) { 435 rcu_nocb_lock(rdp); 436 if (!rcu_nocb_flush_bypass(rdp, rhp, j)) { 437 *was_alldone = !rcu_segcblist_pend_cbs(&rdp->cblist); 438 if (*was_alldone) 439 trace_rcu_nocb_wake(rcu_state.name, rdp->cpu, 440 TPS("FirstQ")); 441 WARN_ON_ONCE(rcu_cblist_n_cbs(&rdp->nocb_bypass)); 442 return false; // Caller must enqueue the callback. 443 } 444 if (j != rdp->nocb_gp_adv_time && 445 rcu_segcblist_nextgp(&rdp->cblist, &cur_gp_seq) && 446 rcu_seq_done(&rdp->mynode->gp_seq, cur_gp_seq)) { 447 rcu_advance_cbs_nowake(rdp->mynode, rdp); 448 rdp->nocb_gp_adv_time = j; 449 } 450 rcu_nocb_unlock_irqrestore(rdp, flags); 451 return true; // Callback already enqueued. 452 } 453 454 // We need to use the bypass. 455 rcu_nocb_wait_contended(rdp); 456 rcu_nocb_bypass_lock(rdp); 457 ncbs = rcu_cblist_n_cbs(&rdp->nocb_bypass); 458 rcu_segcblist_inc_len(&rdp->cblist); /* Must precede enqueue. */ 459 rcu_cblist_enqueue(&rdp->nocb_bypass, rhp); 460 if (!ncbs) { 461 WRITE_ONCE(rdp->nocb_bypass_first, j); 462 trace_rcu_nocb_wake(rcu_state.name, rdp->cpu, TPS("FirstBQ")); 463 } 464 rcu_nocb_bypass_unlock(rdp); 465 smp_mb(); /* Order enqueue before wake. */ 466 if (ncbs) { 467 local_irq_restore(flags); 468 } else { 469 // No-CBs GP kthread might be indefinitely asleep, if so, wake. 470 rcu_nocb_lock(rdp); // Rare during call_rcu() flood. 471 if (!rcu_segcblist_pend_cbs(&rdp->cblist)) { 472 trace_rcu_nocb_wake(rcu_state.name, rdp->cpu, 473 TPS("FirstBQwake")); 474 __call_rcu_nocb_wake(rdp, true, flags); 475 } else { 476 trace_rcu_nocb_wake(rcu_state.name, rdp->cpu, 477 TPS("FirstBQnoWake")); 478 rcu_nocb_unlock_irqrestore(rdp, flags); 479 } 480 } 481 return true; // Callback already enqueued. 482 } 483 484 /* 485 * Awaken the no-CBs grace-period kthread if needed, either due to it 486 * legitimately being asleep or due to overload conditions. 487 * 488 * If warranted, also wake up the kthread servicing this CPUs queues. 489 */ 490 static void __call_rcu_nocb_wake(struct rcu_data *rdp, bool was_alldone, 491 unsigned long flags) 492 __releases(rdp->nocb_lock) 493 { 494 unsigned long cur_gp_seq; 495 unsigned long j; 496 long len; 497 struct task_struct *t; 498 499 // If we are being polled or there is no kthread, just leave. 500 t = READ_ONCE(rdp->nocb_gp_kthread); 501 if (rcu_nocb_poll || !t) { 502 rcu_nocb_unlock_irqrestore(rdp, flags); 503 trace_rcu_nocb_wake(rcu_state.name, rdp->cpu, 504 TPS("WakeNotPoll")); 505 return; 506 } 507 // Need to actually to a wakeup. 508 len = rcu_segcblist_n_cbs(&rdp->cblist); 509 if (was_alldone) { 510 rdp->qlen_last_fqs_check = len; 511 if (!irqs_disabled_flags(flags)) { 512 /* ... if queue was empty ... */ 513 rcu_nocb_unlock_irqrestore(rdp, flags); 514 wake_nocb_gp(rdp, false); 515 trace_rcu_nocb_wake(rcu_state.name, rdp->cpu, 516 TPS("WakeEmpty")); 517 } else { 518 rcu_nocb_unlock_irqrestore(rdp, flags); 519 wake_nocb_gp_defer(rdp, RCU_NOCB_WAKE, 520 TPS("WakeEmptyIsDeferred")); 521 } 522 } else if (len > rdp->qlen_last_fqs_check + qhimark) { 523 /* ... or if many callbacks queued. */ 524 rdp->qlen_last_fqs_check = len; 525 j = jiffies; 526 if (j != rdp->nocb_gp_adv_time && 527 rcu_segcblist_nextgp(&rdp->cblist, &cur_gp_seq) && 528 rcu_seq_done(&rdp->mynode->gp_seq, cur_gp_seq)) { 529 rcu_advance_cbs_nowake(rdp->mynode, rdp); 530 rdp->nocb_gp_adv_time = j; 531 } 532 smp_mb(); /* Enqueue before timer_pending(). */ 533 if ((rdp->nocb_cb_sleep || 534 !rcu_segcblist_ready_cbs(&rdp->cblist)) && 535 !timer_pending(&rdp->nocb_timer)) { 536 rcu_nocb_unlock_irqrestore(rdp, flags); 537 wake_nocb_gp_defer(rdp, RCU_NOCB_WAKE_FORCE, 538 TPS("WakeOvfIsDeferred")); 539 } else { 540 rcu_nocb_unlock_irqrestore(rdp, flags); 541 trace_rcu_nocb_wake(rcu_state.name, rdp->cpu, TPS("WakeNot")); 542 } 543 } else { 544 rcu_nocb_unlock_irqrestore(rdp, flags); 545 trace_rcu_nocb_wake(rcu_state.name, rdp->cpu, TPS("WakeNot")); 546 } 547 } 548 549 /* 550 * Check if we ignore this rdp. 551 * 552 * We check that without holding the nocb lock but 553 * we make sure not to miss a freshly offloaded rdp 554 * with the current ordering: 555 * 556 * rdp_offload_toggle() nocb_gp_enabled_cb() 557 * ------------------------- ---------------------------- 558 * WRITE flags LOCK nocb_gp_lock 559 * LOCK nocb_gp_lock READ/WRITE nocb_gp_sleep 560 * READ/WRITE nocb_gp_sleep UNLOCK nocb_gp_lock 561 * UNLOCK nocb_gp_lock READ flags 562 */ 563 static inline bool nocb_gp_enabled_cb(struct rcu_data *rdp) 564 { 565 u8 flags = SEGCBLIST_OFFLOADED | SEGCBLIST_KTHREAD_GP; 566 567 return rcu_segcblist_test_flags(&rdp->cblist, flags); 568 } 569 570 static inline bool nocb_gp_update_state_deoffloading(struct rcu_data *rdp, 571 bool *needwake_state) 572 { 573 struct rcu_segcblist *cblist = &rdp->cblist; 574 575 if (rcu_segcblist_test_flags(cblist, SEGCBLIST_OFFLOADED)) { 576 if (!rcu_segcblist_test_flags(cblist, SEGCBLIST_KTHREAD_GP)) { 577 rcu_segcblist_set_flags(cblist, SEGCBLIST_KTHREAD_GP); 578 if (rcu_segcblist_test_flags(cblist, SEGCBLIST_KTHREAD_CB)) 579 *needwake_state = true; 580 } 581 return false; 582 } 583 584 /* 585 * De-offloading. Clear our flag and notify the de-offload worker. 586 * We will ignore this rdp until it ever gets re-offloaded. 587 */ 588 WARN_ON_ONCE(!rcu_segcblist_test_flags(cblist, SEGCBLIST_KTHREAD_GP)); 589 rcu_segcblist_clear_flags(cblist, SEGCBLIST_KTHREAD_GP); 590 if (!rcu_segcblist_test_flags(cblist, SEGCBLIST_KTHREAD_CB)) 591 *needwake_state = true; 592 return true; 593 } 594 595 596 /* 597 * No-CBs GP kthreads come here to wait for additional callbacks to show up 598 * or for grace periods to end. 599 */ 600 static void nocb_gp_wait(struct rcu_data *my_rdp) 601 { 602 bool bypass = false; 603 long bypass_ncbs; 604 int __maybe_unused cpu = my_rdp->cpu; 605 unsigned long cur_gp_seq; 606 unsigned long flags; 607 bool gotcbs = false; 608 unsigned long j = jiffies; 609 bool needwait_gp = false; // This prevents actual uninitialized use. 610 bool needwake; 611 bool needwake_gp; 612 struct rcu_data *rdp; 613 struct rcu_node *rnp; 614 unsigned long wait_gp_seq = 0; // Suppress "use uninitialized" warning. 615 bool wasempty = false; 616 617 /* 618 * Each pass through the following loop checks for CBs and for the 619 * nearest grace period (if any) to wait for next. The CB kthreads 620 * and the global grace-period kthread are awakened if needed. 621 */ 622 WARN_ON_ONCE(my_rdp->nocb_gp_rdp != my_rdp); 623 /* 624 * An rcu_data structure is removed from the list after its 625 * CPU is de-offloaded and added to the list before that CPU is 626 * (re-)offloaded. If the following loop happens to be referencing 627 * that rcu_data structure during the time that the corresponding 628 * CPU is de-offloaded and then immediately re-offloaded, this 629 * loop's rdp pointer will be carried to the end of the list by 630 * the resulting pair of list operations. This can cause the loop 631 * to skip over some of the rcu_data structures that were supposed 632 * to have been scanned. Fortunately a new iteration through the 633 * entire loop is forced after a given CPU's rcu_data structure 634 * is added to the list, so the skipped-over rcu_data structures 635 * won't be ignored for long. 636 */ 637 list_for_each_entry_rcu(rdp, &my_rdp->nocb_head_rdp, nocb_entry_rdp, 1) { 638 bool needwake_state = false; 639 640 if (!nocb_gp_enabled_cb(rdp)) 641 continue; 642 trace_rcu_nocb_wake(rcu_state.name, rdp->cpu, TPS("Check")); 643 rcu_nocb_lock_irqsave(rdp, flags); 644 if (nocb_gp_update_state_deoffloading(rdp, &needwake_state)) { 645 rcu_nocb_unlock_irqrestore(rdp, flags); 646 if (needwake_state) 647 swake_up_one(&rdp->nocb_state_wq); 648 continue; 649 } 650 bypass_ncbs = rcu_cblist_n_cbs(&rdp->nocb_bypass); 651 if (bypass_ncbs && 652 (time_after(j, READ_ONCE(rdp->nocb_bypass_first) + 1) || 653 bypass_ncbs > 2 * qhimark)) { 654 // Bypass full or old, so flush it. 655 (void)rcu_nocb_try_flush_bypass(rdp, j); 656 bypass_ncbs = rcu_cblist_n_cbs(&rdp->nocb_bypass); 657 } else if (!bypass_ncbs && rcu_segcblist_empty(&rdp->cblist)) { 658 rcu_nocb_unlock_irqrestore(rdp, flags); 659 if (needwake_state) 660 swake_up_one(&rdp->nocb_state_wq); 661 continue; /* No callbacks here, try next. */ 662 } 663 if (bypass_ncbs) { 664 trace_rcu_nocb_wake(rcu_state.name, rdp->cpu, 665 TPS("Bypass")); 666 bypass = true; 667 } 668 rnp = rdp->mynode; 669 670 // Advance callbacks if helpful and low contention. 671 needwake_gp = false; 672 if (!rcu_segcblist_restempty(&rdp->cblist, 673 RCU_NEXT_READY_TAIL) || 674 (rcu_segcblist_nextgp(&rdp->cblist, &cur_gp_seq) && 675 rcu_seq_done(&rnp->gp_seq, cur_gp_seq))) { 676 raw_spin_lock_rcu_node(rnp); /* irqs disabled. */ 677 needwake_gp = rcu_advance_cbs(rnp, rdp); 678 wasempty = rcu_segcblist_restempty(&rdp->cblist, 679 RCU_NEXT_READY_TAIL); 680 raw_spin_unlock_rcu_node(rnp); /* irqs disabled. */ 681 } 682 // Need to wait on some grace period? 683 WARN_ON_ONCE(wasempty && 684 !rcu_segcblist_restempty(&rdp->cblist, 685 RCU_NEXT_READY_TAIL)); 686 if (rcu_segcblist_nextgp(&rdp->cblist, &cur_gp_seq)) { 687 if (!needwait_gp || 688 ULONG_CMP_LT(cur_gp_seq, wait_gp_seq)) 689 wait_gp_seq = cur_gp_seq; 690 needwait_gp = true; 691 trace_rcu_nocb_wake(rcu_state.name, rdp->cpu, 692 TPS("NeedWaitGP")); 693 } 694 if (rcu_segcblist_ready_cbs(&rdp->cblist)) { 695 needwake = rdp->nocb_cb_sleep; 696 WRITE_ONCE(rdp->nocb_cb_sleep, false); 697 smp_mb(); /* CB invocation -after- GP end. */ 698 } else { 699 needwake = false; 700 } 701 rcu_nocb_unlock_irqrestore(rdp, flags); 702 if (needwake) { 703 swake_up_one(&rdp->nocb_cb_wq); 704 gotcbs = true; 705 } 706 if (needwake_gp) 707 rcu_gp_kthread_wake(); 708 if (needwake_state) 709 swake_up_one(&rdp->nocb_state_wq); 710 } 711 712 my_rdp->nocb_gp_bypass = bypass; 713 my_rdp->nocb_gp_gp = needwait_gp; 714 my_rdp->nocb_gp_seq = needwait_gp ? wait_gp_seq : 0; 715 716 if (bypass && !rcu_nocb_poll) { 717 // At least one child with non-empty ->nocb_bypass, so set 718 // timer in order to avoid stranding its callbacks. 719 wake_nocb_gp_defer(my_rdp, RCU_NOCB_WAKE_BYPASS, 720 TPS("WakeBypassIsDeferred")); 721 } 722 if (rcu_nocb_poll) { 723 /* Polling, so trace if first poll in the series. */ 724 if (gotcbs) 725 trace_rcu_nocb_wake(rcu_state.name, cpu, TPS("Poll")); 726 schedule_timeout_idle(1); 727 } else if (!needwait_gp) { 728 /* Wait for callbacks to appear. */ 729 trace_rcu_nocb_wake(rcu_state.name, cpu, TPS("Sleep")); 730 swait_event_interruptible_exclusive(my_rdp->nocb_gp_wq, 731 !READ_ONCE(my_rdp->nocb_gp_sleep)); 732 trace_rcu_nocb_wake(rcu_state.name, cpu, TPS("EndSleep")); 733 } else { 734 rnp = my_rdp->mynode; 735 trace_rcu_this_gp(rnp, my_rdp, wait_gp_seq, TPS("StartWait")); 736 swait_event_interruptible_exclusive( 737 rnp->nocb_gp_wq[rcu_seq_ctr(wait_gp_seq) & 0x1], 738 rcu_seq_done(&rnp->gp_seq, wait_gp_seq) || 739 !READ_ONCE(my_rdp->nocb_gp_sleep)); 740 trace_rcu_this_gp(rnp, my_rdp, wait_gp_seq, TPS("EndWait")); 741 } 742 if (!rcu_nocb_poll) { 743 raw_spin_lock_irqsave(&my_rdp->nocb_gp_lock, flags); 744 if (my_rdp->nocb_defer_wakeup > RCU_NOCB_WAKE_NOT) { 745 WRITE_ONCE(my_rdp->nocb_defer_wakeup, RCU_NOCB_WAKE_NOT); 746 del_timer(&my_rdp->nocb_timer); 747 } 748 WRITE_ONCE(my_rdp->nocb_gp_sleep, true); 749 raw_spin_unlock_irqrestore(&my_rdp->nocb_gp_lock, flags); 750 } 751 my_rdp->nocb_gp_seq = -1; 752 WARN_ON(signal_pending(current)); 753 } 754 755 /* 756 * No-CBs grace-period-wait kthread. There is one of these per group 757 * of CPUs, but only once at least one CPU in that group has come online 758 * at least once since boot. This kthread checks for newly posted 759 * callbacks from any of the CPUs it is responsible for, waits for a 760 * grace period, then awakens all of the rcu_nocb_cb_kthread() instances 761 * that then have callback-invocation work to do. 762 */ 763 static int rcu_nocb_gp_kthread(void *arg) 764 { 765 struct rcu_data *rdp = arg; 766 767 for (;;) { 768 WRITE_ONCE(rdp->nocb_gp_loops, rdp->nocb_gp_loops + 1); 769 nocb_gp_wait(rdp); 770 cond_resched_tasks_rcu_qs(); 771 } 772 return 0; 773 } 774 775 static inline bool nocb_cb_can_run(struct rcu_data *rdp) 776 { 777 u8 flags = SEGCBLIST_OFFLOADED | SEGCBLIST_KTHREAD_CB; 778 779 return rcu_segcblist_test_flags(&rdp->cblist, flags); 780 } 781 782 static inline bool nocb_cb_wait_cond(struct rcu_data *rdp) 783 { 784 return nocb_cb_can_run(rdp) && !READ_ONCE(rdp->nocb_cb_sleep); 785 } 786 787 /* 788 * Invoke any ready callbacks from the corresponding no-CBs CPU, 789 * then, if there are no more, wait for more to appear. 790 */ 791 static void nocb_cb_wait(struct rcu_data *rdp) 792 { 793 struct rcu_segcblist *cblist = &rdp->cblist; 794 unsigned long cur_gp_seq; 795 unsigned long flags; 796 bool needwake_state = false; 797 bool needwake_gp = false; 798 bool can_sleep = true; 799 struct rcu_node *rnp = rdp->mynode; 800 801 do { 802 swait_event_interruptible_exclusive(rdp->nocb_cb_wq, 803 nocb_cb_wait_cond(rdp)); 804 805 // VVV Ensure CB invocation follows _sleep test. 806 if (smp_load_acquire(&rdp->nocb_cb_sleep)) { // ^^^ 807 WARN_ON(signal_pending(current)); 808 trace_rcu_nocb_wake(rcu_state.name, rdp->cpu, TPS("WokeEmpty")); 809 } 810 } while (!nocb_cb_can_run(rdp)); 811 812 813 local_irq_save(flags); 814 rcu_momentary_dyntick_idle(); 815 local_irq_restore(flags); 816 /* 817 * Disable BH to provide the expected environment. Also, when 818 * transitioning to/from NOCB mode, a self-requeuing callback might 819 * be invoked from softirq. A short grace period could cause both 820 * instances of this callback would execute concurrently. 821 */ 822 local_bh_disable(); 823 rcu_do_batch(rdp); 824 local_bh_enable(); 825 lockdep_assert_irqs_enabled(); 826 rcu_nocb_lock_irqsave(rdp, flags); 827 if (rcu_segcblist_nextgp(cblist, &cur_gp_seq) && 828 rcu_seq_done(&rnp->gp_seq, cur_gp_seq) && 829 raw_spin_trylock_rcu_node(rnp)) { /* irqs already disabled. */ 830 needwake_gp = rcu_advance_cbs(rdp->mynode, rdp); 831 raw_spin_unlock_rcu_node(rnp); /* irqs remain disabled. */ 832 } 833 834 if (rcu_segcblist_test_flags(cblist, SEGCBLIST_OFFLOADED)) { 835 if (!rcu_segcblist_test_flags(cblist, SEGCBLIST_KTHREAD_CB)) { 836 rcu_segcblist_set_flags(cblist, SEGCBLIST_KTHREAD_CB); 837 if (rcu_segcblist_test_flags(cblist, SEGCBLIST_KTHREAD_GP)) 838 needwake_state = true; 839 } 840 if (rcu_segcblist_ready_cbs(cblist)) 841 can_sleep = false; 842 } else { 843 /* 844 * De-offloading. Clear our flag and notify the de-offload worker. 845 * We won't touch the callbacks and keep sleeping until we ever 846 * get re-offloaded. 847 */ 848 WARN_ON_ONCE(!rcu_segcblist_test_flags(cblist, SEGCBLIST_KTHREAD_CB)); 849 rcu_segcblist_clear_flags(cblist, SEGCBLIST_KTHREAD_CB); 850 if (!rcu_segcblist_test_flags(cblist, SEGCBLIST_KTHREAD_GP)) 851 needwake_state = true; 852 } 853 854 WRITE_ONCE(rdp->nocb_cb_sleep, can_sleep); 855 856 if (rdp->nocb_cb_sleep) 857 trace_rcu_nocb_wake(rcu_state.name, rdp->cpu, TPS("CBSleep")); 858 859 rcu_nocb_unlock_irqrestore(rdp, flags); 860 if (needwake_gp) 861 rcu_gp_kthread_wake(); 862 863 if (needwake_state) 864 swake_up_one(&rdp->nocb_state_wq); 865 } 866 867 /* 868 * Per-rcu_data kthread, but only for no-CBs CPUs. Repeatedly invoke 869 * nocb_cb_wait() to do the dirty work. 870 */ 871 static int rcu_nocb_cb_kthread(void *arg) 872 { 873 struct rcu_data *rdp = arg; 874 875 // Each pass through this loop does one callback batch, and, 876 // if there are no more ready callbacks, waits for them. 877 for (;;) { 878 nocb_cb_wait(rdp); 879 cond_resched_tasks_rcu_qs(); 880 } 881 return 0; 882 } 883 884 /* Is a deferred wakeup of rcu_nocb_kthread() required? */ 885 static int rcu_nocb_need_deferred_wakeup(struct rcu_data *rdp, int level) 886 { 887 return READ_ONCE(rdp->nocb_defer_wakeup) >= level; 888 } 889 890 /* Do a deferred wakeup of rcu_nocb_kthread(). */ 891 static bool do_nocb_deferred_wakeup_common(struct rcu_data *rdp_gp, 892 struct rcu_data *rdp, int level, 893 unsigned long flags) 894 __releases(rdp_gp->nocb_gp_lock) 895 { 896 int ndw; 897 int ret; 898 899 if (!rcu_nocb_need_deferred_wakeup(rdp_gp, level)) { 900 raw_spin_unlock_irqrestore(&rdp_gp->nocb_gp_lock, flags); 901 return false; 902 } 903 904 ndw = rdp_gp->nocb_defer_wakeup; 905 ret = __wake_nocb_gp(rdp_gp, rdp, ndw == RCU_NOCB_WAKE_FORCE, flags); 906 trace_rcu_nocb_wake(rcu_state.name, rdp->cpu, TPS("DeferredWake")); 907 908 return ret; 909 } 910 911 /* Do a deferred wakeup of rcu_nocb_kthread() from a timer handler. */ 912 static void do_nocb_deferred_wakeup_timer(struct timer_list *t) 913 { 914 unsigned long flags; 915 struct rcu_data *rdp = from_timer(rdp, t, nocb_timer); 916 917 WARN_ON_ONCE(rdp->nocb_gp_rdp != rdp); 918 trace_rcu_nocb_wake(rcu_state.name, rdp->cpu, TPS("Timer")); 919 920 raw_spin_lock_irqsave(&rdp->nocb_gp_lock, flags); 921 smp_mb__after_spinlock(); /* Timer expire before wakeup. */ 922 do_nocb_deferred_wakeup_common(rdp, rdp, RCU_NOCB_WAKE_BYPASS, flags); 923 } 924 925 /* 926 * Do a deferred wakeup of rcu_nocb_kthread() from fastpath. 927 * This means we do an inexact common-case check. Note that if 928 * we miss, ->nocb_timer will eventually clean things up. 929 */ 930 static bool do_nocb_deferred_wakeup(struct rcu_data *rdp) 931 { 932 unsigned long flags; 933 struct rcu_data *rdp_gp = rdp->nocb_gp_rdp; 934 935 if (!rdp_gp || !rcu_nocb_need_deferred_wakeup(rdp_gp, RCU_NOCB_WAKE)) 936 return false; 937 938 raw_spin_lock_irqsave(&rdp_gp->nocb_gp_lock, flags); 939 return do_nocb_deferred_wakeup_common(rdp_gp, rdp, RCU_NOCB_WAKE, flags); 940 } 941 942 void rcu_nocb_flush_deferred_wakeup(void) 943 { 944 do_nocb_deferred_wakeup(this_cpu_ptr(&rcu_data)); 945 } 946 EXPORT_SYMBOL_GPL(rcu_nocb_flush_deferred_wakeup); 947 948 static int rdp_offload_toggle(struct rcu_data *rdp, 949 bool offload, unsigned long flags) 950 __releases(rdp->nocb_lock) 951 { 952 struct rcu_segcblist *cblist = &rdp->cblist; 953 struct rcu_data *rdp_gp = rdp->nocb_gp_rdp; 954 bool wake_gp = false; 955 956 rcu_segcblist_offload(cblist, offload); 957 958 if (rdp->nocb_cb_sleep) 959 rdp->nocb_cb_sleep = false; 960 rcu_nocb_unlock_irqrestore(rdp, flags); 961 962 /* 963 * Ignore former value of nocb_cb_sleep and force wake up as it could 964 * have been spuriously set to false already. 965 */ 966 swake_up_one(&rdp->nocb_cb_wq); 967 968 raw_spin_lock_irqsave(&rdp_gp->nocb_gp_lock, flags); 969 if (rdp_gp->nocb_gp_sleep) { 970 rdp_gp->nocb_gp_sleep = false; 971 wake_gp = true; 972 } 973 raw_spin_unlock_irqrestore(&rdp_gp->nocb_gp_lock, flags); 974 975 if (wake_gp) 976 wake_up_process(rdp_gp->nocb_gp_kthread); 977 978 return 0; 979 } 980 981 static long rcu_nocb_rdp_deoffload(void *arg) 982 { 983 struct rcu_data *rdp = arg; 984 struct rcu_segcblist *cblist = &rdp->cblist; 985 unsigned long flags; 986 int ret; 987 988 WARN_ON_ONCE(rdp->cpu != raw_smp_processor_id()); 989 990 pr_info("De-offloading %d\n", rdp->cpu); 991 992 rcu_nocb_lock_irqsave(rdp, flags); 993 /* 994 * Flush once and for all now. This suffices because we are 995 * running on the target CPU holding ->nocb_lock (thus having 996 * interrupts disabled), and because rdp_offload_toggle() 997 * invokes rcu_segcblist_offload(), which clears SEGCBLIST_OFFLOADED. 998 * Thus future calls to rcu_segcblist_completely_offloaded() will 999 * return false, which means that future calls to rcu_nocb_try_bypass() 1000 * will refuse to put anything into the bypass. 1001 */ 1002 WARN_ON_ONCE(!rcu_nocb_flush_bypass(rdp, NULL, jiffies)); 1003 /* 1004 * Start with invoking rcu_core() early. This way if the current thread 1005 * happens to preempt an ongoing call to rcu_core() in the middle, 1006 * leaving some work dismissed because rcu_core() still thinks the rdp is 1007 * completely offloaded, we are guaranteed a nearby future instance of 1008 * rcu_core() to catch up. 1009 */ 1010 rcu_segcblist_set_flags(cblist, SEGCBLIST_RCU_CORE); 1011 invoke_rcu_core(); 1012 ret = rdp_offload_toggle(rdp, false, flags); 1013 swait_event_exclusive(rdp->nocb_state_wq, 1014 !rcu_segcblist_test_flags(cblist, SEGCBLIST_KTHREAD_CB | 1015 SEGCBLIST_KTHREAD_GP)); 1016 /* Stop nocb_gp_wait() from iterating over this structure. */ 1017 list_del_rcu(&rdp->nocb_entry_rdp); 1018 /* 1019 * Lock one last time to acquire latest callback updates from kthreads 1020 * so we can later handle callbacks locally without locking. 1021 */ 1022 rcu_nocb_lock_irqsave(rdp, flags); 1023 /* 1024 * Theoretically we could clear SEGCBLIST_LOCKING after the nocb 1025 * lock is released but how about being paranoid for once? 1026 */ 1027 rcu_segcblist_clear_flags(cblist, SEGCBLIST_LOCKING); 1028 /* 1029 * Without SEGCBLIST_LOCKING, we can't use 1030 * rcu_nocb_unlock_irqrestore() anymore. 1031 */ 1032 raw_spin_unlock_irqrestore(&rdp->nocb_lock, flags); 1033 1034 /* Sanity check */ 1035 WARN_ON_ONCE(rcu_cblist_n_cbs(&rdp->nocb_bypass)); 1036 1037 1038 return ret; 1039 } 1040 1041 int rcu_nocb_cpu_deoffload(int cpu) 1042 { 1043 struct rcu_data *rdp = per_cpu_ptr(&rcu_data, cpu); 1044 int ret = 0; 1045 1046 mutex_lock(&rcu_state.barrier_mutex); 1047 cpus_read_lock(); 1048 if (rcu_rdp_is_offloaded(rdp)) { 1049 if (cpu_online(cpu)) { 1050 ret = work_on_cpu(cpu, rcu_nocb_rdp_deoffload, rdp); 1051 if (!ret) 1052 cpumask_clear_cpu(cpu, rcu_nocb_mask); 1053 } else { 1054 pr_info("NOCB: Can't CB-deoffload an offline CPU\n"); 1055 ret = -EINVAL; 1056 } 1057 } 1058 cpus_read_unlock(); 1059 mutex_unlock(&rcu_state.barrier_mutex); 1060 1061 return ret; 1062 } 1063 EXPORT_SYMBOL_GPL(rcu_nocb_cpu_deoffload); 1064 1065 static long rcu_nocb_rdp_offload(void *arg) 1066 { 1067 struct rcu_data *rdp = arg; 1068 struct rcu_segcblist *cblist = &rdp->cblist; 1069 unsigned long flags; 1070 int ret; 1071 1072 WARN_ON_ONCE(rdp->cpu != raw_smp_processor_id()); 1073 /* 1074 * For now we only support re-offload, ie: the rdp must have been 1075 * offloaded on boot first. 1076 */ 1077 if (!rdp->nocb_gp_rdp) 1078 return -EINVAL; 1079 1080 pr_info("Offloading %d\n", rdp->cpu); 1081 1082 /* 1083 * Cause future nocb_gp_wait() invocations to iterate over 1084 * structure, resetting ->nocb_gp_sleep and waking up the related 1085 * "rcuog". Since nocb_gp_wait() in turn locks ->nocb_gp_lock 1086 * before setting ->nocb_gp_sleep again, we are guaranteed to 1087 * iterate this newly added structure before "rcuog" goes to 1088 * sleep again. 1089 */ 1090 list_add_tail_rcu(&rdp->nocb_entry_rdp, &rdp->nocb_gp_rdp->nocb_head_rdp); 1091 1092 /* 1093 * Can't use rcu_nocb_lock_irqsave() before SEGCBLIST_LOCKING 1094 * is set. 1095 */ 1096 raw_spin_lock_irqsave(&rdp->nocb_lock, flags); 1097 1098 /* 1099 * We didn't take the nocb lock while working on the 1100 * rdp->cblist with SEGCBLIST_LOCKING cleared (pure softirq/rcuc mode). 1101 * Every modifications that have been done previously on 1102 * rdp->cblist must be visible remotely by the nocb kthreads 1103 * upon wake up after reading the cblist flags. 1104 * 1105 * The layout against nocb_lock enforces that ordering: 1106 * 1107 * __rcu_nocb_rdp_offload() nocb_cb_wait()/nocb_gp_wait() 1108 * ------------------------- ---------------------------- 1109 * WRITE callbacks rcu_nocb_lock() 1110 * rcu_nocb_lock() READ flags 1111 * WRITE flags READ callbacks 1112 * rcu_nocb_unlock() rcu_nocb_unlock() 1113 */ 1114 ret = rdp_offload_toggle(rdp, true, flags); 1115 swait_event_exclusive(rdp->nocb_state_wq, 1116 rcu_segcblist_test_flags(cblist, SEGCBLIST_KTHREAD_CB) && 1117 rcu_segcblist_test_flags(cblist, SEGCBLIST_KTHREAD_GP)); 1118 1119 /* 1120 * All kthreads are ready to work, we can finally relieve rcu_core() and 1121 * enable nocb bypass. 1122 */ 1123 rcu_nocb_lock_irqsave(rdp, flags); 1124 rcu_segcblist_clear_flags(cblist, SEGCBLIST_RCU_CORE); 1125 rcu_nocb_unlock_irqrestore(rdp, flags); 1126 1127 return ret; 1128 } 1129 1130 int rcu_nocb_cpu_offload(int cpu) 1131 { 1132 struct rcu_data *rdp = per_cpu_ptr(&rcu_data, cpu); 1133 int ret = 0; 1134 1135 mutex_lock(&rcu_state.barrier_mutex); 1136 cpus_read_lock(); 1137 if (!rcu_rdp_is_offloaded(rdp)) { 1138 if (cpu_online(cpu)) { 1139 ret = work_on_cpu(cpu, rcu_nocb_rdp_offload, rdp); 1140 if (!ret) 1141 cpumask_set_cpu(cpu, rcu_nocb_mask); 1142 } else { 1143 pr_info("NOCB: Can't CB-offload an offline CPU\n"); 1144 ret = -EINVAL; 1145 } 1146 } 1147 cpus_read_unlock(); 1148 mutex_unlock(&rcu_state.barrier_mutex); 1149 1150 return ret; 1151 } 1152 EXPORT_SYMBOL_GPL(rcu_nocb_cpu_offload); 1153 1154 void __init rcu_init_nohz(void) 1155 { 1156 int cpu; 1157 bool need_rcu_nocb_mask = false; 1158 struct rcu_data *rdp; 1159 1160 #if defined(CONFIG_NO_HZ_FULL) 1161 if (tick_nohz_full_running && !cpumask_empty(tick_nohz_full_mask)) 1162 need_rcu_nocb_mask = true; 1163 #endif /* #if defined(CONFIG_NO_HZ_FULL) */ 1164 1165 if (need_rcu_nocb_mask) { 1166 if (!cpumask_available(rcu_nocb_mask)) { 1167 if (!zalloc_cpumask_var(&rcu_nocb_mask, GFP_KERNEL)) { 1168 pr_info("rcu_nocb_mask allocation failed, callback offloading disabled.\n"); 1169 return; 1170 } 1171 } 1172 rcu_state.nocb_is_setup = true; 1173 } 1174 1175 if (!rcu_state.nocb_is_setup) 1176 return; 1177 1178 #if defined(CONFIG_NO_HZ_FULL) 1179 if (tick_nohz_full_running) 1180 cpumask_or(rcu_nocb_mask, rcu_nocb_mask, tick_nohz_full_mask); 1181 #endif /* #if defined(CONFIG_NO_HZ_FULL) */ 1182 1183 if (!cpumask_subset(rcu_nocb_mask, cpu_possible_mask)) { 1184 pr_info("\tNote: kernel parameter 'rcu_nocbs=', 'nohz_full', or 'isolcpus=' contains nonexistent CPUs.\n"); 1185 cpumask_and(rcu_nocb_mask, cpu_possible_mask, 1186 rcu_nocb_mask); 1187 } 1188 if (cpumask_empty(rcu_nocb_mask)) 1189 pr_info("\tOffload RCU callbacks from CPUs: (none).\n"); 1190 else 1191 pr_info("\tOffload RCU callbacks from CPUs: %*pbl.\n", 1192 cpumask_pr_args(rcu_nocb_mask)); 1193 if (rcu_nocb_poll) 1194 pr_info("\tPoll for callbacks from no-CBs CPUs.\n"); 1195 1196 for_each_cpu(cpu, rcu_nocb_mask) { 1197 rdp = per_cpu_ptr(&rcu_data, cpu); 1198 if (rcu_segcblist_empty(&rdp->cblist)) 1199 rcu_segcblist_init(&rdp->cblist); 1200 rcu_segcblist_offload(&rdp->cblist, true); 1201 rcu_segcblist_set_flags(&rdp->cblist, SEGCBLIST_KTHREAD_CB | SEGCBLIST_KTHREAD_GP); 1202 rcu_segcblist_clear_flags(&rdp->cblist, SEGCBLIST_RCU_CORE); 1203 } 1204 rcu_organize_nocb_kthreads(); 1205 } 1206 1207 /* Initialize per-rcu_data variables for no-CBs CPUs. */ 1208 static void __init rcu_boot_init_nocb_percpu_data(struct rcu_data *rdp) 1209 { 1210 init_swait_queue_head(&rdp->nocb_cb_wq); 1211 init_swait_queue_head(&rdp->nocb_gp_wq); 1212 init_swait_queue_head(&rdp->nocb_state_wq); 1213 raw_spin_lock_init(&rdp->nocb_lock); 1214 raw_spin_lock_init(&rdp->nocb_bypass_lock); 1215 raw_spin_lock_init(&rdp->nocb_gp_lock); 1216 timer_setup(&rdp->nocb_timer, do_nocb_deferred_wakeup_timer, 0); 1217 rcu_cblist_init(&rdp->nocb_bypass); 1218 mutex_init(&rdp->nocb_gp_kthread_mutex); 1219 } 1220 1221 /* 1222 * If the specified CPU is a no-CBs CPU that does not already have its 1223 * rcuo CB kthread, spawn it. Additionally, if the rcuo GP kthread 1224 * for this CPU's group has not yet been created, spawn it as well. 1225 */ 1226 static void rcu_spawn_cpu_nocb_kthread(int cpu) 1227 { 1228 struct rcu_data *rdp = per_cpu_ptr(&rcu_data, cpu); 1229 struct rcu_data *rdp_gp; 1230 struct task_struct *t; 1231 struct sched_param sp; 1232 1233 if (!rcu_scheduler_fully_active || !rcu_state.nocb_is_setup) 1234 return; 1235 1236 /* If there already is an rcuo kthread, then nothing to do. */ 1237 if (rdp->nocb_cb_kthread) 1238 return; 1239 1240 /* If we didn't spawn the GP kthread first, reorganize! */ 1241 sp.sched_priority = kthread_prio; 1242 rdp_gp = rdp->nocb_gp_rdp; 1243 mutex_lock(&rdp_gp->nocb_gp_kthread_mutex); 1244 if (!rdp_gp->nocb_gp_kthread) { 1245 t = kthread_run(rcu_nocb_gp_kthread, rdp_gp, 1246 "rcuog/%d", rdp_gp->cpu); 1247 if (WARN_ONCE(IS_ERR(t), "%s: Could not start rcuo GP kthread, OOM is now expected behavior\n", __func__)) { 1248 mutex_unlock(&rdp_gp->nocb_gp_kthread_mutex); 1249 return; 1250 } 1251 WRITE_ONCE(rdp_gp->nocb_gp_kthread, t); 1252 if (kthread_prio) 1253 sched_setscheduler_nocheck(t, SCHED_FIFO, &sp); 1254 } 1255 mutex_unlock(&rdp_gp->nocb_gp_kthread_mutex); 1256 1257 /* Spawn the kthread for this CPU. */ 1258 t = kthread_run(rcu_nocb_cb_kthread, rdp, 1259 "rcuo%c/%d", rcu_state.abbr, cpu); 1260 if (WARN_ONCE(IS_ERR(t), "%s: Could not start rcuo CB kthread, OOM is now expected behavior\n", __func__)) 1261 return; 1262 1263 if (kthread_prio) 1264 sched_setscheduler_nocheck(t, SCHED_FIFO, &sp); 1265 WRITE_ONCE(rdp->nocb_cb_kthread, t); 1266 WRITE_ONCE(rdp->nocb_gp_kthread, rdp_gp->nocb_gp_kthread); 1267 } 1268 1269 /* How many CB CPU IDs per GP kthread? Default of -1 for sqrt(nr_cpu_ids). */ 1270 static int rcu_nocb_gp_stride = -1; 1271 module_param(rcu_nocb_gp_stride, int, 0444); 1272 1273 /* 1274 * Initialize GP-CB relationships for all no-CBs CPU. 1275 */ 1276 static void __init rcu_organize_nocb_kthreads(void) 1277 { 1278 int cpu; 1279 bool firsttime = true; 1280 bool gotnocbs = false; 1281 bool gotnocbscbs = true; 1282 int ls = rcu_nocb_gp_stride; 1283 int nl = 0; /* Next GP kthread. */ 1284 struct rcu_data *rdp; 1285 struct rcu_data *rdp_gp = NULL; /* Suppress misguided gcc warn. */ 1286 1287 if (!cpumask_available(rcu_nocb_mask)) 1288 return; 1289 if (ls == -1) { 1290 ls = nr_cpu_ids / int_sqrt(nr_cpu_ids); 1291 rcu_nocb_gp_stride = ls; 1292 } 1293 1294 /* 1295 * Each pass through this loop sets up one rcu_data structure. 1296 * Should the corresponding CPU come online in the future, then 1297 * we will spawn the needed set of rcu_nocb_kthread() kthreads. 1298 */ 1299 for_each_possible_cpu(cpu) { 1300 rdp = per_cpu_ptr(&rcu_data, cpu); 1301 if (rdp->cpu >= nl) { 1302 /* New GP kthread, set up for CBs & next GP. */ 1303 gotnocbs = true; 1304 nl = DIV_ROUND_UP(rdp->cpu + 1, ls) * ls; 1305 rdp_gp = rdp; 1306 INIT_LIST_HEAD(&rdp->nocb_head_rdp); 1307 if (dump_tree) { 1308 if (!firsttime) 1309 pr_cont("%s\n", gotnocbscbs 1310 ? "" : " (self only)"); 1311 gotnocbscbs = false; 1312 firsttime = false; 1313 pr_alert("%s: No-CB GP kthread CPU %d:", 1314 __func__, cpu); 1315 } 1316 } else { 1317 /* Another CB kthread, link to previous GP kthread. */ 1318 gotnocbscbs = true; 1319 if (dump_tree) 1320 pr_cont(" %d", cpu); 1321 } 1322 rdp->nocb_gp_rdp = rdp_gp; 1323 if (cpumask_test_cpu(cpu, rcu_nocb_mask)) 1324 list_add_tail(&rdp->nocb_entry_rdp, &rdp_gp->nocb_head_rdp); 1325 } 1326 if (gotnocbs && dump_tree) 1327 pr_cont("%s\n", gotnocbscbs ? "" : " (self only)"); 1328 } 1329 1330 /* 1331 * Bind the current task to the offloaded CPUs. If there are no offloaded 1332 * CPUs, leave the task unbound. Splat if the bind attempt fails. 1333 */ 1334 void rcu_bind_current_to_nocb(void) 1335 { 1336 if (cpumask_available(rcu_nocb_mask) && !cpumask_empty(rcu_nocb_mask)) 1337 WARN_ON(sched_setaffinity(current->pid, rcu_nocb_mask)); 1338 } 1339 EXPORT_SYMBOL_GPL(rcu_bind_current_to_nocb); 1340 1341 // The ->on_cpu field is available only in CONFIG_SMP=y, so... 1342 #ifdef CONFIG_SMP 1343 static char *show_rcu_should_be_on_cpu(struct task_struct *tsp) 1344 { 1345 return tsp && task_is_running(tsp) && !tsp->on_cpu ? "!" : ""; 1346 } 1347 #else // #ifdef CONFIG_SMP 1348 static char *show_rcu_should_be_on_cpu(struct task_struct *tsp) 1349 { 1350 return ""; 1351 } 1352 #endif // #else #ifdef CONFIG_SMP 1353 1354 /* 1355 * Dump out nocb grace-period kthread state for the specified rcu_data 1356 * structure. 1357 */ 1358 static void show_rcu_nocb_gp_state(struct rcu_data *rdp) 1359 { 1360 struct rcu_node *rnp = rdp->mynode; 1361 1362 pr_info("nocb GP %d %c%c%c%c%c %c[%c%c] %c%c:%ld rnp %d:%d %lu %c CPU %d%s\n", 1363 rdp->cpu, 1364 "kK"[!!rdp->nocb_gp_kthread], 1365 "lL"[raw_spin_is_locked(&rdp->nocb_gp_lock)], 1366 "dD"[!!rdp->nocb_defer_wakeup], 1367 "tT"[timer_pending(&rdp->nocb_timer)], 1368 "sS"[!!rdp->nocb_gp_sleep], 1369 ".W"[swait_active(&rdp->nocb_gp_wq)], 1370 ".W"[swait_active(&rnp->nocb_gp_wq[0])], 1371 ".W"[swait_active(&rnp->nocb_gp_wq[1])], 1372 ".B"[!!rdp->nocb_gp_bypass], 1373 ".G"[!!rdp->nocb_gp_gp], 1374 (long)rdp->nocb_gp_seq, 1375 rnp->grplo, rnp->grphi, READ_ONCE(rdp->nocb_gp_loops), 1376 rdp->nocb_gp_kthread ? task_state_to_char(rdp->nocb_gp_kthread) : '.', 1377 rdp->nocb_cb_kthread ? (int)task_cpu(rdp->nocb_gp_kthread) : -1, 1378 show_rcu_should_be_on_cpu(rdp->nocb_cb_kthread)); 1379 } 1380 1381 /* Dump out nocb kthread state for the specified rcu_data structure. */ 1382 static void show_rcu_nocb_state(struct rcu_data *rdp) 1383 { 1384 char bufw[20]; 1385 char bufr[20]; 1386 struct rcu_data *nocb_next_rdp; 1387 struct rcu_segcblist *rsclp = &rdp->cblist; 1388 bool waslocked; 1389 bool wassleep; 1390 1391 if (rdp->nocb_gp_rdp == rdp) 1392 show_rcu_nocb_gp_state(rdp); 1393 1394 nocb_next_rdp = list_next_or_null_rcu(&rdp->nocb_gp_rdp->nocb_head_rdp, 1395 &rdp->nocb_entry_rdp, 1396 typeof(*rdp), 1397 nocb_entry_rdp); 1398 1399 sprintf(bufw, "%ld", rsclp->gp_seq[RCU_WAIT_TAIL]); 1400 sprintf(bufr, "%ld", rsclp->gp_seq[RCU_NEXT_READY_TAIL]); 1401 pr_info(" CB %d^%d->%d %c%c%c%c%c%c F%ld L%ld C%d %c%c%s%c%s%c%c q%ld %c CPU %d%s\n", 1402 rdp->cpu, rdp->nocb_gp_rdp->cpu, 1403 nocb_next_rdp ? nocb_next_rdp->cpu : -1, 1404 "kK"[!!rdp->nocb_cb_kthread], 1405 "bB"[raw_spin_is_locked(&rdp->nocb_bypass_lock)], 1406 "cC"[!!atomic_read(&rdp->nocb_lock_contended)], 1407 "lL"[raw_spin_is_locked(&rdp->nocb_lock)], 1408 "sS"[!!rdp->nocb_cb_sleep], 1409 ".W"[swait_active(&rdp->nocb_cb_wq)], 1410 jiffies - rdp->nocb_bypass_first, 1411 jiffies - rdp->nocb_nobypass_last, 1412 rdp->nocb_nobypass_count, 1413 ".D"[rcu_segcblist_ready_cbs(rsclp)], 1414 ".W"[!rcu_segcblist_segempty(rsclp, RCU_WAIT_TAIL)], 1415 rcu_segcblist_segempty(rsclp, RCU_WAIT_TAIL) ? "" : bufw, 1416 ".R"[!rcu_segcblist_segempty(rsclp, RCU_NEXT_READY_TAIL)], 1417 rcu_segcblist_segempty(rsclp, RCU_NEXT_READY_TAIL) ? "" : bufr, 1418 ".N"[!rcu_segcblist_segempty(rsclp, RCU_NEXT_TAIL)], 1419 ".B"[!!rcu_cblist_n_cbs(&rdp->nocb_bypass)], 1420 rcu_segcblist_n_cbs(&rdp->cblist), 1421 rdp->nocb_cb_kthread ? task_state_to_char(rdp->nocb_cb_kthread) : '.', 1422 rdp->nocb_cb_kthread ? (int)task_cpu(rdp->nocb_gp_kthread) : -1, 1423 show_rcu_should_be_on_cpu(rdp->nocb_cb_kthread)); 1424 1425 /* It is OK for GP kthreads to have GP state. */ 1426 if (rdp->nocb_gp_rdp == rdp) 1427 return; 1428 1429 waslocked = raw_spin_is_locked(&rdp->nocb_gp_lock); 1430 wassleep = swait_active(&rdp->nocb_gp_wq); 1431 if (!rdp->nocb_gp_sleep && !waslocked && !wassleep) 1432 return; /* Nothing untoward. */ 1433 1434 pr_info(" nocb GP activity on CB-only CPU!!! %c%c%c %c\n", 1435 "lL"[waslocked], 1436 "dD"[!!rdp->nocb_defer_wakeup], 1437 "sS"[!!rdp->nocb_gp_sleep], 1438 ".W"[wassleep]); 1439 } 1440 1441 #else /* #ifdef CONFIG_RCU_NOCB_CPU */ 1442 1443 static inline int rcu_lockdep_is_held_nocb(struct rcu_data *rdp) 1444 { 1445 return 0; 1446 } 1447 1448 static inline bool rcu_current_is_nocb_kthread(struct rcu_data *rdp) 1449 { 1450 return false; 1451 } 1452 1453 /* No ->nocb_lock to acquire. */ 1454 static void rcu_nocb_lock(struct rcu_data *rdp) 1455 { 1456 } 1457 1458 /* No ->nocb_lock to release. */ 1459 static void rcu_nocb_unlock(struct rcu_data *rdp) 1460 { 1461 } 1462 1463 /* No ->nocb_lock to release. */ 1464 static void rcu_nocb_unlock_irqrestore(struct rcu_data *rdp, 1465 unsigned long flags) 1466 { 1467 local_irq_restore(flags); 1468 } 1469 1470 /* Lockdep check that ->cblist may be safely accessed. */ 1471 static void rcu_lockdep_assert_cblist_protected(struct rcu_data *rdp) 1472 { 1473 lockdep_assert_irqs_disabled(); 1474 } 1475 1476 static void rcu_nocb_gp_cleanup(struct swait_queue_head *sq) 1477 { 1478 } 1479 1480 static struct swait_queue_head *rcu_nocb_gp_get(struct rcu_node *rnp) 1481 { 1482 return NULL; 1483 } 1484 1485 static void rcu_init_one_nocb(struct rcu_node *rnp) 1486 { 1487 } 1488 1489 static bool rcu_nocb_flush_bypass(struct rcu_data *rdp, struct rcu_head *rhp, 1490 unsigned long j) 1491 { 1492 return true; 1493 } 1494 1495 static bool rcu_nocb_try_bypass(struct rcu_data *rdp, struct rcu_head *rhp, 1496 bool *was_alldone, unsigned long flags) 1497 { 1498 return false; 1499 } 1500 1501 static void __call_rcu_nocb_wake(struct rcu_data *rdp, bool was_empty, 1502 unsigned long flags) 1503 { 1504 WARN_ON_ONCE(1); /* Should be dead code! */ 1505 } 1506 1507 static void __init rcu_boot_init_nocb_percpu_data(struct rcu_data *rdp) 1508 { 1509 } 1510 1511 static int rcu_nocb_need_deferred_wakeup(struct rcu_data *rdp, int level) 1512 { 1513 return false; 1514 } 1515 1516 static bool do_nocb_deferred_wakeup(struct rcu_data *rdp) 1517 { 1518 return false; 1519 } 1520 1521 static void rcu_spawn_cpu_nocb_kthread(int cpu) 1522 { 1523 } 1524 1525 static void show_rcu_nocb_state(struct rcu_data *rdp) 1526 { 1527 } 1528 1529 #endif /* #else #ifdef CONFIG_RCU_NOCB_CPU */ 1530