blob: fe343a06b7da3278ba68c9c15c261ea14afde685 [file] [log] [blame]
Len Brown4f86d3a2007-10-03 18:58:00 -04001/*
2 * menu.c - the menu idle governor
3 *
4 * Copyright (C) 2006-2007 Adam Belay <abelay@novell.com>
Arjan van de Ven69d25872009-09-21 17:04:08 -07005 * Copyright (C) 2009 Intel Corporation
6 * Author:
7 * Arjan van de Ven <arjan@linux.intel.com>
Len Brown4f86d3a2007-10-03 18:58:00 -04008 *
Arjan van de Ven69d25872009-09-21 17:04:08 -07009 * This code is licenced under the GPL version 2 as described
10 * in the COPYING file that acompanies the Linux Kernel.
Len Brown4f86d3a2007-10-03 18:58:00 -040011 */
12
13#include <linux/kernel.h>
14#include <linux/cpuidle.h>
Jean Pihete8db0be2011-08-25 15:35:03 +020015#include <linux/pm_qos.h>
Len Brown4f86d3a2007-10-03 18:58:00 -040016#include <linux/time.h>
17#include <linux/ktime.h>
18#include <linux/hrtimer.h>
19#include <linux/tick.h>
Arjan van de Ven69d25872009-09-21 17:04:08 -070020#include <linux/sched.h>
Stephen Hemminger57875362010-01-08 14:43:08 -080021#include <linux/math64.h>
Paul Gortmaker884b17e2011-08-29 17:52:39 -040022#include <linux/module.h>
Len Brown4f86d3a2007-10-03 18:58:00 -040023
Arjan van de Ven69d25872009-09-21 17:04:08 -070024#define BUCKETS 12
Arjan van de Ven1f85f872010-05-24 14:32:59 -070025#define INTERVALS 8
Arjan van de Ven69d25872009-09-21 17:04:08 -070026#define RESOLUTION 1024
Arjan van de Ven1f85f872010-05-24 14:32:59 -070027#define DECAY 8
Arjan van de Ven69d25872009-09-21 17:04:08 -070028#define MAX_INTERESTING 50000
Arjan van de Ven1f85f872010-05-24 14:32:59 -070029#define STDDEV_THRESH 400
30
Youquan Song69a37be2012-10-26 12:26:41 +020031/* 60 * 60 > STDDEV_THRESH * INTERVALS = 400 * 8 */
32#define MAX_DEVIATION 60
33
34static DEFINE_PER_CPU(struct hrtimer, menu_hrtimer);
35static DEFINE_PER_CPU(int, hrtimer_status);
36/* menu hrtimer mode */
Youquan Songe11538d2012-10-26 12:26:50 +020037enum {MENU_HRTIMER_STOP, MENU_HRTIMER_REPEAT, MENU_HRTIMER_GENERAL};
Arjan van de Ven69d25872009-09-21 17:04:08 -070038
39/*
40 * Concepts and ideas behind the menu governor
41 *
42 * For the menu governor, there are 3 decision factors for picking a C
43 * state:
44 * 1) Energy break even point
45 * 2) Performance impact
46 * 3) Latency tolerance (from pmqos infrastructure)
47 * These these three factors are treated independently.
48 *
49 * Energy break even point
50 * -----------------------
51 * C state entry and exit have an energy cost, and a certain amount of time in
52 * the C state is required to actually break even on this cost. CPUIDLE
53 * provides us this duration in the "target_residency" field. So all that we
54 * need is a good prediction of how long we'll be idle. Like the traditional
55 * menu governor, we start with the actual known "next timer event" time.
56 *
57 * Since there are other source of wakeups (interrupts for example) than
58 * the next timer event, this estimation is rather optimistic. To get a
59 * more realistic estimate, a correction factor is applied to the estimate,
60 * that is based on historic behavior. For example, if in the past the actual
61 * duration always was 50% of the next timer tick, the correction factor will
62 * be 0.5.
63 *
64 * menu uses a running average for this correction factor, however it uses a
65 * set of factors, not just a single factor. This stems from the realization
66 * that the ratio is dependent on the order of magnitude of the expected
67 * duration; if we expect 500 milliseconds of idle time the likelihood of
68 * getting an interrupt very early is much higher than if we expect 50 micro
69 * seconds of idle time. A second independent factor that has big impact on
70 * the actual factor is if there is (disk) IO outstanding or not.
71 * (as a special twist, we consider every sleep longer than 50 milliseconds
72 * as perfect; there are no power gains for sleeping longer than this)
73 *
74 * For these two reasons we keep an array of 12 independent factors, that gets
75 * indexed based on the magnitude of the expected duration as well as the
76 * "is IO outstanding" property.
77 *
Arjan van de Ven1f85f872010-05-24 14:32:59 -070078 * Repeatable-interval-detector
79 * ----------------------------
80 * There are some cases where "next timer" is a completely unusable predictor:
81 * Those cases where the interval is fixed, for example due to hardware
82 * interrupt mitigation, but also due to fixed transfer rate devices such as
83 * mice.
84 * For this, we use a different predictor: We track the duration of the last 8
85 * intervals and if the stand deviation of these 8 intervals is below a
86 * threshold value, we use the average of these intervals as prediction.
87 *
Arjan van de Ven69d25872009-09-21 17:04:08 -070088 * Limiting Performance Impact
89 * ---------------------------
90 * C states, especially those with large exit latencies, can have a real
Lucas De Marchi20e33412010-09-07 12:53:49 -040091 * noticeable impact on workloads, which is not acceptable for most sysadmins,
Arjan van de Ven69d25872009-09-21 17:04:08 -070092 * and in addition, less performance has a power price of its own.
93 *
94 * As a general rule of thumb, menu assumes that the following heuristic
95 * holds:
96 * The busier the system, the less impact of C states is acceptable
97 *
98 * This rule-of-thumb is implemented using a performance-multiplier:
99 * If the exit latency times the performance multiplier is longer than
100 * the predicted duration, the C state is not considered a candidate
101 * for selection due to a too high performance impact. So the higher
102 * this multiplier is, the longer we need to be idle to pick a deep C
103 * state, and thus the less likely a busy CPU will hit such a deep
104 * C state.
105 *
106 * Two factors are used in determing this multiplier:
107 * a value of 10 is added for each point of "per cpu load average" we have.
108 * a value of 5 points is added for each process that is waiting for
109 * IO on this CPU.
110 * (these values are experimentally determined)
111 *
112 * The load average factor gives a longer term (few seconds) input to the
113 * decision, while the iowait value gives a cpu local instantanious input.
114 * The iowait factor may look low, but realize that this is also already
115 * represented in the system load average.
116 *
117 */
Len Brown4f86d3a2007-10-03 18:58:00 -0400118
Youquan Songe11538d2012-10-26 12:26:50 +0200119/*
120 * The C-state residency is so long that is is worthwhile to exit
121 * from the shallow C-state and re-enter into a deeper C-state.
122 */
123static unsigned int perfect_cstate_ms __read_mostly = 30;
124module_param(perfect_cstate_ms, uint, 0000);
125
Len Brown4f86d3a2007-10-03 18:58:00 -0400126struct menu_device {
127 int last_state_idx;
Corrado Zoccolo672917d2009-09-21 17:04:09 -0700128 int needs_update;
Len Brown4f86d3a2007-10-03 18:58:00 -0400129
130 unsigned int expected_us;
Richard Kennedy56e6943b42010-03-05 13:42:30 -0800131 u64 predicted_us;
Arjan van de Ven69d25872009-09-21 17:04:08 -0700132 unsigned int exit_us;
133 unsigned int bucket;
134 u64 correction_factor[BUCKETS];
Arjan van de Ven1f85f872010-05-24 14:32:59 -0700135 u32 intervals[INTERVALS];
136 int interval_ptr;
Len Brown4f86d3a2007-10-03 18:58:00 -0400137};
138
Arjan van de Ven69d25872009-09-21 17:04:08 -0700139
140#define LOAD_INT(x) ((x) >> FSHIFT)
141#define LOAD_FRAC(x) LOAD_INT(((x) & (FIXED_1-1)) * 100)
142
143static int get_loadavg(void)
144{
145 unsigned long this = this_cpu_load();
146
147
148 return LOAD_INT(this) * 10 + LOAD_FRAC(this) / 10;
149}
150
151static inline int which_bucket(unsigned int duration)
152{
153 int bucket = 0;
154
155 /*
156 * We keep two groups of stats; one with no
157 * IO pending, one without.
158 * This allows us to calculate
159 * E(duration)|iowait
160 */
Peter Zijlstra8c215bd2010-07-01 09:07:17 +0200161 if (nr_iowait_cpu(smp_processor_id()))
Arjan van de Ven69d25872009-09-21 17:04:08 -0700162 bucket = BUCKETS/2;
163
164 if (duration < 10)
165 return bucket;
166 if (duration < 100)
167 return bucket + 1;
168 if (duration < 1000)
169 return bucket + 2;
170 if (duration < 10000)
171 return bucket + 3;
172 if (duration < 100000)
173 return bucket + 4;
174 return bucket + 5;
175}
176
177/*
178 * Return a multiplier for the exit latency that is intended
179 * to take performance requirements into account.
180 * The more performance critical we estimate the system
181 * to be, the higher this multiplier, and thus the higher
182 * the barrier to go to an expensive C state.
183 */
184static inline int performance_multiplier(void)
185{
186 int mult = 1;
187
188 /* for higher loadavg, we are more reluctant */
189
190 mult += 2 * get_loadavg();
191
192 /* for IO wait tasks (per cpu!) we add 5x each */
Peter Zijlstra8c215bd2010-07-01 09:07:17 +0200193 mult += 10 * nr_iowait_cpu(smp_processor_id());
Arjan van de Ven69d25872009-09-21 17:04:08 -0700194
195 return mult;
196}
197
Len Brown4f86d3a2007-10-03 18:58:00 -0400198static DEFINE_PER_CPU(struct menu_device, menu_devices);
199
Deepthi Dharwar46bcfad2011-10-28 16:20:42 +0530200static void menu_update(struct cpuidle_driver *drv, struct cpuidle_device *dev);
Corrado Zoccolo672917d2009-09-21 17:04:09 -0700201
Stephen Hemminger57875362010-01-08 14:43:08 -0800202/* This implements DIV_ROUND_CLOSEST but avoids 64 bit division */
203static u64 div_round64(u64 dividend, u32 divisor)
204{
205 return div_u64(dividend + (divisor / 2), divisor);
206}
207
Youquan Song69a37be2012-10-26 12:26:41 +0200208/* Cancel the hrtimer if it is not triggered yet */
209void menu_hrtimer_cancel(void)
210{
211 int cpu = smp_processor_id();
212 struct hrtimer *hrtmr = &per_cpu(menu_hrtimer, cpu);
213
214 /* The timer is still not time out*/
215 if (per_cpu(hrtimer_status, cpu)) {
216 hrtimer_cancel(hrtmr);
217 per_cpu(hrtimer_status, cpu) = MENU_HRTIMER_STOP;
218 }
219}
220EXPORT_SYMBOL_GPL(menu_hrtimer_cancel);
221
222/* Call back for hrtimer is triggered */
223static enum hrtimer_restart menu_hrtimer_notify(struct hrtimer *hrtimer)
224{
225 int cpu = smp_processor_id();
Youquan Songe11538d2012-10-26 12:26:50 +0200226 struct menu_device *data = &per_cpu(menu_devices, cpu);
227
228 /* In general case, the expected residency is much larger than
229 * deepest C-state target residency, but prediction logic still
230 * predicts a small predicted residency, so the prediction
231 * history is totally broken if the timer is triggered.
232 * So reset the correction factor.
233 */
234 if (per_cpu(hrtimer_status, cpu) == MENU_HRTIMER_GENERAL)
235 data->correction_factor[data->bucket] = RESOLUTION * DECAY;
Youquan Song69a37be2012-10-26 12:26:41 +0200236
237 per_cpu(hrtimer_status, cpu) = MENU_HRTIMER_STOP;
238
239 return HRTIMER_NORESTART;
240}
241
Arjan van de Ven1f85f872010-05-24 14:32:59 -0700242/*
243 * Try detecting repeating patterns by keeping track of the last 8
244 * intervals, and checking if the standard deviation of that set
245 * of points is below a threshold. If it is... then use the
246 * average of these 8 points as the estimated value.
247 */
Youquan Songc96ca4f2012-10-26 12:27:07 +0200248static u32 get_typical_interval(struct menu_device *data)
Arjan van de Ven1f85f872010-05-24 14:32:59 -0700249{
Youquan Songc96ca4f2012-10-26 12:27:07 +0200250 int i = 0, divisor = 0;
251 uint64_t max = 0, avg = 0, stddev = 0;
252 int64_t thresh = LLONG_MAX; /* Discard outliers above this value. */
253 unsigned int ret = 0;
254
255again:
Arjan van de Ven1f85f872010-05-24 14:32:59 -0700256
257 /* first calculate average and standard deviation of the past */
Youquan Songc96ca4f2012-10-26 12:27:07 +0200258 max = avg = divisor = stddev = 0;
259 for (i = 0; i < INTERVALS; i++) {
260 int64_t value = data->intervals[i];
261 if (value <= thresh) {
262 avg += value;
263 divisor++;
264 if (value > max)
265 max = value;
266 }
267 }
268 do_div(avg, divisor);
Arjan van de Ven1f85f872010-05-24 14:32:59 -0700269
Youquan Songc96ca4f2012-10-26 12:27:07 +0200270 for (i = 0; i < INTERVALS; i++) {
271 int64_t value = data->intervals[i];
272 if (value <= thresh) {
273 int64_t diff = value - avg;
274 stddev += diff * diff;
275 }
276 }
277 do_div(stddev, divisor);
278 stddev = int_sqrt(stddev);
Arjan van de Ven1f85f872010-05-24 14:32:59 -0700279 /*
Youquan Songc96ca4f2012-10-26 12:27:07 +0200280 * If we have outliers to the upside in our distribution, discard
281 * those by setting the threshold to exclude these outliers, then
282 * calculate the average and standard deviation again. Once we get
283 * down to the bottom 3/4 of our samples, stop excluding samples.
284 *
285 * This can deal with workloads that have long pauses interspersed
286 * with sporadic activity with a bunch of short pauses.
287 *
288 * The typical interval is obtained when standard deviation is small
289 * or standard deviation is small compared to the average interval.
Arjan van de Ven1f85f872010-05-24 14:32:59 -0700290 */
Youquan Songc96ca4f2012-10-26 12:27:07 +0200291 if (((avg > stddev * 6) && (divisor * 4 >= INTERVALS * 3))
292 || stddev <= 20) {
Arjan van de Ven1f85f872010-05-24 14:32:59 -0700293 data->predicted_us = avg;
Youquan Song69a37be2012-10-26 12:26:41 +0200294 ret = 1;
Youquan Songc96ca4f2012-10-26 12:27:07 +0200295 return ret;
296
297 } else if ((divisor * 4) > INTERVALS * 3) {
298 /* Exclude the max interval */
299 thresh = max - 1;
300 goto again;
Youquan Song69a37be2012-10-26 12:26:41 +0200301 }
302
303 return ret;
Arjan van de Ven1f85f872010-05-24 14:32:59 -0700304}
305
Len Brown4f86d3a2007-10-03 18:58:00 -0400306/**
307 * menu_select - selects the next idle state to enter
Deepthi Dharwar46bcfad2011-10-28 16:20:42 +0530308 * @drv: cpuidle driver containing state data
Len Brown4f86d3a2007-10-03 18:58:00 -0400309 * @dev: the CPU
310 */
Deepthi Dharwar46bcfad2011-10-28 16:20:42 +0530311static int menu_select(struct cpuidle_driver *drv, struct cpuidle_device *dev)
Len Brown4f86d3a2007-10-03 18:58:00 -0400312{
313 struct menu_device *data = &__get_cpu_var(menu_devices);
Mark Grossed771342010-05-06 01:59:26 +0200314 int latency_req = pm_qos_request(PM_QOS_CPU_DMA_LATENCY);
Len Brown4f86d3a2007-10-03 18:58:00 -0400315 int i;
Arjan van de Ven69d25872009-09-21 17:04:08 -0700316 int multiplier;
Tero Kristo74675712011-02-24 17:19:23 +0200317 struct timespec t;
Youquan Song69a37be2012-10-26 12:26:41 +0200318 int repeat = 0, low_predicted = 0;
319 int cpu = smp_processor_id();
320 struct hrtimer *hrtmr = &per_cpu(menu_hrtimer, cpu);
Arjan van de Ven69d25872009-09-21 17:04:08 -0700321
Corrado Zoccolo672917d2009-09-21 17:04:09 -0700322 if (data->needs_update) {
Deepthi Dharwar46bcfad2011-10-28 16:20:42 +0530323 menu_update(drv, dev);
Corrado Zoccolo672917d2009-09-21 17:04:09 -0700324 data->needs_update = 0;
325 }
326
Arjan van de Ven1c6fe032010-05-08 15:47:37 -0700327 data->last_state_idx = 0;
328 data->exit_us = 0;
329
venkatesh.pallipadi@intel.coma2bd92022008-07-30 19:21:42 -0700330 /* Special case when user has set very strict latency requirement */
Arjan van de Ven69d25872009-09-21 17:04:08 -0700331 if (unlikely(latency_req == 0))
venkatesh.pallipadi@intel.coma2bd92022008-07-30 19:21:42 -0700332 return 0;
venkatesh.pallipadi@intel.coma2bd92022008-07-30 19:21:42 -0700333
Arjan van de Ven69d25872009-09-21 17:04:08 -0700334 /* determine the expected residency time, round up */
Tero Kristo74675712011-02-24 17:19:23 +0200335 t = ktime_to_timespec(tick_nohz_get_sleep_length());
Len Brown4f86d3a2007-10-03 18:58:00 -0400336 data->expected_us =
Tero Kristo74675712011-02-24 17:19:23 +0200337 t.tv_sec * USEC_PER_SEC + t.tv_nsec / NSEC_PER_USEC;
Len Brown4f86d3a2007-10-03 18:58:00 -0400338
Arjan van de Ven69d25872009-09-21 17:04:08 -0700339
340 data->bucket = which_bucket(data->expected_us);
341
342 multiplier = performance_multiplier();
343
344 /*
345 * if the correction factor is 0 (eg first time init or cpu hotplug
346 * etc), we actually want to start out with a unity factor.
347 */
348 if (data->correction_factor[data->bucket] == 0)
349 data->correction_factor[data->bucket] = RESOLUTION * DECAY;
350
351 /* Make sure to round up for half microseconds */
Stephen Hemminger57875362010-01-08 14:43:08 -0800352 data->predicted_us = div_round64(data->expected_us * data->correction_factor[data->bucket],
353 RESOLUTION * DECAY);
Arjan van de Ven69d25872009-09-21 17:04:08 -0700354
Youquan Songc96ca4f2012-10-26 12:27:07 +0200355 repeat = get_typical_interval(data);
Arjan van de Ven1f85f872010-05-24 14:32:59 -0700356
Arjan van de Ven69d25872009-09-21 17:04:08 -0700357 /*
358 * We want to default to C1 (hlt), not to busy polling
359 * unless the timer is happening really really soon.
360 */
ShuoX Liu3a533962012-03-28 15:19:11 -0700361 if (data->expected_us > 5 &&
Rafael J. Wysockicbc9ef02012-07-03 19:07:42 +0200362 !drv->states[CPUIDLE_DRIVER_STATE_START].disabled &&
ShuoX Liudc7fd272012-07-03 19:05:31 +0200363 dev->states_usage[CPUIDLE_DRIVER_STATE_START].disable == 0)
Arjan van de Ven69d25872009-09-21 17:04:08 -0700364 data->last_state_idx = CPUIDLE_DRIVER_STATE_START;
365
Ai Li71abbbf2010-08-09 17:20:13 -0700366 /*
367 * Find the idle state with the lowest power while satisfying
368 * our constraints.
369 */
Deepthi Dharwar46bcfad2011-10-28 16:20:42 +0530370 for (i = CPUIDLE_DRIVER_STATE_START; i < drv->state_count; i++) {
371 struct cpuidle_state *s = &drv->states[i];
ShuoX Liudc7fd272012-07-03 19:05:31 +0200372 struct cpuidle_state_usage *su = &dev->states_usage[i];
Len Brown4f86d3a2007-10-03 18:58:00 -0400373
Rafael J. Wysockicbc9ef02012-07-03 19:07:42 +0200374 if (s->disabled || su->disable)
ShuoX Liu3a533962012-03-28 15:19:11 -0700375 continue;
Youquan Song69a37be2012-10-26 12:26:41 +0200376 if (s->target_residency > data->predicted_us) {
377 low_predicted = 1;
Ai Li71abbbf2010-08-09 17:20:13 -0700378 continue;
Youquan Song69a37be2012-10-26 12:26:41 +0200379 }
venkatesh.pallipadi@intel.coma2bd92022008-07-30 19:21:42 -0700380 if (s->exit_latency > latency_req)
Ai Li71abbbf2010-08-09 17:20:13 -0700381 continue;
Arjan van de Ven69d25872009-09-21 17:04:08 -0700382 if (s->exit_latency * multiplier > data->predicted_us)
Ai Li71abbbf2010-08-09 17:20:13 -0700383 continue;
384
Daniel Lezcano8aef33a2013-01-15 14:18:04 +0100385 data->last_state_idx = i;
386 data->exit_us = s->exit_latency;
Len Brown4f86d3a2007-10-03 18:58:00 -0400387 }
388
Youquan Song69a37be2012-10-26 12:26:41 +0200389 /* not deepest C-state chosen for low predicted residency */
390 if (low_predicted) {
391 unsigned int timer_us = 0;
Youquan Songe11538d2012-10-26 12:26:50 +0200392 unsigned int perfect_us = 0;
Youquan Song69a37be2012-10-26 12:26:41 +0200393
394 /*
395 * Set a timer to detect whether this sleep is much
396 * longer than repeat mode predicted. If the timer
397 * triggers, the code will evaluate whether to put
398 * the CPU into a deeper C-state.
399 * The timer is cancelled on CPU wakeup.
400 */
401 timer_us = 2 * (data->predicted_us + MAX_DEVIATION);
402
Youquan Songe11538d2012-10-26 12:26:50 +0200403 perfect_us = perfect_cstate_ms * 1000;
404
Youquan Song69a37be2012-10-26 12:26:41 +0200405 if (repeat && (4 * timer_us < data->expected_us)) {
Li Zhonga093b932012-11-23 00:05:03 +0100406 RCU_NONIDLE(hrtimer_start(hrtmr,
407 ns_to_ktime(1000 * timer_us),
408 HRTIMER_MODE_REL_PINNED));
Youquan Song69a37be2012-10-26 12:26:41 +0200409 /* In repeat case, menu hrtimer is started */
410 per_cpu(hrtimer_status, cpu) = MENU_HRTIMER_REPEAT;
Youquan Songe11538d2012-10-26 12:26:50 +0200411 } else if (perfect_us < data->expected_us) {
412 /*
413 * The next timer is long. This could be because
414 * we did not make a useful prediction.
415 * In that case, it makes sense to re-enter
416 * into a deeper C-state after some time.
417 */
Li Zhonga093b932012-11-23 00:05:03 +0100418 RCU_NONIDLE(hrtimer_start(hrtmr,
419 ns_to_ktime(1000 * timer_us),
420 HRTIMER_MODE_REL_PINNED));
Youquan Songe11538d2012-10-26 12:26:50 +0200421 /* In general case, menu hrtimer is started */
422 per_cpu(hrtimer_status, cpu) = MENU_HRTIMER_GENERAL;
Youquan Song69a37be2012-10-26 12:26:41 +0200423 }
Youquan Songe11538d2012-10-26 12:26:50 +0200424
Youquan Song69a37be2012-10-26 12:26:41 +0200425 }
426
Arjan van de Ven69d25872009-09-21 17:04:08 -0700427 return data->last_state_idx;
Len Brown4f86d3a2007-10-03 18:58:00 -0400428}
429
430/**
Corrado Zoccolo672917d2009-09-21 17:04:09 -0700431 * menu_reflect - records that data structures need update
Len Brown4f86d3a2007-10-03 18:58:00 -0400432 * @dev: the CPU
Deepthi Dharware978aa72011-10-28 16:20:09 +0530433 * @index: the index of actual entered state
Len Brown4f86d3a2007-10-03 18:58:00 -0400434 *
435 * NOTE: it's important to be fast here because this operation will add to
436 * the overall exit latency.
437 */
Deepthi Dharware978aa72011-10-28 16:20:09 +0530438static void menu_reflect(struct cpuidle_device *dev, int index)
Len Brown4f86d3a2007-10-03 18:58:00 -0400439{
440 struct menu_device *data = &__get_cpu_var(menu_devices);
Deepthi Dharware978aa72011-10-28 16:20:09 +0530441 data->last_state_idx = index;
442 if (index >= 0)
443 data->needs_update = 1;
Corrado Zoccolo672917d2009-09-21 17:04:09 -0700444}
445
446/**
447 * menu_update - attempts to guess what happened after entry
Deepthi Dharwar46bcfad2011-10-28 16:20:42 +0530448 * @drv: cpuidle driver containing state data
Corrado Zoccolo672917d2009-09-21 17:04:09 -0700449 * @dev: the CPU
450 */
Deepthi Dharwar46bcfad2011-10-28 16:20:42 +0530451static void menu_update(struct cpuidle_driver *drv, struct cpuidle_device *dev)
Corrado Zoccolo672917d2009-09-21 17:04:09 -0700452{
453 struct menu_device *data = &__get_cpu_var(menu_devices);
Len Brown4f86d3a2007-10-03 18:58:00 -0400454 int last_idx = data->last_state_idx;
venkatesh.pallipadi@intel.com320eee72008-07-30 19:21:43 -0700455 unsigned int last_idle_us = cpuidle_get_last_residency(dev);
Deepthi Dharwar46bcfad2011-10-28 16:20:42 +0530456 struct cpuidle_state *target = &drv->states[last_idx];
venkatesh.pallipadi@intel.com320eee72008-07-30 19:21:43 -0700457 unsigned int measured_us;
Arjan van de Ven69d25872009-09-21 17:04:08 -0700458 u64 new_factor;
Len Brown4f86d3a2007-10-03 18:58:00 -0400459
460 /*
461 * Ugh, this idle state doesn't support residency measurements, so we
462 * are basically lost in the dark. As a compromise, assume we slept
Arjan van de Ven69d25872009-09-21 17:04:08 -0700463 * for the whole expected time.
Len Brown4f86d3a2007-10-03 18:58:00 -0400464 */
venkatesh.pallipadi@intel.com320eee72008-07-30 19:21:43 -0700465 if (unlikely(!(target->flags & CPUIDLE_FLAG_TIME_VALID)))
Arjan van de Ven69d25872009-09-21 17:04:08 -0700466 last_idle_us = data->expected_us;
467
468
469 measured_us = last_idle_us;
Len Brown4f86d3a2007-10-03 18:58:00 -0400470
venkatesh.pallipadi@intel.com320eee72008-07-30 19:21:43 -0700471 /*
Arjan van de Ven69d25872009-09-21 17:04:08 -0700472 * We correct for the exit latency; we are assuming here that the
473 * exit latency happens after the event that we're interested in.
venkatesh.pallipadi@intel.com320eee72008-07-30 19:21:43 -0700474 */
Arjan van de Ven69d25872009-09-21 17:04:08 -0700475 if (measured_us > data->exit_us)
476 measured_us -= data->exit_us;
477
478
479 /* update our correction ratio */
480
481 new_factor = data->correction_factor[data->bucket]
482 * (DECAY - 1) / DECAY;
483
Arjan van de Ven1c6fe032010-05-08 15:47:37 -0700484 if (data->expected_us > 0 && measured_us < MAX_INTERESTING)
Arjan van de Ven69d25872009-09-21 17:04:08 -0700485 new_factor += RESOLUTION * measured_us / data->expected_us;
venkatesh.pallipadi@intel.com320eee72008-07-30 19:21:43 -0700486 else
Arjan van de Ven69d25872009-09-21 17:04:08 -0700487 /*
488 * we were idle so long that we count it as a perfect
489 * prediction
490 */
491 new_factor += RESOLUTION;
venkatesh.pallipadi@intel.com320eee72008-07-30 19:21:43 -0700492
Arjan van de Ven69d25872009-09-21 17:04:08 -0700493 /*
494 * We don't want 0 as factor; we always want at least
495 * a tiny bit of estimated time.
496 */
497 if (new_factor == 0)
498 new_factor = 1;
venkatesh.pallipadi@intel.com320eee72008-07-30 19:21:43 -0700499
Arjan van de Ven69d25872009-09-21 17:04:08 -0700500 data->correction_factor[data->bucket] = new_factor;
Arjan van de Ven1f85f872010-05-24 14:32:59 -0700501
502 /* update the repeating-pattern data */
503 data->intervals[data->interval_ptr++] = last_idle_us;
504 if (data->interval_ptr >= INTERVALS)
505 data->interval_ptr = 0;
Len Brown4f86d3a2007-10-03 18:58:00 -0400506}
507
508/**
509 * menu_enable_device - scans a CPU's states and does setup
Deepthi Dharwar46bcfad2011-10-28 16:20:42 +0530510 * @drv: cpuidle driver
Len Brown4f86d3a2007-10-03 18:58:00 -0400511 * @dev: the CPU
512 */
Deepthi Dharwar46bcfad2011-10-28 16:20:42 +0530513static int menu_enable_device(struct cpuidle_driver *drv,
514 struct cpuidle_device *dev)
Len Brown4f86d3a2007-10-03 18:58:00 -0400515{
516 struct menu_device *data = &per_cpu(menu_devices, dev->cpu);
Youquan Song69a37be2012-10-26 12:26:41 +0200517 struct hrtimer *t = &per_cpu(menu_hrtimer, dev->cpu);
518 hrtimer_init(t, CLOCK_MONOTONIC, HRTIMER_MODE_REL);
519 t->function = menu_hrtimer_notify;
Len Brown4f86d3a2007-10-03 18:58:00 -0400520
521 memset(data, 0, sizeof(struct menu_device));
522
523 return 0;
524}
525
526static struct cpuidle_governor menu_governor = {
527 .name = "menu",
528 .rating = 20,
529 .enable = menu_enable_device,
530 .select = menu_select,
531 .reflect = menu_reflect,
532 .owner = THIS_MODULE,
533};
534
535/**
536 * init_menu - initializes the governor
537 */
538static int __init init_menu(void)
539{
540 return cpuidle_register_governor(&menu_governor);
541}
542
543/**
544 * exit_menu - exits the governor
545 */
546static void __exit exit_menu(void)
547{
548 cpuidle_unregister_governor(&menu_governor);
549}
550
551MODULE_LICENSE("GPL");
552module_init(init_menu);
553module_exit(exit_menu);