Blame view
kernel/sched/fair.c
141 KB
bf0f6f24a sched: cfs core, ... |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 |
/* * Completely Fair Scheduling (CFS) Class (SCHED_NORMAL/SCHED_BATCH) * * Copyright (C) 2007 Red Hat, Inc., Ingo Molnar <mingo@redhat.com> * * Interactivity improvements by Mike Galbraith * (C) 2007 Mike Galbraith <efault@gmx.de> * * Various enhancements by Dmitry Adamushko. * (C) 2007 Dmitry Adamushko <dmitry.adamushko@gmail.com> * * Group scheduling enhancements by Srivatsa Vaddagiri * Copyright IBM Corporation, 2007 * Author: Srivatsa Vaddagiri <vatsa@linux.vnet.ibm.com> * * Scaled math optimizations by Thomas Gleixner * Copyright (C) 2007, Thomas Gleixner <tglx@linutronix.de> |
218050855 sched: adaptive s... |
18 19 20 |
* * Adaptive scheduling granularity, math enhancements by Peter Zijlstra * Copyright (C) 2007 Red Hat, Inc., Peter Zijlstra <pzijlstr@redhat.com> |
bf0f6f24a sched: cfs core, ... |
21 |
*/ |
9745512ce sched: latencytop... |
22 |
#include <linux/latencytop.h> |
1983a922a sched: Make tunab... |
23 |
#include <linux/sched.h> |
3436ae129 sched: Fix rebala... |
24 |
#include <linux/cpumask.h> |
029632fbb sched: Make separ... |
25 26 27 28 29 30 31 |
#include <linux/slab.h> #include <linux/profile.h> #include <linux/interrupt.h> #include <trace/events/sched.h> #include "sched.h" |
9745512ce sched: latencytop... |
32 |
|
bf0f6f24a sched: cfs core, ... |
33 |
/* |
218050855 sched: adaptive s... |
34 |
* Targeted preemption latency for CPU-bound tasks: |
864616ee6 sched: Comment up... |
35 |
* (default: 6ms * (1 + ilog(ncpus)), units: nanoseconds) |
bf0f6f24a sched: cfs core, ... |
36 |
* |
218050855 sched: adaptive s... |
37 |
* NOTE: this latency value is not the same as the concept of |
d274a4cee sched: update com... |
38 39 40 |
* 'timeslice length' - timeslices in CFS are of variable length * and have no persistent notion like in traditional, time-slice * based scheduling concepts. |
bf0f6f24a sched: cfs core, ... |
41 |
* |
d274a4cee sched: update com... |
42 43 |
* (to see the precise effective timeslice length of your workload, * run vmstat and monitor the context-switches (cs) field) |
bf0f6f24a sched: cfs core, ... |
44 |
*/ |
21406928a sched: Tweak sche... |
45 46 |
unsigned int sysctl_sched_latency = 6000000ULL; unsigned int normalized_sysctl_sched_latency = 6000000ULL; |
2bd8e6d42 sched: use consta... |
47 48 |
/* |
1983a922a sched: Make tunab... |
49 50 51 52 53 54 55 56 57 58 59 60 |
* The initial- and re-scaling of tunables is configurable * (default SCHED_TUNABLESCALING_LOG = *(1+ilog(ncpus)) * * Options are: * SCHED_TUNABLESCALING_NONE - unscaled, always *1 * SCHED_TUNABLESCALING_LOG - scaled logarithmical, *1+ilog(ncpus) * SCHED_TUNABLESCALING_LINEAR - scaled linear, *ncpus */ enum sched_tunable_scaling sysctl_sched_tunable_scaling = SCHED_TUNABLESCALING_LOG; /* |
b2be5e96d sched: reintroduc... |
61 |
* Minimal preemption granularity for CPU-bound tasks: |
864616ee6 sched: Comment up... |
62 |
* (default: 0.75 msec * (1 + ilog(ncpus)), units: nanoseconds) |
2bd8e6d42 sched: use consta... |
63 |
*/ |
0bf377bbb sched: Improve la... |
64 65 |
unsigned int sysctl_sched_min_granularity = 750000ULL; unsigned int normalized_sysctl_sched_min_granularity = 750000ULL; |
218050855 sched: adaptive s... |
66 67 |
/* |
b2be5e96d sched: reintroduc... |
68 69 |
* is kept at sysctl_sched_latency / sysctl_sched_min_granularity */ |
0bf377bbb sched: Improve la... |
70 |
static unsigned int sched_nr_latency = 8; |
b2be5e96d sched: reintroduc... |
71 72 |
/* |
2bba22c50 sched: Turn off c... |
73 |
* After fork, child runs first. If set to 0 (default) then |
b2be5e96d sched: reintroduc... |
74 |
* parent will (try to) run first. |
218050855 sched: adaptive s... |
75 |
*/ |
2bba22c50 sched: Turn off c... |
76 |
unsigned int sysctl_sched_child_runs_first __read_mostly; |
bf0f6f24a sched: cfs core, ... |
77 78 |
/* |
bf0f6f24a sched: cfs core, ... |
79 |
* SCHED_OTHER wake-up granularity. |
172e082a9 sched: Re-tune th... |
80 |
* (default: 1 msec * (1 + ilog(ncpus)), units: nanoseconds) |
bf0f6f24a sched: cfs core, ... |
81 82 83 84 85 |
* * This option delays the preemption effects of decoupled workloads * and reduces their over-scheduling. Synchronous workloads will still * have immediate wakeup/sleep latencies. */ |
172e082a9 sched: Re-tune th... |
86 |
unsigned int sysctl_sched_wakeup_granularity = 1000000UL; |
0bcdcf28c sched: Fix missin... |
87 |
unsigned int normalized_sysctl_sched_wakeup_granularity = 1000000UL; |
bf0f6f24a sched: cfs core, ... |
88 |
|
da84d9617 sched: reintroduc... |
89 |
const_debug unsigned int sysctl_sched_migration_cost = 500000UL; |
a7a4f8a75 sched: Add sysctl... |
90 91 92 93 94 95 |
/* * The exponential sliding window over which load is averaged for shares * distribution. * (default: 10msec) */ unsigned int __read_mostly sysctl_sched_shares_window = 10000000UL; |
ec12cb7f3 sched: Accumulate... |
96 97 98 99 100 101 102 103 104 105 106 107 108 |
#ifdef CONFIG_CFS_BANDWIDTH /* * Amount of runtime to allocate from global (tg) to local (per-cfs_rq) pool * each time a cfs_rq requests quota. * * Note: in the case that the slice exceeds the runtime remaining (either due * to consumption or the quota being specified to be smaller than the slice) * we will always only issue the remaining available time. * * default: 5 msec, units: microseconds */ unsigned int sysctl_sched_cfs_bandwidth_slice = 5000UL; #endif |
029632fbb sched: Make separ... |
109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 |
/* * Increase the granularity value when there are more CPUs, * because with more CPUs the 'effective latency' as visible * to users decreases. But the relationship is not linear, * so pick a second-best guess by going with the log2 of the * number of CPUs. * * This idea comes from the SD scheduler of Con Kolivas: */ static int get_update_sysctl_factor(void) { unsigned int cpus = min_t(int, num_online_cpus(), 8); unsigned int factor; switch (sysctl_sched_tunable_scaling) { case SCHED_TUNABLESCALING_NONE: factor = 1; break; case SCHED_TUNABLESCALING_LINEAR: factor = cpus; break; case SCHED_TUNABLESCALING_LOG: default: factor = 1 + ilog2(cpus); break; } return factor; } static void update_sysctl(void) { unsigned int factor = get_update_sysctl_factor(); #define SET_SYSCTL(name) \ (sysctl_##name = (factor) * normalized_sysctl_##name) SET_SYSCTL(sched_min_granularity); SET_SYSCTL(sched_latency); SET_SYSCTL(sched_wakeup_granularity); #undef SET_SYSCTL } void sched_init_granularity(void) { update_sysctl(); } #if BITS_PER_LONG == 32 # define WMULT_CONST (~0UL) #else # define WMULT_CONST (1UL << 32) #endif #define WMULT_SHIFT 32 /* * Shift right and round: */ #define SRR(x, y) (((x) + (1UL << ((y) - 1))) >> (y)) /* * delta *= weight / lw */ static unsigned long calc_delta_mine(unsigned long delta_exec, unsigned long weight, struct load_weight *lw) { u64 tmp; /* * weight can be less than 2^SCHED_LOAD_RESOLUTION for task group sched * entities since MIN_SHARES = 2. Treat weight as 1 if less than * 2^SCHED_LOAD_RESOLUTION. */ if (likely(weight > (1UL << SCHED_LOAD_RESOLUTION))) tmp = (u64)delta_exec * scale_load_down(weight); else tmp = (u64)delta_exec; if (!lw->inv_weight) { unsigned long w = scale_load_down(lw->weight); if (BITS_PER_LONG > 32 && unlikely(w >= WMULT_CONST)) lw->inv_weight = 1; else if (unlikely(!w)) lw->inv_weight = WMULT_CONST; else lw->inv_weight = WMULT_CONST / w; } /* * Check whether we'd overflow the 64-bit multiplication: */ if (unlikely(tmp > WMULT_CONST)) tmp = SRR(SRR(tmp, WMULT_SHIFT/2) * lw->inv_weight, WMULT_SHIFT/2); else tmp = SRR(tmp * lw->inv_weight, WMULT_SHIFT); return (unsigned long)min(tmp, (u64)(unsigned long)LONG_MAX); } const struct sched_class fair_sched_class; |
a4c2f00f5 sched: fair sched... |
213 |
|
bf0f6f24a sched: cfs core, ... |
214 215 216 |
/************************************************************** * CFS operations on generic schedulable entities: */ |
62160e3f4 sched: track cfs_... |
217 |
#ifdef CONFIG_FAIR_GROUP_SCHED |
bf0f6f24a sched: cfs core, ... |
218 |
|
62160e3f4 sched: track cfs_... |
219 |
/* cpu runqueue to which this cfs_rq is attached */ |
bf0f6f24a sched: cfs core, ... |
220 221 |
static inline struct rq *rq_of(struct cfs_rq *cfs_rq) { |
62160e3f4 sched: track cfs_... |
222 |
return cfs_rq->rq; |
bf0f6f24a sched: cfs core, ... |
223 |
} |
62160e3f4 sched: track cfs_... |
224 225 |
/* An entity is a task if it doesn't "own" a runqueue */ #define entity_is_task(se) (!se->my_q) |
bf0f6f24a sched: cfs core, ... |
226 |
|
8f48894fc sched: Add debug ... |
227 228 229 230 231 232 233 |
static inline struct task_struct *task_of(struct sched_entity *se) { #ifdef CONFIG_SCHED_DEBUG WARN_ON_ONCE(!entity_is_task(se)); #endif return container_of(se, struct task_struct, se); } |
b758149c0 sched: prepatory ... |
234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 |
/* Walk up scheduling entities hierarchy */ #define for_each_sched_entity(se) \ for (; se; se = se->parent) static inline struct cfs_rq *task_cfs_rq(struct task_struct *p) { return p->se.cfs_rq; } /* runqueue on which this entity is (to be) queued */ static inline struct cfs_rq *cfs_rq_of(struct sched_entity *se) { return se->cfs_rq; } /* runqueue "owned" by this group */ static inline struct cfs_rq *group_cfs_rq(struct sched_entity *grp) { return grp->my_q; } |
3d4b47b4b sched: Implement ... |
254 255 256 |
static inline void list_add_leaf_cfs_rq(struct cfs_rq *cfs_rq) { if (!cfs_rq->on_list) { |
67e86250f sched: Introduce ... |
257 258 259 260 261 262 263 264 265 266 267 268 |
/* * Ensure we either appear before our parent (if already * enqueued) or force our parent to appear after us when it is * enqueued. The fact that we always enqueue bottom-up * reduces this to two cases. */ if (cfs_rq->tg->parent && cfs_rq->tg->parent->cfs_rq[cpu_of(rq_of(cfs_rq))]->on_list) { list_add_rcu(&cfs_rq->leaf_cfs_rq_list, &rq_of(cfs_rq)->leaf_cfs_rq_list); } else { list_add_tail_rcu(&cfs_rq->leaf_cfs_rq_list, |
3d4b47b4b sched: Implement ... |
269 |
&rq_of(cfs_rq)->leaf_cfs_rq_list); |
67e86250f sched: Introduce ... |
270 |
} |
3d4b47b4b sched: Implement ... |
271 272 273 274 275 276 277 278 279 280 281 282 |
cfs_rq->on_list = 1; } } static inline void list_del_leaf_cfs_rq(struct cfs_rq *cfs_rq) { if (cfs_rq->on_list) { list_del_rcu(&cfs_rq->leaf_cfs_rq_list); cfs_rq->on_list = 0; } } |
b758149c0 sched: prepatory ... |
283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 |
/* Iterate thr' all leaf cfs_rq's on a runqueue */ #define for_each_leaf_cfs_rq(rq, cfs_rq) \ list_for_each_entry_rcu(cfs_rq, &rq->leaf_cfs_rq_list, leaf_cfs_rq_list) /* Do the two (enqueued) entities belong to the same group ? */ static inline int is_same_group(struct sched_entity *se, struct sched_entity *pse) { if (se->cfs_rq == pse->cfs_rq) return 1; return 0; } static inline struct sched_entity *parent_entity(struct sched_entity *se) { return se->parent; } |
464b75273 sched: re-instate... |
301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 |
/* return depth at which a sched entity is present in the hierarchy */ static inline int depth_se(struct sched_entity *se) { int depth = 0; for_each_sched_entity(se) depth++; return depth; } static void find_matching_se(struct sched_entity **se, struct sched_entity **pse) { int se_depth, pse_depth; /* * preemption test can be made between sibling entities who are in the * same cfs_rq i.e who have a common parent. Walk up the hierarchy of * both tasks until we find their ancestors who are siblings of common * parent. */ /* First walk up until both entities are at same depth */ se_depth = depth_se(*se); pse_depth = depth_se(*pse); while (se_depth > pse_depth) { se_depth--; *se = parent_entity(*se); } while (pse_depth > se_depth) { pse_depth--; *pse = parent_entity(*pse); } while (!is_same_group(*se, *pse)) { *se = parent_entity(*se); *pse = parent_entity(*pse); } } |
8f48894fc sched: Add debug ... |
343 344 345 346 347 348 |
#else /* !CONFIG_FAIR_GROUP_SCHED */ static inline struct task_struct *task_of(struct sched_entity *se) { return container_of(se, struct task_struct, se); } |
bf0f6f24a sched: cfs core, ... |
349 |
|
62160e3f4 sched: track cfs_... |
350 351 352 |
static inline struct rq *rq_of(struct cfs_rq *cfs_rq) { return container_of(cfs_rq, struct rq, cfs); |
bf0f6f24a sched: cfs core, ... |
353 354 355 |
} #define entity_is_task(se) 1 |
b758149c0 sched: prepatory ... |
356 357 |
#define for_each_sched_entity(se) \ for (; se; se = NULL) |
bf0f6f24a sched: cfs core, ... |
358 |
|
b758149c0 sched: prepatory ... |
359 |
static inline struct cfs_rq *task_cfs_rq(struct task_struct *p) |
bf0f6f24a sched: cfs core, ... |
360 |
{ |
b758149c0 sched: prepatory ... |
361 |
return &task_rq(p)->cfs; |
bf0f6f24a sched: cfs core, ... |
362 |
} |
b758149c0 sched: prepatory ... |
363 364 365 366 367 368 369 370 371 372 373 374 375 |
static inline struct cfs_rq *cfs_rq_of(struct sched_entity *se) { struct task_struct *p = task_of(se); struct rq *rq = task_rq(p); return &rq->cfs; } /* runqueue "owned" by this group */ static inline struct cfs_rq *group_cfs_rq(struct sched_entity *grp) { return NULL; } |
3d4b47b4b sched: Implement ... |
376 377 378 379 380 381 382 |
static inline void list_add_leaf_cfs_rq(struct cfs_rq *cfs_rq) { } static inline void list_del_leaf_cfs_rq(struct cfs_rq *cfs_rq) { } |
b758149c0 sched: prepatory ... |
383 384 385 386 387 388 389 390 391 392 393 394 395 |
#define for_each_leaf_cfs_rq(rq, cfs_rq) \ for (cfs_rq = &rq->cfs; cfs_rq; cfs_rq = NULL) static inline int is_same_group(struct sched_entity *se, struct sched_entity *pse) { return 1; } static inline struct sched_entity *parent_entity(struct sched_entity *se) { return NULL; } |
464b75273 sched: re-instate... |
396 397 398 399 |
static inline void find_matching_se(struct sched_entity **se, struct sched_entity **pse) { } |
b758149c0 sched: prepatory ... |
400 |
#endif /* CONFIG_FAIR_GROUP_SCHED */ |
ec12cb7f3 sched: Accumulate... |
401 402 |
static void account_cfs_rq_runtime(struct cfs_rq *cfs_rq, unsigned long delta_exec); |
bf0f6f24a sched: cfs core, ... |
403 404 405 406 |
/************************************************************** * Scheduling class tree data structure manipulation methods: */ |
0702e3ebc sched: cleanup: f... |
407 |
static inline u64 max_vruntime(u64 min_vruntime, u64 vruntime) |
02e0431a3 sched: better min... |
408 |
{ |
368059a97 sched: max_vrunti... |
409 410 |
s64 delta = (s64)(vruntime - min_vruntime); if (delta > 0) |
02e0431a3 sched: better min... |
411 412 413 414 |
min_vruntime = vruntime; return min_vruntime; } |
0702e3ebc sched: cleanup: f... |
415 |
static inline u64 min_vruntime(u64 min_vruntime, u64 vruntime) |
b0ffd246e sched: clean up m... |
416 417 418 419 420 421 422 |
{ s64 delta = (s64)(vruntime - min_vruntime); if (delta < 0) min_vruntime = vruntime; return min_vruntime; } |
54fdc5816 sched: Account fo... |
423 424 425 426 427 |
static inline int entity_before(struct sched_entity *a, struct sched_entity *b) { return (s64)(a->vruntime - b->vruntime) < 0; } |
1af5f730f sched: more accur... |
428 429 430 431 432 433 434 435 436 437 438 |
static void update_min_vruntime(struct cfs_rq *cfs_rq) { u64 vruntime = cfs_rq->min_vruntime; if (cfs_rq->curr) vruntime = cfs_rq->curr->vruntime; if (cfs_rq->rb_leftmost) { struct sched_entity *se = rb_entry(cfs_rq->rb_leftmost, struct sched_entity, run_node); |
e17036dac sched: fix update... |
439 |
if (!cfs_rq->curr) |
1af5f730f sched: more accur... |
440 441 442 443 444 445 |
vruntime = se->vruntime; else vruntime = min_vruntime(vruntime, se->vruntime); } cfs_rq->min_vruntime = max_vruntime(cfs_rq->min_vruntime, vruntime); |
3fe1698b7 sched: Deal with ... |
446 447 448 449 |
#ifndef CONFIG_64BIT smp_wmb(); cfs_rq->min_vruntime_copy = cfs_rq->min_vruntime; #endif |
1af5f730f sched: more accur... |
450 |
} |
bf0f6f24a sched: cfs core, ... |
451 452 453 |
/* * Enqueue an entity into the rb-tree: */ |
0702e3ebc sched: cleanup: f... |
454 |
static void __enqueue_entity(struct cfs_rq *cfs_rq, struct sched_entity *se) |
bf0f6f24a sched: cfs core, ... |
455 456 457 458 |
{ struct rb_node **link = &cfs_rq->tasks_timeline.rb_node; struct rb_node *parent = NULL; struct sched_entity *entry; |
bf0f6f24a sched: cfs core, ... |
459 460 461 462 463 464 465 466 467 468 469 470 |
int leftmost = 1; /* * Find the right place in the rbtree: */ while (*link) { parent = *link; entry = rb_entry(parent, struct sched_entity, run_node); /* * We dont care about collisions. Nodes with * the same key stay together. */ |
2bd2d6f2d sched: Replace us... |
471 |
if (entity_before(se, entry)) { |
bf0f6f24a sched: cfs core, ... |
472 473 474 475 476 477 478 479 480 481 482 |
link = &parent->rb_left; } else { link = &parent->rb_right; leftmost = 0; } } /* * Maintain a cache of leftmost tree entries (it is frequently * used): */ |
1af5f730f sched: more accur... |
483 |
if (leftmost) |
57cb499df sched: remove set... |
484 |
cfs_rq->rb_leftmost = &se->run_node; |
bf0f6f24a sched: cfs core, ... |
485 486 487 |
rb_link_node(&se->run_node, parent, link); rb_insert_color(&se->run_node, &cfs_rq->tasks_timeline); |
bf0f6f24a sched: cfs core, ... |
488 |
} |
0702e3ebc sched: cleanup: f... |
489 |
static void __dequeue_entity(struct cfs_rq *cfs_rq, struct sched_entity *se) |
bf0f6f24a sched: cfs core, ... |
490 |
{ |
3fe69747d sched: min_vrunti... |
491 492 |
if (cfs_rq->rb_leftmost == &se->run_node) { struct rb_node *next_node; |
3fe69747d sched: min_vrunti... |
493 494 495 |
next_node = rb_next(&se->run_node); cfs_rq->rb_leftmost = next_node; |
3fe69747d sched: min_vrunti... |
496 |
} |
e9acbff64 sched: introduce ... |
497 |
|
bf0f6f24a sched: cfs core, ... |
498 |
rb_erase(&se->run_node, &cfs_rq->tasks_timeline); |
bf0f6f24a sched: cfs core, ... |
499 |
} |
029632fbb sched: Make separ... |
500 |
struct sched_entity *__pick_first_entity(struct cfs_rq *cfs_rq) |
bf0f6f24a sched: cfs core, ... |
501 |
{ |
f4b6755fb sched: cleanup fa... |
502 503 504 505 506 507 |
struct rb_node *left = cfs_rq->rb_leftmost; if (!left) return NULL; return rb_entry(left, struct sched_entity, run_node); |
bf0f6f24a sched: cfs core, ... |
508 |
} |
ac53db596 sched: Use a budd... |
509 510 511 512 513 514 515 516 517 518 519 |
static struct sched_entity *__pick_next_entity(struct sched_entity *se) { struct rb_node *next = rb_next(&se->run_node); if (!next) return NULL; return rb_entry(next, struct sched_entity, run_node); } #ifdef CONFIG_SCHED_DEBUG |
029632fbb sched: Make separ... |
520 |
struct sched_entity *__pick_last_entity(struct cfs_rq *cfs_rq) |
aeb73b040 sched: clean up n... |
521 |
{ |
7eee3e677 sched: clean up _... |
522 |
struct rb_node *last = rb_last(&cfs_rq->tasks_timeline); |
aeb73b040 sched: clean up n... |
523 |
|
70eee74b7 sched: remove dup... |
524 525 |
if (!last) return NULL; |
7eee3e677 sched: clean up _... |
526 527 |
return rb_entry(last, struct sched_entity, run_node); |
aeb73b040 sched: clean up n... |
528 |
} |
bf0f6f24a sched: cfs core, ... |
529 530 531 |
/************************************************************** * Scheduling class statistics methods: */ |
acb4a848d sched: Update nor... |
532 |
int sched_proc_update_handler(struct ctl_table *table, int write, |
8d65af789 sysctl: remove "s... |
533 |
void __user *buffer, size_t *lenp, |
b2be5e96d sched: reintroduc... |
534 535 |
loff_t *ppos) { |
8d65af789 sysctl: remove "s... |
536 |
int ret = proc_dointvec_minmax(table, write, buffer, lenp, ppos); |
acb4a848d sched: Update nor... |
537 |
int factor = get_update_sysctl_factor(); |
b2be5e96d sched: reintroduc... |
538 539 540 541 542 543 |
if (ret || !write) return ret; sched_nr_latency = DIV_ROUND_UP(sysctl_sched_latency, sysctl_sched_min_granularity); |
acb4a848d sched: Update nor... |
544 545 546 547 548 |
#define WRT_SYSCTL(name) \ (normalized_sysctl_##name = sysctl_##name / (factor)) WRT_SYSCTL(sched_min_granularity); WRT_SYSCTL(sched_latency); WRT_SYSCTL(sched_wakeup_granularity); |
acb4a848d sched: Update nor... |
549 |
#undef WRT_SYSCTL |
b2be5e96d sched: reintroduc... |
550 551 552 |
return 0; } #endif |
647e7cac2 sched: vslice fix... |
553 554 |
/* |
f9c0b0950 sched: revert bac... |
555 |
* delta /= w |
a7be37ac8 sched: revert the... |
556 557 558 559 |
*/ static inline unsigned long calc_delta_fair(unsigned long delta, struct sched_entity *se) { |
f9c0b0950 sched: revert bac... |
560 561 |
if (unlikely(se->load.weight != NICE_0_LOAD)) delta = calc_delta_mine(delta, NICE_0_LOAD, &se->load); |
a7be37ac8 sched: revert the... |
562 563 564 565 566 |
return delta; } /* |
647e7cac2 sched: vslice fix... |
567 568 569 570 571 572 573 |
* The idea is to set a period in which each task runs once. * * When there are too many tasks (sysctl_sched_nr_latency) we have to stretch * this period because otherwise the slices get too small. * * p = (nr <= nl) ? l : l*nr/nl */ |
4d78e7b65 sched: new task p... |
574 575 576 |
static u64 __sched_period(unsigned long nr_running) { u64 period = sysctl_sched_latency; |
b2be5e96d sched: reintroduc... |
577 |
unsigned long nr_latency = sched_nr_latency; |
4d78e7b65 sched: new task p... |
578 579 |
if (unlikely(nr_running > nr_latency)) { |
4bf0b7715 sched: remove do_... |
580 |
period = sysctl_sched_min_granularity; |
4d78e7b65 sched: new task p... |
581 |
period *= nr_running; |
4d78e7b65 sched: new task p... |
582 583 584 585 |
} return period; } |
647e7cac2 sched: vslice fix... |
586 587 588 589 |
/* * We calculate the wall-time slice from the period by taking a part * proportional to the weight. * |
f9c0b0950 sched: revert bac... |
590 |
* s = p*P[w/rw] |
647e7cac2 sched: vslice fix... |
591 |
*/ |
6d0f0ebd0 sched: simplify a... |
592 |
static u64 sched_slice(struct cfs_rq *cfs_rq, struct sched_entity *se) |
218050855 sched: adaptive s... |
593 |
{ |
0a582440f sched: fix sched_... |
594 |
u64 slice = __sched_period(cfs_rq->nr_running + !se->on_rq); |
f9c0b0950 sched: revert bac... |
595 |
|
0a582440f sched: fix sched_... |
596 |
for_each_sched_entity(se) { |
6272d68cc sched: sched_slic... |
597 |
struct load_weight *load; |
3104bf03a sched: Fix out of... |
598 |
struct load_weight lw; |
6272d68cc sched: sched_slic... |
599 600 601 |
cfs_rq = cfs_rq_of(se); load = &cfs_rq->load; |
f9c0b0950 sched: revert bac... |
602 |
|
0a582440f sched: fix sched_... |
603 |
if (unlikely(!se->on_rq)) { |
3104bf03a sched: Fix out of... |
604 |
lw = cfs_rq->load; |
0a582440f sched: fix sched_... |
605 606 607 608 609 610 611 |
update_load_add(&lw, se->load.weight); load = &lw; } slice = calc_delta_mine(slice, se->load.weight, load); } return slice; |
bf0f6f24a sched: cfs core, ... |
612 |
} |
647e7cac2 sched: vslice fix... |
613 |
/* |
ac884dec6 sched: fair-group... |
614 |
* We calculate the vruntime slice of a to be inserted task |
647e7cac2 sched: vslice fix... |
615 |
* |
f9c0b0950 sched: revert bac... |
616 |
* vs = s/w |
647e7cac2 sched: vslice fix... |
617 |
*/ |
f9c0b0950 sched: revert bac... |
618 |
static u64 sched_vslice(struct cfs_rq *cfs_rq, struct sched_entity *se) |
67e9fb2a3 sched: add vslice |
619 |
{ |
f9c0b0950 sched: revert bac... |
620 |
return calc_delta_fair(sched_slice(cfs_rq, se), se); |
a7be37ac8 sched: revert the... |
621 |
} |
d6b559182 sched: Allow upda... |
622 |
static void update_cfs_load(struct cfs_rq *cfs_rq, int global_update); |
6d5ab2932 sched: Simplify u... |
623 |
static void update_cfs_shares(struct cfs_rq *cfs_rq); |
3b3d190ec sched: Implement ... |
624 |
|
a7be37ac8 sched: revert the... |
625 |
/* |
bf0f6f24a sched: cfs core, ... |
626 627 628 629 |
* Update the current task's runtime statistics. Skip current tasks that * are not in our scheduling class. */ static inline void |
8ebc91d93 sched: remove sta... |
630 631 |
__update_curr(struct cfs_rq *cfs_rq, struct sched_entity *curr, unsigned long delta_exec) |
bf0f6f24a sched: cfs core, ... |
632 |
{ |
bbdba7c0e sched: remove wai... |
633 |
unsigned long delta_exec_weighted; |
bf0f6f24a sched: cfs core, ... |
634 |
|
41acab885 sched: Implement ... |
635 636 |
schedstat_set(curr->statistics.exec_max, max((u64)delta_exec, curr->statistics.exec_max)); |
bf0f6f24a sched: cfs core, ... |
637 638 |
curr->sum_exec_runtime += delta_exec; |
7a62eabc4 sched: debug: upd... |
639 |
schedstat_add(cfs_rq, exec_clock, delta_exec); |
a7be37ac8 sched: revert the... |
640 |
delta_exec_weighted = calc_delta_fair(delta_exec, curr); |
88ec22d3e sched: Remove the... |
641 |
|
e9acbff64 sched: introduce ... |
642 |
curr->vruntime += delta_exec_weighted; |
1af5f730f sched: more accur... |
643 |
update_min_vruntime(cfs_rq); |
3b3d190ec sched: Implement ... |
644 |
|
70caf8a6c sched: Fix UP bui... |
645 |
#if defined CONFIG_SMP && defined CONFIG_FAIR_GROUP_SCHED |
3b3d190ec sched: Implement ... |
646 |
cfs_rq->load_unacc_exec_time += delta_exec; |
3b3d190ec sched: Implement ... |
647 |
#endif |
bf0f6f24a sched: cfs core, ... |
648 |
} |
b7cc08965 sched: remove the... |
649 |
static void update_curr(struct cfs_rq *cfs_rq) |
bf0f6f24a sched: cfs core, ... |
650 |
{ |
429d43bcc sched: cleanup: s... |
651 |
struct sched_entity *curr = cfs_rq->curr; |
305e6835e sched: Do not acc... |
652 |
u64 now = rq_of(cfs_rq)->clock_task; |
bf0f6f24a sched: cfs core, ... |
653 654 655 656 657 658 659 660 661 662 |
unsigned long delta_exec; if (unlikely(!curr)) return; /* * Get the amount of time the current task was running * since the last time we changed load (this cannot * overflow on 32 bits): */ |
8ebc91d93 sched: remove sta... |
663 |
delta_exec = (unsigned long)(now - curr->exec_start); |
34f28ecd0 sched: optimize u... |
664 665 |
if (!delta_exec) return; |
bf0f6f24a sched: cfs core, ... |
666 |
|
8ebc91d93 sched: remove sta... |
667 668 |
__update_curr(cfs_rq, curr, delta_exec); curr->exec_start = now; |
d842de871 sched: cpu accoun... |
669 670 671 |
if (entity_is_task(curr)) { struct task_struct *curtask = task_of(curr); |
f977bb493 perf_counter, sch... |
672 |
trace_sched_stat_runtime(curtask, delta_exec, curr->vruntime); |
d842de871 sched: cpu accoun... |
673 |
cpuacct_charge(curtask, delta_exec); |
f06febc96 timers: fix itime... |
674 |
account_group_exec_runtime(curtask, delta_exec); |
d842de871 sched: cpu accoun... |
675 |
} |
ec12cb7f3 sched: Accumulate... |
676 677 |
account_cfs_rq_runtime(cfs_rq, delta_exec); |
bf0f6f24a sched: cfs core, ... |
678 679 680 |
} static inline void |
5870db5b8 sched: remove the... |
681 |
update_stats_wait_start(struct cfs_rq *cfs_rq, struct sched_entity *se) |
bf0f6f24a sched: cfs core, ... |
682 |
{ |
41acab885 sched: Implement ... |
683 |
schedstat_set(se->statistics.wait_start, rq_of(cfs_rq)->clock); |
bf0f6f24a sched: cfs core, ... |
684 |
} |
bf0f6f24a sched: cfs core, ... |
685 686 687 |
/* * Task is being enqueued - update stats: */ |
d2417e5a3 sched: remove the... |
688 |
static void update_stats_enqueue(struct cfs_rq *cfs_rq, struct sched_entity *se) |
bf0f6f24a sched: cfs core, ... |
689 |
{ |
bf0f6f24a sched: cfs core, ... |
690 691 692 693 |
/* * Are we enqueueing a waiting task? (for current tasks * a dequeue/enqueue event is a NOP) */ |
429d43bcc sched: cleanup: s... |
694 |
if (se != cfs_rq->curr) |
5870db5b8 sched: remove the... |
695 |
update_stats_wait_start(cfs_rq, se); |
bf0f6f24a sched: cfs core, ... |
696 |
} |
bf0f6f24a sched: cfs core, ... |
697 |
static void |
9ef0a9615 sched: remove the... |
698 |
update_stats_wait_end(struct cfs_rq *cfs_rq, struct sched_entity *se) |
bf0f6f24a sched: cfs core, ... |
699 |
{ |
41acab885 sched: Implement ... |
700 701 702 703 704 |
schedstat_set(se->statistics.wait_max, max(se->statistics.wait_max, rq_of(cfs_rq)->clock - se->statistics.wait_start)); schedstat_set(se->statistics.wait_count, se->statistics.wait_count + 1); schedstat_set(se->statistics.wait_sum, se->statistics.wait_sum + rq_of(cfs_rq)->clock - se->statistics.wait_start); |
768d0c272 sched: Add wait, ... |
705 706 707 |
#ifdef CONFIG_SCHEDSTATS if (entity_is_task(se)) { trace_sched_stat_wait(task_of(se), |
41acab885 sched: Implement ... |
708 |
rq_of(cfs_rq)->clock - se->statistics.wait_start); |
768d0c272 sched: Add wait, ... |
709 710 |
} #endif |
41acab885 sched: Implement ... |
711 |
schedstat_set(se->statistics.wait_start, 0); |
bf0f6f24a sched: cfs core, ... |
712 713 714 |
} static inline void |
19b6a2e37 sched: remove the... |
715 |
update_stats_dequeue(struct cfs_rq *cfs_rq, struct sched_entity *se) |
bf0f6f24a sched: cfs core, ... |
716 |
{ |
bf0f6f24a sched: cfs core, ... |
717 718 719 720 |
/* * Mark the end of the wait period if dequeueing a * waiting task: */ |
429d43bcc sched: cleanup: s... |
721 |
if (se != cfs_rq->curr) |
9ef0a9615 sched: remove the... |
722 |
update_stats_wait_end(cfs_rq, se); |
bf0f6f24a sched: cfs core, ... |
723 724 725 726 727 728 |
} /* * We are picking a new current task - update its stats: */ static inline void |
79303e9e0 sched: remove the... |
729 |
update_stats_curr_start(struct cfs_rq *cfs_rq, struct sched_entity *se) |
bf0f6f24a sched: cfs core, ... |
730 731 732 733 |
{ /* * We are starting a new run period: */ |
305e6835e sched: Do not acc... |
734 |
se->exec_start = rq_of(cfs_rq)->clock_task; |
bf0f6f24a sched: cfs core, ... |
735 |
} |
bf0f6f24a sched: cfs core, ... |
736 737 738 |
/************************************************** * Scheduling class queueing methods: */ |
c09595f63 sched: revert rev... |
739 740 741 742 743 744 745 746 747 748 749 750 |
#if defined CONFIG_SMP && defined CONFIG_FAIR_GROUP_SCHED static void add_cfs_task_weight(struct cfs_rq *cfs_rq, unsigned long weight) { cfs_rq->task_weight += weight; } #else static inline void add_cfs_task_weight(struct cfs_rq *cfs_rq, unsigned long weight) { } #endif |
30cfdcfc5 sched: do not kee... |
751 752 753 754 |
static void account_entity_enqueue(struct cfs_rq *cfs_rq, struct sched_entity *se) { update_load_add(&cfs_rq->load, se->load.weight); |
c09595f63 sched: revert rev... |
755 |
if (!parent_entity(se)) |
029632fbb sched: Make separ... |
756 |
update_load_add(&rq_of(cfs_rq)->load, se->load.weight); |
b87f17242 sched: maintain o... |
757 |
if (entity_is_task(se)) { |
c09595f63 sched: revert rev... |
758 |
add_cfs_task_weight(cfs_rq, se->load.weight); |
b87f17242 sched: maintain o... |
759 760 |
list_add(&se->group_node, &cfs_rq->tasks); } |
30cfdcfc5 sched: do not kee... |
761 |
cfs_rq->nr_running++; |
30cfdcfc5 sched: do not kee... |
762 763 764 765 766 767 |
} static void account_entity_dequeue(struct cfs_rq *cfs_rq, struct sched_entity *se) { update_load_sub(&cfs_rq->load, se->load.weight); |
c09595f63 sched: revert rev... |
768 |
if (!parent_entity(se)) |
029632fbb sched: Make separ... |
769 |
update_load_sub(&rq_of(cfs_rq)->load, se->load.weight); |
b87f17242 sched: maintain o... |
770 |
if (entity_is_task(se)) { |
c09595f63 sched: revert rev... |
771 |
add_cfs_task_weight(cfs_rq, -se->load.weight); |
b87f17242 sched: maintain o... |
772 773 |
list_del_init(&se->group_node); } |
30cfdcfc5 sched: do not kee... |
774 |
cfs_rq->nr_running--; |
30cfdcfc5 sched: do not kee... |
775 |
} |
3ff6dcac7 sched: Fix poor i... |
776 |
#ifdef CONFIG_FAIR_GROUP_SCHED |
64660c864 sched: Prevent in... |
777 778 |
/* we need this in update_cfs_load and load-balance functions below */ static inline int throttled_hierarchy(struct cfs_rq *cfs_rq); |
3ff6dcac7 sched: Fix poor i... |
779 |
# ifdef CONFIG_SMP |
d6b559182 sched: Allow upda... |
780 781 782 783 784 785 786 787 788 789 790 791 792 793 794 795 |
static void update_cfs_rq_load_contribution(struct cfs_rq *cfs_rq, int global_update) { struct task_group *tg = cfs_rq->tg; long load_avg; load_avg = div64_u64(cfs_rq->load_avg, cfs_rq->load_period+1); load_avg -= cfs_rq->load_contribution; if (global_update || abs(load_avg) > cfs_rq->load_contribution / 8) { atomic_add(load_avg, &tg->load_weight); cfs_rq->load_contribution += load_avg; } } static void update_cfs_load(struct cfs_rq *cfs_rq, int global_update) |
2069dd75c sched: Rewrite tg... |
796 |
{ |
a7a4f8a75 sched: Add sysctl... |
797 |
u64 period = sysctl_sched_shares_window; |
2069dd75c sched: Rewrite tg... |
798 |
u64 now, delta; |
e33078baa sched: Fix update... |
799 |
unsigned long load = cfs_rq->load.weight; |
2069dd75c sched: Rewrite tg... |
800 |
|
64660c864 sched: Prevent in... |
801 |
if (cfs_rq->tg == &root_task_group || throttled_hierarchy(cfs_rq)) |
2069dd75c sched: Rewrite tg... |
802 |
return; |
05ca62c6c sched: Use rq->cl... |
803 |
now = rq_of(cfs_rq)->clock_task; |
2069dd75c sched: Rewrite tg... |
804 |
delta = now - cfs_rq->load_stamp; |
e33078baa sched: Fix update... |
805 806 807 808 809 |
/* truncate load history at 4 idle periods */ if (cfs_rq->load_stamp > cfs_rq->load_last && now - cfs_rq->load_last > 4 * period) { cfs_rq->load_period = 0; cfs_rq->load_avg = 0; |
f07333bf6 sched: Avoid expe... |
810 |
delta = period - 1; |
e33078baa sched: Fix update... |
811 |
} |
2069dd75c sched: Rewrite tg... |
812 |
cfs_rq->load_stamp = now; |
3b3d190ec sched: Implement ... |
813 |
cfs_rq->load_unacc_exec_time = 0; |
2069dd75c sched: Rewrite tg... |
814 |
cfs_rq->load_period += delta; |
e33078baa sched: Fix update... |
815 816 817 818 |
if (load) { cfs_rq->load_last = now; cfs_rq->load_avg += delta * load; } |
2069dd75c sched: Rewrite tg... |
819 |
|
d6b559182 sched: Allow upda... |
820 821 822 823 |
/* consider updating load contribution on each fold or truncate */ if (global_update || cfs_rq->load_period > period || !cfs_rq->load_period) update_cfs_rq_load_contribution(cfs_rq, global_update); |
2069dd75c sched: Rewrite tg... |
824 825 826 827 828 829 830 831 832 833 |
while (cfs_rq->load_period > period) { /* * Inline assembly required to prevent the compiler * optimising this loop into a divmod call. * See __iter_div_u64_rem() for another example of this. */ asm("" : "+rm" (cfs_rq->load_period)); cfs_rq->load_period /= 2; cfs_rq->load_avg /= 2; } |
3d4b47b4b sched: Implement ... |
834 |
|
e33078baa sched: Fix update... |
835 836 |
if (!cfs_rq->curr && !cfs_rq->nr_running && !cfs_rq->load_avg) list_del_leaf_cfs_rq(cfs_rq); |
2069dd75c sched: Rewrite tg... |
837 |
} |
cf5f0acf3 sched: Add a comm... |
838 839 840 841 842 843 844 845 846 847 848 849 850 851 852 |
static inline long calc_tg_weight(struct task_group *tg, struct cfs_rq *cfs_rq) { long tg_weight; /* * Use this CPU's actual weight instead of the last load_contribution * to gain a more accurate current total weight. See * update_cfs_rq_load_contribution(). */ tg_weight = atomic_read(&tg->load_weight); tg_weight -= cfs_rq->load_contribution; tg_weight += cfs_rq->load.weight; return tg_weight; } |
6d5ab2932 sched: Simplify u... |
853 |
static long calc_cfs_shares(struct cfs_rq *cfs_rq, struct task_group *tg) |
3ff6dcac7 sched: Fix poor i... |
854 |
{ |
cf5f0acf3 sched: Add a comm... |
855 |
long tg_weight, load, shares; |
3ff6dcac7 sched: Fix poor i... |
856 |
|
cf5f0acf3 sched: Add a comm... |
857 |
tg_weight = calc_tg_weight(tg, cfs_rq); |
6d5ab2932 sched: Simplify u... |
858 |
load = cfs_rq->load.weight; |
3ff6dcac7 sched: Fix poor i... |
859 |
|
3ff6dcac7 sched: Fix poor i... |
860 |
shares = (tg->shares * load); |
cf5f0acf3 sched: Add a comm... |
861 862 |
if (tg_weight) shares /= tg_weight; |
3ff6dcac7 sched: Fix poor i... |
863 864 865 866 867 868 869 870 871 872 873 874 875 |
if (shares < MIN_SHARES) shares = MIN_SHARES; if (shares > tg->shares) shares = tg->shares; return shares; } static void update_entity_shares_tick(struct cfs_rq *cfs_rq) { if (cfs_rq->load_unacc_exec_time > sysctl_sched_shares_window) { update_cfs_load(cfs_rq, 0); |
6d5ab2932 sched: Simplify u... |
876 |
update_cfs_shares(cfs_rq); |
3ff6dcac7 sched: Fix poor i... |
877 878 879 880 881 882 |
} } # else /* CONFIG_SMP */ static void update_cfs_load(struct cfs_rq *cfs_rq, int global_update) { } |
6d5ab2932 sched: Simplify u... |
883 |
static inline long calc_cfs_shares(struct cfs_rq *cfs_rq, struct task_group *tg) |
3ff6dcac7 sched: Fix poor i... |
884 885 886 887 888 889 890 891 |
{ return tg->shares; } static inline void update_entity_shares_tick(struct cfs_rq *cfs_rq) { } # endif /* CONFIG_SMP */ |
2069dd75c sched: Rewrite tg... |
892 893 894 |
static void reweight_entity(struct cfs_rq *cfs_rq, struct sched_entity *se, unsigned long weight) { |
19e5eebb8 sched: Fix intera... |
895 896 897 898 |
if (se->on_rq) { /* commit outstanding execution time */ if (cfs_rq->curr == se) update_curr(cfs_rq); |
2069dd75c sched: Rewrite tg... |
899 |
account_entity_dequeue(cfs_rq, se); |
19e5eebb8 sched: Fix intera... |
900 |
} |
2069dd75c sched: Rewrite tg... |
901 902 903 904 905 906 |
update_load_set(&se->load, weight); if (se->on_rq) account_entity_enqueue(cfs_rq, se); } |
6d5ab2932 sched: Simplify u... |
907 |
static void update_cfs_shares(struct cfs_rq *cfs_rq) |
2069dd75c sched: Rewrite tg... |
908 909 910 |
{ struct task_group *tg; struct sched_entity *se; |
3ff6dcac7 sched: Fix poor i... |
911 |
long shares; |
2069dd75c sched: Rewrite tg... |
912 |
|
2069dd75c sched: Rewrite tg... |
913 914 |
tg = cfs_rq->tg; se = tg->se[cpu_of(rq_of(cfs_rq))]; |
64660c864 sched: Prevent in... |
915 |
if (!se || throttled_hierarchy(cfs_rq)) |
2069dd75c sched: Rewrite tg... |
916 |
return; |
3ff6dcac7 sched: Fix poor i... |
917 918 919 920 |
#ifndef CONFIG_SMP if (likely(se->load.weight == tg->shares)) return; #endif |
6d5ab2932 sched: Simplify u... |
921 |
shares = calc_cfs_shares(cfs_rq, tg); |
2069dd75c sched: Rewrite tg... |
922 923 924 925 |
reweight_entity(cfs_rq_of(se), se, shares); } #else /* CONFIG_FAIR_GROUP_SCHED */ |
d6b559182 sched: Allow upda... |
926 |
static void update_cfs_load(struct cfs_rq *cfs_rq, int global_update) |
2069dd75c sched: Rewrite tg... |
927 928 |
{ } |
6d5ab2932 sched: Simplify u... |
929 |
static inline void update_cfs_shares(struct cfs_rq *cfs_rq) |
2069dd75c sched: Rewrite tg... |
930 931 |
{ } |
43365bd7f sched: Move perio... |
932 933 934 935 |
static inline void update_entity_shares_tick(struct cfs_rq *cfs_rq) { } |
2069dd75c sched: Rewrite tg... |
936 |
#endif /* CONFIG_FAIR_GROUP_SCHED */ |
2396af69b sched: remove the... |
937 |
static void enqueue_sleeper(struct cfs_rq *cfs_rq, struct sched_entity *se) |
bf0f6f24a sched: cfs core, ... |
938 |
{ |
bf0f6f24a sched: cfs core, ... |
939 |
#ifdef CONFIG_SCHEDSTATS |
e414314cc sched: Fix latenc... |
940 941 942 943 |
struct task_struct *tsk = NULL; if (entity_is_task(se)) tsk = task_of(se); |
41acab885 sched: Implement ... |
944 945 |
if (se->statistics.sleep_start) { u64 delta = rq_of(cfs_rq)->clock - se->statistics.sleep_start; |
bf0f6f24a sched: cfs core, ... |
946 947 948 |
if ((s64)delta < 0) delta = 0; |
41acab885 sched: Implement ... |
949 950 |
if (unlikely(delta > se->statistics.sleep_max)) se->statistics.sleep_max = delta; |
bf0f6f24a sched: cfs core, ... |
951 |
|
41acab885 sched: Implement ... |
952 |
se->statistics.sum_sleep_runtime += delta; |
9745512ce sched: latencytop... |
953 |
|
768d0c272 sched: Add wait, ... |
954 |
if (tsk) { |
e414314cc sched: Fix latenc... |
955 |
account_scheduler_latency(tsk, delta >> 10, 1); |
768d0c272 sched: Add wait, ... |
956 957 |
trace_sched_stat_sleep(tsk, delta); } |
bf0f6f24a sched: cfs core, ... |
958 |
} |
41acab885 sched: Implement ... |
959 960 |
if (se->statistics.block_start) { u64 delta = rq_of(cfs_rq)->clock - se->statistics.block_start; |
bf0f6f24a sched: cfs core, ... |
961 962 963 |
if ((s64)delta < 0) delta = 0; |
41acab885 sched: Implement ... |
964 965 |
if (unlikely(delta > se->statistics.block_max)) se->statistics.block_max = delta; |
bf0f6f24a sched: cfs core, ... |
966 |
|
41acab885 sched: Implement ... |
967 |
se->statistics.sum_sleep_runtime += delta; |
30084fbd1 sched: fix profil... |
968 |
|
e414314cc sched: Fix latenc... |
969 |
if (tsk) { |
8f0dfc34e sched: Provide io... |
970 |
if (tsk->in_iowait) { |
41acab885 sched: Implement ... |
971 972 |
se->statistics.iowait_sum += delta; se->statistics.iowait_count++; |
768d0c272 sched: Add wait, ... |
973 |
trace_sched_stat_iowait(tsk, delta); |
8f0dfc34e sched: Provide io... |
974 |
} |
b781a602a events, sched: Ad... |
975 |
trace_sched_stat_blocked(tsk, delta); |
e414314cc sched: Fix latenc... |
976 977 978 979 980 981 982 983 984 985 986 |
/* * Blocking time is in units of nanosecs, so shift by * 20 to get a milliseconds-range estimation of the * amount of time that the task spent sleeping: */ if (unlikely(prof_on == SLEEP_PROFILING)) { profile_hits(SLEEP_PROFILING, (void *)get_wchan(tsk), delta >> 20); } account_scheduler_latency(tsk, delta >> 10, 0); |
30084fbd1 sched: fix profil... |
987 |
} |
bf0f6f24a sched: cfs core, ... |
988 989 990 |
} #endif } |
ddc972975 sched debug: chec... |
991 992 993 994 995 996 997 998 999 1000 1001 1002 |
static void check_spread(struct cfs_rq *cfs_rq, struct sched_entity *se) { #ifdef CONFIG_SCHED_DEBUG s64 d = se->vruntime - cfs_rq->min_vruntime; if (d < 0) d = -d; if (d > 3*sysctl_sched_latency) schedstat_inc(cfs_rq, nr_spread_over); #endif } |
bf0f6f24a sched: cfs core, ... |
1003 |
static void |
aeb73b040 sched: clean up n... |
1004 1005 |
place_entity(struct cfs_rq *cfs_rq, struct sched_entity *se, int initial) { |
1af5f730f sched: more accur... |
1006 |
u64 vruntime = cfs_rq->min_vruntime; |
94dfb5e75 sched: add tree b... |
1007 |
|
2cb8600e6 sched: documentat... |
1008 1009 1010 1011 1012 1013 |
/* * The 'current' period is already promised to the current tasks, * however the extra weight of the new task will slow them down a * little, place the new task so that it fits in the slot that * stays open at the end. */ |
94dfb5e75 sched: add tree b... |
1014 |
if (initial && sched_feat(START_DEBIT)) |
f9c0b0950 sched: revert bac... |
1015 |
vruntime += sched_vslice(cfs_rq, se); |
aeb73b040 sched: clean up n... |
1016 |
|
a2e7a7eb2 sched: Remove unn... |
1017 |
/* sleeps up to a single latency don't count. */ |
5ca9880c6 sched: Remove FAI... |
1018 |
if (!initial) { |
a2e7a7eb2 sched: Remove unn... |
1019 |
unsigned long thresh = sysctl_sched_latency; |
a7be37ac8 sched: revert the... |
1020 |
|
a2e7a7eb2 sched: Remove unn... |
1021 |
/* |
a2e7a7eb2 sched: Remove unn... |
1022 1023 1024 1025 1026 |
* Halve their sleep time's effect, to allow * for a gentler effect of sleepers: */ if (sched_feat(GENTLE_FAIR_SLEEPERS)) thresh >>= 1; |
51e0304ce sched: Implement ... |
1027 |
|
a2e7a7eb2 sched: Remove unn... |
1028 |
vruntime -= thresh; |
aeb73b040 sched: clean up n... |
1029 |
} |
b5d9d734a sched: Ensure tha... |
1030 1031 |
/* ensure we never gain time by being placed backwards. */ vruntime = max_vruntime(se->vruntime, vruntime); |
67e9fb2a3 sched: add vslice |
1032 |
se->vruntime = vruntime; |
aeb73b040 sched: clean up n... |
1033 |
} |
d3d9dc330 sched: Throttle e... |
1034 |
static void check_enqueue_throttle(struct cfs_rq *cfs_rq); |
aeb73b040 sched: clean up n... |
1035 |
static void |
88ec22d3e sched: Remove the... |
1036 |
enqueue_entity(struct cfs_rq *cfs_rq, struct sched_entity *se, int flags) |
bf0f6f24a sched: cfs core, ... |
1037 1038 |
{ /* |
88ec22d3e sched: Remove the... |
1039 1040 1041 |
* Update the normalized vruntime before updating min_vruntime * through callig update_curr(). */ |
371fd7e7a sched: Add enqueu... |
1042 |
if (!(flags & ENQUEUE_WAKEUP) || (flags & ENQUEUE_WAKING)) |
88ec22d3e sched: Remove the... |
1043 1044 1045 |
se->vruntime += cfs_rq->min_vruntime; /* |
a2a2d6807 sched: cleanup, m... |
1046 |
* Update run-time statistics of the 'current'. |
bf0f6f24a sched: cfs core, ... |
1047 |
*/ |
b7cc08965 sched: remove the... |
1048 |
update_curr(cfs_rq); |
d6b559182 sched: Allow upda... |
1049 |
update_cfs_load(cfs_rq, 0); |
a992241de sched: fix normal... |
1050 |
account_entity_enqueue(cfs_rq, se); |
6d5ab2932 sched: Simplify u... |
1051 |
update_cfs_shares(cfs_rq); |
bf0f6f24a sched: cfs core, ... |
1052 |
|
88ec22d3e sched: Remove the... |
1053 |
if (flags & ENQUEUE_WAKEUP) { |
aeb73b040 sched: clean up n... |
1054 |
place_entity(cfs_rq, se, 0); |
2396af69b sched: remove the... |
1055 |
enqueue_sleeper(cfs_rq, se); |
e9acbff64 sched: introduce ... |
1056 |
} |
bf0f6f24a sched: cfs core, ... |
1057 |
|
d2417e5a3 sched: remove the... |
1058 |
update_stats_enqueue(cfs_rq, se); |
ddc972975 sched debug: chec... |
1059 |
check_spread(cfs_rq, se); |
83b699ed2 sched: revert rec... |
1060 1061 |
if (se != cfs_rq->curr) __enqueue_entity(cfs_rq, se); |
2069dd75c sched: Rewrite tg... |
1062 |
se->on_rq = 1; |
3d4b47b4b sched: Implement ... |
1063 |
|
d3d9dc330 sched: Throttle e... |
1064 |
if (cfs_rq->nr_running == 1) { |
3d4b47b4b sched: Implement ... |
1065 |
list_add_leaf_cfs_rq(cfs_rq); |
d3d9dc330 sched: Throttle e... |
1066 1067 |
check_enqueue_throttle(cfs_rq); } |
bf0f6f24a sched: cfs core, ... |
1068 |
} |
2c13c919d sched: Limit the ... |
1069 |
static void __clear_buddies_last(struct sched_entity *se) |
2002c6959 sched: release bu... |
1070 |
{ |
2c13c919d sched: Limit the ... |
1071 1072 1073 1074 1075 1076 1077 1078 |
for_each_sched_entity(se) { struct cfs_rq *cfs_rq = cfs_rq_of(se); if (cfs_rq->last == se) cfs_rq->last = NULL; else break; } } |
2002c6959 sched: release bu... |
1079 |
|
2c13c919d sched: Limit the ... |
1080 1081 1082 1083 1084 1085 1086 1087 1088 |
static void __clear_buddies_next(struct sched_entity *se) { for_each_sched_entity(se) { struct cfs_rq *cfs_rq = cfs_rq_of(se); if (cfs_rq->next == se) cfs_rq->next = NULL; else break; } |
2002c6959 sched: release bu... |
1089 |
} |
ac53db596 sched: Use a budd... |
1090 1091 1092 1093 1094 1095 1096 1097 1098 1099 |
static void __clear_buddies_skip(struct sched_entity *se) { for_each_sched_entity(se) { struct cfs_rq *cfs_rq = cfs_rq_of(se); if (cfs_rq->skip == se) cfs_rq->skip = NULL; else break; } } |
a571bbeaf sched: fix buddie... |
1100 1101 |
static void clear_buddies(struct cfs_rq *cfs_rq, struct sched_entity *se) { |
2c13c919d sched: Limit the ... |
1102 1103 1104 1105 1106 |
if (cfs_rq->last == se) __clear_buddies_last(se); if (cfs_rq->next == se) __clear_buddies_next(se); |
ac53db596 sched: Use a budd... |
1107 1108 1109 |
if (cfs_rq->skip == se) __clear_buddies_skip(se); |
a571bbeaf sched: fix buddie... |
1110 |
} |
d8b4986d3 sched: Return unu... |
1111 |
static void return_cfs_rq_runtime(struct cfs_rq *cfs_rq); |
bf0f6f24a sched: cfs core, ... |
1112 |
static void |
371fd7e7a sched: Add enqueu... |
1113 |
dequeue_entity(struct cfs_rq *cfs_rq, struct sched_entity *se, int flags) |
bf0f6f24a sched: cfs core, ... |
1114 |
{ |
a2a2d6807 sched: cleanup, m... |
1115 1116 1117 1118 |
/* * Update run-time statistics of the 'current'. */ update_curr(cfs_rq); |
19b6a2e37 sched: remove the... |
1119 |
update_stats_dequeue(cfs_rq, se); |
371fd7e7a sched: Add enqueu... |
1120 |
if (flags & DEQUEUE_SLEEP) { |
67e9fb2a3 sched: add vslice |
1121 |
#ifdef CONFIG_SCHEDSTATS |
bf0f6f24a sched: cfs core, ... |
1122 1123 1124 1125 |
if (entity_is_task(se)) { struct task_struct *tsk = task_of(se); if (tsk->state & TASK_INTERRUPTIBLE) |
41acab885 sched: Implement ... |
1126 |
se->statistics.sleep_start = rq_of(cfs_rq)->clock; |
bf0f6f24a sched: cfs core, ... |
1127 |
if (tsk->state & TASK_UNINTERRUPTIBLE) |
41acab885 sched: Implement ... |
1128 |
se->statistics.block_start = rq_of(cfs_rq)->clock; |
bf0f6f24a sched: cfs core, ... |
1129 |
} |
db36cc7d6 sched: clean up s... |
1130 |
#endif |
67e9fb2a3 sched: add vslice |
1131 |
} |
2002c6959 sched: release bu... |
1132 |
clear_buddies(cfs_rq, se); |
4793241be sched: backward l... |
1133 |
|
83b699ed2 sched: revert rec... |
1134 |
if (se != cfs_rq->curr) |
30cfdcfc5 sched: do not kee... |
1135 |
__dequeue_entity(cfs_rq, se); |
2069dd75c sched: Rewrite tg... |
1136 |
se->on_rq = 0; |
d6b559182 sched: Allow upda... |
1137 |
update_cfs_load(cfs_rq, 0); |
30cfdcfc5 sched: do not kee... |
1138 |
account_entity_dequeue(cfs_rq, se); |
88ec22d3e sched: Remove the... |
1139 1140 1141 1142 1143 1144 |
/* * Normalize the entity after updating the min_vruntime because the * update can refer to the ->curr item and we need to reflect this * movement in our normalized position. */ |
371fd7e7a sched: Add enqueu... |
1145 |
if (!(flags & DEQUEUE_SLEEP)) |
88ec22d3e sched: Remove the... |
1146 |
se->vruntime -= cfs_rq->min_vruntime; |
1e8762317 sched: Fix ->min_... |
1147 |
|
d8b4986d3 sched: Return unu... |
1148 1149 |
/* return excess runtime on last dequeue */ return_cfs_rq_runtime(cfs_rq); |
1e8762317 sched: Fix ->min_... |
1150 1151 |
update_min_vruntime(cfs_rq); update_cfs_shares(cfs_rq); |
bf0f6f24a sched: cfs core, ... |
1152 1153 1154 1155 1156 |
} /* * Preempt the current task with a newly woken task if needed: */ |
7c92e54f6 sched: simplify _... |
1157 |
static void |
2e09bf556 sched: wakeup gra... |
1158 |
check_preempt_tick(struct cfs_rq *cfs_rq, struct sched_entity *curr) |
bf0f6f24a sched: cfs core, ... |
1159 |
{ |
116978308 sched: fix ideal_... |
1160 |
unsigned long ideal_runtime, delta_exec; |
f4cfb33ed sched: Remove red... |
1161 1162 |
struct sched_entity *se; s64 delta; |
116978308 sched: fix ideal_... |
1163 |
|
6d0f0ebd0 sched: simplify a... |
1164 |
ideal_runtime = sched_slice(cfs_rq, curr); |
116978308 sched: fix ideal_... |
1165 |
delta_exec = curr->sum_exec_runtime - curr->prev_sum_exec_runtime; |
a9f3e2b54 sched: clear budd... |
1166 |
if (delta_exec > ideal_runtime) { |
bf0f6f24a sched: cfs core, ... |
1167 |
resched_task(rq_of(cfs_rq)->curr); |
a9f3e2b54 sched: clear budd... |
1168 1169 1170 1171 1172 |
/* * The current task ran long enough, ensure it doesn't get * re-elected due to buddy favours. */ clear_buddies(cfs_rq, curr); |
f685ceaca sched: Strengthen... |
1173 1174 1175 1176 1177 1178 1179 1180 |
return; } /* * Ensure that a task that missed wakeup preemption by a * narrow margin doesn't have to wait for a full slice. * This also mitigates buddy induced latencies under load. */ |
f685ceaca sched: Strengthen... |
1181 1182 |
if (delta_exec < sysctl_sched_min_granularity) return; |
f4cfb33ed sched: Remove red... |
1183 1184 |
se = __pick_first_entity(cfs_rq); delta = curr->vruntime - se->vruntime; |
f685ceaca sched: Strengthen... |
1185 |
|
f4cfb33ed sched: Remove red... |
1186 1187 |
if (delta < 0) return; |
d7d829441 sched: Fix signed... |
1188 |
|
f4cfb33ed sched: Remove red... |
1189 1190 |
if (delta > ideal_runtime) resched_task(rq_of(cfs_rq)->curr); |
bf0f6f24a sched: cfs core, ... |
1191 |
} |
83b699ed2 sched: revert rec... |
1192 |
static void |
8494f412e sched: remove the... |
1193 |
set_next_entity(struct cfs_rq *cfs_rq, struct sched_entity *se) |
bf0f6f24a sched: cfs core, ... |
1194 |
{ |
83b699ed2 sched: revert rec... |
1195 1196 1197 1198 1199 1200 1201 1202 1203 1204 |
/* 'current' is not kept within the tree. */ if (se->on_rq) { /* * Any task has to be enqueued before it get to execute on * a CPU. So account for the time it spent waiting on the * runqueue. */ update_stats_wait_end(cfs_rq, se); __dequeue_entity(cfs_rq, se); } |
79303e9e0 sched: remove the... |
1205 |
update_stats_curr_start(cfs_rq, se); |
429d43bcc sched: cleanup: s... |
1206 |
cfs_rq->curr = se; |
eba1ed4b7 sched: debug: tra... |
1207 1208 1209 1210 1211 1212 |
#ifdef CONFIG_SCHEDSTATS /* * Track our maximum slice length, if the CPU's load is at * least twice that of our own weight (i.e. dont track it * when there are only lesser-weight tasks around): */ |
495eca494 sched: clean up s... |
1213 |
if (rq_of(cfs_rq)->load.weight >= 2*se->load.weight) { |
41acab885 sched: Implement ... |
1214 |
se->statistics.slice_max = max(se->statistics.slice_max, |
eba1ed4b7 sched: debug: tra... |
1215 1216 1217 |
se->sum_exec_runtime - se->prev_sum_exec_runtime); } #endif |
4a55b4503 sched: improve pr... |
1218 |
se->prev_sum_exec_runtime = se->sum_exec_runtime; |
bf0f6f24a sched: cfs core, ... |
1219 |
} |
3f3a49048 sched: virtual ti... |
1220 1221 |
static int wakeup_preempt_entity(struct sched_entity *curr, struct sched_entity *se); |
ac53db596 sched: Use a budd... |
1222 1223 1224 1225 1226 1227 1228 |
/* * Pick the next process, keeping these things in mind, in this order: * 1) keep things fair between processes/task groups * 2) pick the "next" process, since someone really wants that to run * 3) pick the "last" process, for cache locality * 4) do not run the "skip" process, if something else is available */ |
f4b6755fb sched: cleanup fa... |
1229 |
static struct sched_entity *pick_next_entity(struct cfs_rq *cfs_rq) |
aa2ac2522 sched: fix overlo... |
1230 |
{ |
ac53db596 sched: Use a budd... |
1231 |
struct sched_entity *se = __pick_first_entity(cfs_rq); |
f685ceaca sched: Strengthen... |
1232 |
struct sched_entity *left = se; |
f4b6755fb sched: cleanup fa... |
1233 |
|
ac53db596 sched: Use a budd... |
1234 1235 1236 1237 1238 1239 1240 1241 1242 |
/* * Avoid running the skip buddy, if running something else can * be done without getting too unfair. */ if (cfs_rq->skip == se) { struct sched_entity *second = __pick_next_entity(se); if (second && wakeup_preempt_entity(second, left) < 1) se = second; } |
aa2ac2522 sched: fix overlo... |
1243 |
|
f685ceaca sched: Strengthen... |
1244 1245 1246 1247 1248 |
/* * Prefer last buddy, try to return the CPU to a preempted task. */ if (cfs_rq->last && wakeup_preempt_entity(cfs_rq->last, left) < 1) se = cfs_rq->last; |
ac53db596 sched: Use a budd... |
1249 1250 1251 1252 1253 |
/* * Someone really wants this to run. If it's not unfair, run it. */ if (cfs_rq->next && wakeup_preempt_entity(cfs_rq->next, left) < 1) se = cfs_rq->next; |
f685ceaca sched: Strengthen... |
1254 |
clear_buddies(cfs_rq, se); |
4793241be sched: backward l... |
1255 1256 |
return se; |
aa2ac2522 sched: fix overlo... |
1257 |
} |
d3d9dc330 sched: Throttle e... |
1258 |
static void check_cfs_rq_runtime(struct cfs_rq *cfs_rq); |
ab6cde269 sched: remove the... |
1259 |
static void put_prev_entity(struct cfs_rq *cfs_rq, struct sched_entity *prev) |
bf0f6f24a sched: cfs core, ... |
1260 1261 1262 1263 1264 1265 |
{ /* * If still on the runqueue then deactivate_task() * was not called and update_curr() has to be done: */ if (prev->on_rq) |
b7cc08965 sched: remove the... |
1266 |
update_curr(cfs_rq); |
bf0f6f24a sched: cfs core, ... |
1267 |
|
d3d9dc330 sched: Throttle e... |
1268 1269 |
/* throttle cfs_rqs exceeding runtime */ check_cfs_rq_runtime(cfs_rq); |
ddc972975 sched debug: chec... |
1270 |
check_spread(cfs_rq, prev); |
30cfdcfc5 sched: do not kee... |
1271 |
if (prev->on_rq) { |
5870db5b8 sched: remove the... |
1272 |
update_stats_wait_start(cfs_rq, prev); |
30cfdcfc5 sched: do not kee... |
1273 1274 1275 |
/* Put 'current' back into the tree. */ __enqueue_entity(cfs_rq, prev); } |
429d43bcc sched: cleanup: s... |
1276 |
cfs_rq->curr = NULL; |
bf0f6f24a sched: cfs core, ... |
1277 |
} |
8f4d37ec0 sched: high-res p... |
1278 1279 |
static void entity_tick(struct cfs_rq *cfs_rq, struct sched_entity *curr, int queued) |
bf0f6f24a sched: cfs core, ... |
1280 |
{ |
bf0f6f24a sched: cfs core, ... |
1281 |
/* |
30cfdcfc5 sched: do not kee... |
1282 |
* Update run-time statistics of the 'current'. |
bf0f6f24a sched: cfs core, ... |
1283 |
*/ |
30cfdcfc5 sched: do not kee... |
1284 |
update_curr(cfs_rq); |
bf0f6f24a sched: cfs core, ... |
1285 |
|
43365bd7f sched: Move perio... |
1286 1287 1288 1289 |
/* * Update share accounting for long-running entities. */ update_entity_shares_tick(cfs_rq); |
8f4d37ec0 sched: high-res p... |
1290 1291 1292 1293 1294 |
#ifdef CONFIG_SCHED_HRTICK /* * queued ticks are scheduled to match the slice, so don't bother * validating it and just reschedule. */ |
983ed7a66 sched: add static... |
1295 1296 1297 1298 |
if (queued) { resched_task(rq_of(cfs_rq)->curr); return; } |
8f4d37ec0 sched: high-res p... |
1299 1300 1301 1302 1303 1304 1305 |
/* * don't let the period tick interfere with the hrtick preemption */ if (!sched_feat(DOUBLE_TICK) && hrtimer_active(&rq_of(cfs_rq)->hrtick_timer)) return; #endif |
2c2efaed9 sched: Kill WAKEU... |
1306 |
if (cfs_rq->nr_running > 1) |
2e09bf556 sched: wakeup gra... |
1307 |
check_preempt_tick(cfs_rq, curr); |
bf0f6f24a sched: cfs core, ... |
1308 |
} |
ab84d31e1 sched: Introduce ... |
1309 1310 1311 1312 1313 1314 |
/************************************************** * CFS bandwidth control machinery */ #ifdef CONFIG_CFS_BANDWIDTH |
029632fbb sched: Make separ... |
1315 1316 1317 1318 1319 1320 1321 1322 1323 1324 1325 1326 1327 1328 1329 1330 1331 1332 1333 1334 1335 1336 1337 1338 1339 |
#ifdef HAVE_JUMP_LABEL static struct jump_label_key __cfs_bandwidth_used; static inline bool cfs_bandwidth_used(void) { return static_branch(&__cfs_bandwidth_used); } void account_cfs_bandwidth_used(int enabled, int was_enabled) { /* only need to count groups transitioning between enabled/!enabled */ if (enabled && !was_enabled) jump_label_inc(&__cfs_bandwidth_used); else if (!enabled && was_enabled) jump_label_dec(&__cfs_bandwidth_used); } #else /* HAVE_JUMP_LABEL */ static bool cfs_bandwidth_used(void) { return true; } void account_cfs_bandwidth_used(int enabled, int was_enabled) {} #endif /* HAVE_JUMP_LABEL */ |
ab84d31e1 sched: Introduce ... |
1340 1341 1342 1343 1344 1345 1346 1347 |
/* * default period for cfs group bandwidth. * default: 0.1s, units: nanoseconds */ static inline u64 default_cfs_period(void) { return 100000000ULL; } |
ec12cb7f3 sched: Accumulate... |
1348 1349 1350 1351 1352 |
static inline u64 sched_cfs_bandwidth_slice(void) { return (u64)sysctl_sched_cfs_bandwidth_slice * NSEC_PER_USEC; } |
a9cf55b28 sched: Expire inv... |
1353 1354 1355 1356 1357 1358 1359 |
/* * Replenish runtime according to assigned quota and update expiration time. * We use sched_clock_cpu directly instead of rq->clock to avoid adding * additional synchronization around rq->lock. * * requires cfs_b->lock */ |
029632fbb sched: Make separ... |
1360 |
void __refill_cfs_bandwidth_runtime(struct cfs_bandwidth *cfs_b) |
a9cf55b28 sched: Expire inv... |
1361 1362 1363 1364 1365 1366 1367 1368 1369 1370 |
{ u64 now; if (cfs_b->quota == RUNTIME_INF) return; now = sched_clock_cpu(smp_processor_id()); cfs_b->runtime = cfs_b->quota; cfs_b->runtime_expires = now + ktime_to_ns(cfs_b->period); } |
029632fbb sched: Make separ... |
1371 1372 1373 1374 |
static inline struct cfs_bandwidth *tg_cfs_bandwidth(struct task_group *tg) { return &tg->cfs_bandwidth; } |
85dac906b sched: Add suppor... |
1375 1376 |
/* returns 0 on failure to allocate runtime */ static int assign_cfs_rq_runtime(struct cfs_rq *cfs_rq) |
ec12cb7f3 sched: Accumulate... |
1377 1378 1379 |
{ struct task_group *tg = cfs_rq->tg; struct cfs_bandwidth *cfs_b = tg_cfs_bandwidth(tg); |
a9cf55b28 sched: Expire inv... |
1380 |
u64 amount = 0, min_amount, expires; |
ec12cb7f3 sched: Accumulate... |
1381 1382 1383 1384 1385 1386 1387 |
/* note: this is a positive sum as runtime_remaining <= 0 */ min_amount = sched_cfs_bandwidth_slice() - cfs_rq->runtime_remaining; raw_spin_lock(&cfs_b->lock); if (cfs_b->quota == RUNTIME_INF) amount = min_amount; |
58088ad01 sched: Add a time... |
1388 |
else { |
a9cf55b28 sched: Expire inv... |
1389 1390 1391 1392 1393 1394 1395 1396 |
/* * If the bandwidth pool has become inactive, then at least one * period must have elapsed since the last consumption. * Refresh the global state and ensure bandwidth timer becomes * active. */ if (!cfs_b->timer_active) { __refill_cfs_bandwidth_runtime(cfs_b); |
58088ad01 sched: Add a time... |
1397 |
__start_cfs_bandwidth(cfs_b); |
a9cf55b28 sched: Expire inv... |
1398 |
} |
58088ad01 sched: Add a time... |
1399 1400 1401 1402 1403 1404 |
if (cfs_b->runtime > 0) { amount = min(cfs_b->runtime, min_amount); cfs_b->runtime -= amount; cfs_b->idle = 0; } |
ec12cb7f3 sched: Accumulate... |
1405 |
} |
a9cf55b28 sched: Expire inv... |
1406 |
expires = cfs_b->runtime_expires; |
ec12cb7f3 sched: Accumulate... |
1407 1408 1409 |
raw_spin_unlock(&cfs_b->lock); cfs_rq->runtime_remaining += amount; |
a9cf55b28 sched: Expire inv... |
1410 1411 1412 1413 1414 1415 1416 |
/* * we may have advanced our local expiration to account for allowed * spread between our sched_clock and the one on which runtime was * issued. */ if ((s64)(expires - cfs_rq->runtime_expires) > 0) cfs_rq->runtime_expires = expires; |
85dac906b sched: Add suppor... |
1417 1418 |
return cfs_rq->runtime_remaining > 0; |
ec12cb7f3 sched: Accumulate... |
1419 |
} |
a9cf55b28 sched: Expire inv... |
1420 1421 1422 1423 1424 |
/* * Note: This depends on the synchronization provided by sched_clock and the * fact that rq->clock snapshots this value. */ static void expire_cfs_rq_runtime(struct cfs_rq *cfs_rq) |
ec12cb7f3 sched: Accumulate... |
1425 |
{ |
a9cf55b28 sched: Expire inv... |
1426 1427 1428 1429 1430 |
struct cfs_bandwidth *cfs_b = tg_cfs_bandwidth(cfs_rq->tg); struct rq *rq = rq_of(cfs_rq); /* if the deadline is ahead of our clock, nothing to do */ if (likely((s64)(rq->clock - cfs_rq->runtime_expires) < 0)) |
ec12cb7f3 sched: Accumulate... |
1431 |
return; |
a9cf55b28 sched: Expire inv... |
1432 1433 1434 1435 1436 1437 1438 1439 1440 1441 1442 1443 1444 1445 1446 1447 1448 1449 1450 1451 1452 1453 1454 1455 1456 |
if (cfs_rq->runtime_remaining < 0) return; /* * If the local deadline has passed we have to consider the * possibility that our sched_clock is 'fast' and the global deadline * has not truly expired. * * Fortunately we can check determine whether this the case by checking * whether the global deadline has advanced. */ if ((s64)(cfs_rq->runtime_expires - cfs_b->runtime_expires) >= 0) { /* extend local deadline, drift is bounded above by 2 ticks */ cfs_rq->runtime_expires += TICK_NSEC; } else { /* global deadline is ahead, expiration has passed */ cfs_rq->runtime_remaining = 0; } } static void __account_cfs_rq_runtime(struct cfs_rq *cfs_rq, unsigned long delta_exec) { /* dock delta_exec before expiring quota (as it could span periods) */ |
ec12cb7f3 sched: Accumulate... |
1457 |
cfs_rq->runtime_remaining -= delta_exec; |
a9cf55b28 sched: Expire inv... |
1458 1459 1460 |
expire_cfs_rq_runtime(cfs_rq); if (likely(cfs_rq->runtime_remaining > 0)) |
ec12cb7f3 sched: Accumulate... |
1461 |
return; |
85dac906b sched: Add suppor... |
1462 1463 1464 1465 1466 1467 |
/* * if we're unable to extend our runtime we resched so that the active * hierarchy can be throttled */ if (!assign_cfs_rq_runtime(cfs_rq) && likely(cfs_rq->curr)) resched_task(rq_of(cfs_rq)->curr); |
ec12cb7f3 sched: Accumulate... |
1468 1469 1470 1471 1472 |
} static __always_inline void account_cfs_rq_runtime(struct cfs_rq *cfs_rq, unsigned long delta_exec) { |
56f570e51 sched: Use jump l... |
1473 |
if (!cfs_bandwidth_used() || !cfs_rq->runtime_enabled) |
ec12cb7f3 sched: Accumulate... |
1474 1475 1476 1477 |
return; __account_cfs_rq_runtime(cfs_rq, delta_exec); } |
85dac906b sched: Add suppor... |
1478 1479 |
static inline int cfs_rq_throttled(struct cfs_rq *cfs_rq) { |
56f570e51 sched: Use jump l... |
1480 |
return cfs_bandwidth_used() && cfs_rq->throttled; |
85dac906b sched: Add suppor... |
1481 |
} |
64660c864 sched: Prevent in... |
1482 1483 1484 |
/* check whether cfs_rq, or any parent, is throttled */ static inline int throttled_hierarchy(struct cfs_rq *cfs_rq) { |
56f570e51 sched: Use jump l... |
1485 |
return cfs_bandwidth_used() && cfs_rq->throttle_count; |
64660c864 sched: Prevent in... |
1486 1487 1488 1489 1490 1491 1492 1493 1494 1495 1496 1497 1498 1499 1500 1501 1502 1503 1504 1505 1506 1507 1508 1509 1510 1511 1512 1513 1514 1515 1516 1517 1518 1519 1520 1521 1522 1523 1524 1525 1526 1527 1528 1529 1530 1531 1532 1533 1534 1535 1536 1537 1538 1539 |
} /* * Ensure that neither of the group entities corresponding to src_cpu or * dest_cpu are members of a throttled hierarchy when performing group * load-balance operations. */ static inline int throttled_lb_pair(struct task_group *tg, int src_cpu, int dest_cpu) { struct cfs_rq *src_cfs_rq, *dest_cfs_rq; src_cfs_rq = tg->cfs_rq[src_cpu]; dest_cfs_rq = tg->cfs_rq[dest_cpu]; return throttled_hierarchy(src_cfs_rq) || throttled_hierarchy(dest_cfs_rq); } /* updated child weight may affect parent so we have to do this bottom up */ static int tg_unthrottle_up(struct task_group *tg, void *data) { struct rq *rq = data; struct cfs_rq *cfs_rq = tg->cfs_rq[cpu_of(rq)]; cfs_rq->throttle_count--; #ifdef CONFIG_SMP if (!cfs_rq->throttle_count) { u64 delta = rq->clock_task - cfs_rq->load_stamp; /* leaving throttled state, advance shares averaging windows */ cfs_rq->load_stamp += delta; cfs_rq->load_last += delta; /* update entity weight now that we are on_rq again */ update_cfs_shares(cfs_rq); } #endif return 0; } static int tg_throttle_down(struct task_group *tg, void *data) { struct rq *rq = data; struct cfs_rq *cfs_rq = tg->cfs_rq[cpu_of(rq)]; /* group is entering throttled state, record last load */ if (!cfs_rq->throttle_count) update_cfs_load(cfs_rq, 0); cfs_rq->throttle_count++; return 0; } |
d3d9dc330 sched: Throttle e... |
1540 |
static void throttle_cfs_rq(struct cfs_rq *cfs_rq) |
85dac906b sched: Add suppor... |
1541 1542 1543 1544 1545 1546 1547 1548 1549 |
{ struct rq *rq = rq_of(cfs_rq); struct cfs_bandwidth *cfs_b = tg_cfs_bandwidth(cfs_rq->tg); struct sched_entity *se; long task_delta, dequeue = 1; se = cfs_rq->tg->se[cpu_of(rq_of(cfs_rq))]; /* account load preceding throttle */ |
64660c864 sched: Prevent in... |
1550 1551 1552 |
rcu_read_lock(); walk_tg_tree_from(cfs_rq->tg, tg_throttle_down, tg_nop, (void *)rq); rcu_read_unlock(); |
85dac906b sched: Add suppor... |
1553 1554 1555 1556 1557 1558 1559 1560 1561 1562 1563 1564 1565 1566 1567 1568 1569 1570 1571 1572 |
task_delta = cfs_rq->h_nr_running; for_each_sched_entity(se) { struct cfs_rq *qcfs_rq = cfs_rq_of(se); /* throttled entity or throttle-on-deactivate */ if (!se->on_rq) break; if (dequeue) dequeue_entity(qcfs_rq, se, DEQUEUE_SLEEP); qcfs_rq->h_nr_running -= task_delta; if (qcfs_rq->load.weight) dequeue = 0; } if (!se) rq->nr_running -= task_delta; cfs_rq->throttled = 1; |
e8da1b18b sched: Add export... |
1573 |
cfs_rq->throttled_timestamp = rq->clock; |
85dac906b sched: Add suppor... |
1574 1575 1576 1577 |
raw_spin_lock(&cfs_b->lock); list_add_tail_rcu(&cfs_rq->throttled_list, &cfs_b->throttled_cfs_rq); raw_spin_unlock(&cfs_b->lock); } |
029632fbb sched: Make separ... |
1578 |
void unthrottle_cfs_rq(struct cfs_rq *cfs_rq) |
671fd9dab sched: Add suppor... |
1579 1580 1581 1582 1583 1584 1585 1586 1587 1588 1589 |
{ struct rq *rq = rq_of(cfs_rq); struct cfs_bandwidth *cfs_b = tg_cfs_bandwidth(cfs_rq->tg); struct sched_entity *se; int enqueue = 1; long task_delta; se = cfs_rq->tg->se[cpu_of(rq_of(cfs_rq))]; cfs_rq->throttled = 0; raw_spin_lock(&cfs_b->lock); |
e8da1b18b sched: Add export... |
1590 |
cfs_b->throttled_time += rq->clock - cfs_rq->throttled_timestamp; |
671fd9dab sched: Add suppor... |
1591 1592 |
list_del_rcu(&cfs_rq->throttled_list); raw_spin_unlock(&cfs_b->lock); |
e8da1b18b sched: Add export... |
1593 |
cfs_rq->throttled_timestamp = 0; |
671fd9dab sched: Add suppor... |
1594 |
|
64660c864 sched: Prevent in... |
1595 1596 1597 |
update_rq_clock(rq); /* update hierarchical throttle state */ walk_tg_tree_from(cfs_rq->tg, tg_nop, tg_unthrottle_up, (void *)rq); |
671fd9dab sched: Add suppor... |
1598 1599 1600 1601 1602 1603 1604 1605 1606 1607 1608 1609 1610 1611 1612 1613 1614 1615 1616 1617 1618 1619 1620 1621 1622 1623 1624 1625 1626 1627 1628 1629 1630 1631 1632 1633 1634 1635 1636 1637 1638 1639 1640 1641 1642 1643 1644 1645 1646 1647 1648 1649 1650 1651 1652 1653 1654 1655 1656 1657 1658 1659 |
if (!cfs_rq->load.weight) return; task_delta = cfs_rq->h_nr_running; for_each_sched_entity(se) { if (se->on_rq) enqueue = 0; cfs_rq = cfs_rq_of(se); if (enqueue) enqueue_entity(cfs_rq, se, ENQUEUE_WAKEUP); cfs_rq->h_nr_running += task_delta; if (cfs_rq_throttled(cfs_rq)) break; } if (!se) rq->nr_running += task_delta; /* determine whether we need to wake up potentially idle cpu */ if (rq->curr == rq->idle && rq->cfs.nr_running) resched_task(rq->curr); } static u64 distribute_cfs_runtime(struct cfs_bandwidth *cfs_b, u64 remaining, u64 expires) { struct cfs_rq *cfs_rq; u64 runtime = remaining; rcu_read_lock(); list_for_each_entry_rcu(cfs_rq, &cfs_b->throttled_cfs_rq, throttled_list) { struct rq *rq = rq_of(cfs_rq); raw_spin_lock(&rq->lock); if (!cfs_rq_throttled(cfs_rq)) goto next; runtime = -cfs_rq->runtime_remaining + 1; if (runtime > remaining) runtime = remaining; remaining -= runtime; cfs_rq->runtime_remaining += runtime; cfs_rq->runtime_expires = expires; /* we check whether we're throttled above */ if (cfs_rq->runtime_remaining > 0) unthrottle_cfs_rq(cfs_rq); next: raw_spin_unlock(&rq->lock); if (!remaining) break; } rcu_read_unlock(); return remaining; } |
58088ad01 sched: Add a time... |
1660 1661 1662 1663 1664 1665 1666 1667 |
/* * Responsible for refilling a task_group's bandwidth and unthrottling its * cfs_rqs as appropriate. If there has been no activity within the last * period the timer is deactivated until scheduling resumes; cfs_b->idle is * used to track this state. */ static int do_sched_cfs_period_timer(struct cfs_bandwidth *cfs_b, int overrun) { |
671fd9dab sched: Add suppor... |
1668 1669 |
u64 runtime, runtime_expires; int idle = 1, throttled; |
58088ad01 sched: Add a time... |
1670 1671 1672 1673 1674 |
raw_spin_lock(&cfs_b->lock); /* no need to continue the timer with no bandwidth constraint */ if (cfs_b->quota == RUNTIME_INF) goto out_unlock; |
671fd9dab sched: Add suppor... |
1675 1676 1677 |
throttled = !list_empty(&cfs_b->throttled_cfs_rq); /* idle depends on !throttled (for the case of a large deficit) */ idle = cfs_b->idle && !throttled; |
e8da1b18b sched: Add export... |
1678 |
cfs_b->nr_periods += overrun; |
671fd9dab sched: Add suppor... |
1679 |
|
a9cf55b28 sched: Expire inv... |
1680 1681 1682 1683 1684 |
/* if we're going inactive then everything else can be deferred */ if (idle) goto out_unlock; __refill_cfs_bandwidth_runtime(cfs_b); |
671fd9dab sched: Add suppor... |
1685 1686 1687 1688 1689 |
if (!throttled) { /* mark as potentially idle for the upcoming period */ cfs_b->idle = 1; goto out_unlock; } |
e8da1b18b sched: Add export... |
1690 1691 |
/* account preceding periods in which throttling occurred */ cfs_b->nr_throttled += overrun; |
671fd9dab sched: Add suppor... |
1692 1693 1694 1695 1696 1697 1698 1699 1700 1701 1702 1703 1704 1705 1706 1707 1708 1709 1710 1711 1712 1713 1714 1715 |
/* * There are throttled entities so we must first use the new bandwidth * to unthrottle them before making it generally available. This * ensures that all existing debts will be paid before a new cfs_rq is * allowed to run. */ runtime = cfs_b->runtime; runtime_expires = cfs_b->runtime_expires; cfs_b->runtime = 0; /* * This check is repeated as we are holding onto the new bandwidth * while we unthrottle. This can potentially race with an unthrottled * group trying to acquire new bandwidth from the global pool. */ while (throttled && runtime > 0) { raw_spin_unlock(&cfs_b->lock); /* we can't nest cfs_b->lock while distributing bandwidth */ runtime = distribute_cfs_runtime(cfs_b, runtime, runtime_expires); raw_spin_lock(&cfs_b->lock); throttled = !list_empty(&cfs_b->throttled_cfs_rq); } |
58088ad01 sched: Add a time... |
1716 |
|
671fd9dab sched: Add suppor... |
1717 1718 1719 1720 1721 1722 1723 1724 1725 |
/* return (any) remaining runtime */ cfs_b->runtime = runtime; /* * While we are ensured activity in the period following an * unthrottle, this also covers the case in which the new bandwidth is * insufficient to cover the existing bandwidth deficit. (Forcing the * timer to remain active while there are any throttled entities.) */ cfs_b->idle = 0; |
58088ad01 sched: Add a time... |
1726 1727 1728 1729 1730 1731 1732 |
out_unlock: if (idle) cfs_b->timer_active = 0; raw_spin_unlock(&cfs_b->lock); return idle; } |
d3d9dc330 sched: Throttle e... |
1733 |
|
d8b4986d3 sched: Return unu... |
1734 1735 1736 1737 1738 1739 1740 1741 1742 1743 1744 1745 1746 1747 1748 1749 1750 1751 1752 1753 1754 1755 1756 1757 1758 1759 1760 1761 1762 1763 1764 1765 1766 1767 1768 1769 1770 1771 1772 1773 1774 1775 1776 1777 1778 1779 1780 1781 1782 1783 1784 1785 1786 1787 1788 1789 1790 1791 1792 1793 1794 1795 1796 1797 |
/* a cfs_rq won't donate quota below this amount */ static const u64 min_cfs_rq_runtime = 1 * NSEC_PER_MSEC; /* minimum remaining period time to redistribute slack quota */ static const u64 min_bandwidth_expiration = 2 * NSEC_PER_MSEC; /* how long we wait to gather additional slack before distributing */ static const u64 cfs_bandwidth_slack_period = 5 * NSEC_PER_MSEC; /* are we near the end of the current quota period? */ static int runtime_refresh_within(struct cfs_bandwidth *cfs_b, u64 min_expire) { struct hrtimer *refresh_timer = &cfs_b->period_timer; u64 remaining; /* if the call-back is running a quota refresh is already occurring */ if (hrtimer_callback_running(refresh_timer)) return 1; /* is a quota refresh about to occur? */ remaining = ktime_to_ns(hrtimer_expires_remaining(refresh_timer)); if (remaining < min_expire) return 1; return 0; } static void start_cfs_slack_bandwidth(struct cfs_bandwidth *cfs_b) { u64 min_left = cfs_bandwidth_slack_period + min_bandwidth_expiration; /* if there's a quota refresh soon don't bother with slack */ if (runtime_refresh_within(cfs_b, min_left)) return; start_bandwidth_timer(&cfs_b->slack_timer, ns_to_ktime(cfs_bandwidth_slack_period)); } /* we know any runtime found here is valid as update_curr() precedes return */ static void __return_cfs_rq_runtime(struct cfs_rq *cfs_rq) { struct cfs_bandwidth *cfs_b = tg_cfs_bandwidth(cfs_rq->tg); s64 slack_runtime = cfs_rq->runtime_remaining - min_cfs_rq_runtime; if (slack_runtime <= 0) return; raw_spin_lock(&cfs_b->lock); if (cfs_b->quota != RUNTIME_INF && cfs_rq->runtime_expires == cfs_b->runtime_expires) { cfs_b->runtime += slack_runtime; /* we are under rq->lock, defer unthrottling using a timer */ if (cfs_b->runtime > sched_cfs_bandwidth_slice() && !list_empty(&cfs_b->throttled_cfs_rq)) start_cfs_slack_bandwidth(cfs_b); } raw_spin_unlock(&cfs_b->lock); /* even if it's not valid for return we don't want to try again */ cfs_rq->runtime_remaining -= slack_runtime; } static __always_inline void return_cfs_rq_runtime(struct cfs_rq *cfs_rq) { |
56f570e51 sched: Use jump l... |
1798 1799 |
if (!cfs_bandwidth_used()) return; |
fccfdc6f0 sched: Fix buglet... |
1800 |
if (!cfs_rq->runtime_enabled || cfs_rq->nr_running) |
d8b4986d3 sched: Return unu... |
1801 1802 1803 1804 1805 1806 1807 1808 1809 1810 1811 1812 1813 1814 1815 1816 1817 1818 1819 1820 1821 1822 1823 1824 1825 1826 1827 1828 1829 1830 1831 1832 1833 1834 1835 1836 |
return; __return_cfs_rq_runtime(cfs_rq); } /* * This is done with a timer (instead of inline with bandwidth return) since * it's necessary to juggle rq->locks to unthrottle their respective cfs_rqs. */ static void do_sched_cfs_slack_timer(struct cfs_bandwidth *cfs_b) { u64 runtime = 0, slice = sched_cfs_bandwidth_slice(); u64 expires; /* confirm we're still not at a refresh boundary */ if (runtime_refresh_within(cfs_b, min_bandwidth_expiration)) return; raw_spin_lock(&cfs_b->lock); if (cfs_b->quota != RUNTIME_INF && cfs_b->runtime > slice) { runtime = cfs_b->runtime; cfs_b->runtime = 0; } expires = cfs_b->runtime_expires; raw_spin_unlock(&cfs_b->lock); if (!runtime) return; runtime = distribute_cfs_runtime(cfs_b, runtime, expires); raw_spin_lock(&cfs_b->lock); if (expires == cfs_b->runtime_expires) cfs_b->runtime = runtime; raw_spin_unlock(&cfs_b->lock); } |
d3d9dc330 sched: Throttle e... |
1837 1838 1839 1840 1841 1842 1843 |
/* * When a group wakes up we want to make sure that its quota is not already * expired/exceeded, otherwise it may be allowed to steal additional ticks of * runtime as update_curr() throttling can not not trigger until it's on-rq. */ static void check_enqueue_throttle(struct cfs_rq *cfs_rq) { |
56f570e51 sched: Use jump l... |
1844 1845 |
if (!cfs_bandwidth_used()) return; |
d3d9dc330 sched: Throttle e... |
1846 1847 1848 1849 1850 1851 1852 1853 1854 1855 1856 1857 1858 1859 1860 1861 1862 |
/* an active group must be handled by the update_curr()->put() path */ if (!cfs_rq->runtime_enabled || cfs_rq->curr) return; /* ensure the group is not already throttled */ if (cfs_rq_throttled(cfs_rq)) return; /* update runtime allocation */ account_cfs_rq_runtime(cfs_rq, 0); if (cfs_rq->runtime_remaining <= 0) throttle_cfs_rq(cfs_rq); } /* conditionally throttle active cfs_rq's from put_prev_entity() */ static void check_cfs_rq_runtime(struct cfs_rq *cfs_rq) { |
56f570e51 sched: Use jump l... |
1863 1864 |
if (!cfs_bandwidth_used()) return; |
d3d9dc330 sched: Throttle e... |
1865 1866 1867 1868 1869 1870 1871 1872 1873 1874 1875 1876 |
if (likely(!cfs_rq->runtime_enabled || cfs_rq->runtime_remaining > 0)) return; /* * it's possible for a throttled entity to be forced into a running * state (e.g. set_curr_task), in this case we're finished. */ if (cfs_rq_throttled(cfs_rq)) return; throttle_cfs_rq(cfs_rq); } |
029632fbb sched: Make separ... |
1877 1878 1879 1880 1881 1882 1883 1884 1885 1886 1887 1888 1889 1890 1891 1892 1893 1894 1895 1896 1897 1898 1899 1900 1901 1902 1903 1904 1905 1906 1907 1908 1909 1910 1911 1912 1913 1914 1915 1916 1917 1918 1919 1920 1921 1922 1923 1924 1925 1926 1927 1928 1929 1930 1931 1932 1933 1934 1935 1936 1937 1938 1939 1940 1941 1942 1943 1944 1945 1946 1947 1948 1949 1950 1951 1952 1953 1954 1955 1956 1957 1958 1959 1960 1961 1962 1963 1964 1965 1966 1967 1968 1969 1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982 |
static inline u64 default_cfs_period(void); static int do_sched_cfs_period_timer(struct cfs_bandwidth *cfs_b, int overrun); static void do_sched_cfs_slack_timer(struct cfs_bandwidth *cfs_b); static enum hrtimer_restart sched_cfs_slack_timer(struct hrtimer *timer) { struct cfs_bandwidth *cfs_b = container_of(timer, struct cfs_bandwidth, slack_timer); do_sched_cfs_slack_timer(cfs_b); return HRTIMER_NORESTART; } static enum hrtimer_restart sched_cfs_period_timer(struct hrtimer *timer) { struct cfs_bandwidth *cfs_b = container_of(timer, struct cfs_bandwidth, period_timer); ktime_t now; int overrun; int idle = 0; for (;;) { now = hrtimer_cb_get_time(timer); overrun = hrtimer_forward(timer, now, cfs_b->period); if (!overrun) break; idle = do_sched_cfs_period_timer(cfs_b, overrun); } return idle ? HRTIMER_NORESTART : HRTIMER_RESTART; } void init_cfs_bandwidth(struct cfs_bandwidth *cfs_b) { raw_spin_lock_init(&cfs_b->lock); cfs_b->runtime = 0; cfs_b->quota = RUNTIME_INF; cfs_b->period = ns_to_ktime(default_cfs_period()); INIT_LIST_HEAD(&cfs_b->throttled_cfs_rq); hrtimer_init(&cfs_b->period_timer, CLOCK_MONOTONIC, HRTIMER_MODE_REL); cfs_b->period_timer.function = sched_cfs_period_timer; hrtimer_init(&cfs_b->slack_timer, CLOCK_MONOTONIC, HRTIMER_MODE_REL); cfs_b->slack_timer.function = sched_cfs_slack_timer; } static void init_cfs_rq_runtime(struct cfs_rq *cfs_rq) { cfs_rq->runtime_enabled = 0; INIT_LIST_HEAD(&cfs_rq->throttled_list); } /* requires cfs_b->lock, may release to reprogram timer */ void __start_cfs_bandwidth(struct cfs_bandwidth *cfs_b) { /* * The timer may be active because we're trying to set a new bandwidth * period or because we're racing with the tear-down path * (timer_active==0 becomes visible before the hrtimer call-back * terminates). In either case we ensure that it's re-programmed */ while (unlikely(hrtimer_active(&cfs_b->period_timer))) { raw_spin_unlock(&cfs_b->lock); /* ensure cfs_b->lock is available while we wait */ hrtimer_cancel(&cfs_b->period_timer); raw_spin_lock(&cfs_b->lock); /* if someone else restarted the timer then we're done */ if (cfs_b->timer_active) return; } cfs_b->timer_active = 1; start_bandwidth_timer(&cfs_b->period_timer, cfs_b->period); } static void destroy_cfs_bandwidth(struct cfs_bandwidth *cfs_b) { hrtimer_cancel(&cfs_b->period_timer); hrtimer_cancel(&cfs_b->slack_timer); } void unthrottle_offline_cfs_rqs(struct rq *rq) { struct cfs_rq *cfs_rq; for_each_leaf_cfs_rq(rq, cfs_rq) { struct cfs_bandwidth *cfs_b = tg_cfs_bandwidth(cfs_rq->tg); if (!cfs_rq->runtime_enabled) continue; /* * clock_task is not advancing so we just need to make sure * there's some valid quota amount */ cfs_rq->runtime_remaining = cfs_b->quota; if (cfs_rq_throttled(cfs_rq)) unthrottle_cfs_rq(cfs_rq); } } #else /* CONFIG_CFS_BANDWIDTH */ |
ec12cb7f3 sched: Accumulate... |
1983 1984 |
static void account_cfs_rq_runtime(struct cfs_rq *cfs_rq, unsigned long delta_exec) {} |
d3d9dc330 sched: Throttle e... |
1985 1986 |
static void check_cfs_rq_runtime(struct cfs_rq *cfs_rq) {} static void check_enqueue_throttle(struct cfs_rq *cfs_rq) {} |
d8b4986d3 sched: Return unu... |
1987 |
static void return_cfs_rq_runtime(struct cfs_rq *cfs_rq) {} |
85dac906b sched: Add suppor... |
1988 1989 1990 1991 1992 |
static inline int cfs_rq_throttled(struct cfs_rq *cfs_rq) { return 0; } |
64660c864 sched: Prevent in... |
1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 |
static inline int throttled_hierarchy(struct cfs_rq *cfs_rq) { return 0; } static inline int throttled_lb_pair(struct task_group *tg, int src_cpu, int dest_cpu) { return 0; } |
029632fbb sched: Make separ... |
2004 2005 2006 2007 2008 |
void init_cfs_bandwidth(struct cfs_bandwidth *cfs_b) {} #ifdef CONFIG_FAIR_GROUP_SCHED static void init_cfs_rq_runtime(struct cfs_rq *cfs_rq) {} |
ab84d31e1 sched: Introduce ... |
2009 |
#endif |
029632fbb sched: Make separ... |
2010 2011 2012 2013 2014 2015 2016 2017 |
static inline struct cfs_bandwidth *tg_cfs_bandwidth(struct task_group *tg) { return NULL; } static inline void destroy_cfs_bandwidth(struct cfs_bandwidth *cfs_b) {} void unthrottle_offline_cfs_rqs(struct rq *rq) {} #endif /* CONFIG_CFS_BANDWIDTH */ |
bf0f6f24a sched: cfs core, ... |
2018 2019 2020 |
/************************************************** * CFS operations on tasks: */ |
8f4d37ec0 sched: high-res p... |
2021 2022 2023 |
#ifdef CONFIG_SCHED_HRTICK static void hrtick_start_fair(struct rq *rq, struct task_struct *p) { |
8f4d37ec0 sched: high-res p... |
2024 2025 2026 2027 |
struct sched_entity *se = &p->se; struct cfs_rq *cfs_rq = cfs_rq_of(se); WARN_ON(task_rq(p) != rq); |
b39e66eaf sched: Save some ... |
2028 |
if (cfs_rq->nr_running > 1) { |
8f4d37ec0 sched: high-res p... |
2029 2030 2031 2032 2033 2034 2035 2036 2037 2038 2039 2040 2041 2042 |
u64 slice = sched_slice(cfs_rq, se); u64 ran = se->sum_exec_runtime - se->prev_sum_exec_runtime; s64 delta = slice - ran; if (delta < 0) { if (rq->curr == p) resched_task(p); return; } /* * Don't schedule slices shorter than 10000ns, that just * doesn't make sense. Rely on vruntime for fairness. */ |
31656519e sched, x86: clean... |
2043 |
if (rq->curr != p) |
157124c11 sched: fix warnin... |
2044 |
delta = max_t(s64, 10000LL, delta); |
8f4d37ec0 sched: high-res p... |
2045 |
|
31656519e sched, x86: clean... |
2046 |
hrtick_start(rq, delta); |
8f4d37ec0 sched: high-res p... |
2047 2048 |
} } |
a4c2f00f5 sched: fair sched... |
2049 2050 2051 2052 2053 2054 2055 2056 2057 |
/* * called from enqueue/dequeue and updates the hrtick when the * current task is from our class and nr_running is low enough * to matter. */ static void hrtick_update(struct rq *rq) { struct task_struct *curr = rq->curr; |
b39e66eaf sched: Save some ... |
2058 |
if (!hrtick_enabled(rq) || curr->sched_class != &fair_sched_class) |
a4c2f00f5 sched: fair sched... |
2059 2060 2061 2062 2063 |
return; if (cfs_rq_of(&curr->se)->nr_running < sched_nr_latency) hrtick_start_fair(rq, curr); } |
55e12e5e7 sched: make sched... |
2064 |
#else /* !CONFIG_SCHED_HRTICK */ |
8f4d37ec0 sched: high-res p... |
2065 2066 2067 2068 |
static inline void hrtick_start_fair(struct rq *rq, struct task_struct *p) { } |
a4c2f00f5 sched: fair sched... |
2069 2070 2071 2072 |
static inline void hrtick_update(struct rq *rq) { } |
8f4d37ec0 sched: high-res p... |
2073 |
#endif |
bf0f6f24a sched: cfs core, ... |
2074 2075 2076 2077 2078 |
/* * The enqueue_task method is called before nr_running is * increased. Here we update the fair scheduling stats and * then put the task into the rbtree: */ |
ea87bb785 sched: Extend enq... |
2079 |
static void |
371fd7e7a sched: Add enqueu... |
2080 |
enqueue_task_fair(struct rq *rq, struct task_struct *p, int flags) |
bf0f6f24a sched: cfs core, ... |
2081 2082 |
{ struct cfs_rq *cfs_rq; |
62fb18513 sched: revert loa... |
2083 |
struct sched_entity *se = &p->se; |
bf0f6f24a sched: cfs core, ... |
2084 2085 |
for_each_sched_entity(se) { |
62fb18513 sched: revert loa... |
2086 |
if (se->on_rq) |
bf0f6f24a sched: cfs core, ... |
2087 2088 |
break; cfs_rq = cfs_rq_of(se); |
88ec22d3e sched: Remove the... |
2089 |
enqueue_entity(cfs_rq, se, flags); |
85dac906b sched: Add suppor... |
2090 2091 2092 2093 2094 2095 2096 2097 2098 |
/* * end evaluation on encountering a throttled cfs_rq * * note: in the case of encountering a throttled cfs_rq we will * post the final h_nr_running increment below. */ if (cfs_rq_throttled(cfs_rq)) break; |
953bfcd10 sched: Implement ... |
2099 |
cfs_rq->h_nr_running++; |
85dac906b sched: Add suppor... |
2100 |
|
88ec22d3e sched: Remove the... |
2101 |
flags = ENQUEUE_WAKEUP; |
bf0f6f24a sched: cfs core, ... |
2102 |
} |
8f4d37ec0 sched: high-res p... |
2103 |
|
2069dd75c sched: Rewrite tg... |
2104 |
for_each_sched_entity(se) { |
0f3171438 sched: Cleanup du... |
2105 |
cfs_rq = cfs_rq_of(se); |
953bfcd10 sched: Implement ... |
2106 |
cfs_rq->h_nr_running++; |
2069dd75c sched: Rewrite tg... |
2107 |
|
85dac906b sched: Add suppor... |
2108 2109 |
if (cfs_rq_throttled(cfs_rq)) break; |
d6b559182 sched: Allow upda... |
2110 |
update_cfs_load(cfs_rq, 0); |
6d5ab2932 sched: Simplify u... |
2111 |
update_cfs_shares(cfs_rq); |
2069dd75c sched: Rewrite tg... |
2112 |
} |
85dac906b sched: Add suppor... |
2113 2114 |
if (!se) inc_nr_running(rq); |
a4c2f00f5 sched: fair sched... |
2115 |
hrtick_update(rq); |
bf0f6f24a sched: cfs core, ... |
2116 |
} |
2f36825b1 sched: Next buddy... |
2117 |
static void set_next_buddy(struct sched_entity *se); |
bf0f6f24a sched: cfs core, ... |
2118 2119 2120 2121 2122 |
/* * The dequeue_task method is called before nr_running is * decreased. We remove the task from the rbtree and * update the fair scheduling stats: */ |
371fd7e7a sched: Add enqueu... |
2123 |
static void dequeue_task_fair(struct rq *rq, struct task_struct *p, int flags) |
bf0f6f24a sched: cfs core, ... |
2124 2125 |
{ struct cfs_rq *cfs_rq; |
62fb18513 sched: revert loa... |
2126 |
struct sched_entity *se = &p->se; |
2f36825b1 sched: Next buddy... |
2127 |
int task_sleep = flags & DEQUEUE_SLEEP; |
bf0f6f24a sched: cfs core, ... |
2128 2129 2130 |
for_each_sched_entity(se) { cfs_rq = cfs_rq_of(se); |
371fd7e7a sched: Add enqueu... |
2131 |
dequeue_entity(cfs_rq, se, flags); |
85dac906b sched: Add suppor... |
2132 2133 2134 2135 2136 2137 2138 2139 2140 |
/* * end evaluation on encountering a throttled cfs_rq * * note: in the case of encountering a throttled cfs_rq we will * post the final h_nr_running decrement below. */ if (cfs_rq_throttled(cfs_rq)) break; |
953bfcd10 sched: Implement ... |
2141 |
cfs_rq->h_nr_running--; |
2069dd75c sched: Rewrite tg... |
2142 |
|
bf0f6f24a sched: cfs core, ... |
2143 |
/* Don't dequeue parent if it has other entities besides us */ |
2f36825b1 sched: Next buddy... |
2144 2145 2146 2147 2148 2149 2150 |
if (cfs_rq->load.weight) { /* * Bias pick_next to pick a task from this cfs_rq, as * p is sleeping when it is within its sched_slice. */ if (task_sleep && parent_entity(se)) set_next_buddy(parent_entity(se)); |
9598c82dc sched: Don't upda... |
2151 2152 2153 |
/* avoid re-evaluating load for this entity */ se = parent_entity(se); |
bf0f6f24a sched: cfs core, ... |
2154 |
break; |
2f36825b1 sched: Next buddy... |
2155 |
} |
371fd7e7a sched: Add enqueu... |
2156 |
flags |= DEQUEUE_SLEEP; |
bf0f6f24a sched: cfs core, ... |
2157 |
} |
8f4d37ec0 sched: high-res p... |
2158 |
|
2069dd75c sched: Rewrite tg... |
2159 |
for_each_sched_entity(se) { |
0f3171438 sched: Cleanup du... |
2160 |
cfs_rq = cfs_rq_of(se); |
953bfcd10 sched: Implement ... |
2161 |
cfs_rq->h_nr_running--; |
2069dd75c sched: Rewrite tg... |
2162 |
|
85dac906b sched: Add suppor... |
2163 2164 |
if (cfs_rq_throttled(cfs_rq)) break; |
d6b559182 sched: Allow upda... |
2165 |
update_cfs_load(cfs_rq, 0); |
6d5ab2932 sched: Simplify u... |
2166 |
update_cfs_shares(cfs_rq); |
2069dd75c sched: Rewrite tg... |
2167 |
} |
85dac906b sched: Add suppor... |
2168 2169 |
if (!se) dec_nr_running(rq); |
a4c2f00f5 sched: fair sched... |
2170 |
hrtick_update(rq); |
bf0f6f24a sched: cfs core, ... |
2171 |
} |
e7693a362 sched: de-SCHED_O... |
2172 |
#ifdef CONFIG_SMP |
029632fbb sched: Make separ... |
2173 2174 2175 2176 2177 2178 2179 2180 2181 2182 2183 2184 2185 2186 2187 2188 2189 2190 2191 2192 2193 2194 2195 2196 2197 2198 2199 2200 2201 2202 2203 2204 2205 2206 2207 2208 2209 2210 2211 2212 2213 2214 2215 2216 2217 2218 2219 2220 2221 2222 2223 2224 2225 2226 |
/* Used instead of source_load when we know the type == 0 */ static unsigned long weighted_cpuload(const int cpu) { return cpu_rq(cpu)->load.weight; } /* * Return a low guess at the load of a migration-source cpu weighted * according to the scheduling class and "nice" value. * * We want to under-estimate the load of migration sources, to * balance conservatively. */ static unsigned long source_load(int cpu, int type) { struct rq *rq = cpu_rq(cpu); unsigned long total = weighted_cpuload(cpu); if (type == 0 || !sched_feat(LB_BIAS)) return total; return min(rq->cpu_load[type-1], total); } /* * Return a high guess at the load of a migration-target cpu weighted * according to the scheduling class and "nice" value. */ static unsigned long target_load(int cpu, int type) { struct rq *rq = cpu_rq(cpu); unsigned long total = weighted_cpuload(cpu); if (type == 0 || !sched_feat(LB_BIAS)) return total; return max(rq->cpu_load[type-1], total); } static unsigned long power_of(int cpu) { return cpu_rq(cpu)->cpu_power; } static unsigned long cpu_avg_load_per_task(int cpu) { struct rq *rq = cpu_rq(cpu); unsigned long nr_running = ACCESS_ONCE(rq->nr_running); if (nr_running) return rq->load.weight / nr_running; return 0; } |
098fb9db2 sched: clean up w... |
2227 |
|
74f8e4b23 sched: Remove rq ... |
2228 |
static void task_waking_fair(struct task_struct *p) |
88ec22d3e sched: Remove the... |
2229 2230 2231 |
{ struct sched_entity *se = &p->se; struct cfs_rq *cfs_rq = cfs_rq_of(se); |
3fe1698b7 sched: Deal with ... |
2232 2233 2234 2235 |
u64 min_vruntime; #ifndef CONFIG_64BIT u64 min_vruntime_copy; |
88ec22d3e sched: Remove the... |
2236 |
|
3fe1698b7 sched: Deal with ... |
2237 2238 2239 2240 2241 2242 2243 2244 |
do { min_vruntime_copy = cfs_rq->min_vruntime_copy; smp_rmb(); min_vruntime = cfs_rq->min_vruntime; } while (min_vruntime != min_vruntime_copy); #else min_vruntime = cfs_rq->min_vruntime; #endif |
88ec22d3e sched: Remove the... |
2245 |
|
3fe1698b7 sched: Deal with ... |
2246 |
se->vruntime -= min_vruntime; |
88ec22d3e sched: Remove the... |
2247 |
} |
bb3469ac9 sched: hierarchic... |
2248 |
#ifdef CONFIG_FAIR_GROUP_SCHED |
f5bfb7d9f sched: bias effec... |
2249 2250 2251 2252 2253 2254 |
/* * effective_load() calculates the load change as seen from the root_task_group * * Adding load to a group doesn't make a group heavier, but can cause movement * of group shares between cpus. Assuming the shares were perfectly aligned one * can calculate the shift in shares. |
cf5f0acf3 sched: Add a comm... |
2255 2256 2257 2258 2259 2260 2261 2262 2263 2264 2265 2266 2267 2268 2269 2270 2271 2272 2273 2274 2275 2276 2277 2278 2279 2280 2281 2282 2283 2284 2285 2286 2287 2288 2289 2290 2291 2292 2293 2294 2295 2296 2297 |
* * Calculate the effective load difference if @wl is added (subtracted) to @tg * on this @cpu and results in a total addition (subtraction) of @wg to the * total group weight. * * Given a runqueue weight distribution (rw_i) we can compute a shares * distribution (s_i) using: * * s_i = rw_i / \Sum rw_j (1) * * Suppose we have 4 CPUs and our @tg is a direct child of the root group and * has 7 equal weight tasks, distributed as below (rw_i), with the resulting * shares distribution (s_i): * * rw_i = { 2, 4, 1, 0 } * s_i = { 2/7, 4/7, 1/7, 0 } * * As per wake_affine() we're interested in the load of two CPUs (the CPU the * task used to run on and the CPU the waker is running on), we need to * compute the effect of waking a task on either CPU and, in case of a sync * wakeup, compute the effect of the current task going to sleep. * * So for a change of @wl to the local @cpu with an overall group weight change * of @wl we can compute the new shares distribution (s'_i) using: * * s'_i = (rw_i + @wl) / (@wg + \Sum rw_j) (2) * * Suppose we're interested in CPUs 0 and 1, and want to compute the load * differences in waking a task to CPU 0. The additional task changes the * weight and shares distributions like: * * rw'_i = { 3, 4, 1, 0 } * s'_i = { 3/8, 4/8, 1/8, 0 } * * We can then compute the difference in effective weight by using: * * dw_i = S * (s'_i - s_i) (3) * * Where 'S' is the group weight as seen by its parent. * * Therefore the effective change in loads on CPU 0 would be 5/56 (3/8 - 2/7) * times the weight of the group. The effect on CPU 1 would be -4/56 (4/8 - * 4/7) times the weight of the group. |
f5bfb7d9f sched: bias effec... |
2298 |
*/ |
2069dd75c sched: Rewrite tg... |
2299 |
static long effective_load(struct task_group *tg, int cpu, long wl, long wg) |
bb3469ac9 sched: hierarchic... |
2300 |
{ |
4be9daaa1 sched: fix task_h... |
2301 |
struct sched_entity *se = tg->se[cpu]; |
f1d239f73 sched: incrementa... |
2302 |
|
cf5f0acf3 sched: Add a comm... |
2303 |
if (!tg->parent) /* the trivial, non-cgroup case */ |
f1d239f73 sched: incrementa... |
2304 |
return wl; |
4be9daaa1 sched: fix task_h... |
2305 |
for_each_sched_entity(se) { |
cf5f0acf3 sched: Add a comm... |
2306 |
long w, W; |
4be9daaa1 sched: fix task_h... |
2307 |
|
977dda7c9 sched: Update eff... |
2308 |
tg = se->my_q->tg; |
bb3469ac9 sched: hierarchic... |
2309 |
|
cf5f0acf3 sched: Add a comm... |
2310 2311 2312 2313 |
/* * W = @wg + \Sum rw_j */ W = wg + calc_tg_weight(tg, se->my_q); |
4be9daaa1 sched: fix task_h... |
2314 |
|
cf5f0acf3 sched: Add a comm... |
2315 2316 2317 2318 |
/* * w = rw_i + @wl */ w = se->my_q->load.weight + wl; |
940959e93 sched: fixlet for... |
2319 |
|
cf5f0acf3 sched: Add a comm... |
2320 2321 2322 2323 2324 |
/* * wl = S * s'_i; see (2) */ if (W > 0 && w < W) wl = (w * tg->shares) / W; |
977dda7c9 sched: Update eff... |
2325 2326 |
else wl = tg->shares; |
940959e93 sched: fixlet for... |
2327 |
|
cf5f0acf3 sched: Add a comm... |
2328 2329 2330 2331 2332 |
/* * Per the above, wl is the new se->load.weight value; since * those are clipped to [MIN_SHARES, ...) do so now. See * calc_cfs_shares(). */ |
977dda7c9 sched: Update eff... |
2333 2334 |
if (wl < MIN_SHARES) wl = MIN_SHARES; |
cf5f0acf3 sched: Add a comm... |
2335 2336 2337 2338 |
/* * wl = dw_i = S * (s'_i - s_i); see (3) */ |
977dda7c9 sched: Update eff... |
2339 |
wl -= se->load.weight; |
cf5f0acf3 sched: Add a comm... |
2340 2341 2342 2343 2344 2345 2346 2347 |
/* * Recursively apply this logic to all parent groups to compute * the final effective load change on the root group. Since * only the @tg group gets extra weight, all parent groups can * only redistribute existing shares. @wl is the shift in shares * resulting from this level per the above. */ |
4be9daaa1 sched: fix task_h... |
2348 |
wg = 0; |
4be9daaa1 sched: fix task_h... |
2349 |
} |
bb3469ac9 sched: hierarchic... |
2350 |
|
4be9daaa1 sched: fix task_h... |
2351 |
return wl; |
bb3469ac9 sched: hierarchic... |
2352 2353 |
} #else |
4be9daaa1 sched: fix task_h... |
2354 |
|
83378269a sched: correct wa... |
2355 2356 |
static inline unsigned long effective_load(struct task_group *tg, int cpu, unsigned long wl, unsigned long wg) |
4be9daaa1 sched: fix task_h... |
2357 |
{ |
83378269a sched: correct wa... |
2358 |
return wl; |
bb3469ac9 sched: hierarchic... |
2359 |
} |
4be9daaa1 sched: fix task_h... |
2360 |
|
bb3469ac9 sched: hierarchic... |
2361 |
#endif |
c88d59108 sched: Merge sele... |
2362 |
static int wake_affine(struct sched_domain *sd, struct task_struct *p, int sync) |
098fb9db2 sched: clean up w... |
2363 |
{ |
e37b6a7b2 sched: Fix sign u... |
2364 |
s64 this_load, load; |
c88d59108 sched: Merge sele... |
2365 |
int idx, this_cpu, prev_cpu; |
098fb9db2 sched: clean up w... |
2366 |
unsigned long tl_per_task; |
c88d59108 sched: Merge sele... |
2367 |
struct task_group *tg; |
83378269a sched: correct wa... |
2368 |
unsigned long weight; |
b3137bc8e sched: stop wake_... |
2369 |
int balanced; |
098fb9db2 sched: clean up w... |
2370 |
|
c88d59108 sched: Merge sele... |
2371 2372 2373 2374 2375 |
idx = sd->wake_idx; this_cpu = smp_processor_id(); prev_cpu = task_cpu(p); load = source_load(prev_cpu, idx); this_load = target_load(this_cpu, idx); |
098fb9db2 sched: clean up w... |
2376 2377 |
/* |
b3137bc8e sched: stop wake_... |
2378 2379 2380 2381 |
* If sync wakeup then subtract the (maximum possible) * effect of the currently running task from the load * of the current CPU: */ |
83378269a sched: correct wa... |
2382 2383 2384 |
if (sync) { tg = task_group(current); weight = current->se.load.weight; |
c88d59108 sched: Merge sele... |
2385 |
this_load += effective_load(tg, this_cpu, -weight, -weight); |
83378269a sched: correct wa... |
2386 2387 |
load += effective_load(tg, prev_cpu, 0, -weight); } |
b3137bc8e sched: stop wake_... |
2388 |
|
83378269a sched: correct wa... |
2389 2390 |
tg = task_group(p); weight = p->se.load.weight; |
b3137bc8e sched: stop wake_... |
2391 |
|
71a29aa7b sched: Deal with ... |
2392 2393 |
/* * In low-load situations, where prev_cpu is idle and this_cpu is idle |
c88d59108 sched: Merge sele... |
2394 2395 2396 |
* due to the sync cause above having dropped this_load to 0, we'll * always have an imbalance, but there's really nothing you can do * about that, so that's good too. |
71a29aa7b sched: Deal with ... |
2397 2398 2399 2400 |
* * Otherwise check if either cpus are near enough in load to allow this * task to be woken on this_cpu. */ |
e37b6a7b2 sched: Fix sign u... |
2401 2402 |
if (this_load > 0) { s64 this_eff_load, prev_eff_load; |
e51fd5e22 sched: Fix wake_a... |
2403 2404 2405 2406 2407 2408 2409 2410 2411 2412 2413 2414 2415 |
this_eff_load = 100; this_eff_load *= power_of(prev_cpu); this_eff_load *= this_load + effective_load(tg, this_cpu, weight, weight); prev_eff_load = 100 + (sd->imbalance_pct - 100) / 2; prev_eff_load *= power_of(this_cpu); prev_eff_load *= load + effective_load(tg, prev_cpu, 0, weight); balanced = this_eff_load <= prev_eff_load; } else balanced = true; |
b3137bc8e sched: stop wake_... |
2416 2417 |
/* |
4ae7d5cef sched: improve af... |
2418 2419 2420 |
* If the currently running task will sleep within * a reasonable amount of time then attract this newly * woken task: |
098fb9db2 sched: clean up w... |
2421 |
*/ |
2fb7635c4 sched: sync wakeu... |
2422 2423 |
if (sync && balanced) return 1; |
098fb9db2 sched: clean up w... |
2424 |
|
41acab885 sched: Implement ... |
2425 |
schedstat_inc(p, se.statistics.nr_wakeups_affine_attempts); |
098fb9db2 sched: clean up w... |
2426 |
tl_per_task = cpu_avg_load_per_task(this_cpu); |
c88d59108 sched: Merge sele... |
2427 2428 2429 |
if (balanced || (this_load <= load && this_load + target_load(prev_cpu, idx) <= tl_per_task)) { |
098fb9db2 sched: clean up w... |
2430 2431 2432 2433 2434 |
/* * This domain has SD_WAKE_AFFINE and * p is cache cold in this domain, and * there is no bad imbalance. */ |
c88d59108 sched: Merge sele... |
2435 |
schedstat_inc(sd, ttwu_move_affine); |
41acab885 sched: Implement ... |
2436 |
schedstat_inc(p, se.statistics.nr_wakeups_affine); |
098fb9db2 sched: clean up w... |
2437 2438 2439 2440 2441 |
return 1; } return 0; } |
aaee1203c sched: Move sched... |
2442 2443 2444 2445 2446 |
/* * find_idlest_group finds and returns the least busy CPU group within the * domain. */ static struct sched_group * |
78e7ed53c sched: Tweak wake... |
2447 |
find_idlest_group(struct sched_domain *sd, struct task_struct *p, |
5158f4e44 sched: Clean up t... |
2448 |
int this_cpu, int load_idx) |
e7693a362 sched: de-SCHED_O... |
2449 |
{ |
b3bd3de66 gcc-4.6: kernel/*... |
2450 |
struct sched_group *idlest = NULL, *group = sd->groups; |
aaee1203c sched: Move sched... |
2451 |
unsigned long min_load = ULONG_MAX, this_load = 0; |
aaee1203c sched: Move sched... |
2452 |
int imbalance = 100 + (sd->imbalance_pct-100)/2; |
e7693a362 sched: de-SCHED_O... |
2453 |
|
aaee1203c sched: Move sched... |
2454 2455 2456 2457 |
do { unsigned long load, avg_load; int local_group; int i; |
e7693a362 sched: de-SCHED_O... |
2458 |
|
aaee1203c sched: Move sched... |
2459 2460 |
/* Skip over this group if it has no CPUs allowed */ if (!cpumask_intersects(sched_group_cpus(group), |
fa17b507f sched: Wrap sched... |
2461 |
tsk_cpus_allowed(p))) |
aaee1203c sched: Move sched... |
2462 2463 2464 2465 2466 2467 2468 2469 2470 2471 2472 2473 2474 2475 2476 2477 2478 2479 2480 |
continue; local_group = cpumask_test_cpu(this_cpu, sched_group_cpus(group)); /* Tally up the load of all CPUs in the group */ avg_load = 0; for_each_cpu(i, sched_group_cpus(group)) { /* Bias balancing toward cpus of our domain */ if (local_group) load = source_load(i, load_idx); else load = target_load(i, load_idx); avg_load += load; } /* Adjust by relative CPU power of the group */ |
9c3f75cbd sched: Break out ... |
2481 |
avg_load = (avg_load * SCHED_POWER_SCALE) / group->sgp->power; |
aaee1203c sched: Move sched... |
2482 2483 2484 |
if (local_group) { this_load = avg_load; |
aaee1203c sched: Move sched... |
2485 2486 2487 2488 2489 2490 2491 2492 2493 2494 2495 2496 2497 2498 2499 2500 2501 2502 2503 2504 2505 2506 |
} else if (avg_load < min_load) { min_load = avg_load; idlest = group; } } while (group = group->next, group != sd->groups); if (!idlest || 100*this_load < imbalance*min_load) return NULL; return idlest; } /* * find_idlest_cpu - find the idlest cpu among the cpus in group. */ static int find_idlest_cpu(struct sched_group *group, struct task_struct *p, int this_cpu) { unsigned long load, min_load = ULONG_MAX; int idlest = -1; int i; /* Traverse only the allowed CPUs */ |
fa17b507f sched: Wrap sched... |
2507 |
for_each_cpu_and(i, sched_group_cpus(group), tsk_cpus_allowed(p)) { |
aaee1203c sched: Move sched... |
2508 2509 2510 2511 2512 |
load = weighted_cpuload(i); if (load < min_load || (load == min_load && i == this_cpu)) { min_load = load; idlest = i; |
e7693a362 sched: de-SCHED_O... |
2513 2514 |
} } |
aaee1203c sched: Move sched... |
2515 2516 |
return idlest; } |
e7693a362 sched: de-SCHED_O... |
2517 |
|
aaee1203c sched: Move sched... |
2518 |
/* |
a50bde513 sched: Cleanup se... |
2519 2520 |
* Try and locate an idle CPU in the sched_domain. */ |
99bd5e2f2 sched: Fix select... |
2521 |
static int select_idle_sibling(struct task_struct *p, int target) |
a50bde513 sched: Cleanup se... |
2522 2523 2524 |
{ int cpu = smp_processor_id(); int prev_cpu = task_cpu(p); |
99bd5e2f2 sched: Fix select... |
2525 |
struct sched_domain *sd; |
4dcfe1025 sched: Avoid SMT ... |
2526 |
struct sched_group *sg; |
77e81365e sched: Clean up d... |
2527 |
int i; |
a50bde513 sched: Cleanup se... |
2528 2529 |
/* |
99bd5e2f2 sched: Fix select... |
2530 2531 |
* If the task is going to be woken-up on this cpu and if it is * already idle, then it is the right target. |
a50bde513 sched: Cleanup se... |
2532 |
*/ |
99bd5e2f2 sched: Fix select... |
2533 2534 2535 2536 2537 2538 2539 2540 |
if (target == cpu && idle_cpu(cpu)) return cpu; /* * If the task is going to be woken-up on the cpu where it previously * ran and if it is currently idle, then it the right target. */ if (target == prev_cpu && idle_cpu(prev_cpu)) |
fe3bcfe1f sched: More gener... |
2541 |
return prev_cpu; |
a50bde513 sched: Cleanup se... |
2542 2543 |
/* |
99bd5e2f2 sched: Fix select... |
2544 |
* Otherwise, iterate the domains and find an elegible idle cpu. |
a50bde513 sched: Cleanup se... |
2545 |
*/ |
dce840a08 sched: Dynamicall... |
2546 |
rcu_read_lock(); |
99bd5e2f2 sched: Fix select... |
2547 |
|
518cd6234 sched: Only queue... |
2548 |
sd = rcu_dereference(per_cpu(sd_llc, target)); |
77e81365e sched: Clean up d... |
2549 |
for_each_lower_domain(sd) { |
4dcfe1025 sched: Avoid SMT ... |
2550 2551 2552 2553 2554 2555 2556 2557 2558 2559 2560 2561 2562 2563 2564 2565 2566 |
sg = sd->groups; do { if (!cpumask_intersects(sched_group_cpus(sg), tsk_cpus_allowed(p))) goto next; for_each_cpu(i, sched_group_cpus(sg)) { if (!idle_cpu(i)) goto next; } target = cpumask_first_and(sched_group_cpus(sg), tsk_cpus_allowed(p)); goto done; next: sg = sg->next; } while (sg != sd->groups); |
a50bde513 sched: Cleanup se... |
2567 |
} |
4dcfe1025 sched: Avoid SMT ... |
2568 |
done: |
dce840a08 sched: Dynamicall... |
2569 |
rcu_read_unlock(); |
a50bde513 sched: Cleanup se... |
2570 2571 2572 2573 2574 |
return target; } /* |
aaee1203c sched: Move sched... |
2575 2576 2577 2578 2579 2580 2581 2582 2583 2584 |
* sched_balance_self: balance the current task (running on cpu) in domains * that have the 'flag' flag set. In practice, this is SD_BALANCE_FORK and * SD_BALANCE_EXEC. * * Balance, ie. select the least loaded group. * * Returns the target CPU number, or the same CPU if no balancing is needed. * * preempt must be disabled. */ |
0017d7350 sched: Fix TASK_W... |
2585 |
static int |
7608dec2c sched: Drop the r... |
2586 |
select_task_rq_fair(struct task_struct *p, int sd_flag, int wake_flags) |
aaee1203c sched: Move sched... |
2587 |
{ |
29cd8bae3 sched: Fix SD_POW... |
2588 |
struct sched_domain *tmp, *affine_sd = NULL, *sd = NULL; |
c88d59108 sched: Merge sele... |
2589 2590 2591 |
int cpu = smp_processor_id(); int prev_cpu = task_cpu(p); int new_cpu = cpu; |
99bd5e2f2 sched: Fix select... |
2592 |
int want_affine = 0; |
29cd8bae3 sched: Fix SD_POW... |
2593 |
int want_sd = 1; |
5158f4e44 sched: Clean up t... |
2594 |
int sync = wake_flags & WF_SYNC; |
c88d59108 sched: Merge sele... |
2595 |
|
76854c7e8 sched: Use rt.nr_... |
2596 2597 |
if (p->rt.nr_cpus_allowed == 1) return prev_cpu; |
0763a660a sched: Rename sel... |
2598 |
if (sd_flag & SD_BALANCE_WAKE) { |
fa17b507f sched: Wrap sched... |
2599 |
if (cpumask_test_cpu(cpu, tsk_cpus_allowed(p))) |
c88d59108 sched: Merge sele... |
2600 2601 2602 |
want_affine = 1; new_cpu = prev_cpu; } |
aaee1203c sched: Move sched... |
2603 |
|
dce840a08 sched: Dynamicall... |
2604 |
rcu_read_lock(); |
aaee1203c sched: Move sched... |
2605 |
for_each_domain(cpu, tmp) { |
e4f428884 sched: Select_tas... |
2606 2607 |
if (!(tmp->flags & SD_LOAD_BALANCE)) continue; |
aaee1203c sched: Move sched... |
2608 |
/* |
ae154be1f sched: Weaken SD_... |
2609 2610 |
* If power savings logic is enabled for a domain, see if we * are not overloaded, if so, don't balance wider. |
aaee1203c sched: Move sched... |
2611 |
*/ |
59abf0264 sched: Add SD_PRE... |
2612 |
if (tmp->flags & (SD_POWERSAVINGS_BALANCE|SD_PREFER_LOCAL)) { |
ae154be1f sched: Weaken SD_... |
2613 2614 2615 2616 2617 2618 2619 2620 2621 |
unsigned long power = 0; unsigned long nr_running = 0; unsigned long capacity; int i; for_each_cpu(i, sched_domain_span(tmp)) { power += power_of(i); nr_running += cpu_rq(i)->cfs.nr_running; } |
1399fa780 sched: Introduce ... |
2622 |
capacity = DIV_ROUND_CLOSEST(power, SCHED_POWER_SCALE); |
ae154be1f sched: Weaken SD_... |
2623 |
|
59abf0264 sched: Add SD_PRE... |
2624 2625 2626 2627 |
if (tmp->flags & SD_POWERSAVINGS_BALANCE) nr_running /= 2; if (nr_running < capacity) |
29cd8bae3 sched: Fix SD_POW... |
2628 |
want_sd = 0; |
ae154be1f sched: Weaken SD_... |
2629 |
} |
aaee1203c sched: Move sched... |
2630 |
|
fe3bcfe1f sched: More gener... |
2631 |
/* |
99bd5e2f2 sched: Fix select... |
2632 2633 |
* If both cpu and prev_cpu are part of this domain, * cpu is a valid SD_WAKE_AFFINE target. |
fe3bcfe1f sched: More gener... |
2634 |
*/ |
99bd5e2f2 sched: Fix select... |
2635 2636 2637 2638 |
if (want_affine && (tmp->flags & SD_WAKE_AFFINE) && cpumask_test_cpu(prev_cpu, sched_domain_span(tmp))) { affine_sd = tmp; want_affine = 0; |
c88d59108 sched: Merge sele... |
2639 |
} |
29cd8bae3 sched: Fix SD_POW... |
2640 2641 |
if (!want_sd && !want_affine) break; |
0763a660a sched: Rename sel... |
2642 |
if (!(tmp->flags & sd_flag)) |
c88d59108 sched: Merge sele... |
2643 |
continue; |
29cd8bae3 sched: Fix SD_POW... |
2644 2645 2646 |
if (want_sd) sd = tmp; } |
8b911acdf sched: Fix select... |
2647 |
if (affine_sd) { |
99bd5e2f2 sched: Fix select... |
2648 |
if (cpu == prev_cpu || wake_affine(affine_sd, p, sync)) |
dce840a08 sched: Dynamicall... |
2649 2650 2651 2652 |
prev_cpu = cpu; new_cpu = select_idle_sibling(p, prev_cpu); goto unlock; |
8b911acdf sched: Fix select... |
2653 |
} |
e7693a362 sched: de-SCHED_O... |
2654 |
|
aaee1203c sched: Move sched... |
2655 |
while (sd) { |
5158f4e44 sched: Clean up t... |
2656 |
int load_idx = sd->forkexec_idx; |
aaee1203c sched: Move sched... |
2657 |
struct sched_group *group; |
c88d59108 sched: Merge sele... |
2658 |
int weight; |
098fb9db2 sched: clean up w... |
2659 |
|
0763a660a sched: Rename sel... |
2660 |
if (!(sd->flags & sd_flag)) { |
aaee1203c sched: Move sched... |
2661 2662 2663 |
sd = sd->child; continue; } |
098fb9db2 sched: clean up w... |
2664 |
|
5158f4e44 sched: Clean up t... |
2665 2666 |
if (sd_flag & SD_BALANCE_WAKE) load_idx = sd->wake_idx; |
098fb9db2 sched: clean up w... |
2667 |
|
5158f4e44 sched: Clean up t... |
2668 |
group = find_idlest_group(sd, p, cpu, load_idx); |
aaee1203c sched: Move sched... |
2669 2670 2671 2672 |
if (!group) { sd = sd->child; continue; } |
4ae7d5cef sched: improve af... |
2673 |
|
d7c33c493 sched: Fix task a... |
2674 |
new_cpu = find_idlest_cpu(group, p, cpu); |
aaee1203c sched: Move sched... |
2675 2676 2677 2678 |
if (new_cpu == -1 || new_cpu == cpu) { /* Now try balancing at a lower domain level of cpu */ sd = sd->child; continue; |
e7693a362 sched: de-SCHED_O... |
2679 |
} |
aaee1203c sched: Move sched... |
2680 2681 2682 |
/* Now try balancing at a lower domain level of new_cpu */ cpu = new_cpu; |
669c55e9f sched: Pre-comput... |
2683 |
weight = sd->span_weight; |
aaee1203c sched: Move sched... |
2684 2685 |
sd = NULL; for_each_domain(cpu, tmp) { |
669c55e9f sched: Pre-comput... |
2686 |
if (weight <= tmp->span_weight) |
aaee1203c sched: Move sched... |
2687 |
break; |
0763a660a sched: Rename sel... |
2688 |
if (tmp->flags & sd_flag) |
aaee1203c sched: Move sched... |
2689 2690 2691 |
sd = tmp; } /* while loop will break here if sd == NULL */ |
e7693a362 sched: de-SCHED_O... |
2692 |
} |
dce840a08 sched: Dynamicall... |
2693 2694 |
unlock: rcu_read_unlock(); |
e7693a362 sched: de-SCHED_O... |
2695 |
|
c88d59108 sched: Merge sele... |
2696 |
return new_cpu; |
e7693a362 sched: de-SCHED_O... |
2697 2698 |
} #endif /* CONFIG_SMP */ |
e52fb7c09 sched: prefer wakers |
2699 2700 |
static unsigned long wakeup_gran(struct sched_entity *curr, struct sched_entity *se) |
0bbd3336e sched: fix wakeup... |
2701 2702 2703 2704 |
{ unsigned long gran = sysctl_sched_wakeup_granularity; /* |
e52fb7c09 sched: prefer wakers |
2705 2706 |
* Since its curr running now, convert the gran from real-time * to virtual-time in his units. |
13814d42e sched: Remove ASY... |
2707 2708 2709 2710 2711 2712 2713 2714 2715 |
* * By using 'se' instead of 'curr' we penalize light tasks, so * they get preempted easier. That is, if 'se' < 'curr' then * the resulting gran will be larger, therefore penalizing the * lighter, if otoh 'se' > 'curr' then the resulting gran will * be smaller, again penalizing the lighter task. * * This is especially important for buddies when the leftmost * task is higher priority than the buddy. |
0bbd3336e sched: fix wakeup... |
2716 |
*/ |
f4ad9bd20 sched: Eliminate ... |
2717 |
return calc_delta_fair(gran, se); |
0bbd3336e sched: fix wakeup... |
2718 2719 2720 |
} /* |
464b75273 sched: re-instate... |
2721 2722 2723 2724 2725 2726 2727 2728 2729 2730 2731 2732 2733 2734 2735 2736 2737 2738 2739 2740 |
* Should 'se' preempt 'curr'. * * |s1 * |s2 * |s3 * g * |<--->|c * * w(c, s1) = -1 * w(c, s2) = 0 * w(c, s3) = 1 * */ static int wakeup_preempt_entity(struct sched_entity *curr, struct sched_entity *se) { s64 gran, vdiff = curr->vruntime - se->vruntime; if (vdiff <= 0) return -1; |
e52fb7c09 sched: prefer wakers |
2741 |
gran = wakeup_gran(curr, se); |
464b75273 sched: re-instate... |
2742 2743 2744 2745 2746 |
if (vdiff > gran) return 1; return 0; } |
02479099c sched: fix buddie... |
2747 2748 |
static void set_last_buddy(struct sched_entity *se) { |
69c80f3e9 sched: Make set_*... |
2749 2750 2751 2752 2753 |
if (entity_is_task(se) && unlikely(task_of(se)->policy == SCHED_IDLE)) return; for_each_sched_entity(se) cfs_rq_of(se)->last = se; |
02479099c sched: fix buddie... |
2754 2755 2756 2757 |
} static void set_next_buddy(struct sched_entity *se) { |
69c80f3e9 sched: Make set_*... |
2758 2759 2760 2761 2762 |
if (entity_is_task(se) && unlikely(task_of(se)->policy == SCHED_IDLE)) return; for_each_sched_entity(se) cfs_rq_of(se)->next = se; |
02479099c sched: fix buddie... |
2763 |
} |
ac53db596 sched: Use a budd... |
2764 2765 |
static void set_skip_buddy(struct sched_entity *se) { |
69c80f3e9 sched: Make set_*... |
2766 2767 |
for_each_sched_entity(se) cfs_rq_of(se)->skip = se; |
ac53db596 sched: Use a budd... |
2768 |
} |
464b75273 sched: re-instate... |
2769 |
/* |
bf0f6f24a sched: cfs core, ... |
2770 2771 |
* Preempt the current task with a newly woken task if needed: */ |
5a9b86f64 sched: Rename fla... |
2772 |
static void check_preempt_wakeup(struct rq *rq, struct task_struct *p, int wake_flags) |
bf0f6f24a sched: cfs core, ... |
2773 2774 |
{ struct task_struct *curr = rq->curr; |
8651a86c3 sched: group sche... |
2775 |
struct sched_entity *se = &curr->se, *pse = &p->se; |
03e89e457 sched: fix wakeup... |
2776 |
struct cfs_rq *cfs_rq = task_cfs_rq(curr); |
f685ceaca sched: Strengthen... |
2777 |
int scale = cfs_rq->nr_running >= sched_nr_latency; |
2f36825b1 sched: Next buddy... |
2778 |
int next_buddy_marked = 0; |
bf0f6f24a sched: cfs core, ... |
2779 |
|
4ae7d5cef sched: improve af... |
2780 2781 |
if (unlikely(se == pse)) return; |
5238cdd38 sched: Prevent bu... |
2782 2783 2784 2785 2786 2787 2788 2789 |
/* * This is possible from callers such as pull_task(), in which we * unconditionally check_prempt_curr() after an enqueue (which may have * lead to a throttle). This both saves work and prevents false * next-buddy nomination below. */ if (unlikely(throttled_hierarchy(cfs_rq_of(pse)))) return; |
2f36825b1 sched: Next buddy... |
2790 |
if (sched_feat(NEXT_BUDDY) && scale && !(wake_flags & WF_FORK)) { |
3cb63d527 sched: Complete b... |
2791 |
set_next_buddy(pse); |
2f36825b1 sched: Next buddy... |
2792 2793 |
next_buddy_marked = 1; } |
57fdc26d4 sched: fixup budd... |
2794 |
|
aec0a5142 sched: call resch... |
2795 2796 2797 |
/* * We can come here with TIF_NEED_RESCHED already set from new task * wake up path. |
5238cdd38 sched: Prevent bu... |
2798 2799 2800 2801 2802 2803 |
* * Note: this also catches the edge-case of curr being in a throttled * group (e.g. via set_curr_task), since update_curr() (in the * enqueue of curr) will have resulted in resched being set. This * prevents us from potentially nominating it as a false LAST_BUDDY * below. |
aec0a5142 sched: call resch... |
2804 2805 2806 |
*/ if (test_tsk_need_resched(curr)) return; |
a2f5c9ab7 sched: Allow SCHE... |
2807 2808 2809 2810 |
/* Idle tasks are by definition preempted by non-idle tasks. */ if (unlikely(curr->policy == SCHED_IDLE) && likely(p->policy != SCHED_IDLE)) goto preempt; |
91c234b4e sched: do not wak... |
2811 |
/* |
a2f5c9ab7 sched: Allow SCHE... |
2812 2813 |
* Batch and idle tasks do not preempt non-idle tasks (their preemption * is driven by the tick): |
91c234b4e sched: do not wak... |
2814 |
*/ |
6bc912b71 sched: SCHED_OTHE... |
2815 |
if (unlikely(p->policy != SCHED_NORMAL)) |
91c234b4e sched: do not wak... |
2816 |
return; |
bf0f6f24a sched: cfs core, ... |
2817 |
|
464b75273 sched: re-instate... |
2818 |
find_matching_se(&se, &pse); |
9bbd73743 sched: update cor... |
2819 |
update_curr(cfs_rq_of(se)); |
002f128b4 sched: remove red... |
2820 |
BUG_ON(!pse); |
2f36825b1 sched: Next buddy... |
2821 2822 2823 2824 2825 2826 2827 |
if (wakeup_preempt_entity(se, pse) == 1) { /* * Bias pick_next to pick the sched entity that is * triggering this preemption. */ if (!next_buddy_marked) set_next_buddy(pse); |
3a7e73a2e sched: Clean up c... |
2828 |
goto preempt; |
2f36825b1 sched: Next buddy... |
2829 |
} |
464b75273 sched: re-instate... |
2830 |
|
3a7e73a2e sched: Clean up c... |
2831 |
return; |
a65ac745e sched: Move updat... |
2832 |
|
3a7e73a2e sched: Clean up c... |
2833 2834 2835 2836 2837 2838 2839 2840 2841 2842 2843 2844 2845 2846 2847 2848 |
preempt: resched_task(curr); /* * Only set the backward buddy when the current task is still * on the rq. This can happen when a wakeup gets interleaved * with schedule on the ->pre_schedule() or idle_balance() * point, either of which can * drop the rq lock. * * Also, during early boot the idle thread is in the fair class, * for obvious reasons its a bad idea to schedule back to it. */ if (unlikely(!se->on_rq || curr == rq->idle)) return; if (sched_feat(LAST_BUDDY) && scale && entity_is_task(se)) set_last_buddy(se); |
bf0f6f24a sched: cfs core, ... |
2849 |
} |
fb8d47240 sched: remove the... |
2850 |
static struct task_struct *pick_next_task_fair(struct rq *rq) |
bf0f6f24a sched: cfs core, ... |
2851 |
{ |
8f4d37ec0 sched: high-res p... |
2852 |
struct task_struct *p; |
bf0f6f24a sched: cfs core, ... |
2853 2854 |
struct cfs_rq *cfs_rq = &rq->cfs; struct sched_entity *se; |
36ace27e3 sched: Optimize b... |
2855 |
if (!cfs_rq->nr_running) |
bf0f6f24a sched: cfs core, ... |
2856 2857 2858 |
return NULL; do { |
9948f4b2a sched: remove the... |
2859 |
se = pick_next_entity(cfs_rq); |
f4b6755fb sched: cleanup fa... |
2860 |
set_next_entity(cfs_rq, se); |
bf0f6f24a sched: cfs core, ... |
2861 2862 |
cfs_rq = group_cfs_rq(se); } while (cfs_rq); |
8f4d37ec0 sched: high-res p... |
2863 |
p = task_of(se); |
b39e66eaf sched: Save some ... |
2864 2865 |
if (hrtick_enabled(rq)) hrtick_start_fair(rq, p); |
8f4d37ec0 sched: high-res p... |
2866 2867 |
return p; |
bf0f6f24a sched: cfs core, ... |
2868 2869 2870 2871 2872 |
} /* * Account for a descheduled task: */ |
31ee529cc sched: remove the... |
2873 |
static void put_prev_task_fair(struct rq *rq, struct task_struct *prev) |
bf0f6f24a sched: cfs core, ... |
2874 2875 2876 2877 2878 2879 |
{ struct sched_entity *se = &prev->se; struct cfs_rq *cfs_rq; for_each_sched_entity(se) { cfs_rq = cfs_rq_of(se); |
ab6cde269 sched: remove the... |
2880 |
put_prev_entity(cfs_rq, se); |
bf0f6f24a sched: cfs core, ... |
2881 2882 |
} } |
ac53db596 sched: Use a budd... |
2883 2884 2885 2886 2887 2888 2889 2890 2891 2892 2893 2894 2895 2896 2897 2898 2899 2900 2901 2902 2903 2904 2905 2906 2907 |
/* * sched_yield() is very simple * * The magic of dealing with the ->skip buddy is in pick_next_entity. */ static void yield_task_fair(struct rq *rq) { struct task_struct *curr = rq->curr; struct cfs_rq *cfs_rq = task_cfs_rq(curr); struct sched_entity *se = &curr->se; /* * Are we the only task in the tree? */ if (unlikely(rq->nr_running == 1)) return; clear_buddies(cfs_rq, se); if (curr->policy != SCHED_BATCH) { update_rq_clock(rq); /* * Update run-time statistics of the 'current'. */ update_curr(cfs_rq); |
916671c08 sched: Set skip_c... |
2908 2909 2910 2911 2912 2913 |
/* * Tell update_rq_clock() that we've just updated, * so we don't do microscopic update in schedule() * and double the fastpath cost. */ rq->skip_clock_update = 1; |
ac53db596 sched: Use a budd... |
2914 2915 2916 2917 |
} set_skip_buddy(se); } |
d95f41220 sched: Add yield_... |
2918 2919 2920 |
static bool yield_to_task_fair(struct rq *rq, struct task_struct *p, bool preempt) { struct sched_entity *se = &p->se; |
5238cdd38 sched: Prevent bu... |
2921 2922 |
/* throttled hierarchies are not runnable */ if (!se->on_rq || throttled_hierarchy(cfs_rq_of(se))) |
d95f41220 sched: Add yield_... |
2923 2924 2925 2926 |
return false; /* Tell the scheduler that we'd really like pse to run next. */ set_next_buddy(se); |
d95f41220 sched: Add yield_... |
2927 2928 2929 2930 |
yield_task_fair(rq); return true; } |
681f3e685 sched: isolate SM... |
2931 |
#ifdef CONFIG_SMP |
bf0f6f24a sched: cfs core, ... |
2932 2933 2934 |
/************************************************** * Fair scheduling class load-balancing methods: */ |
1e3c88bde sched: Move load ... |
2935 2936 2937 2938 2939 2940 2941 2942 2943 2944 2945 2946 2947 2948 |
/* * pull_task - move a task from a remote runqueue to the local runqueue. * Both runqueues must be locked. */ static void pull_task(struct rq *src_rq, struct task_struct *p, struct rq *this_rq, int this_cpu) { deactivate_task(src_rq, p, 0); set_task_cpu(p, this_cpu); activate_task(this_rq, p, 0); check_preempt_curr(this_rq, p, 0); } /* |
029632fbb sched: Make separ... |
2949 2950 2951 2952 2953 2954 2955 2956 2957 2958 2959 2960 2961 2962 2963 2964 2965 2966 2967 2968 2969 2970 2971 2972 2973 2974 2975 2976 2977 2978 |
* Is this task likely cache-hot: */ static int task_hot(struct task_struct *p, u64 now, struct sched_domain *sd) { s64 delta; if (p->sched_class != &fair_sched_class) return 0; if (unlikely(p->policy == SCHED_IDLE)) return 0; /* * Buddy candidates are cache hot: */ if (sched_feat(CACHE_HOT_BUDDY) && this_rq()->nr_running && (&p->se == cfs_rq_of(&p->se)->next || &p->se == cfs_rq_of(&p->se)->last)) return 1; if (sysctl_sched_migration_cost == -1) return 1; if (sysctl_sched_migration_cost == 0) return 0; delta = now - p->se.exec_start; return delta < (s64)sysctl_sched_migration_cost; } |
5b54b56be sched: Replace al... |
2979 |
#define LBF_ALL_PINNED 0x01 |
bced76aea sched: Fix lockup... |
2980 2981 2982 2983 |
#define LBF_NEED_BREAK 0x02 /* clears into HAD_BREAK */ #define LBF_HAD_BREAK 0x04 #define LBF_HAD_BREAKS 0x0C /* count HAD_BREAKs overflows into ABORT */ #define LBF_ABORT 0x10 |
5b54b56be sched: Replace al... |
2984 |
|
029632fbb sched: Make separ... |
2985 |
/* |
1e3c88bde sched: Move load ... |
2986 2987 2988 2989 2990 |
* can_migrate_task - may task p from runqueue rq be migrated to this_cpu? */ static int can_migrate_task(struct task_struct *p, struct rq *rq, int this_cpu, struct sched_domain *sd, enum cpu_idle_type idle, |
5b54b56be sched: Replace al... |
2991 |
int *lb_flags) |
1e3c88bde sched: Move load ... |
2992 2993 2994 2995 2996 2997 2998 2999 |
{ int tsk_cache_hot = 0; /* * We do not migrate tasks that are: * 1) running (obviously), or * 2) cannot be migrated to this CPU due to cpus_allowed, or * 3) are cache-hot on their current CPU. */ |
fa17b507f sched: Wrap sched... |
3000 |
if (!cpumask_test_cpu(this_cpu, tsk_cpus_allowed(p))) { |
41acab885 sched: Implement ... |
3001 |
schedstat_inc(p, se.statistics.nr_failed_migrations_affine); |
1e3c88bde sched: Move load ... |
3002 3003 |
return 0; } |
5b54b56be sched: Replace al... |
3004 |
*lb_flags &= ~LBF_ALL_PINNED; |
1e3c88bde sched: Move load ... |
3005 3006 |
if (task_running(rq, p)) { |
41acab885 sched: Implement ... |
3007 |
schedstat_inc(p, se.statistics.nr_failed_migrations_running); |
1e3c88bde sched: Move load ... |
3008 3009 3010 3011 3012 3013 3014 3015 |
return 0; } /* * Aggressive migration if: * 1) task is cache cold, or * 2) too many balance attempts have failed. */ |
305e6835e sched: Do not acc... |
3016 |
tsk_cache_hot = task_hot(p, rq->clock_task, sd); |
1e3c88bde sched: Move load ... |
3017 3018 3019 3020 3021 |
if (!tsk_cache_hot || sd->nr_balance_failed > sd->cache_nice_tries) { #ifdef CONFIG_SCHEDSTATS if (tsk_cache_hot) { schedstat_inc(sd, lb_hot_gained[idle]); |
41acab885 sched: Implement ... |
3022 |
schedstat_inc(p, se.statistics.nr_forced_migrations); |
1e3c88bde sched: Move load ... |
3023 3024 3025 3026 3027 3028 |
} #endif return 1; } if (tsk_cache_hot) { |
41acab885 sched: Implement ... |
3029 |
schedstat_inc(p, se.statistics.nr_failed_migrations_hot); |
1e3c88bde sched: Move load ... |
3030 3031 3032 3033 |
return 0; } return 1; } |
897c395f4 sched: Remove rq_... |
3034 3035 3036 3037 3038 3039 3040 3041 3042 3043 3044 3045 3046 3047 3048 3049 3050 |
/* * move_one_task tries to move exactly one task from busiest to this_rq, as * part of active balancing operations within "domain". * Returns 1 if successful and 0 otherwise. * * Called with both runqueues locked. */ static int move_one_task(struct rq *this_rq, int this_cpu, struct rq *busiest, struct sched_domain *sd, enum cpu_idle_type idle) { struct task_struct *p, *n; struct cfs_rq *cfs_rq; int pinned = 0; for_each_leaf_cfs_rq(busiest, cfs_rq) { list_for_each_entry_safe(p, n, &cfs_rq->tasks, se.group_node) { |
64660c864 sched: Prevent in... |
3051 3052 3053 |
if (throttled_lb_pair(task_group(p), busiest->cpu, this_cpu)) break; |
897c395f4 sched: Remove rq_... |
3054 3055 3056 3057 3058 3059 3060 3061 3062 3063 3064 3065 3066 3067 3068 3069 3070 3071 |
if (!can_migrate_task(p, busiest, this_cpu, sd, idle, &pinned)) continue; pull_task(busiest, p, this_rq, this_cpu); /* * Right now, this is only the second place pull_task() * is called, so we can safely collect pull_task() * stats here rather than inside pull_task(). */ schedstat_inc(sd, lb_gained[idle]); return 1; } } return 0; } |
1e3c88bde sched: Move load ... |
3072 3073 3074 |
static unsigned long balance_tasks(struct rq *this_rq, int this_cpu, struct rq *busiest, unsigned long max_load_move, struct sched_domain *sd, |
5b54b56be sched: Replace al... |
3075 |
enum cpu_idle_type idle, int *lb_flags, |
931aeeda0 sched: Remove unu... |
3076 |
struct cfs_rq *busiest_cfs_rq) |
1e3c88bde sched: Move load ... |
3077 |
{ |
b30aef17f sched: Fix errone... |
3078 |
int loops = 0, pulled = 0; |
1e3c88bde sched: Move load ... |
3079 |
long rem_load_move = max_load_move; |
ee00e66ff sched: Remove rq_... |
3080 |
struct task_struct *p, *n; |
1e3c88bde sched: Move load ... |
3081 3082 3083 |
if (max_load_move == 0) goto out; |
ee00e66ff sched: Remove rq_... |
3084 |
list_for_each_entry_safe(p, n, &busiest_cfs_rq->tasks, se.group_node) { |
a195f004e sched: Fix load-b... |
3085 3086 |
if (loops++ > sysctl_sched_nr_migrate) { *lb_flags |= LBF_NEED_BREAK; |
ee00e66ff sched: Remove rq_... |
3087 |
break; |
a195f004e sched: Fix load-b... |
3088 |
} |
1e3c88bde sched: Move load ... |
3089 |
|
ee00e66ff sched: Remove rq_... |
3090 |
if ((p->se.load.weight >> 1) > rem_load_move || |
b30aef17f sched: Fix errone... |
3091 |
!can_migrate_task(p, busiest, this_cpu, sd, idle, |
5b54b56be sched: Replace al... |
3092 |
lb_flags)) |
ee00e66ff sched: Remove rq_... |
3093 |
continue; |
1e3c88bde sched: Move load ... |
3094 |
|
ee00e66ff sched: Remove rq_... |
3095 3096 3097 |
pull_task(busiest, p, this_rq, this_cpu); pulled++; rem_load_move -= p->se.load.weight; |
1e3c88bde sched: Move load ... |
3098 3099 |
#ifdef CONFIG_PREEMPT |
ee00e66ff sched: Remove rq_... |
3100 3101 3102 3103 3104 |
/* * NEWIDLE balancing is a source of latency, so preemptible * kernels will stop after the first task is pulled to minimize * the critical section. */ |
a195f004e sched: Fix load-b... |
3105 3106 |
if (idle == CPU_NEWLY_IDLE) { *lb_flags |= LBF_ABORT; |
ee00e66ff sched: Remove rq_... |
3107 |
break; |
a195f004e sched: Fix load-b... |
3108 |
} |
1e3c88bde sched: Move load ... |
3109 |
#endif |
ee00e66ff sched: Remove rq_... |
3110 3111 3112 3113 3114 3115 |
/* * We only want to steal up to the prescribed amount of * weighted load. */ if (rem_load_move <= 0) break; |
1e3c88bde sched: Move load ... |
3116 3117 3118 3119 3120 3121 3122 3123 |
} out: /* * Right now, this is one of only two places pull_task() is called, * so we can safely collect pull_task() stats here rather than * inside pull_task(). */ schedstat_add(sd, lb_gained[idle], pulled); |
1e3c88bde sched: Move load ... |
3124 3125 |
return max_load_move - rem_load_move; } |
230059de7 sched: Remove fro... |
3126 |
#ifdef CONFIG_FAIR_GROUP_SCHED |
9e3081ca6 sched: Make tg_sh... |
3127 3128 3129 |
/* * update tg->load_weight by folding this cpu's load_avg */ |
67e86250f sched: Introduce ... |
3130 |
static int update_shares_cpu(struct task_group *tg, int cpu) |
9e3081ca6 sched: Make tg_sh... |
3131 3132 3133 3134 |
{ struct cfs_rq *cfs_rq; unsigned long flags; struct rq *rq; |
9e3081ca6 sched: Make tg_sh... |
3135 3136 3137 3138 3139 3140 3141 3142 3143 3144 |
if (!tg->se[cpu]) return 0; rq = cpu_rq(cpu); cfs_rq = tg->cfs_rq[cpu]; raw_spin_lock_irqsave(&rq->lock, flags); update_rq_clock(rq); |
d6b559182 sched: Allow upda... |
3145 |
update_cfs_load(cfs_rq, 1); |
9e3081ca6 sched: Make tg_sh... |
3146 3147 3148 3149 3150 |
/* * We need to update shares after updating tg->load_weight in * order to adjust the weight of groups with long running tasks. */ |
6d5ab2932 sched: Simplify u... |
3151 |
update_cfs_shares(cfs_rq); |
9e3081ca6 sched: Make tg_sh... |
3152 3153 3154 3155 3156 3157 3158 3159 3160 3161 3162 3163 |
raw_spin_unlock_irqrestore(&rq->lock, flags); return 0; } static void update_shares(int cpu) { struct cfs_rq *cfs_rq; struct rq *rq = cpu_rq(cpu); rcu_read_lock(); |
9763b67fb sched, cgroup: Op... |
3164 3165 3166 3167 |
/* * Iterates the task_group tree in a bottom up fashion, see * list_add_leaf_cfs_rq() for details. */ |
64660c864 sched: Prevent in... |
3168 3169 3170 3171 |
for_each_leaf_cfs_rq(rq, cfs_rq) { /* throttled entities do not contribute to load */ if (throttled_hierarchy(cfs_rq)) continue; |
67e86250f sched: Introduce ... |
3172 |
update_shares_cpu(cfs_rq->tg, cpu); |
64660c864 sched: Prevent in... |
3173 |
} |
9e3081ca6 sched: Make tg_sh... |
3174 3175 |
rcu_read_unlock(); } |
9763b67fb sched, cgroup: Op... |
3176 3177 3178 3179 3180 3181 3182 3183 3184 3185 3186 3187 3188 3189 3190 3191 3192 3193 3194 3195 3196 3197 3198 3199 3200 3201 3202 |
/* * Compute the cpu's hierarchical load factor for each task group. * This needs to be done in a top-down fashion because the load of a child * group is a fraction of its parents load. */ static int tg_load_down(struct task_group *tg, void *data) { unsigned long load; long cpu = (long)data; if (!tg->parent) { load = cpu_rq(cpu)->load.weight; } else { load = tg->parent->cfs_rq[cpu]->h_load; load *= tg->se[cpu]->load.weight; load /= tg->parent->cfs_rq[cpu]->load.weight + 1; } tg->cfs_rq[cpu]->h_load = load; return 0; } static void update_h_load(long cpu) { walk_tg_tree(tg_load_down, tg_nop, (void *)cpu); } |
230059de7 sched: Remove fro... |
3203 3204 3205 3206 |
static unsigned long load_balance_fair(struct rq *this_rq, int this_cpu, struct rq *busiest, unsigned long max_load_move, struct sched_domain *sd, enum cpu_idle_type idle, |
5b54b56be sched: Replace al... |
3207 |
int *lb_flags) |
230059de7 sched: Remove fro... |
3208 3209 |
{ long rem_load_move = max_load_move; |
9763b67fb sched, cgroup: Op... |
3210 |
struct cfs_rq *busiest_cfs_rq; |
230059de7 sched: Remove fro... |
3211 3212 |
rcu_read_lock(); |
9763b67fb sched, cgroup: Op... |
3213 |
update_h_load(cpu_of(busiest)); |
230059de7 sched: Remove fro... |
3214 |
|
9763b67fb sched, cgroup: Op... |
3215 |
for_each_leaf_cfs_rq(busiest, busiest_cfs_rq) { |
230059de7 sched: Remove fro... |
3216 3217 3218 |
unsigned long busiest_h_load = busiest_cfs_rq->h_load; unsigned long busiest_weight = busiest_cfs_rq->load.weight; u64 rem_load, moved_load; |
a195f004e sched: Fix load-b... |
3219 3220 |
if (*lb_flags & (LBF_NEED_BREAK|LBF_ABORT)) break; |
230059de7 sched: Remove fro... |
3221 |
/* |
64660c864 sched: Prevent in... |
3222 |
* empty group or part of a throttled hierarchy |
230059de7 sched: Remove fro... |
3223 |
*/ |
64660c864 sched: Prevent in... |
3224 3225 |
if (!busiest_cfs_rq->task_weight || throttled_lb_pair(busiest_cfs_rq->tg, cpu_of(busiest), this_cpu)) |
230059de7 sched: Remove fro... |
3226 3227 3228 3229 3230 3231 |
continue; rem_load = (u64)rem_load_move * busiest_weight; rem_load = div_u64(rem_load, busiest_h_load + 1); moved_load = balance_tasks(this_rq, this_cpu, busiest, |
5b54b56be sched: Replace al... |
3232 |
rem_load, sd, idle, lb_flags, |
230059de7 sched: Remove fro... |
3233 3234 3235 3236 3237 3238 3239 3240 3241 3242 3243 3244 3245 3246 3247 3248 3249 |
busiest_cfs_rq); if (!moved_load) continue; moved_load *= busiest_h_load; moved_load = div_u64(moved_load, busiest_weight + 1); rem_load_move -= moved_load; if (rem_load_move < 0) break; } rcu_read_unlock(); return max_load_move - rem_load_move; } #else |
9e3081ca6 sched: Make tg_sh... |
3250 3251 3252 |
static inline void update_shares(int cpu) { } |
230059de7 sched: Remove fro... |
3253 3254 3255 3256 |
static unsigned long load_balance_fair(struct rq *this_rq, int this_cpu, struct rq *busiest, unsigned long max_load_move, struct sched_domain *sd, enum cpu_idle_type idle, |
5b54b56be sched: Replace al... |
3257 |
int *lb_flags) |
230059de7 sched: Remove fro... |
3258 3259 |
{ return balance_tasks(this_rq, this_cpu, busiest, |
5b54b56be sched: Replace al... |
3260 |
max_load_move, sd, idle, lb_flags, |
931aeeda0 sched: Remove unu... |
3261 |
&busiest->cfs); |
230059de7 sched: Remove fro... |
3262 3263 |
} #endif |
1e3c88bde sched: Move load ... |
3264 3265 3266 3267 3268 3269 3270 3271 3272 3273 |
/* * move_tasks tries to move up to max_load_move weighted load from busiest to * this_rq, as part of a balancing operation within domain "sd". * Returns 1 if successful and 0 otherwise. * * Called with both runqueues locked. */ static int move_tasks(struct rq *this_rq, int this_cpu, struct rq *busiest, unsigned long max_load_move, struct sched_domain *sd, enum cpu_idle_type idle, |
5b54b56be sched: Replace al... |
3274 |
int *lb_flags) |
1e3c88bde sched: Move load ... |
3275 |
{ |
3d45fd804 sched: Remove the... |
3276 |
unsigned long total_load_moved = 0, load_moved; |
1e3c88bde sched: Move load ... |
3277 3278 |
do { |
3d45fd804 sched: Remove the... |
3279 |
load_moved = load_balance_fair(this_rq, this_cpu, busiest, |
1e3c88bde sched: Move load ... |
3280 |
max_load_move - total_load_moved, |
5b54b56be sched: Replace al... |
3281 |
sd, idle, lb_flags); |
3d45fd804 sched: Remove the... |
3282 3283 |
total_load_moved += load_moved; |
1e3c88bde sched: Move load ... |
3284 |
|
a195f004e sched: Fix load-b... |
3285 3286 |
if (*lb_flags & (LBF_NEED_BREAK|LBF_ABORT)) break; |
1e3c88bde sched: Move load ... |
3287 3288 3289 3290 3291 3292 |
#ifdef CONFIG_PREEMPT /* * NEWIDLE balancing is a source of latency, so preemptible * kernels will stop after the first task is pulled to minimize * the critical section. */ |
a195f004e sched: Fix load-b... |
3293 3294 |
if (idle == CPU_NEWLY_IDLE && this_rq->nr_running) { *lb_flags |= LBF_ABORT; |
baa8c1102 sched: Add a lock... |
3295 |
break; |
a195f004e sched: Fix load-b... |
3296 |
} |
1e3c88bde sched: Move load ... |
3297 |
#endif |
3d45fd804 sched: Remove the... |
3298 |
} while (load_moved && max_load_move > total_load_moved); |
1e3c88bde sched: Move load ... |
3299 3300 3301 |
return total_load_moved > 0; } |
1e3c88bde sched: Move load ... |
3302 3303 3304 3305 3306 3307 3308 3309 3310 3311 3312 3313 3314 3315 3316 3317 |
/********** Helpers for find_busiest_group ************************/ /* * sd_lb_stats - Structure to store the statistics of a sched_domain * during load balancing. */ struct sd_lb_stats { struct sched_group *busiest; /* Busiest group in this sd */ struct sched_group *this; /* Local group in this sd */ unsigned long total_load; /* Total load of all groups in sd */ unsigned long total_pwr; /* Total power of all groups in sd */ unsigned long avg_load; /* Average load across all groups in sd */ /** Statistics of this group */ unsigned long this_load; unsigned long this_load_per_task; unsigned long this_nr_running; |
fab476228 sched: Force bala... |
3318 |
unsigned long this_has_capacity; |
aae6d3ddd sched: Use group ... |
3319 |
unsigned int this_idle_cpus; |
1e3c88bde sched: Move load ... |
3320 3321 |
/* Statistics of the busiest group */ |
aae6d3ddd sched: Use group ... |
3322 |
unsigned int busiest_idle_cpus; |
1e3c88bde sched: Move load ... |
3323 3324 3325 |
unsigned long max_load; unsigned long busiest_load_per_task; unsigned long busiest_nr_running; |
dd5feea14 sched: Fix SCHED_... |
3326 |
unsigned long busiest_group_capacity; |
fab476228 sched: Force bala... |
3327 |
unsigned long busiest_has_capacity; |
aae6d3ddd sched: Use group ... |
3328 |
unsigned int busiest_group_weight; |
1e3c88bde sched: Move load ... |
3329 3330 3331 3332 3333 3334 3335 3336 3337 3338 3339 3340 3341 3342 3343 3344 3345 3346 3347 3348 3349 |
int group_imb; /* Is there imbalance in this sd */ #if defined(CONFIG_SCHED_MC) || defined(CONFIG_SCHED_SMT) int power_savings_balance; /* Is powersave balance needed for this sd */ struct sched_group *group_min; /* Least loaded group in sd */ struct sched_group *group_leader; /* Group which relieves group_min */ unsigned long min_load_per_task; /* load_per_task in group_min */ unsigned long leader_nr_running; /* Nr running of group_leader */ unsigned long min_nr_running; /* Nr running of group_min */ #endif }; /* * sg_lb_stats - stats of a sched_group required for load_balancing */ struct sg_lb_stats { unsigned long avg_load; /*Avg load across the CPUs of the group */ unsigned long group_load; /* Total load over the CPUs of the group */ unsigned long sum_nr_running; /* Nr tasks running in the group */ unsigned long sum_weighted_load; /* Weighted load of group's tasks */ unsigned long group_capacity; |
aae6d3ddd sched: Use group ... |
3350 3351 |
unsigned long idle_cpus; unsigned long group_weight; |
1e3c88bde sched: Move load ... |
3352 |
int group_imb; /* Is there an imbalance in the group ? */ |
fab476228 sched: Force bala... |
3353 |
int group_has_capacity; /* Is there extra capacity in the group? */ |
1e3c88bde sched: Move load ... |
3354 3355 3356 |
}; /** |
1e3c88bde sched: Move load ... |
3357 3358 3359 3360 3361 3362 3363 3364 3365 3366 3367 3368 3369 3370 3371 3372 3373 3374 3375 3376 3377 3378 3379 3380 3381 3382 3383 3384 3385 3386 3387 3388 3389 3390 3391 3392 3393 3394 3395 3396 3397 3398 3399 3400 3401 3402 3403 3404 3405 3406 3407 3408 3409 3410 3411 3412 3413 3414 3415 3416 3417 3418 3419 3420 3421 3422 3423 3424 3425 3426 3427 3428 3429 3430 3431 3432 3433 3434 3435 3436 3437 3438 3439 3440 3441 3442 3443 3444 3445 3446 3447 3448 3449 3450 3451 3452 3453 3454 3455 3456 3457 3458 3459 3460 3461 3462 3463 3464 3465 3466 3467 3468 3469 3470 3471 3472 3473 3474 3475 3476 3477 3478 3479 3480 3481 3482 3483 3484 3485 3486 3487 3488 3489 3490 3491 3492 3493 3494 3495 3496 3497 3498 3499 3500 3501 3502 3503 3504 3505 3506 3507 3508 3509 3510 3511 3512 3513 3514 3515 3516 3517 3518 3519 3520 3521 3522 3523 3524 3525 |
* get_sd_load_idx - Obtain the load index for a given sched domain. * @sd: The sched_domain whose load_idx is to be obtained. * @idle: The Idle status of the CPU for whose sd load_icx is obtained. */ static inline int get_sd_load_idx(struct sched_domain *sd, enum cpu_idle_type idle) { int load_idx; switch (idle) { case CPU_NOT_IDLE: load_idx = sd->busy_idx; break; case CPU_NEWLY_IDLE: load_idx = sd->newidle_idx; break; default: load_idx = sd->idle_idx; break; } return load_idx; } #if defined(CONFIG_SCHED_MC) || defined(CONFIG_SCHED_SMT) /** * init_sd_power_savings_stats - Initialize power savings statistics for * the given sched_domain, during load balancing. * * @sd: Sched domain whose power-savings statistics are to be initialized. * @sds: Variable containing the statistics for sd. * @idle: Idle status of the CPU at which we're performing load-balancing. */ static inline void init_sd_power_savings_stats(struct sched_domain *sd, struct sd_lb_stats *sds, enum cpu_idle_type idle) { /* * Busy processors will not participate in power savings * balance. */ if (idle == CPU_NOT_IDLE || !(sd->flags & SD_POWERSAVINGS_BALANCE)) sds->power_savings_balance = 0; else { sds->power_savings_balance = 1; sds->min_nr_running = ULONG_MAX; sds->leader_nr_running = 0; } } /** * update_sd_power_savings_stats - Update the power saving stats for a * sched_domain while performing load balancing. * * @group: sched_group belonging to the sched_domain under consideration. * @sds: Variable containing the statistics of the sched_domain * @local_group: Does group contain the CPU for which we're performing * load balancing ? * @sgs: Variable containing the statistics of the group. */ static inline void update_sd_power_savings_stats(struct sched_group *group, struct sd_lb_stats *sds, int local_group, struct sg_lb_stats *sgs) { if (!sds->power_savings_balance) return; /* * If the local group is idle or completely loaded * no need to do power savings balance at this domain */ if (local_group && (sds->this_nr_running >= sgs->group_capacity || !sds->this_nr_running)) sds->power_savings_balance = 0; /* * If a group is already running at full capacity or idle, * don't include that group in power savings calculations */ if (!sds->power_savings_balance || sgs->sum_nr_running >= sgs->group_capacity || !sgs->sum_nr_running) return; /* * Calculate the group which has the least non-idle load. * This is the group from where we need to pick up the load * for saving power */ if ((sgs->sum_nr_running < sds->min_nr_running) || (sgs->sum_nr_running == sds->min_nr_running && group_first_cpu(group) > group_first_cpu(sds->group_min))) { sds->group_min = group; sds->min_nr_running = sgs->sum_nr_running; sds->min_load_per_task = sgs->sum_weighted_load / sgs->sum_nr_running; } /* * Calculate the group which is almost near its * capacity but still has some space to pick up some load * from other group and save more power */ if (sgs->sum_nr_running + 1 > sgs->group_capacity) return; if (sgs->sum_nr_running > sds->leader_nr_running || (sgs->sum_nr_running == sds->leader_nr_running && group_first_cpu(group) < group_first_cpu(sds->group_leader))) { sds->group_leader = group; sds->leader_nr_running = sgs->sum_nr_running; } } /** * check_power_save_busiest_group - see if there is potential for some power-savings balance * @sds: Variable containing the statistics of the sched_domain * under consideration. * @this_cpu: Cpu at which we're currently performing load-balancing. * @imbalance: Variable to store the imbalance. * * Description: * Check if we have potential to perform some power-savings balance. * If yes, set the busiest group to be the least loaded group in the * sched_domain, so that it's CPUs can be put to idle. * * Returns 1 if there is potential to perform power-savings balance. * Else returns 0. */ static inline int check_power_save_busiest_group(struct sd_lb_stats *sds, int this_cpu, unsigned long *imbalance) { if (!sds->power_savings_balance) return 0; if (sds->this != sds->group_leader || sds->group_leader == sds->group_min) return 0; *imbalance = sds->min_load_per_task; sds->busiest = sds->group_min; return 1; } #else /* CONFIG_SCHED_MC || CONFIG_SCHED_SMT */ static inline void init_sd_power_savings_stats(struct sched_domain *sd, struct sd_lb_stats *sds, enum cpu_idle_type idle) { return; } static inline void update_sd_power_savings_stats(struct sched_group *group, struct sd_lb_stats *sds, int local_group, struct sg_lb_stats *sgs) { return; } static inline int check_power_save_busiest_group(struct sd_lb_stats *sds, int this_cpu, unsigned long *imbalance) { return 0; } #endif /* CONFIG_SCHED_MC || CONFIG_SCHED_SMT */ unsigned long default_scale_freq_power(struct sched_domain *sd, int cpu) { |
1399fa780 sched: Introduce ... |
3526 |
return SCHED_POWER_SCALE; |
1e3c88bde sched: Move load ... |
3527 3528 3529 3530 3531 3532 3533 3534 3535 |
} unsigned long __weak arch_scale_freq_power(struct sched_domain *sd, int cpu) { return default_scale_freq_power(sd, cpu); } unsigned long default_scale_smt_power(struct sched_domain *sd, int cpu) { |
669c55e9f sched: Pre-comput... |
3536 |
unsigned long weight = sd->span_weight; |
1e3c88bde sched: Move load ... |
3537 3538 3539 3540 3541 3542 3543 3544 3545 3546 3547 3548 3549 3550 3551 3552 |
unsigned long smt_gain = sd->smt_gain; smt_gain /= weight; return smt_gain; } unsigned long __weak arch_scale_smt_power(struct sched_domain *sd, int cpu) { return default_scale_smt_power(sd, cpu); } unsigned long scale_rt_power(int cpu) { struct rq *rq = cpu_rq(cpu); u64 total, available; |
1e3c88bde sched: Move load ... |
3553 |
total = sched_avg_period() + (rq->clock - rq->age_stamp); |
aa4838085 sched: Remove irq... |
3554 3555 3556 3557 3558 3559 3560 |
if (unlikely(total < rq->rt_avg)) { /* Ensures that power won't end up being negative */ available = 0; } else { available = total - rq->rt_avg; } |
1e3c88bde sched: Move load ... |
3561 |
|
1399fa780 sched: Introduce ... |
3562 3563 |
if (unlikely((s64)total < SCHED_POWER_SCALE)) total = SCHED_POWER_SCALE; |
1e3c88bde sched: Move load ... |
3564 |
|
1399fa780 sched: Introduce ... |
3565 |
total >>= SCHED_POWER_SHIFT; |
1e3c88bde sched: Move load ... |
3566 3567 3568 3569 3570 3571 |
return div_u64(available, total); } static void update_cpu_power(struct sched_domain *sd, int cpu) { |
669c55e9f sched: Pre-comput... |
3572 |
unsigned long weight = sd->span_weight; |
1399fa780 sched: Introduce ... |
3573 |
unsigned long power = SCHED_POWER_SCALE; |
1e3c88bde sched: Move load ... |
3574 |
struct sched_group *sdg = sd->groups; |
1e3c88bde sched: Move load ... |
3575 3576 3577 3578 3579 |
if ((sd->flags & SD_SHARE_CPUPOWER) && weight > 1) { if (sched_feat(ARCH_POWER)) power *= arch_scale_smt_power(sd, cpu); else power *= default_scale_smt_power(sd, cpu); |
1399fa780 sched: Introduce ... |
3580 |
power >>= SCHED_POWER_SHIFT; |
1e3c88bde sched: Move load ... |
3581 |
} |
9c3f75cbd sched: Break out ... |
3582 |
sdg->sgp->power_orig = power; |
9d5efe05e sched: Fix capaci... |
3583 3584 3585 3586 3587 |
if (sched_feat(ARCH_POWER)) power *= arch_scale_freq_power(sd, cpu); else power *= default_scale_freq_power(sd, cpu); |
1399fa780 sched: Introduce ... |
3588 |
power >>= SCHED_POWER_SHIFT; |
9d5efe05e sched: Fix capaci... |
3589 |
|
1e3c88bde sched: Move load ... |
3590 |
power *= scale_rt_power(cpu); |
1399fa780 sched: Introduce ... |
3591 |
power >>= SCHED_POWER_SHIFT; |
1e3c88bde sched: Move load ... |
3592 3593 3594 |
if (!power) power = 1; |
e51fd5e22 sched: Fix wake_a... |
3595 |
cpu_rq(cpu)->cpu_power = power; |
9c3f75cbd sched: Break out ... |
3596 |
sdg->sgp->power = power; |
1e3c88bde sched: Move load ... |
3597 |
} |
029632fbb sched: Make separ... |
3598 |
void update_group_power(struct sched_domain *sd, int cpu) |
1e3c88bde sched: Move load ... |
3599 3600 3601 3602 3603 3604 3605 3606 3607 3608 3609 3610 3611 3612 |
{ struct sched_domain *child = sd->child; struct sched_group *group, *sdg = sd->groups; unsigned long power; if (!child) { update_cpu_power(sd, cpu); return; } power = 0; group = child->groups; do { |
9c3f75cbd sched: Break out ... |
3613 |
power += group->sgp->power; |
1e3c88bde sched: Move load ... |
3614 3615 |
group = group->next; } while (group != child->groups); |
9c3f75cbd sched: Break out ... |
3616 |
sdg->sgp->power = power; |
1e3c88bde sched: Move load ... |
3617 |
} |
9d5efe05e sched: Fix capaci... |
3618 3619 3620 3621 3622 3623 3624 3625 3626 3627 3628 |
/* * Try and fix up capacity for tiny siblings, this is needed when * things like SD_ASYM_PACKING need f_b_g to select another sibling * which on its own isn't powerful enough. * * See update_sd_pick_busiest() and check_asym_packing(). */ static inline int fix_small_capacity(struct sched_domain *sd, struct sched_group *group) { /* |
1399fa780 sched: Introduce ... |
3629 |
* Only siblings can have significantly less than SCHED_POWER_SCALE |
9d5efe05e sched: Fix capaci... |
3630 |
*/ |
a6c75f2f8 sched: Avoid usin... |
3631 |
if (!(sd->flags & SD_SHARE_CPUPOWER)) |
9d5efe05e sched: Fix capaci... |
3632 3633 3634 3635 3636 |
return 0; /* * If ~90% of the cpu_power is still there, we're good. */ |
9c3f75cbd sched: Break out ... |
3637 |
if (group->sgp->power * 32 > group->sgp->power_orig * 29) |
9d5efe05e sched: Fix capaci... |
3638 3639 3640 3641 |
return 1; return 0; } |
1e3c88bde sched: Move load ... |
3642 3643 3644 3645 3646 3647 3648 |
/** * update_sg_lb_stats - Update sched_group's statistics for load balancing. * @sd: The sched_domain whose statistics are to be updated. * @group: sched_group whose statistics are to be updated. * @this_cpu: Cpu for which load balance is currently performed. * @idle: Idle status of this_cpu * @load_idx: Load index of sched_domain of this_cpu for load calc. |
1e3c88bde sched: Move load ... |
3649 3650 3651 3652 3653 3654 3655 |
* @local_group: Does group contain this_cpu. * @cpus: Set of cpus considered for load balancing. * @balance: Should we balance. * @sgs: variable to hold the statistics for this group. */ static inline void update_sg_lb_stats(struct sched_domain *sd, struct sched_group *group, int this_cpu, |
46e49b383 sched: Wholesale ... |
3656 |
enum cpu_idle_type idle, int load_idx, |
1e3c88bde sched: Move load ... |
3657 3658 3659 |
int local_group, const struct cpumask *cpus, int *balance, struct sg_lb_stats *sgs) { |
2582f0eba sched: Set group_... |
3660 |
unsigned long load, max_cpu_load, min_cpu_load, max_nr_running; |
1e3c88bde sched: Move load ... |
3661 3662 |
int i; unsigned int balance_cpu = -1, first_idle_cpu = 0; |
dd5feea14 sched: Fix SCHED_... |
3663 |
unsigned long avg_load_per_task = 0; |
1e3c88bde sched: Move load ... |
3664 |
|
871e35bc9 sched: Fix the pl... |
3665 |
if (local_group) |
1e3c88bde sched: Move load ... |
3666 |
balance_cpu = group_first_cpu(group); |
1e3c88bde sched: Move load ... |
3667 3668 |
/* Tally up the load of all CPUs in the group */ |
1e3c88bde sched: Move load ... |
3669 3670 |
max_cpu_load = 0; min_cpu_load = ~0UL; |
2582f0eba sched: Set group_... |
3671 |
max_nr_running = 0; |
1e3c88bde sched: Move load ... |
3672 3673 3674 |
for_each_cpu_and(i, sched_group_cpus(group), cpus) { struct rq *rq = cpu_rq(i); |
1e3c88bde sched: Move load ... |
3675 3676 3677 3678 3679 3680 3681 3682 3683 3684 |
/* Bias balancing toward cpus of our domain */ if (local_group) { if (idle_cpu(i) && !first_idle_cpu) { first_idle_cpu = 1; balance_cpu = i; } load = target_load(i, load_idx); } else { load = source_load(i, load_idx); |
2582f0eba sched: Set group_... |
3685 |
if (load > max_cpu_load) { |
1e3c88bde sched: Move load ... |
3686 |
max_cpu_load = load; |
2582f0eba sched: Set group_... |
3687 3688 |
max_nr_running = rq->nr_running; } |
1e3c88bde sched: Move load ... |
3689 3690 3691 3692 3693 3694 3695 |
if (min_cpu_load > load) min_cpu_load = load; } sgs->group_load += load; sgs->sum_nr_running += rq->nr_running; sgs->sum_weighted_load += weighted_cpuload(i); |
aae6d3ddd sched: Use group ... |
3696 3697 |
if (idle_cpu(i)) sgs->idle_cpus++; |
1e3c88bde sched: Move load ... |
3698 3699 3700 3701 3702 3703 3704 3705 |
} /* * First idle cpu or the first cpu(busiest) in this sched group * is eligible for doing load balancing at this and above * domains. In the newly idle case, we will allow all the cpu's * to do the newly idle load balance. */ |
bbc8cb5ba sched: Reduce upd... |
3706 3707 3708 3709 3710 3711 |
if (idle != CPU_NEWLY_IDLE && local_group) { if (balance_cpu != this_cpu) { *balance = 0; return; } update_group_power(sd, this_cpu); |
1e3c88bde sched: Move load ... |
3712 3713 3714 |
} /* Adjust by relative CPU power of the group */ |
9c3f75cbd sched: Break out ... |
3715 |
sgs->avg_load = (sgs->group_load*SCHED_POWER_SCALE) / group->sgp->power; |
1e3c88bde sched: Move load ... |
3716 |
|
1e3c88bde sched: Move load ... |
3717 3718 |
/* * Consider the group unbalanced when the imbalance is larger |
866ab43ef sched: Fix the gr... |
3719 |
* than the average weight of a task. |
1e3c88bde sched: Move load ... |
3720 3721 3722 3723 3724 3725 |
* * APZ: with cgroup the avg task weight can vary wildly and * might not be a suitable number - should we keep a * normalized nr_running number somewhere that negates * the hierarchy? */ |
dd5feea14 sched: Fix SCHED_... |
3726 3727 |
if (sgs->sum_nr_running) avg_load_per_task = sgs->sum_weighted_load / sgs->sum_nr_running; |
1e3c88bde sched: Move load ... |
3728 |
|
866ab43ef sched: Fix the gr... |
3729 |
if ((max_cpu_load - min_cpu_load) >= avg_load_per_task && max_nr_running > 1) |
1e3c88bde sched: Move load ... |
3730 |
sgs->group_imb = 1; |
9c3f75cbd sched: Break out ... |
3731 |
sgs->group_capacity = DIV_ROUND_CLOSEST(group->sgp->power, |
1399fa780 sched: Introduce ... |
3732 |
SCHED_POWER_SCALE); |
9d5efe05e sched: Fix capaci... |
3733 3734 |
if (!sgs->group_capacity) sgs->group_capacity = fix_small_capacity(sd, group); |
aae6d3ddd sched: Use group ... |
3735 |
sgs->group_weight = group->group_weight; |
fab476228 sched: Force bala... |
3736 3737 3738 |
if (sgs->group_capacity > sgs->sum_nr_running) sgs->group_has_capacity = 1; |
1e3c88bde sched: Move load ... |
3739 3740 3741 |
} /** |
532cb4c40 sched: Add asymme... |
3742 3743 3744 3745 |
* update_sd_pick_busiest - return 1 on busiest group * @sd: sched_domain whose statistics are to be checked * @sds: sched_domain statistics * @sg: sched_group candidate to be checked for being the busiest |
b6b122944 sched: Fix commen... |
3746 3747 |
* @sgs: sched_group statistics * @this_cpu: the current cpu |
532cb4c40 sched: Add asymme... |
3748 3749 3750 3751 3752 3753 3754 3755 3756 3757 3758 3759 3760 3761 3762 3763 3764 3765 3766 3767 3768 3769 3770 3771 3772 3773 3774 3775 3776 3777 3778 3779 3780 3781 3782 3783 3784 |
* * Determine if @sg is a busier group than the previously selected * busiest group. */ static bool update_sd_pick_busiest(struct sched_domain *sd, struct sd_lb_stats *sds, struct sched_group *sg, struct sg_lb_stats *sgs, int this_cpu) { if (sgs->avg_load <= sds->max_load) return false; if (sgs->sum_nr_running > sgs->group_capacity) return true; if (sgs->group_imb) return true; /* * ASYM_PACKING needs to move all the work to the lowest * numbered CPUs in the group, therefore mark all groups * higher than ourself as busy. */ if ((sd->flags & SD_ASYM_PACKING) && sgs->sum_nr_running && this_cpu < group_first_cpu(sg)) { if (!sds->busiest) return true; if (group_first_cpu(sds->busiest) > group_first_cpu(sg)) return true; } return false; } /** |
461819ac8 sched_fair: Fix a... |
3785 |
* update_sd_lb_stats - Update sched_domain's statistics for load balancing. |
1e3c88bde sched: Move load ... |
3786 3787 3788 |
* @sd: sched_domain whose statistics are to be updated. * @this_cpu: Cpu for which load balance is currently performed. * @idle: Idle status of this_cpu |
1e3c88bde sched: Move load ... |
3789 3790 3791 3792 3793 |
* @cpus: Set of cpus considered for load balancing. * @balance: Should we balance. * @sds: variable to hold the statistics for this sched_domain. */ static inline void update_sd_lb_stats(struct sched_domain *sd, int this_cpu, |
46e49b383 sched: Wholesale ... |
3794 3795 |
enum cpu_idle_type idle, const struct cpumask *cpus, int *balance, struct sd_lb_stats *sds) |
1e3c88bde sched: Move load ... |
3796 3797 |
{ struct sched_domain *child = sd->child; |
532cb4c40 sched: Add asymme... |
3798 |
struct sched_group *sg = sd->groups; |
1e3c88bde sched: Move load ... |
3799 3800 3801 3802 3803 3804 3805 3806 3807 3808 3809 |
struct sg_lb_stats sgs; int load_idx, prefer_sibling = 0; if (child && child->flags & SD_PREFER_SIBLING) prefer_sibling = 1; init_sd_power_savings_stats(sd, sds, idle); load_idx = get_sd_load_idx(sd, idle); do { int local_group; |
532cb4c40 sched: Add asymme... |
3810 |
local_group = cpumask_test_cpu(this_cpu, sched_group_cpus(sg)); |
1e3c88bde sched: Move load ... |
3811 |
memset(&sgs, 0, sizeof(sgs)); |
46e49b383 sched: Wholesale ... |
3812 |
update_sg_lb_stats(sd, sg, this_cpu, idle, load_idx, |
1e3c88bde sched: Move load ... |
3813 |
local_group, cpus, balance, &sgs); |
8f190fb3f sched: Assume *ba... |
3814 |
if (local_group && !(*balance)) |
1e3c88bde sched: Move load ... |
3815 3816 3817 |
return; sds->total_load += sgs.group_load; |
9c3f75cbd sched: Break out ... |
3818 |
sds->total_pwr += sg->sgp->power; |
1e3c88bde sched: Move load ... |
3819 3820 3821 |
/* * In case the child domain prefers tasks go to siblings |
532cb4c40 sched: Add asymme... |
3822 |
* first, lower the sg capacity to one so that we'll try |
75dd321d7 sched: Drop group... |
3823 3824 3825 3826 3827 3828 |
* and move all the excess tasks away. We lower the capacity * of a group only if the local group has the capacity to fit * these excess tasks, i.e. nr_running < group_capacity. The * extra check prevents the case where you always pull from the * heaviest group when it is already under-utilized (possible * with a large weight task outweighs the tasks on the system). |
1e3c88bde sched: Move load ... |
3829 |
*/ |
75dd321d7 sched: Drop group... |
3830 |
if (prefer_sibling && !local_group && sds->this_has_capacity) |
1e3c88bde sched: Move load ... |
3831 3832 3833 3834 |
sgs.group_capacity = min(sgs.group_capacity, 1UL); if (local_group) { sds->this_load = sgs.avg_load; |
532cb4c40 sched: Add asymme... |
3835 |
sds->this = sg; |
1e3c88bde sched: Move load ... |
3836 3837 |
sds->this_nr_running = sgs.sum_nr_running; sds->this_load_per_task = sgs.sum_weighted_load; |
fab476228 sched: Force bala... |
3838 |
sds->this_has_capacity = sgs.group_has_capacity; |
aae6d3ddd sched: Use group ... |
3839 |
sds->this_idle_cpus = sgs.idle_cpus; |
532cb4c40 sched: Add asymme... |
3840 |
} else if (update_sd_pick_busiest(sd, sds, sg, &sgs, this_cpu)) { |
1e3c88bde sched: Move load ... |
3841 |
sds->max_load = sgs.avg_load; |
532cb4c40 sched: Add asymme... |
3842 |
sds->busiest = sg; |
1e3c88bde sched: Move load ... |
3843 |
sds->busiest_nr_running = sgs.sum_nr_running; |
aae6d3ddd sched: Use group ... |
3844 |
sds->busiest_idle_cpus = sgs.idle_cpus; |
dd5feea14 sched: Fix SCHED_... |
3845 |
sds->busiest_group_capacity = sgs.group_capacity; |
1e3c88bde sched: Move load ... |
3846 |
sds->busiest_load_per_task = sgs.sum_weighted_load; |
fab476228 sched: Force bala... |
3847 |
sds->busiest_has_capacity = sgs.group_has_capacity; |
aae6d3ddd sched: Use group ... |
3848 |
sds->busiest_group_weight = sgs.group_weight; |
1e3c88bde sched: Move load ... |
3849 3850 |
sds->group_imb = sgs.group_imb; } |
532cb4c40 sched: Add asymme... |
3851 3852 3853 3854 |
update_sd_power_savings_stats(sg, sds, local_group, &sgs); sg = sg->next; } while (sg != sd->groups); } |
532cb4c40 sched: Add asymme... |
3855 3856 3857 3858 3859 3860 3861 3862 3863 3864 3865 3866 3867 3868 3869 3870 3871 |
/** * check_asym_packing - Check to see if the group is packed into the * sched doman. * * This is primarily intended to used at the sibling level. Some * cores like POWER7 prefer to use lower numbered SMT threads. In the * case of POWER7, it can move to lower SMT modes only when higher * threads are idle. When in lower SMT modes, the threads will * perform better since they share less core resources. Hence when we * have idle threads, we want them to be the higher ones. * * This packing function is run on idle threads. It checks to see if * the busiest CPU in this domain (core in the P7 case) has a higher * CPU number than the packing function is being run on. Here we are * assuming lower CPU number will be equivalent to lower a SMT thread * number. * |
b6b122944 sched: Fix commen... |
3872 3873 3874 |
* Returns 1 when packing is required and a task should be moved to * this CPU. The amount of the imbalance is returned in *imbalance. * |
532cb4c40 sched: Add asymme... |
3875 3876 3877 3878 |
* @sd: The sched_domain whose packing is to be checked. * @sds: Statistics of the sched_domain which is to be packed * @this_cpu: The cpu at whose sched_domain we're performing load-balance. * @imbalance: returns amount of imbalanced due to packing. |
532cb4c40 sched: Add asymme... |
3879 3880 3881 3882 3883 3884 3885 3886 3887 3888 3889 3890 3891 3892 3893 3894 |
*/ static int check_asym_packing(struct sched_domain *sd, struct sd_lb_stats *sds, int this_cpu, unsigned long *imbalance) { int busiest_cpu; if (!(sd->flags & SD_ASYM_PACKING)) return 0; if (!sds->busiest) return 0; busiest_cpu = group_first_cpu(sds->busiest); if (this_cpu > busiest_cpu) return 0; |
9c3f75cbd sched: Break out ... |
3895 |
*imbalance = DIV_ROUND_CLOSEST(sds->max_load * sds->busiest->sgp->power, |
1399fa780 sched: Introduce ... |
3896 |
SCHED_POWER_SCALE); |
532cb4c40 sched: Add asymme... |
3897 |
return 1; |
1e3c88bde sched: Move load ... |
3898 3899 3900 3901 3902 3903 3904 3905 3906 3907 3908 3909 3910 3911 3912 |
} /** * fix_small_imbalance - Calculate the minor imbalance that exists * amongst the groups of a sched_domain, during * load balancing. * @sds: Statistics of the sched_domain whose imbalance is to be calculated. * @this_cpu: The cpu at whose sched_domain we're performing load-balance. * @imbalance: Variable to store the imbalance. */ static inline void fix_small_imbalance(struct sd_lb_stats *sds, int this_cpu, unsigned long *imbalance) { unsigned long tmp, pwr_now = 0, pwr_move = 0; unsigned int imbn = 2; |
dd5feea14 sched: Fix SCHED_... |
3913 |
unsigned long scaled_busy_load_per_task; |
1e3c88bde sched: Move load ... |
3914 3915 3916 3917 3918 3919 3920 3921 3922 |
if (sds->this_nr_running) { sds->this_load_per_task /= sds->this_nr_running; if (sds->busiest_load_per_task > sds->this_load_per_task) imbn = 1; } else sds->this_load_per_task = cpu_avg_load_per_task(this_cpu); |
dd5feea14 sched: Fix SCHED_... |
3923 |
scaled_busy_load_per_task = sds->busiest_load_per_task |
1399fa780 sched: Introduce ... |
3924 |
* SCHED_POWER_SCALE; |
9c3f75cbd sched: Break out ... |
3925 |
scaled_busy_load_per_task /= sds->busiest->sgp->power; |
dd5feea14 sched: Fix SCHED_... |
3926 3927 3928 |
if (sds->max_load - sds->this_load + scaled_busy_load_per_task >= (scaled_busy_load_per_task * imbn)) { |
1e3c88bde sched: Move load ... |
3929 3930 3931 3932 3933 3934 3935 3936 3937 |
*imbalance = sds->busiest_load_per_task; return; } /* * OK, we don't have enough imbalance to justify moving tasks, * however we may be able to increase total CPU power used by * moving them. */ |
9c3f75cbd sched: Break out ... |
3938 |
pwr_now += sds->busiest->sgp->power * |
1e3c88bde sched: Move load ... |
3939 |
min(sds->busiest_load_per_task, sds->max_load); |
9c3f75cbd sched: Break out ... |
3940 |
pwr_now += sds->this->sgp->power * |
1e3c88bde sched: Move load ... |
3941 |
min(sds->this_load_per_task, sds->this_load); |
1399fa780 sched: Introduce ... |
3942 |
pwr_now /= SCHED_POWER_SCALE; |
1e3c88bde sched: Move load ... |
3943 3944 |
/* Amount of load we'd subtract */ |
1399fa780 sched: Introduce ... |
3945 |
tmp = (sds->busiest_load_per_task * SCHED_POWER_SCALE) / |
9c3f75cbd sched: Break out ... |
3946 |
sds->busiest->sgp->power; |
1e3c88bde sched: Move load ... |
3947 |
if (sds->max_load > tmp) |
9c3f75cbd sched: Break out ... |
3948 |
pwr_move += sds->busiest->sgp->power * |
1e3c88bde sched: Move load ... |
3949 3950 3951 |
min(sds->busiest_load_per_task, sds->max_load - tmp); /* Amount of load we'd add */ |
9c3f75cbd sched: Break out ... |
3952 |
if (sds->max_load * sds->busiest->sgp->power < |
1399fa780 sched: Introduce ... |
3953 |
sds->busiest_load_per_task * SCHED_POWER_SCALE) |
9c3f75cbd sched: Break out ... |
3954 3955 |
tmp = (sds->max_load * sds->busiest->sgp->power) / sds->this->sgp->power; |
1e3c88bde sched: Move load ... |
3956 |
else |
1399fa780 sched: Introduce ... |
3957 |
tmp = (sds->busiest_load_per_task * SCHED_POWER_SCALE) / |
9c3f75cbd sched: Break out ... |
3958 3959 |
sds->this->sgp->power; pwr_move += sds->this->sgp->power * |
1e3c88bde sched: Move load ... |
3960 |
min(sds->this_load_per_task, sds->this_load + tmp); |
1399fa780 sched: Introduce ... |
3961 |
pwr_move /= SCHED_POWER_SCALE; |
1e3c88bde sched: Move load ... |
3962 3963 3964 3965 3966 3967 3968 3969 3970 3971 3972 3973 3974 3975 3976 3977 |
/* Move if we gain throughput */ if (pwr_move > pwr_now) *imbalance = sds->busiest_load_per_task; } /** * calculate_imbalance - Calculate the amount of imbalance present within the * groups of a given sched_domain during load balance. * @sds: statistics of the sched_domain whose imbalance is to be calculated. * @this_cpu: Cpu for which currently load balance is being performed. * @imbalance: The variable to store the imbalance. */ static inline void calculate_imbalance(struct sd_lb_stats *sds, int this_cpu, unsigned long *imbalance) { |
dd5feea14 sched: Fix SCHED_... |
3978 3979 3980 3981 3982 3983 3984 |
unsigned long max_pull, load_above_capacity = ~0UL; sds->busiest_load_per_task /= sds->busiest_nr_running; if (sds->group_imb) { sds->busiest_load_per_task = min(sds->busiest_load_per_task, sds->avg_load); } |
1e3c88bde sched: Move load ... |
3985 3986 3987 3988 3989 3990 3991 3992 3993 |
/* * In the presence of smp nice balancing, certain scenarios can have * max load less than avg load(as we skip the groups at or below * its cpu_power, while calculating max_load..) */ if (sds->max_load < sds->avg_load) { *imbalance = 0; return fix_small_imbalance(sds, this_cpu, imbalance); } |
dd5feea14 sched: Fix SCHED_... |
3994 3995 3996 3997 3998 3999 |
if (!sds->group_imb) { /* * Don't want to pull so many tasks that a group would go idle. */ load_above_capacity = (sds->busiest_nr_running - sds->busiest_group_capacity); |
1399fa780 sched: Introduce ... |
4000 |
load_above_capacity *= (SCHED_LOAD_SCALE * SCHED_POWER_SCALE); |
dd5feea14 sched: Fix SCHED_... |
4001 |
|
9c3f75cbd sched: Break out ... |
4002 |
load_above_capacity /= sds->busiest->sgp->power; |
dd5feea14 sched: Fix SCHED_... |
4003 4004 4005 4006 4007 4008 4009 4010 4011 4012 4013 4014 4015 |
} /* * We're trying to get all the cpus to the average_load, so we don't * want to push ourselves above the average load, nor do we wish to * reduce the max loaded cpu below the average load. At the same time, * we also don't want to reduce the group load below the group capacity * (so that we can implement power-savings policies etc). Thus we look * for the minimum possible imbalance. * Be careful of negative numbers as they'll appear as very large values * with unsigned longs. */ max_pull = min(sds->max_load - sds->avg_load, load_above_capacity); |
1e3c88bde sched: Move load ... |
4016 4017 |
/* How much load to actually move to equalise the imbalance */ |
9c3f75cbd sched: Break out ... |
4018 4019 |
*imbalance = min(max_pull * sds->busiest->sgp->power, (sds->avg_load - sds->this_load) * sds->this->sgp->power) |
1399fa780 sched: Introduce ... |
4020 |
/ SCHED_POWER_SCALE; |
1e3c88bde sched: Move load ... |
4021 4022 4023 |
/* * if *imbalance is less than the average load per runnable task |
25985edce Fix common misspe... |
4024 |
* there is no guarantee that any tasks will be moved so we'll have |
1e3c88bde sched: Move load ... |
4025 4026 4027 4028 4029 4030 4031 |
* a think about bumping its value to force at least one task to be * moved */ if (*imbalance < sds->busiest_load_per_task) return fix_small_imbalance(sds, this_cpu, imbalance); } |
fab476228 sched: Force bala... |
4032 |
|
1e3c88bde sched: Move load ... |
4033 4034 4035 4036 4037 4038 4039 4040 4041 4042 4043 4044 4045 4046 4047 4048 4049 |
/******* find_busiest_group() helpers end here *********************/ /** * find_busiest_group - Returns the busiest group within the sched_domain * if there is an imbalance. If there isn't an imbalance, and * the user has opted for power-savings, it returns a group whose * CPUs can be put to idle by rebalancing those tasks elsewhere, if * such a group exists. * * Also calculates the amount of weighted load which should be moved * to restore balance. * * @sd: The sched_domain whose busiest group is to be returned. * @this_cpu: The cpu for which load balancing is currently being performed. * @imbalance: Variable which stores amount of weighted load which should * be moved to restore balance/put a group to idle. * @idle: The idle status of this_cpu. |
1e3c88bde sched: Move load ... |
4050 4051 4052 4053 4054 4055 4056 4057 4058 4059 4060 4061 |
* @cpus: The set of CPUs under consideration for load-balancing. * @balance: Pointer to a variable indicating if this_cpu * is the appropriate cpu to perform load balancing at this_level. * * Returns: - the busiest group if imbalance exists. * - If no imbalance and user has opted for power-savings balance, * return the least loaded group whose CPUs can be * put to idle by rebalancing its tasks onto our group. */ static struct sched_group * find_busiest_group(struct sched_domain *sd, int this_cpu, unsigned long *imbalance, enum cpu_idle_type idle, |
46e49b383 sched: Wholesale ... |
4062 |
const struct cpumask *cpus, int *balance) |
1e3c88bde sched: Move load ... |
4063 4064 4065 4066 4067 4068 4069 4070 4071 |
{ struct sd_lb_stats sds; memset(&sds, 0, sizeof(sds)); /* * Compute the various statistics relavent for load balancing at * this level. */ |
46e49b383 sched: Wholesale ... |
4072 |
update_sd_lb_stats(sd, this_cpu, idle, cpus, balance, &sds); |
1e3c88bde sched: Move load ... |
4073 |
|
cc57aa8f4 sched: Clean up s... |
4074 4075 4076 |
/* * this_cpu is not the appropriate cpu to perform load balancing at * this level. |
1e3c88bde sched: Move load ... |
4077 |
*/ |
8f190fb3f sched: Assume *ba... |
4078 |
if (!(*balance)) |
1e3c88bde sched: Move load ... |
4079 |
goto ret; |
532cb4c40 sched: Add asymme... |
4080 4081 4082 |
if ((idle == CPU_IDLE || idle == CPU_NEWLY_IDLE) && check_asym_packing(sd, &sds, this_cpu, imbalance)) return sds.busiest; |
cc57aa8f4 sched: Clean up s... |
4083 |
/* There is no busy sibling group to pull tasks from */ |
1e3c88bde sched: Move load ... |
4084 4085 |
if (!sds.busiest || sds.busiest_nr_running == 0) goto out_balanced; |
1399fa780 sched: Introduce ... |
4086 |
sds.avg_load = (SCHED_POWER_SCALE * sds.total_load) / sds.total_pwr; |
b0432d8f1 sched: Fix sched-... |
4087 |
|
866ab43ef sched: Fix the gr... |
4088 4089 4090 4091 4092 4093 4094 |
/* * If the busiest group is imbalanced the below checks don't * work because they assumes all things are equal, which typically * isn't true due to cpus_allowed constraints and the like. */ if (sds.group_imb) goto force_balance; |
cc57aa8f4 sched: Clean up s... |
4095 |
/* SD_BALANCE_NEWIDLE trumps SMP nice when underutilized */ |
fab476228 sched: Force bala... |
4096 4097 4098 |
if (idle == CPU_NEWLY_IDLE && sds.this_has_capacity && !sds.busiest_has_capacity) goto force_balance; |
cc57aa8f4 sched: Clean up s... |
4099 4100 4101 4102 |
/* * If the local group is more busy than the selected busiest group * don't try and pull any tasks. */ |
1e3c88bde sched: Move load ... |
4103 4104 |
if (sds.this_load >= sds.max_load) goto out_balanced; |
cc57aa8f4 sched: Clean up s... |
4105 4106 4107 4108 |
/* * Don't pull any tasks if this group is already above the domain * average load. */ |
1e3c88bde sched: Move load ... |
4109 4110 |
if (sds.this_load >= sds.avg_load) goto out_balanced; |
c186fafe9 sched: Clean up r... |
4111 |
if (idle == CPU_IDLE) { |
aae6d3ddd sched: Use group ... |
4112 4113 4114 4115 4116 4117 |
/* * This cpu is idle. If the busiest group load doesn't * have more tasks than the number of available cpu's and * there is no imbalance between this and busiest group * wrt to idle cpu's, it is balanced. */ |
c186fafe9 sched: Clean up r... |
4118 |
if ((sds.this_idle_cpus <= sds.busiest_idle_cpus + 1) && |
aae6d3ddd sched: Use group ... |
4119 4120 |
sds.busiest_nr_running <= sds.busiest_group_weight) goto out_balanced; |
c186fafe9 sched: Clean up r... |
4121 4122 4123 4124 4125 4126 4127 |
} else { /* * In the CPU_NEWLY_IDLE, CPU_NOT_IDLE cases, use * imbalance_pct to be conservative. */ if (100 * sds.max_load <= sd->imbalance_pct * sds.this_load) goto out_balanced; |
aae6d3ddd sched: Use group ... |
4128 |
} |
1e3c88bde sched: Move load ... |
4129 |
|
fab476228 sched: Force bala... |
4130 |
force_balance: |
1e3c88bde sched: Move load ... |
4131 4132 4133 4134 4135 4136 4137 4138 4139 4140 4141 4142 4143 4144 4145 4146 4147 4148 4149 4150 |
/* Looks like there is an imbalance. Compute it */ calculate_imbalance(&sds, this_cpu, imbalance); return sds.busiest; out_balanced: /* * There is no obvious imbalance. But check if we can do some balancing * to save power. */ if (check_power_save_busiest_group(&sds, this_cpu, imbalance)) return sds.busiest; ret: *imbalance = 0; return NULL; } /* * find_busiest_queue - find the busiest runqueue among the cpus in group. */ static struct rq * |
9d5efe05e sched: Fix capaci... |
4151 4152 4153 |
find_busiest_queue(struct sched_domain *sd, struct sched_group *group, enum cpu_idle_type idle, unsigned long imbalance, const struct cpumask *cpus) |
1e3c88bde sched: Move load ... |
4154 4155 4156 4157 4158 4159 4160 |
{ struct rq *busiest = NULL, *rq; unsigned long max_load = 0; int i; for_each_cpu(i, sched_group_cpus(group)) { unsigned long power = power_of(i); |
1399fa780 sched: Introduce ... |
4161 4162 |
unsigned long capacity = DIV_ROUND_CLOSEST(power, SCHED_POWER_SCALE); |
1e3c88bde sched: Move load ... |
4163 |
unsigned long wl; |
9d5efe05e sched: Fix capaci... |
4164 4165 |
if (!capacity) capacity = fix_small_capacity(sd, group); |
1e3c88bde sched: Move load ... |
4166 4167 4168 4169 |
if (!cpumask_test_cpu(i, cpus)) continue; rq = cpu_rq(i); |
6e40f5bbb Merge branch 'sch... |
4170 |
wl = weighted_cpuload(i); |
1e3c88bde sched: Move load ... |
4171 |
|
6e40f5bbb Merge branch 'sch... |
4172 4173 4174 4175 |
/* * When comparing with imbalance, use weighted_cpuload() * which is not scaled with the cpu power. */ |
1e3c88bde sched: Move load ... |
4176 4177 |
if (capacity && rq->nr_running == 1 && wl > imbalance) continue; |
6e40f5bbb Merge branch 'sch... |
4178 4179 4180 4181 4182 4183 |
/* * For the load comparisons with the other cpu's, consider * the weighted_cpuload() scaled with the cpu power, so that * the load can be moved away from the cpu that is potentially * running at a lower capacity. */ |
1399fa780 sched: Introduce ... |
4184 |
wl = (wl * SCHED_POWER_SCALE) / power; |
6e40f5bbb Merge branch 'sch... |
4185 |
|
1e3c88bde sched: Move load ... |
4186 4187 4188 4189 4190 4191 4192 4193 4194 4195 4196 4197 4198 4199 4200 4201 |
if (wl > max_load) { max_load = wl; busiest = rq; } } return busiest; } /* * Max backoff if we encounter pinned tasks. Pretty arbitrary value, but * so long as it is large enough. */ #define MAX_PINNED_INTERVAL 512 /* Working cpumask for load_balance and load_balance_newidle. */ |
029632fbb sched: Make separ... |
4202 |
DEFINE_PER_CPU(cpumask_var_t, load_balance_tmpmask); |
1e3c88bde sched: Move load ... |
4203 |
|
46e49b383 sched: Wholesale ... |
4204 |
static int need_active_balance(struct sched_domain *sd, int idle, |
532cb4c40 sched: Add asymme... |
4205 |
int busiest_cpu, int this_cpu) |
1af3ed3dd sched: Unify load... |
4206 4207 |
{ if (idle == CPU_NEWLY_IDLE) { |
532cb4c40 sched: Add asymme... |
4208 4209 4210 4211 4212 4213 4214 4215 |
/* * ASYM_PACKING needs to force migrate tasks from busy but * higher numbered CPUs in order to pack all tasks in the * lowest numbered CPUs. */ if ((sd->flags & SD_ASYM_PACKING) && busiest_cpu > this_cpu) return 1; |
1af3ed3dd sched: Unify load... |
4216 4217 4218 4219 4220 4221 4222 4223 4224 4225 4226 4227 4228 4229 4230 4231 4232 4233 4234 |
/* * The only task running in a non-idle cpu can be moved to this * cpu in an attempt to completely freeup the other CPU * package. * * The package power saving logic comes from * find_busiest_group(). If there are no imbalance, then * f_b_g() will return NULL. However when sched_mc={1,2} then * f_b_g() will select a group from which a running task may be * pulled to this cpu in order to make the other package idle. * If there is no opportunity to make a package idle and if * there are no imbalance, then f_b_g() will return NULL and no * action will be taken in load_balance_newidle(). * * Under normal task pull operation due to imbalance, there * will be more than one task in the source run queue and * move_tasks() will succeed. ld_moved will be true and this * active balance code will not be triggered. */ |
1af3ed3dd sched: Unify load... |
4235 4236 4237 4238 4239 4240 |
if (sched_mc_power_savings < POWERSAVINGS_BALANCE_WAKEUP) return 0; } return unlikely(sd->nr_balance_failed > sd->cache_nice_tries+2); } |
969c79215 sched: replace mi... |
4241 |
static int active_load_balance_cpu_stop(void *data); |
1e3c88bde sched: Move load ... |
4242 4243 4244 4245 4246 4247 4248 4249 |
/* * Check this_cpu to ensure it is balanced within domain. Attempt to move * tasks if there is an imbalance. */ static int load_balance(int this_cpu, struct rq *this_rq, struct sched_domain *sd, enum cpu_idle_type idle, int *balance) { |
5b54b56be sched: Replace al... |
4250 |
int ld_moved, lb_flags = 0, active_balance = 0; |
1e3c88bde sched: Move load ... |
4251 4252 4253 4254 4255 4256 4257 |
struct sched_group *group; unsigned long imbalance; struct rq *busiest; unsigned long flags; struct cpumask *cpus = __get_cpu_var(load_balance_tmpmask); cpumask_copy(cpus, cpu_active_mask); |
1e3c88bde sched: Move load ... |
4258 4259 4260 |
schedstat_inc(sd, lb_count[idle]); redo: |
46e49b383 sched: Wholesale ... |
4261 |
group = find_busiest_group(sd, this_cpu, &imbalance, idle, |
1e3c88bde sched: Move load ... |
4262 4263 4264 4265 4266 4267 4268 4269 4270 |
cpus, balance); if (*balance == 0) goto out_balanced; if (!group) { schedstat_inc(sd, lb_nobusyg[idle]); goto out_balanced; } |
9d5efe05e sched: Fix capaci... |
4271 |
busiest = find_busiest_queue(sd, group, idle, imbalance, cpus); |
1e3c88bde sched: Move load ... |
4272 4273 4274 4275 4276 4277 4278 4279 4280 4281 4282 4283 4284 4285 4286 4287 4288 |
if (!busiest) { schedstat_inc(sd, lb_nobusyq[idle]); goto out_balanced; } BUG_ON(busiest == this_rq); schedstat_add(sd, lb_imbalance[idle], imbalance); ld_moved = 0; if (busiest->nr_running > 1) { /* * Attempt to move tasks. If find_busiest_group has found * an imbalance but busiest->nr_running <= 1, the group is * still unbalanced. ld_moved simply stays zero, so it is * correctly treated as an imbalance. */ |
5b54b56be sched: Replace al... |
4289 |
lb_flags |= LBF_ALL_PINNED; |
1e3c88bde sched: Move load ... |
4290 4291 4292 |
local_irq_save(flags); double_rq_lock(this_rq, busiest); ld_moved = move_tasks(this_rq, this_cpu, busiest, |
5b54b56be sched: Replace al... |
4293 |
imbalance, sd, idle, &lb_flags); |
1e3c88bde sched: Move load ... |
4294 4295 4296 4297 4298 4299 4300 4301 |
double_rq_unlock(this_rq, busiest); local_irq_restore(flags); /* * some other cpu did the load balance for us. */ if (ld_moved && this_cpu != smp_processor_id()) resched_cpu(this_cpu); |
a195f004e sched: Fix load-b... |
4302 4303 4304 4305 |
if (lb_flags & LBF_ABORT) goto out_balanced; if (lb_flags & LBF_NEED_BREAK) { |
bced76aea sched: Fix lockup... |
4306 4307 4308 |
lb_flags += LBF_HAD_BREAK - LBF_NEED_BREAK; if (lb_flags & LBF_ABORT) goto out_balanced; |
a195f004e sched: Fix load-b... |
4309 4310 |
goto redo; } |
1e3c88bde sched: Move load ... |
4311 |
/* All tasks on this runqueue were pinned by CPU affinity */ |
5b54b56be sched: Replace al... |
4312 |
if (unlikely(lb_flags & LBF_ALL_PINNED)) { |
1e3c88bde sched: Move load ... |
4313 4314 4315 4316 4317 4318 4319 4320 4321 |
cpumask_clear_cpu(cpu_of(busiest), cpus); if (!cpumask_empty(cpus)) goto redo; goto out_balanced; } } if (!ld_moved) { schedstat_inc(sd, lb_failed[idle]); |
58b26c4c0 sched: Increment ... |
4322 4323 4324 4325 4326 4327 4328 4329 |
/* * Increment the failure counter only on periodic balance. * We do not want newidle balance, which can be very * frequent, pollute the failure counter causing * excessive cache_hot migrations and active balances. */ if (idle != CPU_NEWLY_IDLE) sd->nr_balance_failed++; |
1e3c88bde sched: Move load ... |
4330 |
|
46e49b383 sched: Wholesale ... |
4331 |
if (need_active_balance(sd, idle, cpu_of(busiest), this_cpu)) { |
1e3c88bde sched: Move load ... |
4332 |
raw_spin_lock_irqsave(&busiest->lock, flags); |
969c79215 sched: replace mi... |
4333 4334 4335 |
/* don't kick the active_load_balance_cpu_stop, * if the curr task on busiest cpu can't be * moved to this_cpu |
1e3c88bde sched: Move load ... |
4336 4337 |
*/ if (!cpumask_test_cpu(this_cpu, |
fa17b507f sched: Wrap sched... |
4338 |
tsk_cpus_allowed(busiest->curr))) { |
1e3c88bde sched: Move load ... |
4339 4340 |
raw_spin_unlock_irqrestore(&busiest->lock, flags); |
5b54b56be sched: Replace al... |
4341 |
lb_flags |= LBF_ALL_PINNED; |
1e3c88bde sched: Move load ... |
4342 4343 |
goto out_one_pinned; } |
969c79215 sched: replace mi... |
4344 4345 4346 4347 4348 |
/* * ->active_balance synchronizes accesses to * ->active_balance_work. Once set, it's cleared * only after active load balance is finished. */ |
1e3c88bde sched: Move load ... |
4349 4350 4351 4352 4353 4354 |
if (!busiest->active_balance) { busiest->active_balance = 1; busiest->push_cpu = this_cpu; active_balance = 1; } raw_spin_unlock_irqrestore(&busiest->lock, flags); |
969c79215 sched: replace mi... |
4355 |
|
1e3c88bde sched: Move load ... |
4356 |
if (active_balance) |
969c79215 sched: replace mi... |
4357 4358 4359 |
stop_one_cpu_nowait(cpu_of(busiest), active_load_balance_cpu_stop, busiest, &busiest->active_balance_work); |
1e3c88bde sched: Move load ... |
4360 4361 4362 4363 4364 4365 4366 4367 4368 4369 4370 4371 4372 4373 4374 4375 4376 4377 4378 4379 4380 4381 4382 |
/* * We've kicked active balancing, reset the failure * counter. */ sd->nr_balance_failed = sd->cache_nice_tries+1; } } else sd->nr_balance_failed = 0; if (likely(!active_balance)) { /* We were unbalanced, so reset the balancing interval */ sd->balance_interval = sd->min_interval; } else { /* * If we've begun active balancing, start to back off. This * case may not be covered by the all_pinned logic if there * is only 1 task on the busy runqueue (because we don't call * move_tasks). */ if (sd->balance_interval < sd->max_interval) sd->balance_interval *= 2; } |
1e3c88bde sched: Move load ... |
4383 4384 4385 4386 4387 4388 4389 4390 4391 |
goto out; out_balanced: schedstat_inc(sd, lb_balanced[idle]); sd->nr_balance_failed = 0; out_one_pinned: /* tune up the balancing interval */ |
5b54b56be sched: Replace al... |
4392 4393 |
if (((lb_flags & LBF_ALL_PINNED) && sd->balance_interval < MAX_PINNED_INTERVAL) || |
1e3c88bde sched: Move load ... |
4394 4395 |
(sd->balance_interval < sd->max_interval)) sd->balance_interval *= 2; |
46e49b383 sched: Wholesale ... |
4396 |
ld_moved = 0; |
1e3c88bde sched: Move load ... |
4397 |
out: |
1e3c88bde sched: Move load ... |
4398 4399 4400 4401 |
return ld_moved; } /* |
1e3c88bde sched: Move load ... |
4402 4403 4404 |
* idle_balance is called by schedule() if this_cpu is about to become * idle. Attempts to pull tasks from other CPUs. */ |
029632fbb sched: Make separ... |
4405 |
void idle_balance(int this_cpu, struct rq *this_rq) |
1e3c88bde sched: Move load ... |
4406 4407 4408 4409 4410 4411 4412 4413 4414 |
{ struct sched_domain *sd; int pulled_task = 0; unsigned long next_balance = jiffies + HZ; this_rq->idle_stamp = this_rq->clock; if (this_rq->avg_idle < sysctl_sched_migration_cost) return; |
f492e12ef sched: Remove loa... |
4415 4416 4417 4418 |
/* * Drop the rq->lock, but keep IRQ/preempt disabled. */ raw_spin_unlock(&this_rq->lock); |
c66eaf619 sched: Update sha... |
4419 |
update_shares(this_cpu); |
dce840a08 sched: Dynamicall... |
4420 |
rcu_read_lock(); |
1e3c88bde sched: Move load ... |
4421 4422 |
for_each_domain(this_cpu, sd) { unsigned long interval; |
f492e12ef sched: Remove loa... |
4423 |
int balance = 1; |
1e3c88bde sched: Move load ... |
4424 4425 4426 |
if (!(sd->flags & SD_LOAD_BALANCE)) continue; |
f492e12ef sched: Remove loa... |
4427 |
if (sd->flags & SD_BALANCE_NEWIDLE) { |
1e3c88bde sched: Move load ... |
4428 |
/* If we've pulled tasks over stop searching: */ |
f492e12ef sched: Remove loa... |
4429 4430 4431 |
pulled_task = load_balance(this_cpu, this_rq, sd, CPU_NEWLY_IDLE, &balance); } |
1e3c88bde sched: Move load ... |
4432 4433 4434 4435 |
interval = msecs_to_jiffies(sd->balance_interval); if (time_after(next_balance, sd->last_balance + interval)) next_balance = sd->last_balance + interval; |
d5ad140bc sched: Fix idle b... |
4436 4437 |
if (pulled_task) { this_rq->idle_stamp = 0; |
1e3c88bde sched: Move load ... |
4438 |
break; |
d5ad140bc sched: Fix idle b... |
4439 |
} |
1e3c88bde sched: Move load ... |
4440 |
} |
dce840a08 sched: Dynamicall... |
4441 |
rcu_read_unlock(); |
f492e12ef sched: Remove loa... |
4442 4443 |
raw_spin_lock(&this_rq->lock); |
1e3c88bde sched: Move load ... |
4444 4445 4446 4447 4448 4449 4450 4451 4452 4453 |
if (pulled_task || time_after(jiffies, this_rq->next_balance)) { /* * We are going idle. next_balance may be set based on * a busy processor. So reset next_balance. */ this_rq->next_balance = next_balance; } } /* |
969c79215 sched: replace mi... |
4454 4455 4456 4457 |
* active_load_balance_cpu_stop is run by cpu stopper. It pushes * running tasks off the busiest CPU onto idle CPUs. It requires at * least 1 task to be running on each physical CPU where possible, and * avoids physical / logical imbalances. |
1e3c88bde sched: Move load ... |
4458 |
*/ |
969c79215 sched: replace mi... |
4459 |
static int active_load_balance_cpu_stop(void *data) |
1e3c88bde sched: Move load ... |
4460 |
{ |
969c79215 sched: replace mi... |
4461 4462 |
struct rq *busiest_rq = data; int busiest_cpu = cpu_of(busiest_rq); |
1e3c88bde sched: Move load ... |
4463 |
int target_cpu = busiest_rq->push_cpu; |
969c79215 sched: replace mi... |
4464 |
struct rq *target_rq = cpu_rq(target_cpu); |
1e3c88bde sched: Move load ... |
4465 |
struct sched_domain *sd; |
969c79215 sched: replace mi... |
4466 4467 4468 4469 4470 4471 4472 |
raw_spin_lock_irq(&busiest_rq->lock); /* make sure the requested cpu hasn't gone down in the meantime */ if (unlikely(busiest_cpu != smp_processor_id() || !busiest_rq->active_balance)) goto out_unlock; |
1e3c88bde sched: Move load ... |
4473 4474 4475 |
/* Is there any task to move? */ if (busiest_rq->nr_running <= 1) |
969c79215 sched: replace mi... |
4476 |
goto out_unlock; |
1e3c88bde sched: Move load ... |
4477 4478 4479 4480 4481 4482 4483 4484 4485 4486 |
/* * This condition is "impossible", if it occurs * we need to fix it. Originally reported by * Bjorn Helgaas on a 128-cpu setup. */ BUG_ON(busiest_rq == target_rq); /* move a task from busiest_rq to target_rq */ double_lock_balance(busiest_rq, target_rq); |
1e3c88bde sched: Move load ... |
4487 4488 |
/* Search for an sd spanning us and the target CPU. */ |
dce840a08 sched: Dynamicall... |
4489 |
rcu_read_lock(); |
1e3c88bde sched: Move load ... |
4490 4491 4492 4493 4494 4495 4496 4497 4498 4499 4500 4501 4502 4503 4504 |
for_each_domain(target_cpu, sd) { if ((sd->flags & SD_LOAD_BALANCE) && cpumask_test_cpu(busiest_cpu, sched_domain_span(sd))) break; } if (likely(sd)) { schedstat_inc(sd, alb_count); if (move_one_task(target_rq, target_cpu, busiest_rq, sd, CPU_IDLE)) schedstat_inc(sd, alb_pushed); else schedstat_inc(sd, alb_failed); } |
dce840a08 sched: Dynamicall... |
4505 |
rcu_read_unlock(); |
1e3c88bde sched: Move load ... |
4506 |
double_unlock_balance(busiest_rq, target_rq); |
969c79215 sched: replace mi... |
4507 4508 4509 4510 |
out_unlock: busiest_rq->active_balance = 0; raw_spin_unlock_irq(&busiest_rq->lock); return 0; |
1e3c88bde sched: Move load ... |
4511 4512 4513 |
} #ifdef CONFIG_NO_HZ |
83cd4fe27 sched: Change noh... |
4514 4515 |
/* * idle load balancing details |
83cd4fe27 sched: Change noh... |
4516 4517 4518 4519 |
* - When one of the busy CPUs notice that there may be an idle rebalancing * needed, they will kick the idle load balancer, which then does idle * load balancing for all the idle CPUs. */ |
1e3c88bde sched: Move load ... |
4520 |
static struct { |
83cd4fe27 sched: Change noh... |
4521 |
cpumask_var_t idle_cpus_mask; |
0b005cf54 sched, nohz: Impl... |
4522 |
atomic_t nr_cpus; |
83cd4fe27 sched: Change noh... |
4523 4524 |
unsigned long next_balance; /* in jiffy units */ } nohz ____cacheline_aligned; |
1e3c88bde sched: Move load ... |
4525 |
|
1e3c88bde sched: Move load ... |
4526 4527 4528 4529 4530 4531 4532 4533 4534 4535 4536 4537 4538 4539 4540 |
#if defined(CONFIG_SCHED_MC) || defined(CONFIG_SCHED_SMT) /** * lowest_flag_domain - Return lowest sched_domain containing flag. * @cpu: The cpu whose lowest level of sched domain is to * be returned. * @flag: The flag to check for the lowest sched_domain * for the given cpu. * * Returns the lowest sched_domain of a cpu which contains the given flag. */ static inline struct sched_domain *lowest_flag_domain(int cpu, int flag) { struct sched_domain *sd; for_each_domain(cpu, sd) |
083547169 sched: Remove noo... |
4541 |
if (sd->flags & flag) |
1e3c88bde sched: Move load ... |
4542 4543 4544 4545 4546 4547 4548 4549 4550 4551 4552 4553 4554 4555 4556 4557 4558 4559 4560 4561 |
break; return sd; } /** * for_each_flag_domain - Iterates over sched_domains containing the flag. * @cpu: The cpu whose domains we're iterating over. * @sd: variable holding the value of the power_savings_sd * for cpu. * @flag: The flag to filter the sched_domains to be iterated. * * Iterates over all the scheduler domains for a given cpu that has the 'flag' * set, starting from the lowest sched_domain to the highest. */ #define for_each_flag_domain(cpu, sd, flag) \ for (sd = lowest_flag_domain(cpu, flag); \ (sd && (sd->flags & flag)); sd = sd->parent) /** |
1e3c88bde sched: Move load ... |
4562 4563 4564 4565 4566 4567 4568 4569 4570 4571 4572 4573 4574 |
* find_new_ilb - Finds the optimum idle load balancer for nomination. * @cpu: The cpu which is nominating a new idle_load_balancer. * * Returns: Returns the id of the idle load balancer if it exists, * Else, returns >= nr_cpu_ids. * * This algorithm picks the idle load balancer such that it belongs to a * semi-idle powersavings sched_domain. The idea is to try and avoid * completely idle packages/cores just for the purpose of idle load balancing * when there are other idle cpu's which are better suited for that job. */ static int find_new_ilb(int cpu) { |
0b005cf54 sched, nohz: Impl... |
4575 |
int ilb = cpumask_first(nohz.idle_cpus_mask); |
786d6dc7a sched, nohz: Clea... |
4576 |
struct sched_group *ilbg; |
1e3c88bde sched: Move load ... |
4577 |
struct sched_domain *sd; |
1e3c88bde sched: Move load ... |
4578 4579 4580 4581 4582 4583 4584 4585 4586 4587 4588 4589 |
/* * Have idle load balancer selection from semi-idle packages only * when power-aware load balancing is enabled */ if (!(sched_smt_power_savings || sched_mc_power_savings)) goto out_done; /* * Optimize for the case when we have no idle CPUs or only one * idle CPU. Don't walk the sched_domain hierarchy in such cases */ |
83cd4fe27 sched: Change noh... |
4590 |
if (cpumask_weight(nohz.idle_cpus_mask) < 2) |
1e3c88bde sched: Move load ... |
4591 |
goto out_done; |
dce840a08 sched: Dynamicall... |
4592 |
rcu_read_lock(); |
1e3c88bde sched: Move load ... |
4593 |
for_each_flag_domain(cpu, sd, SD_POWERSAVINGS_BALANCE) { |
786d6dc7a sched, nohz: Clea... |
4594 |
ilbg = sd->groups; |
1e3c88bde sched: Move load ... |
4595 4596 |
do { |
786d6dc7a sched, nohz: Clea... |
4597 4598 4599 4600 |
if (ilbg->group_weight != atomic_read(&ilbg->sgp->nr_busy_cpus)) { ilb = cpumask_first_and(nohz.idle_cpus_mask, sched_group_cpus(ilbg)); |
dce840a08 sched: Dynamicall... |
4601 4602 |
goto unlock; } |
1e3c88bde sched: Move load ... |
4603 |
|
786d6dc7a sched, nohz: Clea... |
4604 |
ilbg = ilbg->next; |
1e3c88bde sched: Move load ... |
4605 |
|
786d6dc7a sched, nohz: Clea... |
4606 |
} while (ilbg != sd->groups); |
1e3c88bde sched: Move load ... |
4607 |
} |
dce840a08 sched: Dynamicall... |
4608 4609 |
unlock: rcu_read_unlock(); |
1e3c88bde sched: Move load ... |
4610 4611 |
out_done: |
786d6dc7a sched, nohz: Clea... |
4612 4613 4614 4615 |
if (ilb < nr_cpu_ids && idle_cpu(ilb)) return ilb; return nr_cpu_ids; |
1e3c88bde sched: Move load ... |
4616 4617 4618 4619 |
} #else /* (CONFIG_SCHED_MC || CONFIG_SCHED_SMT) */ static inline int find_new_ilb(int call_cpu) { |
83cd4fe27 sched: Change noh... |
4620 |
return nr_cpu_ids; |
1e3c88bde sched: Move load ... |
4621 4622 4623 4624 |
} #endif /* |
83cd4fe27 sched: Change noh... |
4625 4626 4627 4628 4629 4630 4631 4632 4633 |
* Kick a CPU to do the nohz balancing, if it is time for it. We pick the * nohz_load_balancer CPU (if there is one) otherwise fallback to any idle * CPU (if there is one). */ static void nohz_balancer_kick(int cpu) { int ilb_cpu; nohz.next_balance++; |
0b005cf54 sched, nohz: Impl... |
4634 |
ilb_cpu = find_new_ilb(cpu); |
83cd4fe27 sched: Change noh... |
4635 |
|
0b005cf54 sched, nohz: Impl... |
4636 4637 |
if (ilb_cpu >= nr_cpu_ids) return; |
83cd4fe27 sched: Change noh... |
4638 |
|
cd490c5b2 sched, nohz: Set ... |
4639 |
if (test_and_set_bit(NOHZ_BALANCE_KICK, nohz_flags(ilb_cpu))) |
1c792db7f sched, nohz: Intr... |
4640 4641 4642 4643 4644 4645 4646 4647 |
return; /* * Use smp_send_reschedule() instead of resched_cpu(). * This way we generate a sched IPI on the target cpu which * is idle. And the softirq performing nohz idle load balance * will be run before returning from the IPI. */ smp_send_reschedule(ilb_cpu); |
83cd4fe27 sched: Change noh... |
4648 4649 |
return; } |
69e1e811d sched, nohz: Trac... |
4650 4651 4652 4653 4654 4655 4656 4657 4658 4659 4660 4661 4662 4663 4664 4665 4666 4667 4668 4669 4670 4671 4672 4673 4674 4675 4676 4677 4678 |
static inline void set_cpu_sd_state_busy(void) { struct sched_domain *sd; int cpu = smp_processor_id(); if (!test_bit(NOHZ_IDLE, nohz_flags(cpu))) return; clear_bit(NOHZ_IDLE, nohz_flags(cpu)); rcu_read_lock(); for_each_domain(cpu, sd) atomic_inc(&sd->groups->sgp->nr_busy_cpus); rcu_read_unlock(); } void set_cpu_sd_state_idle(void) { struct sched_domain *sd; int cpu = smp_processor_id(); if (test_bit(NOHZ_IDLE, nohz_flags(cpu))) return; set_bit(NOHZ_IDLE, nohz_flags(cpu)); rcu_read_lock(); for_each_domain(cpu, sd) atomic_dec(&sd->groups->sgp->nr_busy_cpus); rcu_read_unlock(); } |
83cd4fe27 sched: Change noh... |
4679 |
/* |
0b005cf54 sched, nohz: Impl... |
4680 4681 |
* This routine will record that this cpu is going idle with tick stopped. * This info will be used in performing idle load balancing in the future. |
1e3c88bde sched: Move load ... |
4682 |
*/ |
83cd4fe27 sched: Change noh... |
4683 |
void select_nohz_load_balancer(int stop_tick) |
1e3c88bde sched: Move load ... |
4684 4685 4686 4687 |
{ int cpu = smp_processor_id(); if (stop_tick) { |
0b005cf54 sched, nohz: Impl... |
4688 |
if (test_bit(NOHZ_TICK_STOPPED, nohz_flags(cpu))) |
83cd4fe27 sched: Change noh... |
4689 |
return; |
1e3c88bde sched: Move load ... |
4690 |
|
83cd4fe27 sched: Change noh... |
4691 |
cpumask_set_cpu(cpu, nohz.idle_cpus_mask); |
0b005cf54 sched, nohz: Impl... |
4692 |
atomic_inc(&nohz.nr_cpus); |
1c792db7f sched, nohz: Intr... |
4693 |
set_bit(NOHZ_TICK_STOPPED, nohz_flags(cpu)); |
1e3c88bde sched: Move load ... |
4694 |
} |
83cd4fe27 sched: Change noh... |
4695 |
return; |
1e3c88bde sched: Move load ... |
4696 4697 4698 4699 |
} #endif static DEFINE_SPINLOCK(balancing); |
49c022e65 sched: Clean up r... |
4700 4701 4702 4703 4704 4705 |
static unsigned long __read_mostly max_load_balance_interval = HZ/10; /* * Scale the max load_balance interval with the number of CPUs in the system. * This trades load-balance latency on larger machines for less cross talk. */ |
029632fbb sched: Make separ... |
4706 |
void update_max_interval(void) |
49c022e65 sched: Clean up r... |
4707 4708 4709 |
{ max_load_balance_interval = HZ*num_online_cpus()/10; } |
1e3c88bde sched: Move load ... |
4710 4711 4712 4713 4714 4715 4716 4717 4718 4719 4720 4721 4722 4723 4724 4725 |
/* * It checks each scheduling domain to see if it is due to be balanced, * and initiates a balancing operation if so. * * Balancing parameters are set up in arch_init_sched_domains. */ static void rebalance_domains(int cpu, enum cpu_idle_type idle) { int balance = 1; struct rq *rq = cpu_rq(cpu); unsigned long interval; struct sched_domain *sd; /* Earliest time when we have to do rebalance again */ unsigned long next_balance = jiffies + 60*HZ; int update_next_balance = 0; int need_serialize; |
2069dd75c sched: Rewrite tg... |
4726 |
update_shares(cpu); |
dce840a08 sched: Dynamicall... |
4727 |
rcu_read_lock(); |
1e3c88bde sched: Move load ... |
4728 4729 4730 4731 4732 4733 4734 4735 4736 4737 |
for_each_domain(cpu, sd) { if (!(sd->flags & SD_LOAD_BALANCE)) continue; interval = sd->balance_interval; if (idle != CPU_IDLE) interval *= sd->busy_factor; /* scale ms to jiffies */ interval = msecs_to_jiffies(interval); |
49c022e65 sched: Clean up r... |
4738 |
interval = clamp(interval, 1UL, max_load_balance_interval); |
1e3c88bde sched: Move load ... |
4739 4740 4741 4742 4743 4744 4745 4746 4747 4748 4749 4750 |
need_serialize = sd->flags & SD_SERIALIZE; if (need_serialize) { if (!spin_trylock(&balancing)) goto out; } if (time_after_eq(jiffies, sd->last_balance + interval)) { if (load_balance(cpu, rq, sd, idle, &balance)) { /* * We've pulled tasks over so either we're no |
c186fafe9 sched: Clean up r... |
4751 |
* longer idle. |
1e3c88bde sched: Move load ... |
4752 4753 4754 4755 4756 4757 4758 4759 4760 4761 4762 4763 4764 4765 4766 4767 4768 4769 4770 4771 4772 |
*/ idle = CPU_NOT_IDLE; } sd->last_balance = jiffies; } if (need_serialize) spin_unlock(&balancing); out: if (time_after(next_balance, sd->last_balance + interval)) { next_balance = sd->last_balance + interval; update_next_balance = 1; } /* * Stop the load balance at this level. There is another * CPU in our sched group which is doing load balancing more * actively. */ if (!balance) break; } |
dce840a08 sched: Dynamicall... |
4773 |
rcu_read_unlock(); |
1e3c88bde sched: Move load ... |
4774 4775 4776 4777 4778 4779 4780 4781 4782 |
/* * next_balance will be updated only when there is a need. * When the cpu is attached to null domain for ex, it will not be * updated. */ if (likely(update_next_balance)) rq->next_balance = next_balance; } |
83cd4fe27 sched: Change noh... |
4783 |
#ifdef CONFIG_NO_HZ |
1e3c88bde sched: Move load ... |
4784 |
/* |
83cd4fe27 sched: Change noh... |
4785 |
* In CONFIG_NO_HZ case, the idle balance kickee will do the |
1e3c88bde sched: Move load ... |
4786 4787 |
* rebalancing for all the cpus for whom scheduler ticks are stopped. */ |
83cd4fe27 sched: Change noh... |
4788 4789 4790 4791 4792 |
static void nohz_idle_balance(int this_cpu, enum cpu_idle_type idle) { struct rq *this_rq = cpu_rq(this_cpu); struct rq *rq; int balance_cpu; |
1c792db7f sched, nohz: Intr... |
4793 4794 4795 |
if (idle != CPU_IDLE || !test_bit(NOHZ_BALANCE_KICK, nohz_flags(this_cpu))) goto end; |
83cd4fe27 sched: Change noh... |
4796 4797 |
for_each_cpu(balance_cpu, nohz.idle_cpus_mask) { |
8a6d42d1b sched, nohz: Fix ... |
4798 |
if (balance_cpu == this_cpu || !idle_cpu(balance_cpu)) |
83cd4fe27 sched: Change noh... |
4799 4800 4801 4802 4803 4804 4805 |
continue; /* * If this cpu gets work to do, stop the load balancing * work being done for other cpus. Next load * balancing owner will pick it up. */ |
1c792db7f sched, nohz: Intr... |
4806 |
if (need_resched()) |
83cd4fe27 sched: Change noh... |
4807 |
break; |
83cd4fe27 sched: Change noh... |
4808 4809 |
raw_spin_lock_irq(&this_rq->lock); |
5343bdb8f sched: Update rq-... |
4810 |
update_rq_clock(this_rq); |
83cd4fe27 sched: Change noh... |
4811 4812 4813 4814 4815 4816 4817 4818 4819 4820 |
update_cpu_load(this_rq); raw_spin_unlock_irq(&this_rq->lock); rebalance_domains(balance_cpu, CPU_IDLE); rq = cpu_rq(balance_cpu); if (time_after(this_rq->next_balance, rq->next_balance)) this_rq->next_balance = rq->next_balance; } nohz.next_balance = this_rq->next_balance; |
1c792db7f sched, nohz: Intr... |
4821 4822 |
end: clear_bit(NOHZ_BALANCE_KICK, nohz_flags(this_cpu)); |
83cd4fe27 sched: Change noh... |
4823 4824 4825 |
} /* |
0b005cf54 sched, nohz: Impl... |
4826 4827 4828 4829 4830 4831 4832 |
* Current heuristic for kicking the idle load balancer in the presence * of an idle cpu is the system. * - This rq has more than one task. * - At any scheduler domain level, this cpu's scheduler group has multiple * busy cpu's exceeding the group's power. * - For SD_ASYM_PACKING, if the lower numbered cpu's in the scheduler * domain span are idle. |
83cd4fe27 sched: Change noh... |
4833 4834 4835 4836 |
*/ static inline int nohz_kick_needed(struct rq *rq, int cpu) { unsigned long now = jiffies; |
0b005cf54 sched, nohz: Impl... |
4837 |
struct sched_domain *sd; |
83cd4fe27 sched: Change noh... |
4838 |
|
1c792db7f sched, nohz: Intr... |
4839 |
if (unlikely(idle_cpu(cpu))) |
83cd4fe27 sched: Change noh... |
4840 |
return 0; |
1c792db7f sched, nohz: Intr... |
4841 4842 4843 4844 |
/* * We may be recently in ticked or tickless idle mode. At the first * busy tick after returning from idle, we will update the busy stats. */ |
69e1e811d sched, nohz: Trac... |
4845 |
set_cpu_sd_state_busy(); |
0b005cf54 sched, nohz: Impl... |
4846 |
if (unlikely(test_bit(NOHZ_TICK_STOPPED, nohz_flags(cpu)))) { |
1c792db7f sched, nohz: Intr... |
4847 |
clear_bit(NOHZ_TICK_STOPPED, nohz_flags(cpu)); |
0b005cf54 sched, nohz: Impl... |
4848 4849 4850 4851 4852 4853 4854 4855 4856 4857 |
cpumask_clear_cpu(cpu, nohz.idle_cpus_mask); atomic_dec(&nohz.nr_cpus); } /* * None are in tickless mode and hence no need for NOHZ idle load * balancing. */ if (likely(!atomic_read(&nohz.nr_cpus))) return 0; |
1c792db7f sched, nohz: Intr... |
4858 4859 |
if (time_before(now, nohz.next_balance)) |
83cd4fe27 sched: Change noh... |
4860 |
return 0; |
0b005cf54 sched, nohz: Impl... |
4861 4862 |
if (rq->nr_running >= 2) goto need_kick; |
83cd4fe27 sched: Change noh... |
4863 |
|
067491b73 sched, nohz: Fix ... |
4864 |
rcu_read_lock(); |
0b005cf54 sched, nohz: Impl... |
4865 4866 4867 4868 |
for_each_domain(cpu, sd) { struct sched_group *sg = sd->groups; struct sched_group_power *sgp = sg->sgp; int nr_busy = atomic_read(&sgp->nr_busy_cpus); |
83cd4fe27 sched: Change noh... |
4869 |
|
0b005cf54 sched, nohz: Impl... |
4870 |
if (sd->flags & SD_SHARE_PKG_RESOURCES && nr_busy > 1) |
067491b73 sched, nohz: Fix ... |
4871 |
goto need_kick_unlock; |
0b005cf54 sched, nohz: Impl... |
4872 4873 4874 4875 |
if (sd->flags & SD_ASYM_PACKING && nr_busy != sg->group_weight && (cpumask_first_and(nohz.idle_cpus_mask, sched_domain_span(sd)) < cpu)) |
067491b73 sched, nohz: Fix ... |
4876 |
goto need_kick_unlock; |
0b005cf54 sched, nohz: Impl... |
4877 4878 4879 |
if (!(sd->flags & (SD_SHARE_PKG_RESOURCES | SD_ASYM_PACKING))) break; |
83cd4fe27 sched: Change noh... |
4880 |
} |
067491b73 sched, nohz: Fix ... |
4881 |
rcu_read_unlock(); |
83cd4fe27 sched: Change noh... |
4882 |
return 0; |
067491b73 sched, nohz: Fix ... |
4883 4884 4885 |
need_kick_unlock: rcu_read_unlock(); |
0b005cf54 sched, nohz: Impl... |
4886 4887 |
need_kick: return 1; |
83cd4fe27 sched: Change noh... |
4888 4889 4890 4891 4892 4893 4894 4895 4896 |
} #else static void nohz_idle_balance(int this_cpu, enum cpu_idle_type idle) { } #endif /* * run_rebalance_domains is triggered when needed from the scheduler tick. * Also triggered for nohz idle balancing (with nohz_balancing_kick set). */ |
1e3c88bde sched: Move load ... |
4897 4898 4899 4900 |
static void run_rebalance_domains(struct softirq_action *h) { int this_cpu = smp_processor_id(); struct rq *this_rq = cpu_rq(this_cpu); |
6eb57e0d6 sched: Request fo... |
4901 |
enum cpu_idle_type idle = this_rq->idle_balance ? |
1e3c88bde sched: Move load ... |
4902 4903 4904 |
CPU_IDLE : CPU_NOT_IDLE; rebalance_domains(this_cpu, idle); |
1e3c88bde sched: Move load ... |
4905 |
/* |
83cd4fe27 sched: Change noh... |
4906 |
* If this cpu has a pending nohz_balance_kick, then do the |
1e3c88bde sched: Move load ... |
4907 4908 4909 |
* balancing on behalf of the other idle cpus whose ticks are * stopped. */ |
83cd4fe27 sched: Change noh... |
4910 |
nohz_idle_balance(this_cpu, idle); |
1e3c88bde sched: Move load ... |
4911 4912 4913 4914 |
} static inline int on_null_domain(int cpu) { |
90a6501f9 sched, rcu: Fix r... |
4915 |
return !rcu_dereference_sched(cpu_rq(cpu)->sd); |
1e3c88bde sched: Move load ... |
4916 4917 4918 4919 |
} /* * Trigger the SCHED_SOFTIRQ if it is time to do periodic load balancing. |
1e3c88bde sched: Move load ... |
4920 |
*/ |
029632fbb sched: Make separ... |
4921 |
void trigger_load_balance(struct rq *rq, int cpu) |
1e3c88bde sched: Move load ... |
4922 |
{ |
1e3c88bde sched: Move load ... |
4923 4924 4925 4926 |
/* Don't need to rebalance while attached to NULL domain */ if (time_after_eq(jiffies, rq->next_balance) && likely(!on_null_domain(cpu))) raise_softirq(SCHED_SOFTIRQ); |
83cd4fe27 sched: Change noh... |
4927 |
#ifdef CONFIG_NO_HZ |
1c792db7f sched, nohz: Intr... |
4928 |
if (nohz_kick_needed(rq, cpu) && likely(!on_null_domain(cpu))) |
83cd4fe27 sched: Change noh... |
4929 4930 |
nohz_balancer_kick(cpu); #endif |
1e3c88bde sched: Move load ... |
4931 |
} |
0bcdcf28c sched: Fix missin... |
4932 4933 4934 4935 4936 4937 4938 4939 4940 |
static void rq_online_fair(struct rq *rq) { update_sysctl(); } static void rq_offline_fair(struct rq *rq) { update_sysctl(); } |
55e12e5e7 sched: make sched... |
4941 |
#endif /* CONFIG_SMP */ |
e1d1484f7 sched: reduce bal... |
4942 |
|
bf0f6f24a sched: cfs core, ... |
4943 4944 4945 |
/* * scheduler tick hitting a task of our scheduling class: */ |
8f4d37ec0 sched: high-res p... |
4946 |
static void task_tick_fair(struct rq *rq, struct task_struct *curr, int queued) |
bf0f6f24a sched: cfs core, ... |
4947 4948 4949 4950 4951 4952 |
{ struct cfs_rq *cfs_rq; struct sched_entity *se = &curr->se; for_each_sched_entity(se) { cfs_rq = cfs_rq_of(se); |
8f4d37ec0 sched: high-res p... |
4953 |
entity_tick(cfs_rq, se, queued); |
bf0f6f24a sched: cfs core, ... |
4954 4955 4956 4957 |
} } /* |
cd29fe6f2 sched: Sanitize f... |
4958 4959 4960 |
* called on fork with the child task as argument from the parent's context * - child not yet on the tasklist * - preemption disabled |
bf0f6f24a sched: cfs core, ... |
4961 |
*/ |
cd29fe6f2 sched: Sanitize f... |
4962 |
static void task_fork_fair(struct task_struct *p) |
bf0f6f24a sched: cfs core, ... |
4963 |
{ |
4fc420c91 sched: Fix cgroup... |
4964 4965 |
struct cfs_rq *cfs_rq; struct sched_entity *se = &p->se, *curr; |
00bf7bfc2 sched: fix: move ... |
4966 |
int this_cpu = smp_processor_id(); |
cd29fe6f2 sched: Sanitize f... |
4967 4968 |
struct rq *rq = this_rq(); unsigned long flags; |
05fa785cf sched: Convert rq... |
4969 |
raw_spin_lock_irqsave(&rq->lock, flags); |
bf0f6f24a sched: cfs core, ... |
4970 |
|
861d034ee sched: Fix rq->cl... |
4971 |
update_rq_clock(rq); |
4fc420c91 sched: Fix cgroup... |
4972 4973 |
cfs_rq = task_cfs_rq(current); curr = cfs_rq->curr; |
b0a0f667a sched: suppress R... |
4974 4975 |
if (unlikely(task_cpu(p) != this_cpu)) { rcu_read_lock(); |
cd29fe6f2 sched: Sanitize f... |
4976 |
__set_task_cpu(p, this_cpu); |
b0a0f667a sched: suppress R... |
4977 4978 |
rcu_read_unlock(); } |
bf0f6f24a sched: cfs core, ... |
4979 |
|
7109c4429 sched: call updat... |
4980 |
update_curr(cfs_rq); |
cd29fe6f2 sched: Sanitize f... |
4981 |
|
b5d9d734a sched: Ensure tha... |
4982 4983 |
if (curr) se->vruntime = curr->vruntime; |
aeb73b040 sched: clean up n... |
4984 |
place_entity(cfs_rq, se, 1); |
4d78e7b65 sched: new task p... |
4985 |
|
cd29fe6f2 sched: Sanitize f... |
4986 |
if (sysctl_sched_child_runs_first && curr && entity_before(curr, se)) { |
87fefa381 sched: optimize t... |
4987 |
/* |
edcb60a30 sched: kernel/sch... |
4988 4989 4990 |
* Upon rescheduling, sched_class::put_prev_task() will place * 'current' within the tree based on its new key value. */ |
4d78e7b65 sched: new task p... |
4991 |
swap(curr->vruntime, se->vruntime); |
aec0a5142 sched: call resch... |
4992 |
resched_task(rq->curr); |
4d78e7b65 sched: new task p... |
4993 |
} |
bf0f6f24a sched: cfs core, ... |
4994 |
|
88ec22d3e sched: Remove the... |
4995 |
se->vruntime -= cfs_rq->min_vruntime; |
05fa785cf sched: Convert rq... |
4996 |
raw_spin_unlock_irqrestore(&rq->lock, flags); |
bf0f6f24a sched: cfs core, ... |
4997 |
} |
cb4698450 sched: RT-balance... |
4998 4999 5000 5001 |
/* * Priority of the task has changed. Check to see if we preempt * the current task. */ |
da7a735e5 sched: Fix switch... |
5002 5003 |
static void prio_changed_fair(struct rq *rq, struct task_struct *p, int oldprio) |
cb4698450 sched: RT-balance... |
5004 |
{ |
da7a735e5 sched: Fix switch... |
5005 5006 |
if (!p->se.on_rq) return; |
cb4698450 sched: RT-balance... |
5007 5008 5009 5010 5011 |
/* * Reschedule if we are currently running on this runqueue and * our priority decreased, or if we are not currently running on * this runqueue and our priority is higher than the current's */ |
da7a735e5 sched: Fix switch... |
5012 |
if (rq->curr == p) { |
cb4698450 sched: RT-balance... |
5013 5014 5015 |
if (p->prio > oldprio) resched_task(rq->curr); } else |
15afe09bf sched: wakeup pre... |
5016 |
check_preempt_curr(rq, p, 0); |
cb4698450 sched: RT-balance... |
5017 |
} |
da7a735e5 sched: Fix switch... |
5018 5019 5020 5021 5022 5023 5024 5025 5026 5027 5028 5029 5030 5031 5032 5033 5034 5035 5036 5037 5038 5039 5040 |
static void switched_from_fair(struct rq *rq, struct task_struct *p) { struct sched_entity *se = &p->se; struct cfs_rq *cfs_rq = cfs_rq_of(se); /* * Ensure the task's vruntime is normalized, so that when its * switched back to the fair class the enqueue_entity(.flags=0) will * do the right thing. * * If it was on_rq, then the dequeue_entity(.flags=0) will already * have normalized the vruntime, if it was !on_rq, then only when * the task is sleeping will it still have non-normalized vruntime. */ if (!se->on_rq && p->state != TASK_RUNNING) { /* * Fix up our vruntime so that the current sleep doesn't * cause 'unlimited' sleep bonus. */ place_entity(cfs_rq, se, 0); se->vruntime -= cfs_rq->min_vruntime; } } |
cb4698450 sched: RT-balance... |
5041 5042 5043 |
/* * We switched to the sched_fair class. */ |
da7a735e5 sched: Fix switch... |
5044 |
static void switched_to_fair(struct rq *rq, struct task_struct *p) |
cb4698450 sched: RT-balance... |
5045 |
{ |
da7a735e5 sched: Fix switch... |
5046 5047 |
if (!p->se.on_rq) return; |
cb4698450 sched: RT-balance... |
5048 5049 5050 5051 5052 |
/* * We were most likely switched from sched_rt, so * kick off the schedule if running, otherwise just see * if we can still preempt the current task. */ |
da7a735e5 sched: Fix switch... |
5053 |
if (rq->curr == p) |
cb4698450 sched: RT-balance... |
5054 5055 |
resched_task(rq->curr); else |
15afe09bf sched: wakeup pre... |
5056 |
check_preempt_curr(rq, p, 0); |
cb4698450 sched: RT-balance... |
5057 |
} |
83b699ed2 sched: revert rec... |
5058 5059 5060 5061 5062 5063 5064 5065 |
/* Account for a task changing its policy or group. * * This routine is mostly called to set cfs_rq->curr field when a task * migrates between groups/classes. */ static void set_curr_task_fair(struct rq *rq) { struct sched_entity *se = &rq->curr->se; |
ec12cb7f3 sched: Accumulate... |
5066 5067 5068 5069 5070 5071 5072 |
for_each_sched_entity(se) { struct cfs_rq *cfs_rq = cfs_rq_of(se); set_next_entity(cfs_rq, se); /* ensure bandwidth has been allocated on our new cfs_rq */ account_cfs_rq_runtime(cfs_rq, 0); } |
83b699ed2 sched: revert rec... |
5073 |
} |
029632fbb sched: Make separ... |
5074 5075 5076 5077 5078 5079 5080 5081 5082 |
void init_cfs_rq(struct cfs_rq *cfs_rq) { cfs_rq->tasks_timeline = RB_ROOT; INIT_LIST_HEAD(&cfs_rq->tasks); cfs_rq->min_vruntime = (u64)(-(1LL << 20)); #ifndef CONFIG_64BIT cfs_rq->min_vruntime_copy = cfs_rq->min_vruntime; #endif } |
810b38179 sched: retain vru... |
5083 |
#ifdef CONFIG_FAIR_GROUP_SCHED |
b2b5ce022 sched, cgroup: Fi... |
5084 |
static void task_move_group_fair(struct task_struct *p, int on_rq) |
810b38179 sched: retain vru... |
5085 |
{ |
b2b5ce022 sched, cgroup: Fi... |
5086 5087 5088 5089 5090 5091 5092 5093 5094 5095 5096 5097 5098 |
/* * If the task was not on the rq at the time of this cgroup movement * it must have been asleep, sleeping tasks keep their ->vruntime * absolute on their old rq until wakeup (needed for the fair sleeper * bonus in place_entity()). * * If it was on the rq, we've just 'preempted' it, which does convert * ->vruntime to a relative base. * * Make sure both cases convert their relative position when migrating * to another cgroup's rq. This does somewhat interfere with the * fair sleeper stuff for the first placement, but who cares. */ |
7ceff013c sched: Fix cgroup... |
5099 5100 5101 5102 5103 5104 |
/* * When !on_rq, vruntime of the task has usually NOT been normalized. * But there are some cases where it has already been normalized: * * - Moving a forked child which is waiting for being woken up by * wake_up_new_task(). |
62af3783e sched: Fix cgroup... |
5105 5106 |
* - Moving a task which has been woken up by try_to_wake_up() and * waiting for actually being woken up by sched_ttwu_pending(). |
7ceff013c sched: Fix cgroup... |
5107 5108 5109 5110 |
* * To prevent boost or penalty in the new cfs_rq caused by delta * min_vruntime between the two cfs_rqs, we skip vruntime adjustment. */ |
62af3783e sched: Fix cgroup... |
5111 |
if (!on_rq && (!p->se.sum_exec_runtime || p->state == TASK_WAKING)) |
7ceff013c sched: Fix cgroup... |
5112 |
on_rq = 1; |
b2b5ce022 sched, cgroup: Fi... |
5113 5114 5115 |
if (!on_rq) p->se.vruntime -= cfs_rq_of(&p->se)->min_vruntime; set_task_rq(p, task_cpu(p)); |
88ec22d3e sched: Remove the... |
5116 |
if (!on_rq) |
b2b5ce022 sched, cgroup: Fi... |
5117 |
p->se.vruntime += cfs_rq_of(&p->se)->min_vruntime; |
810b38179 sched: retain vru... |
5118 |
} |
029632fbb sched: Make separ... |
5119 5120 5121 5122 5123 5124 5125 5126 5127 5128 5129 5130 5131 5132 5133 5134 5135 5136 5137 5138 5139 5140 5141 5142 5143 5144 5145 5146 5147 5148 5149 5150 5151 5152 5153 5154 5155 5156 5157 5158 5159 5160 5161 5162 5163 5164 5165 5166 5167 5168 5169 5170 5171 5172 5173 5174 5175 5176 5177 5178 5179 5180 5181 5182 5183 5184 5185 5186 5187 5188 5189 5190 5191 5192 5193 5194 5195 5196 5197 5198 5199 5200 5201 5202 5203 5204 |
void free_fair_sched_group(struct task_group *tg) { int i; destroy_cfs_bandwidth(tg_cfs_bandwidth(tg)); for_each_possible_cpu(i) { if (tg->cfs_rq) kfree(tg->cfs_rq[i]); if (tg->se) kfree(tg->se[i]); } kfree(tg->cfs_rq); kfree(tg->se); } int alloc_fair_sched_group(struct task_group *tg, struct task_group *parent) { struct cfs_rq *cfs_rq; struct sched_entity *se; int i; tg->cfs_rq = kzalloc(sizeof(cfs_rq) * nr_cpu_ids, GFP_KERNEL); if (!tg->cfs_rq) goto err; tg->se = kzalloc(sizeof(se) * nr_cpu_ids, GFP_KERNEL); if (!tg->se) goto err; tg->shares = NICE_0_LOAD; init_cfs_bandwidth(tg_cfs_bandwidth(tg)); for_each_possible_cpu(i) { cfs_rq = kzalloc_node(sizeof(struct cfs_rq), GFP_KERNEL, cpu_to_node(i)); if (!cfs_rq) goto err; se = kzalloc_node(sizeof(struct sched_entity), GFP_KERNEL, cpu_to_node(i)); if (!se) goto err_free_rq; init_cfs_rq(cfs_rq); init_tg_cfs_entry(tg, cfs_rq, se, i, parent->se[i]); } return 1; err_free_rq: kfree(cfs_rq); err: return 0; } void unregister_fair_sched_group(struct task_group *tg, int cpu) { struct rq *rq = cpu_rq(cpu); unsigned long flags; /* * Only empty task groups can be destroyed; so we can speculatively * check on_list without danger of it being re-added. */ if (!tg->cfs_rq[cpu]->on_list) return; raw_spin_lock_irqsave(&rq->lock, flags); list_del_leaf_cfs_rq(tg->cfs_rq[cpu]); raw_spin_unlock_irqrestore(&rq->lock, flags); } void init_tg_cfs_entry(struct task_group *tg, struct cfs_rq *cfs_rq, struct sched_entity *se, int cpu, struct sched_entity *parent) { struct rq *rq = cpu_rq(cpu); cfs_rq->tg = tg; cfs_rq->rq = rq; #ifdef CONFIG_SMP /* allow initial update_cfs_load() to truncate */ cfs_rq->load_stamp = 1; |
810b38179 sched: retain vru... |
5205 |
#endif |
029632fbb sched: Make separ... |
5206 5207 5208 5209 5210 5211 5212 5213 5214 5215 5216 5217 5218 5219 5220 5221 5222 5223 5224 5225 5226 5227 5228 5229 5230 5231 5232 5233 5234 5235 5236 5237 5238 5239 5240 5241 5242 5243 5244 5245 5246 5247 5248 5249 5250 5251 5252 5253 5254 5255 5256 5257 5258 5259 5260 5261 5262 5263 5264 5265 5266 5267 5268 5269 5270 5271 5272 |
init_cfs_rq_runtime(cfs_rq); tg->cfs_rq[cpu] = cfs_rq; tg->se[cpu] = se; /* se could be NULL for root_task_group */ if (!se) return; if (!parent) se->cfs_rq = &rq->cfs; else se->cfs_rq = parent->my_q; se->my_q = cfs_rq; update_load_set(&se->load, 0); se->parent = parent; } static DEFINE_MUTEX(shares_mutex); int sched_group_set_shares(struct task_group *tg, unsigned long shares) { int i; unsigned long flags; /* * We can't change the weight of the root cgroup. */ if (!tg->se[0]) return -EINVAL; shares = clamp(shares, scale_load(MIN_SHARES), scale_load(MAX_SHARES)); mutex_lock(&shares_mutex); if (tg->shares == shares) goto done; tg->shares = shares; for_each_possible_cpu(i) { struct rq *rq = cpu_rq(i); struct sched_entity *se; se = tg->se[i]; /* Propagate contribution to hierarchy */ raw_spin_lock_irqsave(&rq->lock, flags); for_each_sched_entity(se) update_cfs_shares(group_cfs_rq(se)); raw_spin_unlock_irqrestore(&rq->lock, flags); } done: mutex_unlock(&shares_mutex); return 0; } #else /* CONFIG_FAIR_GROUP_SCHED */ void free_fair_sched_group(struct task_group *tg) { } int alloc_fair_sched_group(struct task_group *tg, struct task_group *parent) { return 1; } void unregister_fair_sched_group(struct task_group *tg, int cpu) { } #endif /* CONFIG_FAIR_GROUP_SCHED */ |
810b38179 sched: retain vru... |
5273 |
|
6d686f456 sched: Don't expo... |
5274 |
static unsigned int get_rr_interval_fair(struct rq *rq, struct task_struct *task) |
0d721cead sched: Simplify s... |
5275 5276 |
{ struct sched_entity *se = &task->se; |
0d721cead sched: Simplify s... |
5277 5278 5279 5280 5281 5282 |
unsigned int rr_interval = 0; /* * Time slice is 0 for SCHED_OTHER tasks that are on an otherwise * idle runqueue: */ |
0d721cead sched: Simplify s... |
5283 5284 |
if (rq->cfs.load.weight) rr_interval = NS_TO_JIFFIES(sched_slice(&rq->cfs, se)); |
0d721cead sched: Simplify s... |
5285 5286 5287 |
return rr_interval; } |
bf0f6f24a sched: cfs core, ... |
5288 5289 5290 |
/* * All the scheduling class methods: */ |
029632fbb sched: Make separ... |
5291 |
const struct sched_class fair_sched_class = { |
5522d5d5f sched: mark sched... |
5292 |
.next = &idle_sched_class, |
bf0f6f24a sched: cfs core, ... |
5293 5294 5295 |
.enqueue_task = enqueue_task_fair, .dequeue_task = dequeue_task_fair, .yield_task = yield_task_fair, |
d95f41220 sched: Add yield_... |
5296 |
.yield_to_task = yield_to_task_fair, |
bf0f6f24a sched: cfs core, ... |
5297 |
|
2e09bf556 sched: wakeup gra... |
5298 |
.check_preempt_curr = check_preempt_wakeup, |
bf0f6f24a sched: cfs core, ... |
5299 5300 5301 |
.pick_next_task = pick_next_task_fair, .put_prev_task = put_prev_task_fair, |
681f3e685 sched: isolate SM... |
5302 |
#ifdef CONFIG_SMP |
4ce72a2c0 sched: add CONFIG... |
5303 |
.select_task_rq = select_task_rq_fair, |
0bcdcf28c sched: Fix missin... |
5304 5305 |
.rq_online = rq_online_fair, .rq_offline = rq_offline_fair, |
88ec22d3e sched: Remove the... |
5306 5307 |
.task_waking = task_waking_fair, |
681f3e685 sched: isolate SM... |
5308 |
#endif |
bf0f6f24a sched: cfs core, ... |
5309 |
|
83b699ed2 sched: revert rec... |
5310 |
.set_curr_task = set_curr_task_fair, |
bf0f6f24a sched: cfs core, ... |
5311 |
.task_tick = task_tick_fair, |
cd29fe6f2 sched: Sanitize f... |
5312 |
.task_fork = task_fork_fair, |
cb4698450 sched: RT-balance... |
5313 5314 |
.prio_changed = prio_changed_fair, |
da7a735e5 sched: Fix switch... |
5315 |
.switched_from = switched_from_fair, |
cb4698450 sched: RT-balance... |
5316 |
.switched_to = switched_to_fair, |
810b38179 sched: retain vru... |
5317 |
|
0d721cead sched: Simplify s... |
5318 |
.get_rr_interval = get_rr_interval_fair, |
810b38179 sched: retain vru... |
5319 |
#ifdef CONFIG_FAIR_GROUP_SCHED |
b2b5ce022 sched, cgroup: Fi... |
5320 |
.task_move_group = task_move_group_fair, |
810b38179 sched: retain vru... |
5321 |
#endif |
bf0f6f24a sched: cfs core, ... |
5322 5323 5324 |
}; #ifdef CONFIG_SCHED_DEBUG |
029632fbb sched: Make separ... |
5325 |
void print_cfs_stats(struct seq_file *m, int cpu) |
bf0f6f24a sched: cfs core, ... |
5326 |
{ |
bf0f6f24a sched: cfs core, ... |
5327 |
struct cfs_rq *cfs_rq; |
5973e5b95 sched: fix: don't... |
5328 |
rcu_read_lock(); |
c3b64f1e4 sched: clean up s... |
5329 |
for_each_leaf_cfs_rq(cpu_rq(cpu), cfs_rq) |
5cef9eca3 sched: remove the... |
5330 |
print_cfs_rq(m, cpu, cfs_rq); |
5973e5b95 sched: fix: don't... |
5331 |
rcu_read_unlock(); |
bf0f6f24a sched: cfs core, ... |
5332 5333 |
} #endif |
029632fbb sched: Make separ... |
5334 5335 5336 5337 5338 5339 5340 5341 |
__init void init_sched_fair_class(void) { #ifdef CONFIG_SMP open_softirq(SCHED_SOFTIRQ, run_rebalance_domains); #ifdef CONFIG_NO_HZ zalloc_cpumask_var(&nohz.idle_cpus_mask, GFP_NOWAIT); |
029632fbb sched: Make separ... |
5342 5343 5344 5345 |
#endif #endif /* SMP */ } |