Loading...
Note: File does not exist in v4.6.
1/*
2 * kernel/sched/proc.c
3 *
4 * Kernel load calculations, forked from sched/core.c
5 */
6
7#include <linux/export.h>
8
9#include "sched.h"
10
11unsigned long this_cpu_load(void)
12{
13 struct rq *this = this_rq();
14 return this->cpu_load[0];
15}
16
17
18/*
19 * Global load-average calculations
20 *
21 * We take a distributed and async approach to calculating the global load-avg
22 * in order to minimize overhead.
23 *
24 * The global load average is an exponentially decaying average of nr_running +
25 * nr_uninterruptible.
26 *
27 * Once every LOAD_FREQ:
28 *
29 * nr_active = 0;
30 * for_each_possible_cpu(cpu)
31 * nr_active += cpu_of(cpu)->nr_running + cpu_of(cpu)->nr_uninterruptible;
32 *
33 * avenrun[n] = avenrun[0] * exp_n + nr_active * (1 - exp_n)
34 *
35 * Due to a number of reasons the above turns in the mess below:
36 *
37 * - for_each_possible_cpu() is prohibitively expensive on machines with
38 * serious number of cpus, therefore we need to take a distributed approach
39 * to calculating nr_active.
40 *
41 * \Sum_i x_i(t) = \Sum_i x_i(t) - x_i(t_0) | x_i(t_0) := 0
42 * = \Sum_i { \Sum_j=1 x_i(t_j) - x_i(t_j-1) }
43 *
44 * So assuming nr_active := 0 when we start out -- true per definition, we
45 * can simply take per-cpu deltas and fold those into a global accumulate
46 * to obtain the same result. See calc_load_fold_active().
47 *
48 * Furthermore, in order to avoid synchronizing all per-cpu delta folding
49 * across the machine, we assume 10 ticks is sufficient time for every
50 * cpu to have completed this task.
51 *
52 * This places an upper-bound on the IRQ-off latency of the machine. Then
53 * again, being late doesn't loose the delta, just wrecks the sample.
54 *
55 * - cpu_rq()->nr_uninterruptible isn't accurately tracked per-cpu because
56 * this would add another cross-cpu cacheline miss and atomic operation
57 * to the wakeup path. Instead we increment on whatever cpu the task ran
58 * when it went into uninterruptible state and decrement on whatever cpu
59 * did the wakeup. This means that only the sum of nr_uninterruptible over
60 * all cpus yields the correct result.
61 *
62 * This covers the NO_HZ=n code, for extra head-aches, see the comment below.
63 */
64
65/* Variables and functions for calc_load */
66atomic_long_t calc_load_tasks;
67unsigned long calc_load_update;
68unsigned long avenrun[3];
69EXPORT_SYMBOL(avenrun); /* should be removed */
70
71/**
72 * get_avenrun - get the load average array
73 * @loads: pointer to dest load array
74 * @offset: offset to add
75 * @shift: shift count to shift the result left
76 *
77 * These values are estimates at best, so no need for locking.
78 */
79void get_avenrun(unsigned long *loads, unsigned long offset, int shift)
80{
81 loads[0] = (avenrun[0] + offset) << shift;
82 loads[1] = (avenrun[1] + offset) << shift;
83 loads[2] = (avenrun[2] + offset) << shift;
84}
85
86long calc_load_fold_active(struct rq *this_rq)
87{
88 long nr_active, delta = 0;
89
90 nr_active = this_rq->nr_running;
91 nr_active += (long) this_rq->nr_uninterruptible;
92
93 if (nr_active != this_rq->calc_load_active) {
94 delta = nr_active - this_rq->calc_load_active;
95 this_rq->calc_load_active = nr_active;
96 }
97
98 return delta;
99}
100
101/*
102 * a1 = a0 * e + a * (1 - e)
103 */
104static unsigned long
105calc_load(unsigned long load, unsigned long exp, unsigned long active)
106{
107 load *= exp;
108 load += active * (FIXED_1 - exp);
109 load += 1UL << (FSHIFT - 1);
110 return load >> FSHIFT;
111}
112
113#ifdef CONFIG_NO_HZ_COMMON
114/*
115 * Handle NO_HZ for the global load-average.
116 *
117 * Since the above described distributed algorithm to compute the global
118 * load-average relies on per-cpu sampling from the tick, it is affected by
119 * NO_HZ.
120 *
121 * The basic idea is to fold the nr_active delta into a global idle-delta upon
122 * entering NO_HZ state such that we can include this as an 'extra' cpu delta
123 * when we read the global state.
124 *
125 * Obviously reality has to ruin such a delightfully simple scheme:
126 *
127 * - When we go NO_HZ idle during the window, we can negate our sample
128 * contribution, causing under-accounting.
129 *
130 * We avoid this by keeping two idle-delta counters and flipping them
131 * when the window starts, thus separating old and new NO_HZ load.
132 *
133 * The only trick is the slight shift in index flip for read vs write.
134 *
135 * 0s 5s 10s 15s
136 * +10 +10 +10 +10
137 * |-|-----------|-|-----------|-|-----------|-|
138 * r:0 0 1 1 0 0 1 1 0
139 * w:0 1 1 0 0 1 1 0 0
140 *
141 * This ensures we'll fold the old idle contribution in this window while
142 * accumlating the new one.
143 *
144 * - When we wake up from NO_HZ idle during the window, we push up our
145 * contribution, since we effectively move our sample point to a known
146 * busy state.
147 *
148 * This is solved by pushing the window forward, and thus skipping the
149 * sample, for this cpu (effectively using the idle-delta for this cpu which
150 * was in effect at the time the window opened). This also solves the issue
151 * of having to deal with a cpu having been in NOHZ idle for multiple
152 * LOAD_FREQ intervals.
153 *
154 * When making the ILB scale, we should try to pull this in as well.
155 */
156static atomic_long_t calc_load_idle[2];
157static int calc_load_idx;
158
159static inline int calc_load_write_idx(void)
160{
161 int idx = calc_load_idx;
162
163 /*
164 * See calc_global_nohz(), if we observe the new index, we also
165 * need to observe the new update time.
166 */
167 smp_rmb();
168
169 /*
170 * If the folding window started, make sure we start writing in the
171 * next idle-delta.
172 */
173 if (!time_before(jiffies, calc_load_update))
174 idx++;
175
176 return idx & 1;
177}
178
179static inline int calc_load_read_idx(void)
180{
181 return calc_load_idx & 1;
182}
183
184void calc_load_enter_idle(void)
185{
186 struct rq *this_rq = this_rq();
187 long delta;
188
189 /*
190 * We're going into NOHZ mode, if there's any pending delta, fold it
191 * into the pending idle delta.
192 */
193 delta = calc_load_fold_active(this_rq);
194 if (delta) {
195 int idx = calc_load_write_idx();
196 atomic_long_add(delta, &calc_load_idle[idx]);
197 }
198}
199
200void calc_load_exit_idle(void)
201{
202 struct rq *this_rq = this_rq();
203
204 /*
205 * If we're still before the sample window, we're done.
206 */
207 if (time_before(jiffies, this_rq->calc_load_update))
208 return;
209
210 /*
211 * We woke inside or after the sample window, this means we're already
212 * accounted through the nohz accounting, so skip the entire deal and
213 * sync up for the next window.
214 */
215 this_rq->calc_load_update = calc_load_update;
216 if (time_before(jiffies, this_rq->calc_load_update + 10))
217 this_rq->calc_load_update += LOAD_FREQ;
218}
219
220static long calc_load_fold_idle(void)
221{
222 int idx = calc_load_read_idx();
223 long delta = 0;
224
225 if (atomic_long_read(&calc_load_idle[idx]))
226 delta = atomic_long_xchg(&calc_load_idle[idx], 0);
227
228 return delta;
229}
230
231/**
232 * fixed_power_int - compute: x^n, in O(log n) time
233 *
234 * @x: base of the power
235 * @frac_bits: fractional bits of @x
236 * @n: power to raise @x to.
237 *
238 * By exploiting the relation between the definition of the natural power
239 * function: x^n := x*x*...*x (x multiplied by itself for n times), and
240 * the binary encoding of numbers used by computers: n := \Sum n_i * 2^i,
241 * (where: n_i \elem {0, 1}, the binary vector representing n),
242 * we find: x^n := x^(\Sum n_i * 2^i) := \Prod x^(n_i * 2^i), which is
243 * of course trivially computable in O(log_2 n), the length of our binary
244 * vector.
245 */
246static unsigned long
247fixed_power_int(unsigned long x, unsigned int frac_bits, unsigned int n)
248{
249 unsigned long result = 1UL << frac_bits;
250
251 if (n) for (;;) {
252 if (n & 1) {
253 result *= x;
254 result += 1UL << (frac_bits - 1);
255 result >>= frac_bits;
256 }
257 n >>= 1;
258 if (!n)
259 break;
260 x *= x;
261 x += 1UL << (frac_bits - 1);
262 x >>= frac_bits;
263 }
264
265 return result;
266}
267
268/*
269 * a1 = a0 * e + a * (1 - e)
270 *
271 * a2 = a1 * e + a * (1 - e)
272 * = (a0 * e + a * (1 - e)) * e + a * (1 - e)
273 * = a0 * e^2 + a * (1 - e) * (1 + e)
274 *
275 * a3 = a2 * e + a * (1 - e)
276 * = (a0 * e^2 + a * (1 - e) * (1 + e)) * e + a * (1 - e)
277 * = a0 * e^3 + a * (1 - e) * (1 + e + e^2)
278 *
279 * ...
280 *
281 * an = a0 * e^n + a * (1 - e) * (1 + e + ... + e^n-1) [1]
282 * = a0 * e^n + a * (1 - e) * (1 - e^n)/(1 - e)
283 * = a0 * e^n + a * (1 - e^n)
284 *
285 * [1] application of the geometric series:
286 *
287 * n 1 - x^(n+1)
288 * S_n := \Sum x^i = -------------
289 * i=0 1 - x
290 */
291static unsigned long
292calc_load_n(unsigned long load, unsigned long exp,
293 unsigned long active, unsigned int n)
294{
295
296 return calc_load(load, fixed_power_int(exp, FSHIFT, n), active);
297}
298
299/*
300 * NO_HZ can leave us missing all per-cpu ticks calling
301 * calc_load_account_active(), but since an idle CPU folds its delta into
302 * calc_load_tasks_idle per calc_load_account_idle(), all we need to do is fold
303 * in the pending idle delta if our idle period crossed a load cycle boundary.
304 *
305 * Once we've updated the global active value, we need to apply the exponential
306 * weights adjusted to the number of cycles missed.
307 */
308static void calc_global_nohz(void)
309{
310 long delta, active, n;
311
312 if (!time_before(jiffies, calc_load_update + 10)) {
313 /*
314 * Catch-up, fold however many we are behind still
315 */
316 delta = jiffies - calc_load_update - 10;
317 n = 1 + (delta / LOAD_FREQ);
318
319 active = atomic_long_read(&calc_load_tasks);
320 active = active > 0 ? active * FIXED_1 : 0;
321
322 avenrun[0] = calc_load_n(avenrun[0], EXP_1, active, n);
323 avenrun[1] = calc_load_n(avenrun[1], EXP_5, active, n);
324 avenrun[2] = calc_load_n(avenrun[2], EXP_15, active, n);
325
326 calc_load_update += n * LOAD_FREQ;
327 }
328
329 /*
330 * Flip the idle index...
331 *
332 * Make sure we first write the new time then flip the index, so that
333 * calc_load_write_idx() will see the new time when it reads the new
334 * index, this avoids a double flip messing things up.
335 */
336 smp_wmb();
337 calc_load_idx++;
338}
339#else /* !CONFIG_NO_HZ_COMMON */
340
341static inline long calc_load_fold_idle(void) { return 0; }
342static inline void calc_global_nohz(void) { }
343
344#endif /* CONFIG_NO_HZ_COMMON */
345
346/*
347 * calc_load - update the avenrun load estimates 10 ticks after the
348 * CPUs have updated calc_load_tasks.
349 */
350void calc_global_load(unsigned long ticks)
351{
352 long active, delta;
353
354 if (time_before(jiffies, calc_load_update + 10))
355 return;
356
357 /*
358 * Fold the 'old' idle-delta to include all NO_HZ cpus.
359 */
360 delta = calc_load_fold_idle();
361 if (delta)
362 atomic_long_add(delta, &calc_load_tasks);
363
364 active = atomic_long_read(&calc_load_tasks);
365 active = active > 0 ? active * FIXED_1 : 0;
366
367 avenrun[0] = calc_load(avenrun[0], EXP_1, active);
368 avenrun[1] = calc_load(avenrun[1], EXP_5, active);
369 avenrun[2] = calc_load(avenrun[2], EXP_15, active);
370
371 calc_load_update += LOAD_FREQ;
372
373 /*
374 * In case we idled for multiple LOAD_FREQ intervals, catch up in bulk.
375 */
376 calc_global_nohz();
377}
378
379/*
380 * Called from update_cpu_load() to periodically update this CPU's
381 * active count.
382 */
383static void calc_load_account_active(struct rq *this_rq)
384{
385 long delta;
386
387 if (time_before(jiffies, this_rq->calc_load_update))
388 return;
389
390 delta = calc_load_fold_active(this_rq);
391 if (delta)
392 atomic_long_add(delta, &calc_load_tasks);
393
394 this_rq->calc_load_update += LOAD_FREQ;
395}
396
397/*
398 * End of global load-average stuff
399 */
400
401/*
402 * The exact cpuload at various idx values, calculated at every tick would be
403 * load = (2^idx - 1) / 2^idx * load + 1 / 2^idx * cur_load
404 *
405 * If a cpu misses updates for n-1 ticks (as it was idle) and update gets called
406 * on nth tick when cpu may be busy, then we have:
407 * load = ((2^idx - 1) / 2^idx)^(n-1) * load
408 * load = (2^idx - 1) / 2^idx) * load + 1 / 2^idx * cur_load
409 *
410 * decay_load_missed() below does efficient calculation of
411 * load = ((2^idx - 1) / 2^idx)^(n-1) * load
412 * avoiding 0..n-1 loop doing load = ((2^idx - 1) / 2^idx) * load
413 *
414 * The calculation is approximated on a 128 point scale.
415 * degrade_zero_ticks is the number of ticks after which load at any
416 * particular idx is approximated to be zero.
417 * degrade_factor is a precomputed table, a row for each load idx.
418 * Each column corresponds to degradation factor for a power of two ticks,
419 * based on 128 point scale.
420 * Example:
421 * row 2, col 3 (=12) says that the degradation at load idx 2 after
422 * 8 ticks is 12/128 (which is an approximation of exact factor 3^8/4^8).
423 *
424 * With this power of 2 load factors, we can degrade the load n times
425 * by looking at 1 bits in n and doing as many mult/shift instead of
426 * n mult/shifts needed by the exact degradation.
427 */
428#define DEGRADE_SHIFT 7
429static const unsigned char
430 degrade_zero_ticks[CPU_LOAD_IDX_MAX] = {0, 8, 32, 64, 128};
431static const unsigned char
432 degrade_factor[CPU_LOAD_IDX_MAX][DEGRADE_SHIFT + 1] = {
433 {0, 0, 0, 0, 0, 0, 0, 0},
434 {64, 32, 8, 0, 0, 0, 0, 0},
435 {96, 72, 40, 12, 1, 0, 0},
436 {112, 98, 75, 43, 15, 1, 0},
437 {120, 112, 98, 76, 45, 16, 2} };
438
439/*
440 * Update cpu_load for any missed ticks, due to tickless idle. The backlog
441 * would be when CPU is idle and so we just decay the old load without
442 * adding any new load.
443 */
444static unsigned long
445decay_load_missed(unsigned long load, unsigned long missed_updates, int idx)
446{
447 int j = 0;
448
449 if (!missed_updates)
450 return load;
451
452 if (missed_updates >= degrade_zero_ticks[idx])
453 return 0;
454
455 if (idx == 1)
456 return load >> missed_updates;
457
458 while (missed_updates) {
459 if (missed_updates % 2)
460 load = (load * degrade_factor[idx][j]) >> DEGRADE_SHIFT;
461
462 missed_updates >>= 1;
463 j++;
464 }
465 return load;
466}
467
468/*
469 * Update rq->cpu_load[] statistics. This function is usually called every
470 * scheduler tick (TICK_NSEC). With tickless idle this will not be called
471 * every tick. We fix it up based on jiffies.
472 */
473static void __update_cpu_load(struct rq *this_rq, unsigned long this_load,
474 unsigned long pending_updates)
475{
476 int i, scale;
477
478 this_rq->nr_load_updates++;
479
480 /* Update our load: */
481 this_rq->cpu_load[0] = this_load; /* Fasttrack for idx 0 */
482 for (i = 1, scale = 2; i < CPU_LOAD_IDX_MAX; i++, scale += scale) {
483 unsigned long old_load, new_load;
484
485 /* scale is effectively 1 << i now, and >> i divides by scale */
486
487 old_load = this_rq->cpu_load[i];
488 old_load = decay_load_missed(old_load, pending_updates - 1, i);
489 new_load = this_load;
490 /*
491 * Round up the averaging division if load is increasing. This
492 * prevents us from getting stuck on 9 if the load is 10, for
493 * example.
494 */
495 if (new_load > old_load)
496 new_load += scale - 1;
497
498 this_rq->cpu_load[i] = (old_load * (scale - 1) + new_load) >> i;
499 }
500
501 sched_avg_update(this_rq);
502}
503
504#ifdef CONFIG_SMP
505static inline unsigned long get_rq_runnable_load(struct rq *rq)
506{
507 return rq->cfs.runnable_load_avg;
508}
509#else
510static inline unsigned long get_rq_runnable_load(struct rq *rq)
511{
512 return rq->load.weight;
513}
514#endif
515
516#ifdef CONFIG_NO_HZ_COMMON
517/*
518 * There is no sane way to deal with nohz on smp when using jiffies because the
519 * cpu doing the jiffies update might drift wrt the cpu doing the jiffy reading
520 * causing off-by-one errors in observed deltas; {0,2} instead of {1,1}.
521 *
522 * Therefore we cannot use the delta approach from the regular tick since that
523 * would seriously skew the load calculation. However we'll make do for those
524 * updates happening while idle (nohz_idle_balance) or coming out of idle
525 * (tick_nohz_idle_exit).
526 *
527 * This means we might still be one tick off for nohz periods.
528 */
529
530/*
531 * Called from nohz_idle_balance() to update the load ratings before doing the
532 * idle balance.
533 */
534void update_idle_cpu_load(struct rq *this_rq)
535{
536 unsigned long curr_jiffies = ACCESS_ONCE(jiffies);
537 unsigned long load = get_rq_runnable_load(this_rq);
538 unsigned long pending_updates;
539
540 /*
541 * bail if there's load or we're actually up-to-date.
542 */
543 if (load || curr_jiffies == this_rq->last_load_update_tick)
544 return;
545
546 pending_updates = curr_jiffies - this_rq->last_load_update_tick;
547 this_rq->last_load_update_tick = curr_jiffies;
548
549 __update_cpu_load(this_rq, load, pending_updates);
550}
551
552/*
553 * Called from tick_nohz_idle_exit() -- try and fix up the ticks we missed.
554 */
555void update_cpu_load_nohz(void)
556{
557 struct rq *this_rq = this_rq();
558 unsigned long curr_jiffies = ACCESS_ONCE(jiffies);
559 unsigned long pending_updates;
560
561 if (curr_jiffies == this_rq->last_load_update_tick)
562 return;
563
564 raw_spin_lock(&this_rq->lock);
565 pending_updates = curr_jiffies - this_rq->last_load_update_tick;
566 if (pending_updates) {
567 this_rq->last_load_update_tick = curr_jiffies;
568 /*
569 * We were idle, this means load 0, the current load might be
570 * !0 due to remote wakeups and the sort.
571 */
572 __update_cpu_load(this_rq, 0, pending_updates);
573 }
574 raw_spin_unlock(&this_rq->lock);
575}
576#endif /* CONFIG_NO_HZ */
577
578/*
579 * Called from scheduler_tick()
580 */
581void update_cpu_load_active(struct rq *this_rq)
582{
583 unsigned long load = get_rq_runnable_load(this_rq);
584 /*
585 * See the mess around update_idle_cpu_load() / update_cpu_load_nohz().
586 */
587 this_rq->last_load_update_tick = jiffies;
588 __update_cpu_load(this_rq, load, 1);
589
590 calc_load_account_active(this_rq);
591}