1 /*- 2 * SPDX-License-Identifier: BSD-2-Clause-FreeBSD 3 * 4 * Copyright (c) 2002-2007, Jeffrey Roberson <jeff@freebsd.org> 5 * All rights reserved. 6 * 7 * Redistribution and use in source and binary forms, with or without 8 * modification, are permitted provided that the following conditions 9 * are met: 10 * 1. Redistributions of source code must retain the above copyright 11 * notice unmodified, this list of conditions, and the following 12 * disclaimer. 13 * 2. Redistributions in binary form must reproduce the above copyright 14 * notice, this list of conditions and the following disclaimer in the 15 * documentation and/or other materials provided with the distribution. 16 * 17 * THIS SOFTWARE IS PROVIDED BY THE AUTHOR ``AS IS'' AND ANY EXPRESS OR 18 * IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES 19 * OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED. 20 * IN NO EVENT SHALL THE AUTHOR BE LIABLE FOR ANY DIRECT, INDIRECT, 21 * INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT 22 * NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, 23 * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY 24 * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT 25 * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF 26 * THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. 27 */ 28 29 /* 30 * This file implements the ULE scheduler. ULE supports independent CPU 31 * run queues and fine grain locking. It has superior interactive 32 * performance under load even on uni-processor systems. 33 * 34 * etymology: 35 * ULE is the last three letters in schedule. It owes its name to a 36 * generic user created for a scheduling system by Paul Mikesell at 37 * Isilon Systems and a general lack of creativity on the part of the author. 38 */ 39 40 #include <sys/cdefs.h> 41 __FBSDID("$FreeBSD$"); 42 43 #include "opt_hwpmc_hooks.h" 44 #include "opt_sched.h" 45 46 #include <sys/param.h> 47 #include <sys/systm.h> 48 #include <sys/kdb.h> 49 #include <sys/kernel.h> 50 #include <sys/ktr.h> 51 #include <sys/limits.h> 52 #include <sys/lock.h> 53 #include <sys/mutex.h> 54 #include <sys/proc.h> 55 #include <sys/resource.h> 56 #include <sys/resourcevar.h> 57 #include <sys/sched.h> 58 #include <sys/sdt.h> 59 #include <sys/smp.h> 60 #include <sys/sx.h> 61 #include <sys/sysctl.h> 62 #include <sys/sysproto.h> 63 #include <sys/turnstile.h> 64 #include <sys/umtx.h> 65 #include <sys/vmmeter.h> 66 #include <sys/cpuset.h> 67 #include <sys/sbuf.h> 68 69 #ifdef HWPMC_HOOKS 70 #include <sys/pmckern.h> 71 #endif 72 73 #ifdef KDTRACE_HOOKS 74 #include <sys/dtrace_bsd.h> 75 int __read_mostly dtrace_vtime_active; 76 dtrace_vtime_switch_func_t dtrace_vtime_switch_func; 77 #endif 78 79 #include <machine/cpu.h> 80 #include <machine/smp.h> 81 82 #define KTR_ULE 0 83 84 #define TS_NAME_LEN (MAXCOMLEN + sizeof(" td ") + sizeof(__XSTRING(UINT_MAX))) 85 #define TDQ_NAME_LEN (sizeof("sched lock ") + sizeof(__XSTRING(MAXCPU))) 86 #define TDQ_LOADNAME_LEN (sizeof("CPU ") + sizeof(__XSTRING(MAXCPU)) - 1 + sizeof(" load")) 87 88 /* 89 * Thread scheduler specific section. All fields are protected 90 * by the thread lock. 91 */ 92 struct td_sched { 93 struct runq *ts_runq; /* Run-queue we're queued on. */ 94 short ts_flags; /* TSF_* flags. */ 95 int ts_cpu; /* CPU that we have affinity for. */ 96 int ts_rltick; /* Real last tick, for affinity. */ 97 int ts_slice; /* Ticks of slice remaining. */ 98 u_int ts_slptime; /* Number of ticks we vol. slept */ 99 u_int ts_runtime; /* Number of ticks we were running */ 100 int ts_ltick; /* Last tick that we were running on */ 101 int ts_ftick; /* First tick that we were running on */ 102 int ts_ticks; /* Tick count */ 103 #ifdef KTR 104 char ts_name[TS_NAME_LEN]; 105 #endif 106 }; 107 /* flags kept in ts_flags */ 108 #define TSF_BOUND 0x0001 /* Thread can not migrate. */ 109 #define TSF_XFERABLE 0x0002 /* Thread was added as transferable. */ 110 111 #define THREAD_CAN_MIGRATE(td) ((td)->td_pinned == 0) 112 #define THREAD_CAN_SCHED(td, cpu) \ 113 CPU_ISSET((cpu), &(td)->td_cpuset->cs_mask) 114 115 _Static_assert(sizeof(struct thread) + sizeof(struct td_sched) <= 116 sizeof(struct thread0_storage), 117 "increase struct thread0_storage.t0st_sched size"); 118 119 /* 120 * Priority ranges used for interactive and non-interactive timeshare 121 * threads. The timeshare priorities are split up into four ranges. 122 * The first range handles interactive threads. The last three ranges 123 * (NHALF, x, and NHALF) handle non-interactive threads with the outer 124 * ranges supporting nice values. 125 */ 126 #define PRI_TIMESHARE_RANGE (PRI_MAX_TIMESHARE - PRI_MIN_TIMESHARE + 1) 127 #define PRI_INTERACT_RANGE ((PRI_TIMESHARE_RANGE - SCHED_PRI_NRESV) / 2) 128 #define PRI_BATCH_RANGE (PRI_TIMESHARE_RANGE - PRI_INTERACT_RANGE) 129 130 #define PRI_MIN_INTERACT PRI_MIN_TIMESHARE 131 #define PRI_MAX_INTERACT (PRI_MIN_TIMESHARE + PRI_INTERACT_RANGE - 1) 132 #define PRI_MIN_BATCH (PRI_MIN_TIMESHARE + PRI_INTERACT_RANGE) 133 #define PRI_MAX_BATCH PRI_MAX_TIMESHARE 134 135 /* 136 * Cpu percentage computation macros and defines. 137 * 138 * SCHED_TICK_SECS: Number of seconds to average the cpu usage across. 139 * SCHED_TICK_TARG: Number of hz ticks to average the cpu usage across. 140 * SCHED_TICK_MAX: Maximum number of ticks before scaling back. 141 * SCHED_TICK_SHIFT: Shift factor to avoid rounding away results. 142 * SCHED_TICK_HZ: Compute the number of hz ticks for a given ticks count. 143 * SCHED_TICK_TOTAL: Gives the amount of time we've been recording ticks. 144 */ 145 #define SCHED_TICK_SECS 10 146 #define SCHED_TICK_TARG (hz * SCHED_TICK_SECS) 147 #define SCHED_TICK_MAX (SCHED_TICK_TARG + hz) 148 #define SCHED_TICK_SHIFT 10 149 #define SCHED_TICK_HZ(ts) ((ts)->ts_ticks >> SCHED_TICK_SHIFT) 150 #define SCHED_TICK_TOTAL(ts) (max((ts)->ts_ltick - (ts)->ts_ftick, hz)) 151 152 /* 153 * These macros determine priorities for non-interactive threads. They are 154 * assigned a priority based on their recent cpu utilization as expressed 155 * by the ratio of ticks to the tick total. NHALF priorities at the start 156 * and end of the MIN to MAX timeshare range are only reachable with negative 157 * or positive nice respectively. 158 * 159 * PRI_RANGE: Priority range for utilization dependent priorities. 160 * PRI_NRESV: Number of nice values. 161 * PRI_TICKS: Compute a priority in PRI_RANGE from the ticks count and total. 162 * PRI_NICE: Determines the part of the priority inherited from nice. 163 */ 164 #define SCHED_PRI_NRESV (PRIO_MAX - PRIO_MIN) 165 #define SCHED_PRI_NHALF (SCHED_PRI_NRESV / 2) 166 #define SCHED_PRI_MIN (PRI_MIN_BATCH + SCHED_PRI_NHALF) 167 #define SCHED_PRI_MAX (PRI_MAX_BATCH - SCHED_PRI_NHALF) 168 #define SCHED_PRI_RANGE (SCHED_PRI_MAX - SCHED_PRI_MIN + 1) 169 #define SCHED_PRI_TICKS(ts) \ 170 (SCHED_TICK_HZ((ts)) / \ 171 (roundup(SCHED_TICK_TOTAL((ts)), SCHED_PRI_RANGE) / SCHED_PRI_RANGE)) 172 #define SCHED_PRI_NICE(nice) (nice) 173 174 /* 175 * These determine the interactivity of a process. Interactivity differs from 176 * cpu utilization in that it expresses the voluntary time slept vs time ran 177 * while cpu utilization includes all time not running. This more accurately 178 * models the intent of the thread. 179 * 180 * SLP_RUN_MAX: Maximum amount of sleep time + run time we'll accumulate 181 * before throttling back. 182 * SLP_RUN_FORK: Maximum slp+run time to inherit at fork time. 183 * INTERACT_MAX: Maximum interactivity value. Smaller is better. 184 * INTERACT_THRESH: Threshold for placement on the current runq. 185 */ 186 #define SCHED_SLP_RUN_MAX ((hz * 5) << SCHED_TICK_SHIFT) 187 #define SCHED_SLP_RUN_FORK ((hz / 2) << SCHED_TICK_SHIFT) 188 #define SCHED_INTERACT_MAX (100) 189 #define SCHED_INTERACT_HALF (SCHED_INTERACT_MAX / 2) 190 #define SCHED_INTERACT_THRESH (30) 191 192 /* 193 * These parameters determine the slice behavior for batch work. 194 */ 195 #define SCHED_SLICE_DEFAULT_DIVISOR 10 /* ~94 ms, 12 stathz ticks. */ 196 #define SCHED_SLICE_MIN_DIVISOR 6 /* DEFAULT/MIN = ~16 ms. */ 197 198 /* Flags kept in td_flags. */ 199 #define TDF_SLICEEND TDF_SCHED2 /* Thread time slice is over. */ 200 201 /* 202 * tickincr: Converts a stathz tick into a hz domain scaled by 203 * the shift factor. Without the shift the error rate 204 * due to rounding would be unacceptably high. 205 * realstathz: stathz is sometimes 0 and run off of hz. 206 * sched_slice: Runtime of each thread before rescheduling. 207 * preempt_thresh: Priority threshold for preemption and remote IPIs. 208 */ 209 static int __read_mostly sched_interact = SCHED_INTERACT_THRESH; 210 static int __read_mostly tickincr = 8 << SCHED_TICK_SHIFT; 211 static int __read_mostly realstathz = 127; /* reset during boot. */ 212 static int __read_mostly sched_slice = 10; /* reset during boot. */ 213 static int __read_mostly sched_slice_min = 1; /* reset during boot. */ 214 #ifdef PREEMPTION 215 #ifdef FULL_PREEMPTION 216 static int __read_mostly preempt_thresh = PRI_MAX_IDLE; 217 #else 218 static int __read_mostly preempt_thresh = PRI_MIN_KERN; 219 #endif 220 #else 221 static int __read_mostly preempt_thresh = 0; 222 #endif 223 static int __read_mostly static_boost = PRI_MIN_BATCH; 224 static int __read_mostly sched_idlespins = 10000; 225 static int __read_mostly sched_idlespinthresh = -1; 226 227 /* 228 * tdq - per processor runqs and statistics. All fields are protected by the 229 * tdq_lock. The load and lowpri may be accessed without to avoid excess 230 * locking in sched_pickcpu(); 231 */ 232 struct tdq { 233 /* 234 * Ordered to improve efficiency of cpu_search() and switch(). 235 * tdq_lock is padded to avoid false sharing with tdq_load and 236 * tdq_cpu_idle. 237 */ 238 struct mtx_padalign tdq_lock; /* run queue lock. */ 239 struct cpu_group *tdq_cg; /* Pointer to cpu topology. */ 240 volatile int tdq_load; /* Aggregate load. */ 241 volatile int tdq_cpu_idle; /* cpu_idle() is active. */ 242 int tdq_sysload; /* For loadavg, !ITHD load. */ 243 volatile int tdq_transferable; /* Transferable thread count. */ 244 volatile short tdq_switchcnt; /* Switches this tick. */ 245 volatile short tdq_oldswitchcnt; /* Switches last tick. */ 246 u_char tdq_lowpri; /* Lowest priority thread. */ 247 u_char tdq_owepreempt; /* Remote preemption pending. */ 248 u_char tdq_idx; /* Current insert index. */ 249 u_char tdq_ridx; /* Current removal index. */ 250 int tdq_id; /* cpuid. */ 251 struct runq tdq_realtime; /* real-time run queue. */ 252 struct runq tdq_timeshare; /* timeshare run queue. */ 253 struct runq tdq_idle; /* Queue of IDLE threads. */ 254 char tdq_name[TDQ_NAME_LEN]; 255 #ifdef KTR 256 char tdq_loadname[TDQ_LOADNAME_LEN]; 257 #endif 258 } __aligned(64); 259 260 /* Idle thread states and config. */ 261 #define TDQ_RUNNING 1 262 #define TDQ_IDLE 2 263 264 #ifdef SMP 265 struct cpu_group __read_mostly *cpu_top; /* CPU topology */ 266 267 #define SCHED_AFFINITY_DEFAULT (max(1, hz / 1000)) 268 #define SCHED_AFFINITY(ts, t) ((ts)->ts_rltick > ticks - ((t) * affinity)) 269 270 /* 271 * Run-time tunables. 272 */ 273 static int rebalance = 1; 274 static int balance_interval = 128; /* Default set in sched_initticks(). */ 275 static int __read_mostly affinity; 276 static int __read_mostly steal_idle = 1; 277 static int __read_mostly steal_thresh = 2; 278 static int __read_mostly always_steal = 0; 279 static int __read_mostly trysteal_limit = 2; 280 281 /* 282 * One thread queue per processor. 283 */ 284 static struct tdq __read_mostly *balance_tdq; 285 static int balance_ticks; 286 DPCPU_DEFINE_STATIC(struct tdq, tdq); 287 DPCPU_DEFINE_STATIC(uint32_t, randomval); 288 289 #define TDQ_SELF() ((struct tdq *)PCPU_GET(sched)) 290 #define TDQ_CPU(x) (DPCPU_ID_PTR((x), tdq)) 291 #define TDQ_ID(x) ((x)->tdq_id) 292 #else /* !SMP */ 293 static struct tdq tdq_cpu; 294 295 #define TDQ_ID(x) (0) 296 #define TDQ_SELF() (&tdq_cpu) 297 #define TDQ_CPU(x) (&tdq_cpu) 298 #endif 299 300 #define TDQ_LOCK_ASSERT(t, type) mtx_assert(TDQ_LOCKPTR((t)), (type)) 301 #define TDQ_LOCK(t) mtx_lock_spin(TDQ_LOCKPTR((t))) 302 #define TDQ_LOCK_FLAGS(t, f) mtx_lock_spin_flags(TDQ_LOCKPTR((t)), (f)) 303 #define TDQ_UNLOCK(t) mtx_unlock_spin(TDQ_LOCKPTR((t))) 304 #define TDQ_LOCKPTR(t) ((struct mtx *)(&(t)->tdq_lock)) 305 306 static void sched_priority(struct thread *); 307 static void sched_thread_priority(struct thread *, u_char); 308 static int sched_interact_score(struct thread *); 309 static void sched_interact_update(struct thread *); 310 static void sched_interact_fork(struct thread *); 311 static void sched_pctcpu_update(struct td_sched *, int); 312 313 /* Operations on per processor queues */ 314 static struct thread *tdq_choose(struct tdq *); 315 static void tdq_setup(struct tdq *, int i); 316 static void tdq_load_add(struct tdq *, struct thread *); 317 static void tdq_load_rem(struct tdq *, struct thread *); 318 static __inline void tdq_runq_add(struct tdq *, struct thread *, int); 319 static __inline void tdq_runq_rem(struct tdq *, struct thread *); 320 static inline int sched_shouldpreempt(int, int, int); 321 void tdq_print(int cpu); 322 static void runq_print(struct runq *rq); 323 static void tdq_add(struct tdq *, struct thread *, int); 324 #ifdef SMP 325 static struct thread *tdq_move(struct tdq *, struct tdq *); 326 static int tdq_idled(struct tdq *); 327 static void tdq_notify(struct tdq *, struct thread *); 328 static struct thread *tdq_steal(struct tdq *, int); 329 static struct thread *runq_steal(struct runq *, int); 330 static int sched_pickcpu(struct thread *, int); 331 static void sched_balance(void); 332 static int sched_balance_pair(struct tdq *, struct tdq *); 333 static inline struct tdq *sched_setcpu(struct thread *, int, int); 334 static inline void thread_unblock_switch(struct thread *, struct mtx *); 335 static int sysctl_kern_sched_topology_spec(SYSCTL_HANDLER_ARGS); 336 static int sysctl_kern_sched_topology_spec_internal(struct sbuf *sb, 337 struct cpu_group *cg, int indent); 338 #endif 339 340 static void sched_setup(void *dummy); 341 SYSINIT(sched_setup, SI_SUB_RUN_QUEUE, SI_ORDER_FIRST, sched_setup, NULL); 342 343 static void sched_initticks(void *dummy); 344 SYSINIT(sched_initticks, SI_SUB_CLOCKS, SI_ORDER_THIRD, sched_initticks, 345 NULL); 346 347 SDT_PROVIDER_DEFINE(sched); 348 349 SDT_PROBE_DEFINE3(sched, , , change__pri, "struct thread *", 350 "struct proc *", "uint8_t"); 351 SDT_PROBE_DEFINE3(sched, , , dequeue, "struct thread *", 352 "struct proc *", "void *"); 353 SDT_PROBE_DEFINE4(sched, , , enqueue, "struct thread *", 354 "struct proc *", "void *", "int"); 355 SDT_PROBE_DEFINE4(sched, , , lend__pri, "struct thread *", 356 "struct proc *", "uint8_t", "struct thread *"); 357 SDT_PROBE_DEFINE2(sched, , , load__change, "int", "int"); 358 SDT_PROBE_DEFINE2(sched, , , off__cpu, "struct thread *", 359 "struct proc *"); 360 SDT_PROBE_DEFINE(sched, , , on__cpu); 361 SDT_PROBE_DEFINE(sched, , , remain__cpu); 362 SDT_PROBE_DEFINE2(sched, , , surrender, "struct thread *", 363 "struct proc *"); 364 365 /* 366 * Print the threads waiting on a run-queue. 367 */ 368 static void 369 runq_print(struct runq *rq) 370 { 371 struct rqhead *rqh; 372 struct thread *td; 373 int pri; 374 int j; 375 int i; 376 377 for (i = 0; i < RQB_LEN; i++) { 378 printf("\t\trunq bits %d 0x%zx\n", 379 i, rq->rq_status.rqb_bits[i]); 380 for (j = 0; j < RQB_BPW; j++) 381 if (rq->rq_status.rqb_bits[i] & (1ul << j)) { 382 pri = j + (i << RQB_L2BPW); 383 rqh = &rq->rq_queues[pri]; 384 TAILQ_FOREACH(td, rqh, td_runq) { 385 printf("\t\t\ttd %p(%s) priority %d rqindex %d pri %d\n", 386 td, td->td_name, td->td_priority, 387 td->td_rqindex, pri); 388 } 389 } 390 } 391 } 392 393 /* 394 * Print the status of a per-cpu thread queue. Should be a ddb show cmd. 395 */ 396 void 397 tdq_print(int cpu) 398 { 399 struct tdq *tdq; 400 401 tdq = TDQ_CPU(cpu); 402 403 printf("tdq %d:\n", TDQ_ID(tdq)); 404 printf("\tlock %p\n", TDQ_LOCKPTR(tdq)); 405 printf("\tLock name: %s\n", tdq->tdq_name); 406 printf("\tload: %d\n", tdq->tdq_load); 407 printf("\tswitch cnt: %d\n", tdq->tdq_switchcnt); 408 printf("\told switch cnt: %d\n", tdq->tdq_oldswitchcnt); 409 printf("\ttimeshare idx: %d\n", tdq->tdq_idx); 410 printf("\ttimeshare ridx: %d\n", tdq->tdq_ridx); 411 printf("\tload transferable: %d\n", tdq->tdq_transferable); 412 printf("\tlowest priority: %d\n", tdq->tdq_lowpri); 413 printf("\trealtime runq:\n"); 414 runq_print(&tdq->tdq_realtime); 415 printf("\ttimeshare runq:\n"); 416 runq_print(&tdq->tdq_timeshare); 417 printf("\tidle runq:\n"); 418 runq_print(&tdq->tdq_idle); 419 } 420 421 static inline int 422 sched_shouldpreempt(int pri, int cpri, int remote) 423 { 424 /* 425 * If the new priority is not better than the current priority there is 426 * nothing to do. 427 */ 428 if (pri >= cpri) 429 return (0); 430 /* 431 * Always preempt idle. 432 */ 433 if (cpri >= PRI_MIN_IDLE) 434 return (1); 435 /* 436 * If preemption is disabled don't preempt others. 437 */ 438 if (preempt_thresh == 0) 439 return (0); 440 /* 441 * Preempt if we exceed the threshold. 442 */ 443 if (pri <= preempt_thresh) 444 return (1); 445 /* 446 * If we're interactive or better and there is non-interactive 447 * or worse running preempt only remote processors. 448 */ 449 if (remote && pri <= PRI_MAX_INTERACT && cpri > PRI_MAX_INTERACT) 450 return (1); 451 return (0); 452 } 453 454 /* 455 * Add a thread to the actual run-queue. Keeps transferable counts up to 456 * date with what is actually on the run-queue. Selects the correct 457 * queue position for timeshare threads. 458 */ 459 static __inline void 460 tdq_runq_add(struct tdq *tdq, struct thread *td, int flags) 461 { 462 struct td_sched *ts; 463 u_char pri; 464 465 TDQ_LOCK_ASSERT(tdq, MA_OWNED); 466 THREAD_LOCK_BLOCKED_ASSERT(td, MA_OWNED); 467 468 pri = td->td_priority; 469 ts = td_get_sched(td); 470 TD_SET_RUNQ(td); 471 if (THREAD_CAN_MIGRATE(td)) { 472 tdq->tdq_transferable++; 473 ts->ts_flags |= TSF_XFERABLE; 474 } 475 if (pri < PRI_MIN_BATCH) { 476 ts->ts_runq = &tdq->tdq_realtime; 477 } else if (pri <= PRI_MAX_BATCH) { 478 ts->ts_runq = &tdq->tdq_timeshare; 479 KASSERT(pri <= PRI_MAX_BATCH && pri >= PRI_MIN_BATCH, 480 ("Invalid priority %d on timeshare runq", pri)); 481 /* 482 * This queue contains only priorities between MIN and MAX 483 * realtime. Use the whole queue to represent these values. 484 */ 485 if ((flags & (SRQ_BORROWING|SRQ_PREEMPTED)) == 0) { 486 pri = RQ_NQS * (pri - PRI_MIN_BATCH) / PRI_BATCH_RANGE; 487 pri = (pri + tdq->tdq_idx) % RQ_NQS; 488 /* 489 * This effectively shortens the queue by one so we 490 * can have a one slot difference between idx and 491 * ridx while we wait for threads to drain. 492 */ 493 if (tdq->tdq_ridx != tdq->tdq_idx && 494 pri == tdq->tdq_ridx) 495 pri = (unsigned char)(pri - 1) % RQ_NQS; 496 } else 497 pri = tdq->tdq_ridx; 498 runq_add_pri(ts->ts_runq, td, pri, flags); 499 return; 500 } else 501 ts->ts_runq = &tdq->tdq_idle; 502 runq_add(ts->ts_runq, td, flags); 503 } 504 505 /* 506 * Remove a thread from a run-queue. This typically happens when a thread 507 * is selected to run. Running threads are not on the queue and the 508 * transferable count does not reflect them. 509 */ 510 static __inline void 511 tdq_runq_rem(struct tdq *tdq, struct thread *td) 512 { 513 struct td_sched *ts; 514 515 ts = td_get_sched(td); 516 TDQ_LOCK_ASSERT(tdq, MA_OWNED); 517 THREAD_LOCK_BLOCKED_ASSERT(td, MA_OWNED); 518 KASSERT(ts->ts_runq != NULL, 519 ("tdq_runq_remove: thread %p null ts_runq", td)); 520 if (ts->ts_flags & TSF_XFERABLE) { 521 tdq->tdq_transferable--; 522 ts->ts_flags &= ~TSF_XFERABLE; 523 } 524 if (ts->ts_runq == &tdq->tdq_timeshare) { 525 if (tdq->tdq_idx != tdq->tdq_ridx) 526 runq_remove_idx(ts->ts_runq, td, &tdq->tdq_ridx); 527 else 528 runq_remove_idx(ts->ts_runq, td, NULL); 529 } else 530 runq_remove(ts->ts_runq, td); 531 } 532 533 /* 534 * Load is maintained for all threads RUNNING and ON_RUNQ. Add the load 535 * for this thread to the referenced thread queue. 536 */ 537 static void 538 tdq_load_add(struct tdq *tdq, struct thread *td) 539 { 540 541 TDQ_LOCK_ASSERT(tdq, MA_OWNED); 542 THREAD_LOCK_BLOCKED_ASSERT(td, MA_OWNED); 543 544 tdq->tdq_load++; 545 if ((td->td_flags & TDF_NOLOAD) == 0) 546 tdq->tdq_sysload++; 547 KTR_COUNTER0(KTR_SCHED, "load", tdq->tdq_loadname, tdq->tdq_load); 548 SDT_PROBE2(sched, , , load__change, (int)TDQ_ID(tdq), tdq->tdq_load); 549 } 550 551 /* 552 * Remove the load from a thread that is transitioning to a sleep state or 553 * exiting. 554 */ 555 static void 556 tdq_load_rem(struct tdq *tdq, struct thread *td) 557 { 558 559 TDQ_LOCK_ASSERT(tdq, MA_OWNED); 560 THREAD_LOCK_BLOCKED_ASSERT(td, MA_OWNED); 561 KASSERT(tdq->tdq_load != 0, 562 ("tdq_load_rem: Removing with 0 load on queue %d", TDQ_ID(tdq))); 563 564 tdq->tdq_load--; 565 if ((td->td_flags & TDF_NOLOAD) == 0) 566 tdq->tdq_sysload--; 567 KTR_COUNTER0(KTR_SCHED, "load", tdq->tdq_loadname, tdq->tdq_load); 568 SDT_PROBE2(sched, , , load__change, (int)TDQ_ID(tdq), tdq->tdq_load); 569 } 570 571 /* 572 * Bound timeshare latency by decreasing slice size as load increases. We 573 * consider the maximum latency as the sum of the threads waiting to run 574 * aside from curthread and target no more than sched_slice latency but 575 * no less than sched_slice_min runtime. 576 */ 577 static inline int 578 tdq_slice(struct tdq *tdq) 579 { 580 int load; 581 582 /* 583 * It is safe to use sys_load here because this is called from 584 * contexts where timeshare threads are running and so there 585 * cannot be higher priority load in the system. 586 */ 587 load = tdq->tdq_sysload - 1; 588 if (load >= SCHED_SLICE_MIN_DIVISOR) 589 return (sched_slice_min); 590 if (load <= 1) 591 return (sched_slice); 592 return (sched_slice / load); 593 } 594 595 /* 596 * Set lowpri to its exact value by searching the run-queue and 597 * evaluating curthread. curthread may be passed as an optimization. 598 */ 599 static void 600 tdq_setlowpri(struct tdq *tdq, struct thread *ctd) 601 { 602 struct thread *td; 603 604 TDQ_LOCK_ASSERT(tdq, MA_OWNED); 605 if (ctd == NULL) 606 ctd = pcpu_find(TDQ_ID(tdq))->pc_curthread; 607 td = tdq_choose(tdq); 608 if (td == NULL || td->td_priority > ctd->td_priority) 609 tdq->tdq_lowpri = ctd->td_priority; 610 else 611 tdq->tdq_lowpri = td->td_priority; 612 } 613 614 #ifdef SMP 615 /* 616 * We need some randomness. Implement a classic Linear Congruential 617 * Generator X_{n+1}=(aX_n+c) mod m. These values are optimized for 618 * m = 2^32, a = 69069 and c = 5. We only return the upper 16 bits 619 * of the random state (in the low bits of our answer) to keep 620 * the maximum randomness. 621 */ 622 static uint32_t 623 sched_random(void) 624 { 625 uint32_t *rndptr; 626 627 rndptr = DPCPU_PTR(randomval); 628 *rndptr = *rndptr * 69069 + 5; 629 630 return (*rndptr >> 16); 631 } 632 633 struct cpu_search { 634 cpuset_t cs_mask; 635 u_int cs_prefer; 636 int cs_pri; /* Min priority for low. */ 637 int cs_limit; /* Max load for low, min load for high. */ 638 int cs_cpu; 639 int cs_load; 640 }; 641 642 #define CPU_SEARCH_LOWEST 0x1 643 #define CPU_SEARCH_HIGHEST 0x2 644 #define CPU_SEARCH_BOTH (CPU_SEARCH_LOWEST|CPU_SEARCH_HIGHEST) 645 646 static __always_inline int cpu_search(const struct cpu_group *cg, 647 struct cpu_search *low, struct cpu_search *high, const int match); 648 int __noinline cpu_search_lowest(const struct cpu_group *cg, 649 struct cpu_search *low); 650 int __noinline cpu_search_highest(const struct cpu_group *cg, 651 struct cpu_search *high); 652 int __noinline cpu_search_both(const struct cpu_group *cg, 653 struct cpu_search *low, struct cpu_search *high); 654 655 /* 656 * Search the tree of cpu_groups for the lowest or highest loaded cpu 657 * according to the match argument. This routine actually compares the 658 * load on all paths through the tree and finds the least loaded cpu on 659 * the least loaded path, which may differ from the least loaded cpu in 660 * the system. This balances work among caches and buses. 661 * 662 * This inline is instantiated in three forms below using constants for the 663 * match argument. It is reduced to the minimum set for each case. It is 664 * also recursive to the depth of the tree. 665 */ 666 static __always_inline int 667 cpu_search(const struct cpu_group *cg, struct cpu_search *low, 668 struct cpu_search *high, const int match) 669 { 670 struct cpu_search lgroup; 671 struct cpu_search hgroup; 672 cpuset_t cpumask; 673 struct cpu_group *child; 674 struct tdq *tdq; 675 int cpu, i, hload, lload, load, total, rnd; 676 677 total = 0; 678 cpumask = cg->cg_mask; 679 if (match & CPU_SEARCH_LOWEST) { 680 lload = INT_MAX; 681 lgroup = *low; 682 } 683 if (match & CPU_SEARCH_HIGHEST) { 684 hload = INT_MIN; 685 hgroup = *high; 686 } 687 688 /* Iterate through the child CPU groups and then remaining CPUs. */ 689 for (i = cg->cg_children, cpu = mp_maxid; ; ) { 690 if (i == 0) { 691 #ifdef HAVE_INLINE_FFSL 692 cpu = CPU_FFS(&cpumask) - 1; 693 #else 694 while (cpu >= 0 && !CPU_ISSET(cpu, &cpumask)) 695 cpu--; 696 #endif 697 if (cpu < 0) 698 break; 699 child = NULL; 700 } else 701 child = &cg->cg_child[i - 1]; 702 703 if (match & CPU_SEARCH_LOWEST) 704 lgroup.cs_cpu = -1; 705 if (match & CPU_SEARCH_HIGHEST) 706 hgroup.cs_cpu = -1; 707 if (child) { /* Handle child CPU group. */ 708 CPU_ANDNOT(&cpumask, &child->cg_mask); 709 switch (match) { 710 case CPU_SEARCH_LOWEST: 711 load = cpu_search_lowest(child, &lgroup); 712 break; 713 case CPU_SEARCH_HIGHEST: 714 load = cpu_search_highest(child, &hgroup); 715 break; 716 case CPU_SEARCH_BOTH: 717 load = cpu_search_both(child, &lgroup, &hgroup); 718 break; 719 } 720 } else { /* Handle child CPU. */ 721 CPU_CLR(cpu, &cpumask); 722 tdq = TDQ_CPU(cpu); 723 load = tdq->tdq_load * 256; 724 rnd = sched_random() % 32; 725 if (match & CPU_SEARCH_LOWEST) { 726 if (cpu == low->cs_prefer) 727 load -= 64; 728 /* If that CPU is allowed and get data. */ 729 if (tdq->tdq_lowpri > lgroup.cs_pri && 730 tdq->tdq_load <= lgroup.cs_limit && 731 CPU_ISSET(cpu, &lgroup.cs_mask)) { 732 lgroup.cs_cpu = cpu; 733 lgroup.cs_load = load - rnd; 734 } 735 } 736 if (match & CPU_SEARCH_HIGHEST) 737 if (tdq->tdq_load >= hgroup.cs_limit && 738 tdq->tdq_transferable && 739 CPU_ISSET(cpu, &hgroup.cs_mask)) { 740 hgroup.cs_cpu = cpu; 741 hgroup.cs_load = load - rnd; 742 } 743 } 744 total += load; 745 746 /* We have info about child item. Compare it. */ 747 if (match & CPU_SEARCH_LOWEST) { 748 if (lgroup.cs_cpu >= 0 && 749 (load < lload || 750 (load == lload && lgroup.cs_load < low->cs_load))) { 751 lload = load; 752 low->cs_cpu = lgroup.cs_cpu; 753 low->cs_load = lgroup.cs_load; 754 } 755 } 756 if (match & CPU_SEARCH_HIGHEST) 757 if (hgroup.cs_cpu >= 0 && 758 (load > hload || 759 (load == hload && hgroup.cs_load > high->cs_load))) { 760 hload = load; 761 high->cs_cpu = hgroup.cs_cpu; 762 high->cs_load = hgroup.cs_load; 763 } 764 if (child) { 765 i--; 766 if (i == 0 && CPU_EMPTY(&cpumask)) 767 break; 768 } 769 #ifndef HAVE_INLINE_FFSL 770 else 771 cpu--; 772 #endif 773 } 774 return (total); 775 } 776 777 /* 778 * cpu_search instantiations must pass constants to maintain the inline 779 * optimization. 780 */ 781 int 782 cpu_search_lowest(const struct cpu_group *cg, struct cpu_search *low) 783 { 784 return cpu_search(cg, low, NULL, CPU_SEARCH_LOWEST); 785 } 786 787 int 788 cpu_search_highest(const struct cpu_group *cg, struct cpu_search *high) 789 { 790 return cpu_search(cg, NULL, high, CPU_SEARCH_HIGHEST); 791 } 792 793 int 794 cpu_search_both(const struct cpu_group *cg, struct cpu_search *low, 795 struct cpu_search *high) 796 { 797 return cpu_search(cg, low, high, CPU_SEARCH_BOTH); 798 } 799 800 /* 801 * Find the cpu with the least load via the least loaded path that has a 802 * lowpri greater than pri pri. A pri of -1 indicates any priority is 803 * acceptable. 804 */ 805 static inline int 806 sched_lowest(const struct cpu_group *cg, cpuset_t mask, int pri, int maxload, 807 int prefer) 808 { 809 struct cpu_search low; 810 811 low.cs_cpu = -1; 812 low.cs_prefer = prefer; 813 low.cs_mask = mask; 814 low.cs_pri = pri; 815 low.cs_limit = maxload; 816 cpu_search_lowest(cg, &low); 817 return low.cs_cpu; 818 } 819 820 /* 821 * Find the cpu with the highest load via the highest loaded path. 822 */ 823 static inline int 824 sched_highest(const struct cpu_group *cg, cpuset_t mask, int minload) 825 { 826 struct cpu_search high; 827 828 high.cs_cpu = -1; 829 high.cs_mask = mask; 830 high.cs_limit = minload; 831 cpu_search_highest(cg, &high); 832 return high.cs_cpu; 833 } 834 835 static void 836 sched_balance_group(struct cpu_group *cg) 837 { 838 struct tdq *tdq; 839 cpuset_t hmask, lmask; 840 int high, low, anylow; 841 842 CPU_FILL(&hmask); 843 for (;;) { 844 high = sched_highest(cg, hmask, 2); 845 /* Stop if there is no more CPU with transferrable threads. */ 846 if (high == -1) 847 break; 848 CPU_CLR(high, &hmask); 849 CPU_COPY(&hmask, &lmask); 850 /* Stop if there is no more CPU left for low. */ 851 if (CPU_EMPTY(&lmask)) 852 break; 853 anylow = 1; 854 tdq = TDQ_CPU(high); 855 nextlow: 856 low = sched_lowest(cg, lmask, -1, tdq->tdq_load - 1, high); 857 /* Stop if we looked well and found no less loaded CPU. */ 858 if (anylow && low == -1) 859 break; 860 /* Go to next high if we found no less loaded CPU. */ 861 if (low == -1) 862 continue; 863 /* Transfer thread from high to low. */ 864 if (sched_balance_pair(tdq, TDQ_CPU(low))) { 865 /* CPU that got thread can no longer be a donor. */ 866 CPU_CLR(low, &hmask); 867 } else { 868 /* 869 * If failed, then there is no threads on high 870 * that can run on this low. Drop low from low 871 * mask and look for different one. 872 */ 873 CPU_CLR(low, &lmask); 874 anylow = 0; 875 goto nextlow; 876 } 877 } 878 } 879 880 static void 881 sched_balance(void) 882 { 883 struct tdq *tdq; 884 885 balance_ticks = max(balance_interval / 2, 1) + 886 (sched_random() % balance_interval); 887 tdq = TDQ_SELF(); 888 TDQ_UNLOCK(tdq); 889 sched_balance_group(cpu_top); 890 TDQ_LOCK(tdq); 891 } 892 893 /* 894 * Lock two thread queues using their address to maintain lock order. 895 */ 896 static void 897 tdq_lock_pair(struct tdq *one, struct tdq *two) 898 { 899 if (one < two) { 900 TDQ_LOCK(one); 901 TDQ_LOCK_FLAGS(two, MTX_DUPOK); 902 } else { 903 TDQ_LOCK(two); 904 TDQ_LOCK_FLAGS(one, MTX_DUPOK); 905 } 906 } 907 908 /* 909 * Unlock two thread queues. Order is not important here. 910 */ 911 static void 912 tdq_unlock_pair(struct tdq *one, struct tdq *two) 913 { 914 TDQ_UNLOCK(one); 915 TDQ_UNLOCK(two); 916 } 917 918 /* 919 * Transfer load between two imbalanced thread queues. 920 */ 921 static int 922 sched_balance_pair(struct tdq *high, struct tdq *low) 923 { 924 struct thread *td; 925 int cpu; 926 927 tdq_lock_pair(high, low); 928 td = NULL; 929 /* 930 * Transfer a thread from high to low. 931 */ 932 if (high->tdq_transferable != 0 && high->tdq_load > low->tdq_load && 933 (td = tdq_move(high, low)) != NULL) { 934 /* 935 * In case the target isn't the current cpu notify it of the 936 * new load, possibly sending an IPI to force it to reschedule. 937 */ 938 cpu = TDQ_ID(low); 939 if (cpu != PCPU_GET(cpuid)) 940 tdq_notify(low, td); 941 } 942 tdq_unlock_pair(high, low); 943 return (td != NULL); 944 } 945 946 /* 947 * Move a thread from one thread queue to another. 948 */ 949 static struct thread * 950 tdq_move(struct tdq *from, struct tdq *to) 951 { 952 struct thread *td; 953 struct tdq *tdq; 954 int cpu; 955 956 TDQ_LOCK_ASSERT(from, MA_OWNED); 957 TDQ_LOCK_ASSERT(to, MA_OWNED); 958 959 tdq = from; 960 cpu = TDQ_ID(to); 961 td = tdq_steal(tdq, cpu); 962 if (td == NULL) 963 return (NULL); 964 965 /* 966 * Although the run queue is locked the thread may be 967 * blocked. We can not set the lock until it is unblocked. 968 */ 969 thread_lock_block_wait(td); 970 sched_rem(td); 971 THREAD_LOCKPTR_ASSERT(td, TDQ_LOCKPTR(from)); 972 td->td_lock = TDQ_LOCKPTR(to); 973 td_get_sched(td)->ts_cpu = cpu; 974 tdq_add(to, td, SRQ_YIELDING); 975 976 return (td); 977 } 978 979 /* 980 * This tdq has idled. Try to steal a thread from another cpu and switch 981 * to it. 982 */ 983 static int 984 tdq_idled(struct tdq *tdq) 985 { 986 struct cpu_group *cg; 987 struct tdq *steal; 988 cpuset_t mask; 989 int cpu, switchcnt; 990 991 if (smp_started == 0 || steal_idle == 0 || tdq->tdq_cg == NULL) 992 return (1); 993 CPU_FILL(&mask); 994 CPU_CLR(PCPU_GET(cpuid), &mask); 995 restart: 996 switchcnt = tdq->tdq_switchcnt + tdq->tdq_oldswitchcnt; 997 for (cg = tdq->tdq_cg; ; ) { 998 cpu = sched_highest(cg, mask, steal_thresh); 999 /* 1000 * We were assigned a thread but not preempted. Returning 1001 * 0 here will cause our caller to switch to it. 1002 */ 1003 if (tdq->tdq_load) 1004 return (0); 1005 if (cpu == -1) { 1006 cg = cg->cg_parent; 1007 if (cg == NULL) 1008 return (1); 1009 continue; 1010 } 1011 steal = TDQ_CPU(cpu); 1012 /* 1013 * The data returned by sched_highest() is stale and 1014 * the chosen CPU no longer has an eligible thread. 1015 * 1016 * Testing this ahead of tdq_lock_pair() only catches 1017 * this situation about 20% of the time on an 8 core 1018 * 16 thread Ryzen 7, but it still helps performance. 1019 */ 1020 if (steal->tdq_load < steal_thresh || 1021 steal->tdq_transferable == 0) 1022 goto restart; 1023 tdq_lock_pair(tdq, steal); 1024 /* 1025 * We were assigned a thread while waiting for the locks. 1026 * Switch to it now instead of stealing a thread. 1027 */ 1028 if (tdq->tdq_load) 1029 break; 1030 /* 1031 * The data returned by sched_highest() is stale and 1032 * the chosen CPU no longer has an eligible thread, or 1033 * we were preempted and the CPU loading info may be out 1034 * of date. The latter is rare. In either case restart 1035 * the search. 1036 */ 1037 if (steal->tdq_load < steal_thresh || 1038 steal->tdq_transferable == 0 || 1039 switchcnt != tdq->tdq_switchcnt + tdq->tdq_oldswitchcnt) { 1040 tdq_unlock_pair(tdq, steal); 1041 goto restart; 1042 } 1043 /* 1044 * Steal the thread and switch to it. 1045 */ 1046 if (tdq_move(steal, tdq) != NULL) 1047 break; 1048 /* 1049 * We failed to acquire a thread even though it looked 1050 * like one was available. This could be due to affinity 1051 * restrictions or for other reasons. Loop again after 1052 * removing this CPU from the set. The restart logic 1053 * above does not restore this CPU to the set due to the 1054 * likelyhood of failing here again. 1055 */ 1056 CPU_CLR(cpu, &mask); 1057 tdq_unlock_pair(tdq, steal); 1058 } 1059 TDQ_UNLOCK(steal); 1060 mi_switch(SW_VOL | SWT_IDLE); 1061 return (0); 1062 } 1063 1064 /* 1065 * Notify a remote cpu of new work. Sends an IPI if criteria are met. 1066 */ 1067 static void 1068 tdq_notify(struct tdq *tdq, struct thread *td) 1069 { 1070 struct thread *ctd; 1071 int pri; 1072 int cpu; 1073 1074 if (tdq->tdq_owepreempt) 1075 return; 1076 cpu = td_get_sched(td)->ts_cpu; 1077 pri = td->td_priority; 1078 ctd = pcpu_find(cpu)->pc_curthread; 1079 if (!sched_shouldpreempt(pri, ctd->td_priority, 1)) 1080 return; 1081 1082 /* 1083 * Make sure that our caller's earlier update to tdq_load is 1084 * globally visible before we read tdq_cpu_idle. Idle thread 1085 * accesses both of them without locks, and the order is important. 1086 */ 1087 atomic_thread_fence_seq_cst(); 1088 1089 if (TD_IS_IDLETHREAD(ctd)) { 1090 /* 1091 * If the MD code has an idle wakeup routine try that before 1092 * falling back to IPI. 1093 */ 1094 if (!tdq->tdq_cpu_idle || cpu_idle_wakeup(cpu)) 1095 return; 1096 } 1097 1098 /* 1099 * The run queues have been updated, so any switch on the remote CPU 1100 * will satisfy the preemption request. 1101 */ 1102 tdq->tdq_owepreempt = 1; 1103 ipi_cpu(cpu, IPI_PREEMPT); 1104 } 1105 1106 /* 1107 * Steals load from a timeshare queue. Honors the rotating queue head 1108 * index. 1109 */ 1110 static struct thread * 1111 runq_steal_from(struct runq *rq, int cpu, u_char start) 1112 { 1113 struct rqbits *rqb; 1114 struct rqhead *rqh; 1115 struct thread *td, *first; 1116 int bit; 1117 int i; 1118 1119 rqb = &rq->rq_status; 1120 bit = start & (RQB_BPW -1); 1121 first = NULL; 1122 again: 1123 for (i = RQB_WORD(start); i < RQB_LEN; bit = 0, i++) { 1124 if (rqb->rqb_bits[i] == 0) 1125 continue; 1126 if (bit == 0) 1127 bit = RQB_FFS(rqb->rqb_bits[i]); 1128 for (; bit < RQB_BPW; bit++) { 1129 if ((rqb->rqb_bits[i] & (1ul << bit)) == 0) 1130 continue; 1131 rqh = &rq->rq_queues[bit + (i << RQB_L2BPW)]; 1132 TAILQ_FOREACH(td, rqh, td_runq) { 1133 if (first && THREAD_CAN_MIGRATE(td) && 1134 THREAD_CAN_SCHED(td, cpu)) 1135 return (td); 1136 first = td; 1137 } 1138 } 1139 } 1140 if (start != 0) { 1141 start = 0; 1142 goto again; 1143 } 1144 1145 if (first && THREAD_CAN_MIGRATE(first) && 1146 THREAD_CAN_SCHED(first, cpu)) 1147 return (first); 1148 return (NULL); 1149 } 1150 1151 /* 1152 * Steals load from a standard linear queue. 1153 */ 1154 static struct thread * 1155 runq_steal(struct runq *rq, int cpu) 1156 { 1157 struct rqhead *rqh; 1158 struct rqbits *rqb; 1159 struct thread *td; 1160 int word; 1161 int bit; 1162 1163 rqb = &rq->rq_status; 1164 for (word = 0; word < RQB_LEN; word++) { 1165 if (rqb->rqb_bits[word] == 0) 1166 continue; 1167 for (bit = 0; bit < RQB_BPW; bit++) { 1168 if ((rqb->rqb_bits[word] & (1ul << bit)) == 0) 1169 continue; 1170 rqh = &rq->rq_queues[bit + (word << RQB_L2BPW)]; 1171 TAILQ_FOREACH(td, rqh, td_runq) 1172 if (THREAD_CAN_MIGRATE(td) && 1173 THREAD_CAN_SCHED(td, cpu)) 1174 return (td); 1175 } 1176 } 1177 return (NULL); 1178 } 1179 1180 /* 1181 * Attempt to steal a thread in priority order from a thread queue. 1182 */ 1183 static struct thread * 1184 tdq_steal(struct tdq *tdq, int cpu) 1185 { 1186 struct thread *td; 1187 1188 TDQ_LOCK_ASSERT(tdq, MA_OWNED); 1189 if ((td = runq_steal(&tdq->tdq_realtime, cpu)) != NULL) 1190 return (td); 1191 if ((td = runq_steal_from(&tdq->tdq_timeshare, 1192 cpu, tdq->tdq_ridx)) != NULL) 1193 return (td); 1194 return (runq_steal(&tdq->tdq_idle, cpu)); 1195 } 1196 1197 /* 1198 * Sets the thread lock and ts_cpu to match the requested cpu. Unlocks the 1199 * current lock and returns with the assigned queue locked. 1200 */ 1201 static inline struct tdq * 1202 sched_setcpu(struct thread *td, int cpu, int flags) 1203 { 1204 1205 struct tdq *tdq; 1206 struct mtx *mtx; 1207 1208 THREAD_LOCK_ASSERT(td, MA_OWNED); 1209 tdq = TDQ_CPU(cpu); 1210 td_get_sched(td)->ts_cpu = cpu; 1211 /* 1212 * If the lock matches just return the queue. 1213 */ 1214 if (td->td_lock == TDQ_LOCKPTR(tdq)) { 1215 KASSERT((flags & SRQ_HOLD) == 0, 1216 ("sched_setcpu: Invalid lock for SRQ_HOLD")); 1217 return (tdq); 1218 } 1219 1220 /* 1221 * The hard case, migration, we need to block the thread first to 1222 * prevent order reversals with other cpus locks. 1223 */ 1224 spinlock_enter(); 1225 mtx = thread_lock_block(td); 1226 if ((flags & SRQ_HOLD) == 0) 1227 mtx_unlock_spin(mtx); 1228 TDQ_LOCK(tdq); 1229 thread_lock_unblock(td, TDQ_LOCKPTR(tdq)); 1230 spinlock_exit(); 1231 return (tdq); 1232 } 1233 1234 SCHED_STAT_DEFINE(pickcpu_intrbind, "Soft interrupt binding"); 1235 SCHED_STAT_DEFINE(pickcpu_idle_affinity, "Picked idle cpu based on affinity"); 1236 SCHED_STAT_DEFINE(pickcpu_affinity, "Picked cpu based on affinity"); 1237 SCHED_STAT_DEFINE(pickcpu_lowest, "Selected lowest load"); 1238 SCHED_STAT_DEFINE(pickcpu_local, "Migrated to current cpu"); 1239 SCHED_STAT_DEFINE(pickcpu_migration, "Selection may have caused migration"); 1240 1241 static int 1242 sched_pickcpu(struct thread *td, int flags) 1243 { 1244 struct cpu_group *cg, *ccg; 1245 struct td_sched *ts; 1246 struct tdq *tdq; 1247 cpuset_t mask; 1248 int cpu, pri, self, intr; 1249 1250 self = PCPU_GET(cpuid); 1251 ts = td_get_sched(td); 1252 KASSERT(!CPU_ABSENT(ts->ts_cpu), ("sched_pickcpu: Start scheduler on " 1253 "absent CPU %d for thread %s.", ts->ts_cpu, td->td_name)); 1254 if (smp_started == 0) 1255 return (self); 1256 /* 1257 * Don't migrate a running thread from sched_switch(). 1258 */ 1259 if ((flags & SRQ_OURSELF) || !THREAD_CAN_MIGRATE(td)) 1260 return (ts->ts_cpu); 1261 /* 1262 * Prefer to run interrupt threads on the processors that generate 1263 * the interrupt. 1264 */ 1265 if (td->td_priority <= PRI_MAX_ITHD && THREAD_CAN_SCHED(td, self) && 1266 curthread->td_intr_nesting_level) { 1267 tdq = TDQ_SELF(); 1268 if (tdq->tdq_lowpri >= PRI_MIN_IDLE) { 1269 SCHED_STAT_INC(pickcpu_idle_affinity); 1270 return (self); 1271 } 1272 ts->ts_cpu = self; 1273 intr = 1; 1274 cg = tdq->tdq_cg; 1275 goto llc; 1276 } else { 1277 intr = 0; 1278 tdq = TDQ_CPU(ts->ts_cpu); 1279 cg = tdq->tdq_cg; 1280 } 1281 /* 1282 * If the thread can run on the last cpu and the affinity has not 1283 * expired and it is idle, run it there. 1284 */ 1285 if (THREAD_CAN_SCHED(td, ts->ts_cpu) && 1286 tdq->tdq_lowpri >= PRI_MIN_IDLE && 1287 SCHED_AFFINITY(ts, CG_SHARE_L2)) { 1288 if (cg->cg_flags & CG_FLAG_THREAD) { 1289 /* Check all SMT threads for being idle. */ 1290 for (cpu = CPU_FFS(&cg->cg_mask) - 1; ; cpu++) { 1291 if (CPU_ISSET(cpu, &cg->cg_mask) && 1292 TDQ_CPU(cpu)->tdq_lowpri < PRI_MIN_IDLE) 1293 break; 1294 if (cpu >= mp_maxid) { 1295 SCHED_STAT_INC(pickcpu_idle_affinity); 1296 return (ts->ts_cpu); 1297 } 1298 } 1299 } else { 1300 SCHED_STAT_INC(pickcpu_idle_affinity); 1301 return (ts->ts_cpu); 1302 } 1303 } 1304 llc: 1305 /* 1306 * Search for the last level cache CPU group in the tree. 1307 * Skip SMT, identical groups and caches with expired affinity. 1308 * Interrupt threads affinity is explicit and never expires. 1309 */ 1310 for (ccg = NULL; cg != NULL; cg = cg->cg_parent) { 1311 if (cg->cg_flags & CG_FLAG_THREAD) 1312 continue; 1313 if (cg->cg_children == 1 || cg->cg_count == 1) 1314 continue; 1315 if (cg->cg_level == CG_SHARE_NONE || 1316 (!intr && !SCHED_AFFINITY(ts, cg->cg_level))) 1317 continue; 1318 ccg = cg; 1319 } 1320 /* Found LLC shared by all CPUs, so do a global search. */ 1321 if (ccg == cpu_top) 1322 ccg = NULL; 1323 cpu = -1; 1324 mask = td->td_cpuset->cs_mask; 1325 pri = td->td_priority; 1326 /* 1327 * Try hard to keep interrupts within found LLC. Search the LLC for 1328 * the least loaded CPU we can run now. For NUMA systems it should 1329 * be within target domain, and it also reduces scheduling overhead. 1330 */ 1331 if (ccg != NULL && intr) { 1332 cpu = sched_lowest(ccg, mask, pri, INT_MAX, ts->ts_cpu); 1333 if (cpu >= 0) 1334 SCHED_STAT_INC(pickcpu_intrbind); 1335 } else 1336 /* Search the LLC for the least loaded idle CPU we can run now. */ 1337 if (ccg != NULL) { 1338 cpu = sched_lowest(ccg, mask, max(pri, PRI_MAX_TIMESHARE), 1339 INT_MAX, ts->ts_cpu); 1340 if (cpu >= 0) 1341 SCHED_STAT_INC(pickcpu_affinity); 1342 } 1343 /* Search globally for the least loaded CPU we can run now. */ 1344 if (cpu < 0) { 1345 cpu = sched_lowest(cpu_top, mask, pri, INT_MAX, ts->ts_cpu); 1346 if (cpu >= 0) 1347 SCHED_STAT_INC(pickcpu_lowest); 1348 } 1349 /* Search globally for the least loaded CPU. */ 1350 if (cpu < 0) { 1351 cpu = sched_lowest(cpu_top, mask, -1, INT_MAX, ts->ts_cpu); 1352 if (cpu >= 0) 1353 SCHED_STAT_INC(pickcpu_lowest); 1354 } 1355 KASSERT(cpu >= 0, ("sched_pickcpu: Failed to find a cpu.")); 1356 KASSERT(!CPU_ABSENT(cpu), ("sched_pickcpu: Picked absent CPU %d.", cpu)); 1357 /* 1358 * Compare the lowest loaded cpu to current cpu. 1359 */ 1360 tdq = TDQ_CPU(cpu); 1361 if (THREAD_CAN_SCHED(td, self) && TDQ_SELF()->tdq_lowpri > pri && 1362 tdq->tdq_lowpri < PRI_MIN_IDLE && 1363 TDQ_SELF()->tdq_load <= tdq->tdq_load + 1) { 1364 SCHED_STAT_INC(pickcpu_local); 1365 cpu = self; 1366 } 1367 if (cpu != ts->ts_cpu) 1368 SCHED_STAT_INC(pickcpu_migration); 1369 return (cpu); 1370 } 1371 #endif 1372 1373 /* 1374 * Pick the highest priority task we have and return it. 1375 */ 1376 static struct thread * 1377 tdq_choose(struct tdq *tdq) 1378 { 1379 struct thread *td; 1380 1381 TDQ_LOCK_ASSERT(tdq, MA_OWNED); 1382 td = runq_choose(&tdq->tdq_realtime); 1383 if (td != NULL) 1384 return (td); 1385 td = runq_choose_from(&tdq->tdq_timeshare, tdq->tdq_ridx); 1386 if (td != NULL) { 1387 KASSERT(td->td_priority >= PRI_MIN_BATCH, 1388 ("tdq_choose: Invalid priority on timeshare queue %d", 1389 td->td_priority)); 1390 return (td); 1391 } 1392 td = runq_choose(&tdq->tdq_idle); 1393 if (td != NULL) { 1394 KASSERT(td->td_priority >= PRI_MIN_IDLE, 1395 ("tdq_choose: Invalid priority on idle queue %d", 1396 td->td_priority)); 1397 return (td); 1398 } 1399 1400 return (NULL); 1401 } 1402 1403 /* 1404 * Initialize a thread queue. 1405 */ 1406 static void 1407 tdq_setup(struct tdq *tdq, int id) 1408 { 1409 1410 if (bootverbose) 1411 printf("ULE: setup cpu %d\n", id); 1412 runq_init(&tdq->tdq_realtime); 1413 runq_init(&tdq->tdq_timeshare); 1414 runq_init(&tdq->tdq_idle); 1415 tdq->tdq_id = id; 1416 snprintf(tdq->tdq_name, sizeof(tdq->tdq_name), 1417 "sched lock %d", (int)TDQ_ID(tdq)); 1418 mtx_init(&tdq->tdq_lock, tdq->tdq_name, "sched lock", MTX_SPIN); 1419 #ifdef KTR 1420 snprintf(tdq->tdq_loadname, sizeof(tdq->tdq_loadname), 1421 "CPU %d load", (int)TDQ_ID(tdq)); 1422 #endif 1423 } 1424 1425 #ifdef SMP 1426 static void 1427 sched_setup_smp(void) 1428 { 1429 struct tdq *tdq; 1430 int i; 1431 1432 cpu_top = smp_topo(); 1433 CPU_FOREACH(i) { 1434 tdq = DPCPU_ID_PTR(i, tdq); 1435 tdq_setup(tdq, i); 1436 tdq->tdq_cg = smp_topo_find(cpu_top, i); 1437 if (tdq->tdq_cg == NULL) 1438 panic("Can't find cpu group for %d\n", i); 1439 } 1440 PCPU_SET(sched, DPCPU_PTR(tdq)); 1441 balance_tdq = TDQ_SELF(); 1442 } 1443 #endif 1444 1445 /* 1446 * Setup the thread queues and initialize the topology based on MD 1447 * information. 1448 */ 1449 static void 1450 sched_setup(void *dummy) 1451 { 1452 struct tdq *tdq; 1453 1454 #ifdef SMP 1455 sched_setup_smp(); 1456 #else 1457 tdq_setup(TDQ_SELF(), 0); 1458 #endif 1459 tdq = TDQ_SELF(); 1460 1461 /* Add thread0's load since it's running. */ 1462 TDQ_LOCK(tdq); 1463 thread0.td_lock = TDQ_LOCKPTR(tdq); 1464 tdq_load_add(tdq, &thread0); 1465 tdq->tdq_lowpri = thread0.td_priority; 1466 TDQ_UNLOCK(tdq); 1467 } 1468 1469 /* 1470 * This routine determines time constants after stathz and hz are setup. 1471 */ 1472 /* ARGSUSED */ 1473 static void 1474 sched_initticks(void *dummy) 1475 { 1476 int incr; 1477 1478 realstathz = stathz ? stathz : hz; 1479 sched_slice = realstathz / SCHED_SLICE_DEFAULT_DIVISOR; 1480 sched_slice_min = sched_slice / SCHED_SLICE_MIN_DIVISOR; 1481 hogticks = imax(1, (2 * hz * sched_slice + realstathz / 2) / 1482 realstathz); 1483 1484 /* 1485 * tickincr is shifted out by 10 to avoid rounding errors due to 1486 * hz not being evenly divisible by stathz on all platforms. 1487 */ 1488 incr = (hz << SCHED_TICK_SHIFT) / realstathz; 1489 /* 1490 * This does not work for values of stathz that are more than 1491 * 1 << SCHED_TICK_SHIFT * hz. In practice this does not happen. 1492 */ 1493 if (incr == 0) 1494 incr = 1; 1495 tickincr = incr; 1496 #ifdef SMP 1497 /* 1498 * Set the default balance interval now that we know 1499 * what realstathz is. 1500 */ 1501 balance_interval = realstathz; 1502 balance_ticks = balance_interval; 1503 affinity = SCHED_AFFINITY_DEFAULT; 1504 #endif 1505 if (sched_idlespinthresh < 0) 1506 sched_idlespinthresh = 2 * max(10000, 6 * hz) / realstathz; 1507 } 1508 1509 /* 1510 * This is the core of the interactivity algorithm. Determines a score based 1511 * on past behavior. It is the ratio of sleep time to run time scaled to 1512 * a [0, 100] integer. This is the voluntary sleep time of a process, which 1513 * differs from the cpu usage because it does not account for time spent 1514 * waiting on a run-queue. Would be prettier if we had floating point. 1515 * 1516 * When a thread's sleep time is greater than its run time the 1517 * calculation is: 1518 * 1519 * scaling factor 1520 * interactivity score = --------------------- 1521 * sleep time / run time 1522 * 1523 * 1524 * When a thread's run time is greater than its sleep time the 1525 * calculation is: 1526 * 1527 * scaling factor 1528 * interactivity score = --------------------- + scaling factor 1529 * run time / sleep time 1530 */ 1531 static int 1532 sched_interact_score(struct thread *td) 1533 { 1534 struct td_sched *ts; 1535 int div; 1536 1537 ts = td_get_sched(td); 1538 /* 1539 * The score is only needed if this is likely to be an interactive 1540 * task. Don't go through the expense of computing it if there's 1541 * no chance. 1542 */ 1543 if (sched_interact <= SCHED_INTERACT_HALF && 1544 ts->ts_runtime >= ts->ts_slptime) 1545 return (SCHED_INTERACT_HALF); 1546 1547 if (ts->ts_runtime > ts->ts_slptime) { 1548 div = max(1, ts->ts_runtime / SCHED_INTERACT_HALF); 1549 return (SCHED_INTERACT_HALF + 1550 (SCHED_INTERACT_HALF - (ts->ts_slptime / div))); 1551 } 1552 if (ts->ts_slptime > ts->ts_runtime) { 1553 div = max(1, ts->ts_slptime / SCHED_INTERACT_HALF); 1554 return (ts->ts_runtime / div); 1555 } 1556 /* runtime == slptime */ 1557 if (ts->ts_runtime) 1558 return (SCHED_INTERACT_HALF); 1559 1560 /* 1561 * This can happen if slptime and runtime are 0. 1562 */ 1563 return (0); 1564 1565 } 1566 1567 /* 1568 * Scale the scheduling priority according to the "interactivity" of this 1569 * process. 1570 */ 1571 static void 1572 sched_priority(struct thread *td) 1573 { 1574 int score; 1575 int pri; 1576 1577 if (PRI_BASE(td->td_pri_class) != PRI_TIMESHARE) 1578 return; 1579 /* 1580 * If the score is interactive we place the thread in the realtime 1581 * queue with a priority that is less than kernel and interrupt 1582 * priorities. These threads are not subject to nice restrictions. 1583 * 1584 * Scores greater than this are placed on the normal timeshare queue 1585 * where the priority is partially decided by the most recent cpu 1586 * utilization and the rest is decided by nice value. 1587 * 1588 * The nice value of the process has a linear effect on the calculated 1589 * score. Negative nice values make it easier for a thread to be 1590 * considered interactive. 1591 */ 1592 score = imax(0, sched_interact_score(td) + td->td_proc->p_nice); 1593 if (score < sched_interact) { 1594 pri = PRI_MIN_INTERACT; 1595 pri += ((PRI_MAX_INTERACT - PRI_MIN_INTERACT + 1) / 1596 sched_interact) * score; 1597 KASSERT(pri >= PRI_MIN_INTERACT && pri <= PRI_MAX_INTERACT, 1598 ("sched_priority: invalid interactive priority %d score %d", 1599 pri, score)); 1600 } else { 1601 pri = SCHED_PRI_MIN; 1602 if (td_get_sched(td)->ts_ticks) 1603 pri += min(SCHED_PRI_TICKS(td_get_sched(td)), 1604 SCHED_PRI_RANGE - 1); 1605 pri += SCHED_PRI_NICE(td->td_proc->p_nice); 1606 KASSERT(pri >= PRI_MIN_BATCH && pri <= PRI_MAX_BATCH, 1607 ("sched_priority: invalid priority %d: nice %d, " 1608 "ticks %d ftick %d ltick %d tick pri %d", 1609 pri, td->td_proc->p_nice, td_get_sched(td)->ts_ticks, 1610 td_get_sched(td)->ts_ftick, td_get_sched(td)->ts_ltick, 1611 SCHED_PRI_TICKS(td_get_sched(td)))); 1612 } 1613 sched_user_prio(td, pri); 1614 1615 return; 1616 } 1617 1618 /* 1619 * This routine enforces a maximum limit on the amount of scheduling history 1620 * kept. It is called after either the slptime or runtime is adjusted. This 1621 * function is ugly due to integer math. 1622 */ 1623 static void 1624 sched_interact_update(struct thread *td) 1625 { 1626 struct td_sched *ts; 1627 u_int sum; 1628 1629 ts = td_get_sched(td); 1630 sum = ts->ts_runtime + ts->ts_slptime; 1631 if (sum < SCHED_SLP_RUN_MAX) 1632 return; 1633 /* 1634 * This only happens from two places: 1635 * 1) We have added an unusual amount of run time from fork_exit. 1636 * 2) We have added an unusual amount of sleep time from sched_sleep(). 1637 */ 1638 if (sum > SCHED_SLP_RUN_MAX * 2) { 1639 if (ts->ts_runtime > ts->ts_slptime) { 1640 ts->ts_runtime = SCHED_SLP_RUN_MAX; 1641 ts->ts_slptime = 1; 1642 } else { 1643 ts->ts_slptime = SCHED_SLP_RUN_MAX; 1644 ts->ts_runtime = 1; 1645 } 1646 return; 1647 } 1648 /* 1649 * If we have exceeded by more than 1/5th then the algorithm below 1650 * will not bring us back into range. Dividing by two here forces 1651 * us into the range of [4/5 * SCHED_INTERACT_MAX, SCHED_INTERACT_MAX] 1652 */ 1653 if (sum > (SCHED_SLP_RUN_MAX / 5) * 6) { 1654 ts->ts_runtime /= 2; 1655 ts->ts_slptime /= 2; 1656 return; 1657 } 1658 ts->ts_runtime = (ts->ts_runtime / 5) * 4; 1659 ts->ts_slptime = (ts->ts_slptime / 5) * 4; 1660 } 1661 1662 /* 1663 * Scale back the interactivity history when a child thread is created. The 1664 * history is inherited from the parent but the thread may behave totally 1665 * differently. For example, a shell spawning a compiler process. We want 1666 * to learn that the compiler is behaving badly very quickly. 1667 */ 1668 static void 1669 sched_interact_fork(struct thread *td) 1670 { 1671 struct td_sched *ts; 1672 int ratio; 1673 int sum; 1674 1675 ts = td_get_sched(td); 1676 sum = ts->ts_runtime + ts->ts_slptime; 1677 if (sum > SCHED_SLP_RUN_FORK) { 1678 ratio = sum / SCHED_SLP_RUN_FORK; 1679 ts->ts_runtime /= ratio; 1680 ts->ts_slptime /= ratio; 1681 } 1682 } 1683 1684 /* 1685 * Called from proc0_init() to setup the scheduler fields. 1686 */ 1687 void 1688 schedinit(void) 1689 { 1690 struct td_sched *ts0; 1691 1692 /* 1693 * Set up the scheduler specific parts of thread0. 1694 */ 1695 ts0 = td_get_sched(&thread0); 1696 ts0->ts_ltick = ticks; 1697 ts0->ts_ftick = ticks; 1698 ts0->ts_slice = 0; 1699 ts0->ts_cpu = curcpu; /* set valid CPU number */ 1700 } 1701 1702 /* 1703 * This is only somewhat accurate since given many processes of the same 1704 * priority they will switch when their slices run out, which will be 1705 * at most sched_slice stathz ticks. 1706 */ 1707 int 1708 sched_rr_interval(void) 1709 { 1710 1711 /* Convert sched_slice from stathz to hz. */ 1712 return (imax(1, (sched_slice * hz + realstathz / 2) / realstathz)); 1713 } 1714 1715 /* 1716 * Update the percent cpu tracking information when it is requested or 1717 * the total history exceeds the maximum. We keep a sliding history of 1718 * tick counts that slowly decays. This is less precise than the 4BSD 1719 * mechanism since it happens with less regular and frequent events. 1720 */ 1721 static void 1722 sched_pctcpu_update(struct td_sched *ts, int run) 1723 { 1724 int t = ticks; 1725 1726 /* 1727 * The signed difference may be negative if the thread hasn't run for 1728 * over half of the ticks rollover period. 1729 */ 1730 if ((u_int)(t - ts->ts_ltick) >= SCHED_TICK_TARG) { 1731 ts->ts_ticks = 0; 1732 ts->ts_ftick = t - SCHED_TICK_TARG; 1733 } else if (t - ts->ts_ftick >= SCHED_TICK_MAX) { 1734 ts->ts_ticks = (ts->ts_ticks / (ts->ts_ltick - ts->ts_ftick)) * 1735 (ts->ts_ltick - (t - SCHED_TICK_TARG)); 1736 ts->ts_ftick = t - SCHED_TICK_TARG; 1737 } 1738 if (run) 1739 ts->ts_ticks += (t - ts->ts_ltick) << SCHED_TICK_SHIFT; 1740 ts->ts_ltick = t; 1741 } 1742 1743 /* 1744 * Adjust the priority of a thread. Move it to the appropriate run-queue 1745 * if necessary. This is the back-end for several priority related 1746 * functions. 1747 */ 1748 static void 1749 sched_thread_priority(struct thread *td, u_char prio) 1750 { 1751 struct td_sched *ts; 1752 struct tdq *tdq; 1753 int oldpri; 1754 1755 KTR_POINT3(KTR_SCHED, "thread", sched_tdname(td), "prio", 1756 "prio:%d", td->td_priority, "new prio:%d", prio, 1757 KTR_ATTR_LINKED, sched_tdname(curthread)); 1758 SDT_PROBE3(sched, , , change__pri, td, td->td_proc, prio); 1759 if (td != curthread && prio < td->td_priority) { 1760 KTR_POINT3(KTR_SCHED, "thread", sched_tdname(curthread), 1761 "lend prio", "prio:%d", td->td_priority, "new prio:%d", 1762 prio, KTR_ATTR_LINKED, sched_tdname(td)); 1763 SDT_PROBE4(sched, , , lend__pri, td, td->td_proc, prio, 1764 curthread); 1765 } 1766 ts = td_get_sched(td); 1767 THREAD_LOCK_ASSERT(td, MA_OWNED); 1768 if (td->td_priority == prio) 1769 return; 1770 /* 1771 * If the priority has been elevated due to priority 1772 * propagation, we may have to move ourselves to a new 1773 * queue. This could be optimized to not re-add in some 1774 * cases. 1775 */ 1776 if (TD_ON_RUNQ(td) && prio < td->td_priority) { 1777 sched_rem(td); 1778 td->td_priority = prio; 1779 sched_add(td, SRQ_BORROWING | SRQ_HOLDTD); 1780 return; 1781 } 1782 /* 1783 * If the thread is currently running we may have to adjust the lowpri 1784 * information so other cpus are aware of our current priority. 1785 */ 1786 if (TD_IS_RUNNING(td)) { 1787 tdq = TDQ_CPU(ts->ts_cpu); 1788 oldpri = td->td_priority; 1789 td->td_priority = prio; 1790 if (prio < tdq->tdq_lowpri) 1791 tdq->tdq_lowpri = prio; 1792 else if (tdq->tdq_lowpri == oldpri) 1793 tdq_setlowpri(tdq, td); 1794 return; 1795 } 1796 td->td_priority = prio; 1797 } 1798 1799 /* 1800 * Update a thread's priority when it is lent another thread's 1801 * priority. 1802 */ 1803 void 1804 sched_lend_prio(struct thread *td, u_char prio) 1805 { 1806 1807 td->td_flags |= TDF_BORROWING; 1808 sched_thread_priority(td, prio); 1809 } 1810 1811 /* 1812 * Restore a thread's priority when priority propagation is 1813 * over. The prio argument is the minimum priority the thread 1814 * needs to have to satisfy other possible priority lending 1815 * requests. If the thread's regular priority is less 1816 * important than prio, the thread will keep a priority boost 1817 * of prio. 1818 */ 1819 void 1820 sched_unlend_prio(struct thread *td, u_char prio) 1821 { 1822 u_char base_pri; 1823 1824 if (td->td_base_pri >= PRI_MIN_TIMESHARE && 1825 td->td_base_pri <= PRI_MAX_TIMESHARE) 1826 base_pri = td->td_user_pri; 1827 else 1828 base_pri = td->td_base_pri; 1829 if (prio >= base_pri) { 1830 td->td_flags &= ~TDF_BORROWING; 1831 sched_thread_priority(td, base_pri); 1832 } else 1833 sched_lend_prio(td, prio); 1834 } 1835 1836 /* 1837 * Standard entry for setting the priority to an absolute value. 1838 */ 1839 void 1840 sched_prio(struct thread *td, u_char prio) 1841 { 1842 u_char oldprio; 1843 1844 /* First, update the base priority. */ 1845 td->td_base_pri = prio; 1846 1847 /* 1848 * If the thread is borrowing another thread's priority, don't 1849 * ever lower the priority. 1850 */ 1851 if (td->td_flags & TDF_BORROWING && td->td_priority < prio) 1852 return; 1853 1854 /* Change the real priority. */ 1855 oldprio = td->td_priority; 1856 sched_thread_priority(td, prio); 1857 1858 /* 1859 * If the thread is on a turnstile, then let the turnstile update 1860 * its state. 1861 */ 1862 if (TD_ON_LOCK(td) && oldprio != prio) 1863 turnstile_adjust(td, oldprio); 1864 } 1865 1866 /* 1867 * Set the base user priority, does not effect current running priority. 1868 */ 1869 void 1870 sched_user_prio(struct thread *td, u_char prio) 1871 { 1872 1873 td->td_base_user_pri = prio; 1874 if (td->td_lend_user_pri <= prio) 1875 return; 1876 td->td_user_pri = prio; 1877 } 1878 1879 void 1880 sched_lend_user_prio(struct thread *td, u_char prio) 1881 { 1882 1883 THREAD_LOCK_ASSERT(td, MA_OWNED); 1884 td->td_lend_user_pri = prio; 1885 td->td_user_pri = min(prio, td->td_base_user_pri); 1886 if (td->td_priority > td->td_user_pri) 1887 sched_prio(td, td->td_user_pri); 1888 else if (td->td_priority != td->td_user_pri) 1889 td->td_flags |= TDF_NEEDRESCHED; 1890 } 1891 1892 /* 1893 * Like the above but first check if there is anything to do. 1894 */ 1895 void 1896 sched_lend_user_prio_cond(struct thread *td, u_char prio) 1897 { 1898 1899 if (td->td_lend_user_pri != prio) 1900 goto lend; 1901 if (td->td_user_pri != min(prio, td->td_base_user_pri)) 1902 goto lend; 1903 if (td->td_priority >= td->td_user_pri) 1904 goto lend; 1905 return; 1906 1907 lend: 1908 thread_lock(td); 1909 sched_lend_user_prio(td, prio); 1910 thread_unlock(td); 1911 } 1912 1913 #ifdef SMP 1914 /* 1915 * This tdq is about to idle. Try to steal a thread from another CPU before 1916 * choosing the idle thread. 1917 */ 1918 static void 1919 tdq_trysteal(struct tdq *tdq) 1920 { 1921 struct cpu_group *cg; 1922 struct tdq *steal; 1923 cpuset_t mask; 1924 int cpu, i; 1925 1926 if (smp_started == 0 || trysteal_limit == 0 || tdq->tdq_cg == NULL) 1927 return; 1928 CPU_FILL(&mask); 1929 CPU_CLR(PCPU_GET(cpuid), &mask); 1930 /* We don't want to be preempted while we're iterating. */ 1931 spinlock_enter(); 1932 TDQ_UNLOCK(tdq); 1933 for (i = 1, cg = tdq->tdq_cg; ; ) { 1934 cpu = sched_highest(cg, mask, steal_thresh); 1935 /* 1936 * If a thread was added while interrupts were disabled don't 1937 * steal one here. 1938 */ 1939 if (tdq->tdq_load > 0) { 1940 TDQ_LOCK(tdq); 1941 break; 1942 } 1943 if (cpu == -1) { 1944 i++; 1945 cg = cg->cg_parent; 1946 if (cg == NULL || i > trysteal_limit) { 1947 TDQ_LOCK(tdq); 1948 break; 1949 } 1950 continue; 1951 } 1952 steal = TDQ_CPU(cpu); 1953 /* 1954 * The data returned by sched_highest() is stale and 1955 * the chosen CPU no longer has an eligible thread. 1956 */ 1957 if (steal->tdq_load < steal_thresh || 1958 steal->tdq_transferable == 0) 1959 continue; 1960 tdq_lock_pair(tdq, steal); 1961 /* 1962 * If we get to this point, unconditonally exit the loop 1963 * to bound the time spent in the critcal section. 1964 * 1965 * If a thread was added while interrupts were disabled don't 1966 * steal one here. 1967 */ 1968 if (tdq->tdq_load > 0) { 1969 TDQ_UNLOCK(steal); 1970 break; 1971 } 1972 /* 1973 * The data returned by sched_highest() is stale and 1974 * the chosen CPU no longer has an eligible thread. 1975 */ 1976 if (steal->tdq_load < steal_thresh || 1977 steal->tdq_transferable == 0) { 1978 TDQ_UNLOCK(steal); 1979 break; 1980 } 1981 /* 1982 * If we fail to acquire one due to affinity restrictions, 1983 * bail out and let the idle thread to a more complete search 1984 * outside of a critical section. 1985 */ 1986 if (tdq_move(steal, tdq) == NULL) { 1987 TDQ_UNLOCK(steal); 1988 break; 1989 } 1990 TDQ_UNLOCK(steal); 1991 break; 1992 } 1993 spinlock_exit(); 1994 } 1995 #endif 1996 1997 /* 1998 * Handle migration from sched_switch(). This happens only for 1999 * cpu binding. 2000 */ 2001 static struct mtx * 2002 sched_switch_migrate(struct tdq *tdq, struct thread *td, int flags) 2003 { 2004 struct tdq *tdn; 2005 2006 KASSERT(THREAD_CAN_MIGRATE(td) || 2007 (td_get_sched(td)->ts_flags & TSF_BOUND) != 0, 2008 ("Thread %p shouldn't migrate", td)); 2009 KASSERT(!CPU_ABSENT(td_get_sched(td)->ts_cpu), ("sched_switch_migrate: " 2010 "thread %s queued on absent CPU %d.", td->td_name, 2011 td_get_sched(td)->ts_cpu)); 2012 tdn = TDQ_CPU(td_get_sched(td)->ts_cpu); 2013 #ifdef SMP 2014 tdq_load_rem(tdq, td); 2015 /* 2016 * Do the lock dance required to avoid LOR. We have an 2017 * extra spinlock nesting from sched_switch() which will 2018 * prevent preemption while we're holding neither run-queue lock. 2019 */ 2020 TDQ_UNLOCK(tdq); 2021 TDQ_LOCK(tdn); 2022 tdq_add(tdn, td, flags); 2023 tdq_notify(tdn, td); 2024 TDQ_UNLOCK(tdn); 2025 TDQ_LOCK(tdq); 2026 #endif 2027 return (TDQ_LOCKPTR(tdn)); 2028 } 2029 2030 /* 2031 * thread_lock_unblock() that does not assume td_lock is blocked. 2032 */ 2033 static inline void 2034 thread_unblock_switch(struct thread *td, struct mtx *mtx) 2035 { 2036 atomic_store_rel_ptr((volatile uintptr_t *)&td->td_lock, 2037 (uintptr_t)mtx); 2038 } 2039 2040 /* 2041 * Switch threads. This function has to handle threads coming in while 2042 * blocked for some reason, running, or idle. It also must deal with 2043 * migrating a thread from one queue to another as running threads may 2044 * be assigned elsewhere via binding. 2045 */ 2046 void 2047 sched_switch(struct thread *td, int flags) 2048 { 2049 struct thread *newtd; 2050 struct tdq *tdq; 2051 struct td_sched *ts; 2052 struct mtx *mtx; 2053 int srqflag; 2054 int cpuid, preempted; 2055 2056 THREAD_LOCK_ASSERT(td, MA_OWNED); 2057 2058 cpuid = PCPU_GET(cpuid); 2059 tdq = TDQ_SELF(); 2060 ts = td_get_sched(td); 2061 sched_pctcpu_update(ts, 1); 2062 ts->ts_rltick = ticks; 2063 td->td_lastcpu = td->td_oncpu; 2064 preempted = (td->td_flags & TDF_SLICEEND) == 0 && 2065 (flags & SW_PREEMPT) != 0; 2066 td->td_flags &= ~(TDF_NEEDRESCHED | TDF_SLICEEND); 2067 td->td_owepreempt = 0; 2068 tdq->tdq_owepreempt = 0; 2069 if (!TD_IS_IDLETHREAD(td)) 2070 tdq->tdq_switchcnt++; 2071 2072 /* 2073 * Always block the thread lock so we can drop the tdq lock early. 2074 */ 2075 mtx = thread_lock_block(td); 2076 spinlock_enter(); 2077 if (TD_IS_IDLETHREAD(td)) { 2078 MPASS(mtx == TDQ_LOCKPTR(tdq)); 2079 TD_SET_CAN_RUN(td); 2080 } else if (TD_IS_RUNNING(td)) { 2081 MPASS(mtx == TDQ_LOCKPTR(tdq)); 2082 srqflag = preempted ? 2083 SRQ_OURSELF|SRQ_YIELDING|SRQ_PREEMPTED : 2084 SRQ_OURSELF|SRQ_YIELDING; 2085 #ifdef SMP 2086 if (THREAD_CAN_MIGRATE(td) && !THREAD_CAN_SCHED(td, ts->ts_cpu)) 2087 ts->ts_cpu = sched_pickcpu(td, 0); 2088 #endif 2089 if (ts->ts_cpu == cpuid) 2090 tdq_runq_add(tdq, td, srqflag); 2091 else 2092 mtx = sched_switch_migrate(tdq, td, srqflag); 2093 } else { 2094 /* This thread must be going to sleep. */ 2095 if (mtx != TDQ_LOCKPTR(tdq)) { 2096 mtx_unlock_spin(mtx); 2097 TDQ_LOCK(tdq); 2098 } 2099 tdq_load_rem(tdq, td); 2100 #ifdef SMP 2101 if (tdq->tdq_load == 0) 2102 tdq_trysteal(tdq); 2103 #endif 2104 } 2105 2106 #if (KTR_COMPILE & KTR_SCHED) != 0 2107 if (TD_IS_IDLETHREAD(td)) 2108 KTR_STATE1(KTR_SCHED, "thread", sched_tdname(td), "idle", 2109 "prio:%d", td->td_priority); 2110 else 2111 KTR_STATE3(KTR_SCHED, "thread", sched_tdname(td), KTDSTATE(td), 2112 "prio:%d", td->td_priority, "wmesg:\"%s\"", td->td_wmesg, 2113 "lockname:\"%s\"", td->td_lockname); 2114 #endif 2115 2116 /* 2117 * We enter here with the thread blocked and assigned to the 2118 * appropriate cpu run-queue or sleep-queue and with the current 2119 * thread-queue locked. 2120 */ 2121 TDQ_LOCK_ASSERT(tdq, MA_OWNED | MA_NOTRECURSED); 2122 newtd = choosethread(); 2123 sched_pctcpu_update(td_get_sched(newtd), 0); 2124 TDQ_UNLOCK(tdq); 2125 2126 /* 2127 * Call the MD code to switch contexts if necessary. 2128 */ 2129 if (td != newtd) { 2130 #ifdef HWPMC_HOOKS 2131 if (PMC_PROC_IS_USING_PMCS(td->td_proc)) 2132 PMC_SWITCH_CONTEXT(td, PMC_FN_CSW_OUT); 2133 #endif 2134 SDT_PROBE2(sched, , , off__cpu, newtd, newtd->td_proc); 2135 2136 #ifdef KDTRACE_HOOKS 2137 /* 2138 * If DTrace has set the active vtime enum to anything 2139 * other than INACTIVE (0), then it should have set the 2140 * function to call. 2141 */ 2142 if (dtrace_vtime_active) 2143 (*dtrace_vtime_switch_func)(newtd); 2144 #endif 2145 td->td_oncpu = NOCPU; 2146 cpu_switch(td, newtd, mtx); 2147 cpuid = td->td_oncpu = PCPU_GET(cpuid); 2148 2149 SDT_PROBE0(sched, , , on__cpu); 2150 #ifdef HWPMC_HOOKS 2151 if (PMC_PROC_IS_USING_PMCS(td->td_proc)) 2152 PMC_SWITCH_CONTEXT(td, PMC_FN_CSW_IN); 2153 #endif 2154 } else { 2155 thread_unblock_switch(td, mtx); 2156 SDT_PROBE0(sched, , , remain__cpu); 2157 } 2158 KASSERT(curthread->td_md.md_spinlock_count == 1, 2159 ("invalid count %d", curthread->td_md.md_spinlock_count)); 2160 2161 KTR_STATE1(KTR_SCHED, "thread", sched_tdname(td), "running", 2162 "prio:%d", td->td_priority); 2163 } 2164 2165 /* 2166 * Adjust thread priorities as a result of a nice request. 2167 */ 2168 void 2169 sched_nice(struct proc *p, int nice) 2170 { 2171 struct thread *td; 2172 2173 PROC_LOCK_ASSERT(p, MA_OWNED); 2174 2175 p->p_nice = nice; 2176 FOREACH_THREAD_IN_PROC(p, td) { 2177 thread_lock(td); 2178 sched_priority(td); 2179 sched_prio(td, td->td_base_user_pri); 2180 thread_unlock(td); 2181 } 2182 } 2183 2184 /* 2185 * Record the sleep time for the interactivity scorer. 2186 */ 2187 void 2188 sched_sleep(struct thread *td, int prio) 2189 { 2190 2191 THREAD_LOCK_ASSERT(td, MA_OWNED); 2192 2193 td->td_slptick = ticks; 2194 if (TD_IS_SUSPENDED(td) || prio >= PSOCK) 2195 td->td_flags |= TDF_CANSWAP; 2196 if (PRI_BASE(td->td_pri_class) != PRI_TIMESHARE) 2197 return; 2198 if (static_boost == 1 && prio) 2199 sched_prio(td, prio); 2200 else if (static_boost && td->td_priority > static_boost) 2201 sched_prio(td, static_boost); 2202 } 2203 2204 /* 2205 * Schedule a thread to resume execution and record how long it voluntarily 2206 * slept. We also update the pctcpu, interactivity, and priority. 2207 * 2208 * Requires the thread lock on entry, drops on exit. 2209 */ 2210 void 2211 sched_wakeup(struct thread *td, int srqflags) 2212 { 2213 struct td_sched *ts; 2214 int slptick; 2215 2216 THREAD_LOCK_ASSERT(td, MA_OWNED); 2217 ts = td_get_sched(td); 2218 td->td_flags &= ~TDF_CANSWAP; 2219 2220 /* 2221 * If we slept for more than a tick update our interactivity and 2222 * priority. 2223 */ 2224 slptick = td->td_slptick; 2225 td->td_slptick = 0; 2226 if (slptick && slptick != ticks) { 2227 ts->ts_slptime += (ticks - slptick) << SCHED_TICK_SHIFT; 2228 sched_interact_update(td); 2229 sched_pctcpu_update(ts, 0); 2230 } 2231 /* 2232 * Reset the slice value since we slept and advanced the round-robin. 2233 */ 2234 ts->ts_slice = 0; 2235 sched_add(td, SRQ_BORING | srqflags); 2236 } 2237 2238 /* 2239 * Penalize the parent for creating a new child and initialize the child's 2240 * priority. 2241 */ 2242 void 2243 sched_fork(struct thread *td, struct thread *child) 2244 { 2245 THREAD_LOCK_ASSERT(td, MA_OWNED); 2246 sched_pctcpu_update(td_get_sched(td), 1); 2247 sched_fork_thread(td, child); 2248 /* 2249 * Penalize the parent and child for forking. 2250 */ 2251 sched_interact_fork(child); 2252 sched_priority(child); 2253 td_get_sched(td)->ts_runtime += tickincr; 2254 sched_interact_update(td); 2255 sched_priority(td); 2256 } 2257 2258 /* 2259 * Fork a new thread, may be within the same process. 2260 */ 2261 void 2262 sched_fork_thread(struct thread *td, struct thread *child) 2263 { 2264 struct td_sched *ts; 2265 struct td_sched *ts2; 2266 struct tdq *tdq; 2267 2268 tdq = TDQ_SELF(); 2269 THREAD_LOCK_ASSERT(td, MA_OWNED); 2270 /* 2271 * Initialize child. 2272 */ 2273 ts = td_get_sched(td); 2274 ts2 = td_get_sched(child); 2275 child->td_oncpu = NOCPU; 2276 child->td_lastcpu = NOCPU; 2277 child->td_lock = TDQ_LOCKPTR(tdq); 2278 child->td_cpuset = cpuset_ref(td->td_cpuset); 2279 child->td_domain.dr_policy = td->td_cpuset->cs_domain; 2280 ts2->ts_cpu = ts->ts_cpu; 2281 ts2->ts_flags = 0; 2282 /* 2283 * Grab our parents cpu estimation information. 2284 */ 2285 ts2->ts_ticks = ts->ts_ticks; 2286 ts2->ts_ltick = ts->ts_ltick; 2287 ts2->ts_ftick = ts->ts_ftick; 2288 /* 2289 * Do not inherit any borrowed priority from the parent. 2290 */ 2291 child->td_priority = child->td_base_pri; 2292 /* 2293 * And update interactivity score. 2294 */ 2295 ts2->ts_slptime = ts->ts_slptime; 2296 ts2->ts_runtime = ts->ts_runtime; 2297 /* Attempt to quickly learn interactivity. */ 2298 ts2->ts_slice = tdq_slice(tdq) - sched_slice_min; 2299 #ifdef KTR 2300 bzero(ts2->ts_name, sizeof(ts2->ts_name)); 2301 #endif 2302 } 2303 2304 /* 2305 * Adjust the priority class of a thread. 2306 */ 2307 void 2308 sched_class(struct thread *td, int class) 2309 { 2310 2311 THREAD_LOCK_ASSERT(td, MA_OWNED); 2312 if (td->td_pri_class == class) 2313 return; 2314 td->td_pri_class = class; 2315 } 2316 2317 /* 2318 * Return some of the child's priority and interactivity to the parent. 2319 */ 2320 void 2321 sched_exit(struct proc *p, struct thread *child) 2322 { 2323 struct thread *td; 2324 2325 KTR_STATE1(KTR_SCHED, "thread", sched_tdname(child), "proc exit", 2326 "prio:%d", child->td_priority); 2327 PROC_LOCK_ASSERT(p, MA_OWNED); 2328 td = FIRST_THREAD_IN_PROC(p); 2329 sched_exit_thread(td, child); 2330 } 2331 2332 /* 2333 * Penalize another thread for the time spent on this one. This helps to 2334 * worsen the priority and interactivity of processes which schedule batch 2335 * jobs such as make. This has little effect on the make process itself but 2336 * causes new processes spawned by it to receive worse scores immediately. 2337 */ 2338 void 2339 sched_exit_thread(struct thread *td, struct thread *child) 2340 { 2341 2342 KTR_STATE1(KTR_SCHED, "thread", sched_tdname(child), "thread exit", 2343 "prio:%d", child->td_priority); 2344 /* 2345 * Give the child's runtime to the parent without returning the 2346 * sleep time as a penalty to the parent. This causes shells that 2347 * launch expensive things to mark their children as expensive. 2348 */ 2349 thread_lock(td); 2350 td_get_sched(td)->ts_runtime += td_get_sched(child)->ts_runtime; 2351 sched_interact_update(td); 2352 sched_priority(td); 2353 thread_unlock(td); 2354 } 2355 2356 void 2357 sched_preempt(struct thread *td) 2358 { 2359 struct tdq *tdq; 2360 int flags; 2361 2362 SDT_PROBE2(sched, , , surrender, td, td->td_proc); 2363 2364 thread_lock(td); 2365 tdq = TDQ_SELF(); 2366 TDQ_LOCK_ASSERT(tdq, MA_OWNED); 2367 if (td->td_priority > tdq->tdq_lowpri) { 2368 if (td->td_critnest == 1) { 2369 flags = SW_INVOL | SW_PREEMPT; 2370 flags |= TD_IS_IDLETHREAD(td) ? SWT_REMOTEWAKEIDLE : 2371 SWT_REMOTEPREEMPT; 2372 mi_switch(flags); 2373 /* Switch dropped thread lock. */ 2374 return; 2375 } 2376 td->td_owepreempt = 1; 2377 } else { 2378 tdq->tdq_owepreempt = 0; 2379 } 2380 thread_unlock(td); 2381 } 2382 2383 /* 2384 * Fix priorities on return to user-space. Priorities may be elevated due 2385 * to static priorities in msleep() or similar. 2386 */ 2387 void 2388 sched_userret_slowpath(struct thread *td) 2389 { 2390 2391 thread_lock(td); 2392 td->td_priority = td->td_user_pri; 2393 td->td_base_pri = td->td_user_pri; 2394 tdq_setlowpri(TDQ_SELF(), td); 2395 thread_unlock(td); 2396 } 2397 2398 /* 2399 * Handle a stathz tick. This is really only relevant for timeshare 2400 * threads. 2401 */ 2402 void 2403 sched_clock(struct thread *td, int cnt) 2404 { 2405 struct tdq *tdq; 2406 struct td_sched *ts; 2407 2408 THREAD_LOCK_ASSERT(td, MA_OWNED); 2409 tdq = TDQ_SELF(); 2410 #ifdef SMP 2411 /* 2412 * We run the long term load balancer infrequently on the first cpu. 2413 */ 2414 if (balance_tdq == tdq && smp_started != 0 && rebalance != 0 && 2415 balance_ticks != 0) { 2416 balance_ticks -= cnt; 2417 if (balance_ticks <= 0) 2418 sched_balance(); 2419 } 2420 #endif 2421 /* 2422 * Save the old switch count so we have a record of the last ticks 2423 * activity. Initialize the new switch count based on our load. 2424 * If there is some activity seed it to reflect that. 2425 */ 2426 tdq->tdq_oldswitchcnt = tdq->tdq_switchcnt; 2427 tdq->tdq_switchcnt = tdq->tdq_load; 2428 /* 2429 * Advance the insert index once for each tick to ensure that all 2430 * threads get a chance to run. 2431 */ 2432 if (tdq->tdq_idx == tdq->tdq_ridx) { 2433 tdq->tdq_idx = (tdq->tdq_idx + 1) % RQ_NQS; 2434 if (TAILQ_EMPTY(&tdq->tdq_timeshare.rq_queues[tdq->tdq_ridx])) 2435 tdq->tdq_ridx = tdq->tdq_idx; 2436 } 2437 ts = td_get_sched(td); 2438 sched_pctcpu_update(ts, 1); 2439 if ((td->td_pri_class & PRI_FIFO_BIT) || TD_IS_IDLETHREAD(td)) 2440 return; 2441 2442 if (PRI_BASE(td->td_pri_class) == PRI_TIMESHARE) { 2443 /* 2444 * We used a tick; charge it to the thread so 2445 * that we can compute our interactivity. 2446 */ 2447 td_get_sched(td)->ts_runtime += tickincr * cnt; 2448 sched_interact_update(td); 2449 sched_priority(td); 2450 } 2451 2452 /* 2453 * Force a context switch if the current thread has used up a full 2454 * time slice (default is 100ms). 2455 */ 2456 ts->ts_slice += cnt; 2457 if (ts->ts_slice >= tdq_slice(tdq)) { 2458 ts->ts_slice = 0; 2459 td->td_flags |= TDF_NEEDRESCHED | TDF_SLICEEND; 2460 } 2461 } 2462 2463 u_int 2464 sched_estcpu(struct thread *td __unused) 2465 { 2466 2467 return (0); 2468 } 2469 2470 /* 2471 * Return whether the current CPU has runnable tasks. Used for in-kernel 2472 * cooperative idle threads. 2473 */ 2474 int 2475 sched_runnable(void) 2476 { 2477 struct tdq *tdq; 2478 int load; 2479 2480 load = 1; 2481 2482 tdq = TDQ_SELF(); 2483 if ((curthread->td_flags & TDF_IDLETD) != 0) { 2484 if (tdq->tdq_load > 0) 2485 goto out; 2486 } else 2487 if (tdq->tdq_load - 1 > 0) 2488 goto out; 2489 load = 0; 2490 out: 2491 return (load); 2492 } 2493 2494 /* 2495 * Choose the highest priority thread to run. The thread is removed from 2496 * the run-queue while running however the load remains. For SMP we set 2497 * the tdq in the global idle bitmask if it idles here. 2498 */ 2499 struct thread * 2500 sched_choose(void) 2501 { 2502 struct thread *td; 2503 struct tdq *tdq; 2504 2505 tdq = TDQ_SELF(); 2506 TDQ_LOCK_ASSERT(tdq, MA_OWNED); 2507 td = tdq_choose(tdq); 2508 if (td) { 2509 tdq_runq_rem(tdq, td); 2510 tdq->tdq_lowpri = td->td_priority; 2511 return (td); 2512 } 2513 tdq->tdq_lowpri = PRI_MAX_IDLE; 2514 return (PCPU_GET(idlethread)); 2515 } 2516 2517 /* 2518 * Set owepreempt if necessary. Preemption never happens directly in ULE, 2519 * we always request it once we exit a critical section. 2520 */ 2521 static inline void 2522 sched_setpreempt(struct thread *td) 2523 { 2524 struct thread *ctd; 2525 int cpri; 2526 int pri; 2527 2528 THREAD_LOCK_ASSERT(curthread, MA_OWNED); 2529 2530 ctd = curthread; 2531 pri = td->td_priority; 2532 cpri = ctd->td_priority; 2533 if (pri < cpri) 2534 ctd->td_flags |= TDF_NEEDRESCHED; 2535 if (KERNEL_PANICKED() || pri >= cpri || cold || TD_IS_INHIBITED(ctd)) 2536 return; 2537 if (!sched_shouldpreempt(pri, cpri, 0)) 2538 return; 2539 ctd->td_owepreempt = 1; 2540 } 2541 2542 /* 2543 * Add a thread to a thread queue. Select the appropriate runq and add the 2544 * thread to it. This is the internal function called when the tdq is 2545 * predetermined. 2546 */ 2547 void 2548 tdq_add(struct tdq *tdq, struct thread *td, int flags) 2549 { 2550 2551 TDQ_LOCK_ASSERT(tdq, MA_OWNED); 2552 THREAD_LOCK_BLOCKED_ASSERT(td, MA_OWNED); 2553 KASSERT((td->td_inhibitors == 0), 2554 ("sched_add: trying to run inhibited thread")); 2555 KASSERT((TD_CAN_RUN(td) || TD_IS_RUNNING(td)), 2556 ("sched_add: bad thread state")); 2557 KASSERT(td->td_flags & TDF_INMEM, 2558 ("sched_add: thread swapped out")); 2559 2560 if (td->td_priority < tdq->tdq_lowpri) 2561 tdq->tdq_lowpri = td->td_priority; 2562 tdq_runq_add(tdq, td, flags); 2563 tdq_load_add(tdq, td); 2564 } 2565 2566 /* 2567 * Select the target thread queue and add a thread to it. Request 2568 * preemption or IPI a remote processor if required. 2569 * 2570 * Requires the thread lock on entry, drops on exit. 2571 */ 2572 void 2573 sched_add(struct thread *td, int flags) 2574 { 2575 struct tdq *tdq; 2576 #ifdef SMP 2577 int cpu; 2578 #endif 2579 2580 KTR_STATE2(KTR_SCHED, "thread", sched_tdname(td), "runq add", 2581 "prio:%d", td->td_priority, KTR_ATTR_LINKED, 2582 sched_tdname(curthread)); 2583 KTR_POINT1(KTR_SCHED, "thread", sched_tdname(curthread), "wokeup", 2584 KTR_ATTR_LINKED, sched_tdname(td)); 2585 SDT_PROBE4(sched, , , enqueue, td, td->td_proc, NULL, 2586 flags & SRQ_PREEMPTED); 2587 THREAD_LOCK_ASSERT(td, MA_OWNED); 2588 /* 2589 * Recalculate the priority before we select the target cpu or 2590 * run-queue. 2591 */ 2592 if (PRI_BASE(td->td_pri_class) == PRI_TIMESHARE) 2593 sched_priority(td); 2594 #ifdef SMP 2595 /* 2596 * Pick the destination cpu and if it isn't ours transfer to the 2597 * target cpu. 2598 */ 2599 cpu = sched_pickcpu(td, flags); 2600 tdq = sched_setcpu(td, cpu, flags); 2601 tdq_add(tdq, td, flags); 2602 if (cpu != PCPU_GET(cpuid)) 2603 tdq_notify(tdq, td); 2604 else if (!(flags & SRQ_YIELDING)) 2605 sched_setpreempt(td); 2606 #else 2607 tdq = TDQ_SELF(); 2608 /* 2609 * Now that the thread is moving to the run-queue, set the lock 2610 * to the scheduler's lock. 2611 */ 2612 if (td->td_lock != TDQ_LOCKPTR(tdq)) { 2613 TDQ_LOCK(tdq); 2614 if ((flags & SRQ_HOLD) != 0) 2615 td->td_lock = TDQ_LOCKPTR(tdq); 2616 else 2617 thread_lock_set(td, TDQ_LOCKPTR(tdq)); 2618 } 2619 tdq_add(tdq, td, flags); 2620 if (!(flags & SRQ_YIELDING)) 2621 sched_setpreempt(td); 2622 #endif 2623 if (!(flags & SRQ_HOLDTD)) 2624 thread_unlock(td); 2625 } 2626 2627 /* 2628 * Remove a thread from a run-queue without running it. This is used 2629 * when we're stealing a thread from a remote queue. Otherwise all threads 2630 * exit by calling sched_exit_thread() and sched_throw() themselves. 2631 */ 2632 void 2633 sched_rem(struct thread *td) 2634 { 2635 struct tdq *tdq; 2636 2637 KTR_STATE1(KTR_SCHED, "thread", sched_tdname(td), "runq rem", 2638 "prio:%d", td->td_priority); 2639 SDT_PROBE3(sched, , , dequeue, td, td->td_proc, NULL); 2640 tdq = TDQ_CPU(td_get_sched(td)->ts_cpu); 2641 TDQ_LOCK_ASSERT(tdq, MA_OWNED); 2642 MPASS(td->td_lock == TDQ_LOCKPTR(tdq)); 2643 KASSERT(TD_ON_RUNQ(td), 2644 ("sched_rem: thread not on run queue")); 2645 tdq_runq_rem(tdq, td); 2646 tdq_load_rem(tdq, td); 2647 TD_SET_CAN_RUN(td); 2648 if (td->td_priority == tdq->tdq_lowpri) 2649 tdq_setlowpri(tdq, NULL); 2650 } 2651 2652 /* 2653 * Fetch cpu utilization information. Updates on demand. 2654 */ 2655 fixpt_t 2656 sched_pctcpu(struct thread *td) 2657 { 2658 fixpt_t pctcpu; 2659 struct td_sched *ts; 2660 2661 pctcpu = 0; 2662 ts = td_get_sched(td); 2663 2664 THREAD_LOCK_ASSERT(td, MA_OWNED); 2665 sched_pctcpu_update(ts, TD_IS_RUNNING(td)); 2666 if (ts->ts_ticks) { 2667 int rtick; 2668 2669 /* How many rtick per second ? */ 2670 rtick = min(SCHED_TICK_HZ(ts) / SCHED_TICK_SECS, hz); 2671 pctcpu = (FSCALE * ((FSCALE * rtick)/hz)) >> FSHIFT; 2672 } 2673 2674 return (pctcpu); 2675 } 2676 2677 /* 2678 * Enforce affinity settings for a thread. Called after adjustments to 2679 * cpumask. 2680 */ 2681 void 2682 sched_affinity(struct thread *td) 2683 { 2684 #ifdef SMP 2685 struct td_sched *ts; 2686 2687 THREAD_LOCK_ASSERT(td, MA_OWNED); 2688 ts = td_get_sched(td); 2689 if (THREAD_CAN_SCHED(td, ts->ts_cpu)) 2690 return; 2691 if (TD_ON_RUNQ(td)) { 2692 sched_rem(td); 2693 sched_add(td, SRQ_BORING | SRQ_HOLDTD); 2694 return; 2695 } 2696 if (!TD_IS_RUNNING(td)) 2697 return; 2698 /* 2699 * Force a switch before returning to userspace. If the 2700 * target thread is not running locally send an ipi to force 2701 * the issue. 2702 */ 2703 td->td_flags |= TDF_NEEDRESCHED; 2704 if (td != curthread) 2705 ipi_cpu(ts->ts_cpu, IPI_PREEMPT); 2706 #endif 2707 } 2708 2709 /* 2710 * Bind a thread to a target cpu. 2711 */ 2712 void 2713 sched_bind(struct thread *td, int cpu) 2714 { 2715 struct td_sched *ts; 2716 2717 THREAD_LOCK_ASSERT(td, MA_OWNED|MA_NOTRECURSED); 2718 KASSERT(td == curthread, ("sched_bind: can only bind curthread")); 2719 ts = td_get_sched(td); 2720 if (ts->ts_flags & TSF_BOUND) 2721 sched_unbind(td); 2722 KASSERT(THREAD_CAN_MIGRATE(td), ("%p must be migratable", td)); 2723 ts->ts_flags |= TSF_BOUND; 2724 sched_pin(); 2725 if (PCPU_GET(cpuid) == cpu) 2726 return; 2727 ts->ts_cpu = cpu; 2728 /* When we return from mi_switch we'll be on the correct cpu. */ 2729 mi_switch(SW_VOL); 2730 thread_lock(td); 2731 } 2732 2733 /* 2734 * Release a bound thread. 2735 */ 2736 void 2737 sched_unbind(struct thread *td) 2738 { 2739 struct td_sched *ts; 2740 2741 THREAD_LOCK_ASSERT(td, MA_OWNED); 2742 KASSERT(td == curthread, ("sched_unbind: can only bind curthread")); 2743 ts = td_get_sched(td); 2744 if ((ts->ts_flags & TSF_BOUND) == 0) 2745 return; 2746 ts->ts_flags &= ~TSF_BOUND; 2747 sched_unpin(); 2748 } 2749 2750 int 2751 sched_is_bound(struct thread *td) 2752 { 2753 THREAD_LOCK_ASSERT(td, MA_OWNED); 2754 return (td_get_sched(td)->ts_flags & TSF_BOUND); 2755 } 2756 2757 /* 2758 * Basic yield call. 2759 */ 2760 void 2761 sched_relinquish(struct thread *td) 2762 { 2763 thread_lock(td); 2764 mi_switch(SW_VOL | SWT_RELINQUISH); 2765 } 2766 2767 /* 2768 * Return the total system load. 2769 */ 2770 int 2771 sched_load(void) 2772 { 2773 #ifdef SMP 2774 int total; 2775 int i; 2776 2777 total = 0; 2778 CPU_FOREACH(i) 2779 total += TDQ_CPU(i)->tdq_sysload; 2780 return (total); 2781 #else 2782 return (TDQ_SELF()->tdq_sysload); 2783 #endif 2784 } 2785 2786 int 2787 sched_sizeof_proc(void) 2788 { 2789 return (sizeof(struct proc)); 2790 } 2791 2792 int 2793 sched_sizeof_thread(void) 2794 { 2795 return (sizeof(struct thread) + sizeof(struct td_sched)); 2796 } 2797 2798 #ifdef SMP 2799 #define TDQ_IDLESPIN(tdq) \ 2800 ((tdq)->tdq_cg != NULL && ((tdq)->tdq_cg->cg_flags & CG_FLAG_THREAD) == 0) 2801 #else 2802 #define TDQ_IDLESPIN(tdq) 1 2803 #endif 2804 2805 /* 2806 * The actual idle process. 2807 */ 2808 void 2809 sched_idletd(void *dummy) 2810 { 2811 struct thread *td; 2812 struct tdq *tdq; 2813 int oldswitchcnt, switchcnt; 2814 int i; 2815 2816 mtx_assert(&Giant, MA_NOTOWNED); 2817 td = curthread; 2818 tdq = TDQ_SELF(); 2819 THREAD_NO_SLEEPING(); 2820 oldswitchcnt = -1; 2821 for (;;) { 2822 if (tdq->tdq_load) { 2823 thread_lock(td); 2824 mi_switch(SW_VOL | SWT_IDLE); 2825 } 2826 switchcnt = tdq->tdq_switchcnt + tdq->tdq_oldswitchcnt; 2827 #ifdef SMP 2828 if (always_steal || switchcnt != oldswitchcnt) { 2829 oldswitchcnt = switchcnt; 2830 if (tdq_idled(tdq) == 0) 2831 continue; 2832 } 2833 switchcnt = tdq->tdq_switchcnt + tdq->tdq_oldswitchcnt; 2834 #else 2835 oldswitchcnt = switchcnt; 2836 #endif 2837 /* 2838 * If we're switching very frequently, spin while checking 2839 * for load rather than entering a low power state that 2840 * may require an IPI. However, don't do any busy 2841 * loops while on SMT machines as this simply steals 2842 * cycles from cores doing useful work. 2843 */ 2844 if (TDQ_IDLESPIN(tdq) && switchcnt > sched_idlespinthresh) { 2845 for (i = 0; i < sched_idlespins; i++) { 2846 if (tdq->tdq_load) 2847 break; 2848 cpu_spinwait(); 2849 } 2850 } 2851 2852 /* If there was context switch during spin, restart it. */ 2853 switchcnt = tdq->tdq_switchcnt + tdq->tdq_oldswitchcnt; 2854 if (tdq->tdq_load != 0 || switchcnt != oldswitchcnt) 2855 continue; 2856 2857 /* Run main MD idle handler. */ 2858 tdq->tdq_cpu_idle = 1; 2859 /* 2860 * Make sure that tdq_cpu_idle update is globally visible 2861 * before cpu_idle() read tdq_load. The order is important 2862 * to avoid race with tdq_notify. 2863 */ 2864 atomic_thread_fence_seq_cst(); 2865 /* 2866 * Checking for again after the fence picks up assigned 2867 * threads often enough to make it worthwhile to do so in 2868 * order to avoid calling cpu_idle(). 2869 */ 2870 if (tdq->tdq_load != 0) { 2871 tdq->tdq_cpu_idle = 0; 2872 continue; 2873 } 2874 cpu_idle(switchcnt * 4 > sched_idlespinthresh); 2875 tdq->tdq_cpu_idle = 0; 2876 2877 /* 2878 * Account thread-less hardware interrupts and 2879 * other wakeup reasons equal to context switches. 2880 */ 2881 switchcnt = tdq->tdq_switchcnt + tdq->tdq_oldswitchcnt; 2882 if (switchcnt != oldswitchcnt) 2883 continue; 2884 tdq->tdq_switchcnt++; 2885 oldswitchcnt++; 2886 } 2887 } 2888 2889 /* 2890 * A CPU is entering for the first time or a thread is exiting. 2891 */ 2892 void 2893 sched_throw(struct thread *td) 2894 { 2895 struct thread *newtd; 2896 struct tdq *tdq; 2897 2898 if (__predict_false(td == NULL)) { 2899 #ifdef SMP 2900 PCPU_SET(sched, DPCPU_PTR(tdq)); 2901 #endif 2902 /* Correct spinlock nesting and acquire the correct lock. */ 2903 tdq = TDQ_SELF(); 2904 TDQ_LOCK(tdq); 2905 spinlock_exit(); 2906 PCPU_SET(switchtime, cpu_ticks()); 2907 PCPU_SET(switchticks, ticks); 2908 PCPU_GET(idlethread)->td_lock = TDQ_LOCKPTR(tdq); 2909 } else { 2910 tdq = TDQ_SELF(); 2911 THREAD_LOCK_ASSERT(td, MA_OWNED); 2912 THREAD_LOCKPTR_ASSERT(td, TDQ_LOCKPTR(tdq)); 2913 tdq_load_rem(tdq, td); 2914 td->td_lastcpu = td->td_oncpu; 2915 td->td_oncpu = NOCPU; 2916 thread_lock_block(td); 2917 } 2918 newtd = choosethread(); 2919 spinlock_enter(); 2920 TDQ_UNLOCK(tdq); 2921 KASSERT(curthread->td_md.md_spinlock_count == 1, 2922 ("invalid count %d", curthread->td_md.md_spinlock_count)); 2923 /* doesn't return */ 2924 if (__predict_false(td == NULL)) 2925 cpu_throw(td, newtd); /* doesn't return */ 2926 else 2927 cpu_switch(td, newtd, TDQ_LOCKPTR(tdq)); 2928 } 2929 2930 /* 2931 * This is called from fork_exit(). Just acquire the correct locks and 2932 * let fork do the rest of the work. 2933 */ 2934 void 2935 sched_fork_exit(struct thread *td) 2936 { 2937 struct tdq *tdq; 2938 int cpuid; 2939 2940 /* 2941 * Finish setting up thread glue so that it begins execution in a 2942 * non-nested critical section with the scheduler lock held. 2943 */ 2944 KASSERT(curthread->td_md.md_spinlock_count == 1, 2945 ("invalid count %d", curthread->td_md.md_spinlock_count)); 2946 cpuid = PCPU_GET(cpuid); 2947 tdq = TDQ_SELF(); 2948 TDQ_LOCK(tdq); 2949 spinlock_exit(); 2950 MPASS(td->td_lock == TDQ_LOCKPTR(tdq)); 2951 td->td_oncpu = cpuid; 2952 KTR_STATE1(KTR_SCHED, "thread", sched_tdname(td), "running", 2953 "prio:%d", td->td_priority); 2954 SDT_PROBE0(sched, , , on__cpu); 2955 } 2956 2957 /* 2958 * Create on first use to catch odd startup conditons. 2959 */ 2960 char * 2961 sched_tdname(struct thread *td) 2962 { 2963 #ifdef KTR 2964 struct td_sched *ts; 2965 2966 ts = td_get_sched(td); 2967 if (ts->ts_name[0] == '\0') 2968 snprintf(ts->ts_name, sizeof(ts->ts_name), 2969 "%s tid %d", td->td_name, td->td_tid); 2970 return (ts->ts_name); 2971 #else 2972 return (td->td_name); 2973 #endif 2974 } 2975 2976 #ifdef KTR 2977 void 2978 sched_clear_tdname(struct thread *td) 2979 { 2980 struct td_sched *ts; 2981 2982 ts = td_get_sched(td); 2983 ts->ts_name[0] = '\0'; 2984 } 2985 #endif 2986 2987 #ifdef SMP 2988 2989 /* 2990 * Build the CPU topology dump string. Is recursively called to collect 2991 * the topology tree. 2992 */ 2993 static int 2994 sysctl_kern_sched_topology_spec_internal(struct sbuf *sb, struct cpu_group *cg, 2995 int indent) 2996 { 2997 char cpusetbuf[CPUSETBUFSIZ]; 2998 int i, first; 2999 3000 sbuf_printf(sb, "%*s<group level=\"%d\" cache-level=\"%d\">\n", indent, 3001 "", 1 + indent / 2, cg->cg_level); 3002 sbuf_printf(sb, "%*s <cpu count=\"%d\" mask=\"%s\">", indent, "", 3003 cg->cg_count, cpusetobj_strprint(cpusetbuf, &cg->cg_mask)); 3004 first = TRUE; 3005 for (i = 0; i < MAXCPU; i++) { 3006 if (CPU_ISSET(i, &cg->cg_mask)) { 3007 if (!first) 3008 sbuf_printf(sb, ", "); 3009 else 3010 first = FALSE; 3011 sbuf_printf(sb, "%d", i); 3012 } 3013 } 3014 sbuf_printf(sb, "</cpu>\n"); 3015 3016 if (cg->cg_flags != 0) { 3017 sbuf_printf(sb, "%*s <flags>", indent, ""); 3018 if ((cg->cg_flags & CG_FLAG_HTT) != 0) 3019 sbuf_printf(sb, "<flag name=\"HTT\">HTT group</flag>"); 3020 if ((cg->cg_flags & CG_FLAG_THREAD) != 0) 3021 sbuf_printf(sb, "<flag name=\"THREAD\">THREAD group</flag>"); 3022 if ((cg->cg_flags & CG_FLAG_SMT) != 0) 3023 sbuf_printf(sb, "<flag name=\"SMT\">SMT group</flag>"); 3024 sbuf_printf(sb, "</flags>\n"); 3025 } 3026 3027 if (cg->cg_children > 0) { 3028 sbuf_printf(sb, "%*s <children>\n", indent, ""); 3029 for (i = 0; i < cg->cg_children; i++) 3030 sysctl_kern_sched_topology_spec_internal(sb, 3031 &cg->cg_child[i], indent+2); 3032 sbuf_printf(sb, "%*s </children>\n", indent, ""); 3033 } 3034 sbuf_printf(sb, "%*s</group>\n", indent, ""); 3035 return (0); 3036 } 3037 3038 /* 3039 * Sysctl handler for retrieving topology dump. It's a wrapper for 3040 * the recursive sysctl_kern_smp_topology_spec_internal(). 3041 */ 3042 static int 3043 sysctl_kern_sched_topology_spec(SYSCTL_HANDLER_ARGS) 3044 { 3045 struct sbuf *topo; 3046 int err; 3047 3048 KASSERT(cpu_top != NULL, ("cpu_top isn't initialized")); 3049 3050 topo = sbuf_new_for_sysctl(NULL, NULL, 512, req); 3051 if (topo == NULL) 3052 return (ENOMEM); 3053 3054 sbuf_printf(topo, "<groups>\n"); 3055 err = sysctl_kern_sched_topology_spec_internal(topo, cpu_top, 1); 3056 sbuf_printf(topo, "</groups>\n"); 3057 3058 if (err == 0) { 3059 err = sbuf_finish(topo); 3060 } 3061 sbuf_delete(topo); 3062 return (err); 3063 } 3064 3065 #endif 3066 3067 static int 3068 sysctl_kern_quantum(SYSCTL_HANDLER_ARGS) 3069 { 3070 int error, new_val, period; 3071 3072 period = 1000000 / realstathz; 3073 new_val = period * sched_slice; 3074 error = sysctl_handle_int(oidp, &new_val, 0, req); 3075 if (error != 0 || req->newptr == NULL) 3076 return (error); 3077 if (new_val <= 0) 3078 return (EINVAL); 3079 sched_slice = imax(1, (new_val + period / 2) / period); 3080 sched_slice_min = sched_slice / SCHED_SLICE_MIN_DIVISOR; 3081 hogticks = imax(1, (2 * hz * sched_slice + realstathz / 2) / 3082 realstathz); 3083 return (0); 3084 } 3085 3086 SYSCTL_NODE(_kern, OID_AUTO, sched, CTLFLAG_RW | CTLFLAG_MPSAFE, 0, 3087 "Scheduler"); 3088 SYSCTL_STRING(_kern_sched, OID_AUTO, name, CTLFLAG_RD, "ULE", 0, 3089 "Scheduler name"); 3090 SYSCTL_PROC(_kern_sched, OID_AUTO, quantum, 3091 CTLTYPE_INT | CTLFLAG_RW | CTLFLAG_MPSAFE, NULL, 0, 3092 sysctl_kern_quantum, "I", 3093 "Quantum for timeshare threads in microseconds"); 3094 SYSCTL_INT(_kern_sched, OID_AUTO, slice, CTLFLAG_RW, &sched_slice, 0, 3095 "Quantum for timeshare threads in stathz ticks"); 3096 SYSCTL_INT(_kern_sched, OID_AUTO, interact, CTLFLAG_RW, &sched_interact, 0, 3097 "Interactivity score threshold"); 3098 SYSCTL_INT(_kern_sched, OID_AUTO, preempt_thresh, CTLFLAG_RW, 3099 &preempt_thresh, 0, 3100 "Maximal (lowest) priority for preemption"); 3101 SYSCTL_INT(_kern_sched, OID_AUTO, static_boost, CTLFLAG_RW, &static_boost, 0, 3102 "Assign static kernel priorities to sleeping threads"); 3103 SYSCTL_INT(_kern_sched, OID_AUTO, idlespins, CTLFLAG_RW, &sched_idlespins, 0, 3104 "Number of times idle thread will spin waiting for new work"); 3105 SYSCTL_INT(_kern_sched, OID_AUTO, idlespinthresh, CTLFLAG_RW, 3106 &sched_idlespinthresh, 0, 3107 "Threshold before we will permit idle thread spinning"); 3108 #ifdef SMP 3109 SYSCTL_INT(_kern_sched, OID_AUTO, affinity, CTLFLAG_RW, &affinity, 0, 3110 "Number of hz ticks to keep thread affinity for"); 3111 SYSCTL_INT(_kern_sched, OID_AUTO, balance, CTLFLAG_RW, &rebalance, 0, 3112 "Enables the long-term load balancer"); 3113 SYSCTL_INT(_kern_sched, OID_AUTO, balance_interval, CTLFLAG_RW, 3114 &balance_interval, 0, 3115 "Average period in stathz ticks to run the long-term balancer"); 3116 SYSCTL_INT(_kern_sched, OID_AUTO, steal_idle, CTLFLAG_RW, &steal_idle, 0, 3117 "Attempts to steal work from other cores before idling"); 3118 SYSCTL_INT(_kern_sched, OID_AUTO, steal_thresh, CTLFLAG_RW, &steal_thresh, 0, 3119 "Minimum load on remote CPU before we'll steal"); 3120 SYSCTL_INT(_kern_sched, OID_AUTO, trysteal_limit, CTLFLAG_RW, &trysteal_limit, 3121 0, "Topological distance limit for stealing threads in sched_switch()"); 3122 SYSCTL_INT(_kern_sched, OID_AUTO, always_steal, CTLFLAG_RW, &always_steal, 0, 3123 "Always run the stealer from the idle thread"); 3124 SYSCTL_PROC(_kern_sched, OID_AUTO, topology_spec, CTLTYPE_STRING | 3125 CTLFLAG_MPSAFE | CTLFLAG_RD, NULL, 0, sysctl_kern_sched_topology_spec, "A", 3126 "XML dump of detected CPU topology"); 3127 #endif 3128 3129 /* ps compat. All cpu percentages from ULE are weighted. */ 3130 static int ccpu = 0; 3131 SYSCTL_INT(_kern, OID_AUTO, ccpu, CTLFLAG_RD, &ccpu, 0, 3132 "Decay factor used for updating %CPU in 4BSD scheduler"); 3133