summaryrefslogtreecommitdiffstats
path: root/drivers/cpufreq/cpufreq_conservative.c
blob: c18a304b3a38f4daf95dbd354daf129ce9db5ea6 (plain)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
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
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
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
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
/*
 *  drivers/cpufreq/cpufreq_conservative.c
 *
 *  Copyright (C)  2001 Russell King
 *            (C)  2003 Venkatesh Pallipadi <venkatesh.pallipadi@intel.com>.
 *                      Jun Nakajima <jun.nakajima@intel.com>
 *            (C)  2009 Alexander Clouter <alex@digriz.org.uk>
 *
 * This program is free software; you can redistribute it and/or modify
 * it under the terms of the GNU General Public License version 2 as
 * published by the Free Software Foundation.
 */

#include <linux/cpufreq.h>
#include <linux/init.h>
#include <linux/kernel.h>
#include <linux/kernel_stat.h>
#include <linux/kobject.h>
#include <linux/module.h>
#include <linux/mutex.h>
#include <linux/notifier.h>
#include <linux/percpu-defs.h>
#include <linux/sysfs.h>
#include <linux/types.h>

#include "cpufreq_governor.h"

/* Conservative governor macors */
#define DEF_FREQUENCY_UP_THRESHOLD		(80)
#define DEF_FREQUENCY_DOWN_THRESHOLD		(20)
#define DEF_SAMPLING_DOWN_FACTOR		(1)
#define MAX_SAMPLING_DOWN_FACTOR		(10)

static struct dbs_data cs_dbs_data;
static DEFINE_PER_CPU(struct cs_cpu_dbs_info_s, cs_cpu_dbs_info);

static struct cs_dbs_tuners cs_tuners = {
	.up_threshold = DEF_FREQUENCY_UP_THRESHOLD,
	.down_threshold = DEF_FREQUENCY_DOWN_THRESHOLD,
	.sampling_down_factor = DEF_SAMPLING_DOWN_FACTOR,
	.ignore_nice = 0,
	.freq_step = 5,
};

/*
 * Every sampling_rate, we check, if current idle time is less than 20%
 * (default), then we try to increase frequency Every sampling_rate *
 * sampling_down_factor, we check, if current idle time is more than 80%, then
 * we try to decrease frequency
 *
 * Any frequency increase takes it to the maximum frequency. Frequency reduction
 * happens at minimum steps of 5% (default) of maximum frequency
 */
static void cs_check_cpu(int cpu, unsigned int load)
{
	struct cs_cpu_dbs_info_s *dbs_info = &per_cpu(cs_cpu_dbs_info, cpu);
	struct cpufreq_policy *policy = dbs_info->cdbs.cur_policy;
	unsigned int freq_target;

	/*
	 * break out if we 'cannot' reduce the speed as the user might
	 * want freq_step to be zero
	 */
	if (cs_tuners.freq_step == 0)
		return;

	/* Check for frequency increase */
	if (load > cs_tuners.up_threshold) {
		dbs_info->down_skip = 0;

		/* if we are already at full speed then break out early */
		if (dbs_info->requested_freq == policy->max)
			return;

		freq_target = (cs_tuners.freq_step * policy->max) / 100;

		/* max freq cannot be less than 100. But who knows.... */
		if (unlikely(freq_target == 0))
			freq_target = 5;

		dbs_info->requested_freq += freq_target;
		if (dbs_info->requested_freq > policy->max)
			dbs_info->requested_freq = policy->max;

		__cpufreq_driver_target(policy, dbs_info->requested_freq,
			CPUFREQ_RELATION_H);
		return;
	}

	/*
	 * The optimal frequency is the frequency that is the lowest that can
	 * support the current CPU usage without triggering the up policy. To be
	 * safe, we focus 10 points under the threshold.
	 */
	if (load < (cs_tuners.down_threshold - 10)) {
		freq_target = (cs_tuners.freq_step * policy->max) / 100;

		dbs_info->requested_freq -= freq_target;
		if (dbs_info->requested_freq < policy->min)
			dbs_info->requested_freq = policy->min;

		/*
		 * if we cannot reduce the frequency anymore, break out early
		 */
		if (policy->cur == policy->min)
			return;

		__cpufreq_driver_target(policy, dbs_info->requested_freq,
				CPUFREQ_RELATION_H);
		return;
	}
}

static void cs_timer_update(struct cs_cpu_dbs_info_s *dbs_info, bool sample,
			    struct delayed_work *dw)
{
	unsigned int cpu = dbs_info->cdbs.cur_policy->cpu;
	int delay = delay_for_sampling_rate(cs_tuners.sampling_rate);

	if (sample)
		dbs_check_cpu(&cs_dbs_data, cpu);

	schedule_delayed_work_on(smp_processor_id(), dw, delay);
}

static void cs_timer_coordinated(struct cs_cpu_dbs_info_s *dbs_info_local,
				 struct delayed_work *dw)
{
	struct cs_cpu_dbs_info_s *dbs_info;
	ktime_t time_now;
	s64 delta_us;
	bool sample = true;

	/* use leader CPU's dbs_info */
	dbs_info = &per_cpu(cs_cpu_dbs_info,
			    dbs_info_local->cdbs.cur_policy->cpu);
	mutex_lock(&dbs_info->cdbs.timer_mutex);

	time_now = ktime_get();
	delta_us = ktime_us_delta(time_now, dbs_info->cdbs.time_stamp);

	/* Do nothing if we recently have sampled */
	if (delta_us < (s64)(cs_tuners.sampling_rate / 2))
		sample = false;
	else
		dbs_info->cdbs.time_stamp = time_now;

	cs_timer_update(dbs_info, sample, dw);
	mutex_unlock(&dbs_info->cdbs.timer_mutex);
}

static void cs_dbs_timer(struct work_struct *work)
{
	struct delayed_work *dw = to_delayed_work(work);
	struct cs_cpu_dbs_info_s *dbs_info = container_of(work,
			struct cs_cpu_dbs_info_s, cdbs.work.work);

	if (policy_is_shared(dbs_info->cdbs.cur_policy)) {
		cs_timer_coordinated(dbs_info, dw);
	} else {
		mutex_lock(&dbs_info->cdbs.timer_mutex);
		cs_timer_update(dbs_info, true, dw);
		mutex_unlock(&dbs_info->cdbs.timer_mutex);
	}
}
static int dbs_cpufreq_notifier(struct notifier_block *nb, unsigned long val,
		void *data)
{
	struct cpufreq_freqs *freq = data;
	struct cs_cpu_dbs_info_s *dbs_info =
					&per_cpu(cs_cpu_dbs_info, freq->cpu);
	struct cpufreq_policy *policy;

	if (!dbs_info->enable)
		return 0;

	policy = dbs_info->cdbs.cur_policy;

	/*
	 * we only care if our internally tracked freq moves outside the 'valid'
	 * ranges of freqency available to us otherwise we do not change it
	*/
	if (dbs_info->requested_freq > policy->max
			|| dbs_info->requested_freq < policy->min)
		dbs_info->requested_freq = freq->new;

	return 0;
}

/************************** sysfs interface ************************/
static ssize_t show_sampling_rate_min(struct kobject *kobj,
				      struct attribute *attr, char *buf)
{
	return sprintf(buf, "%u\n", cs_dbs_data.min_sampling_rate);
}

static ssize_t store_sampling_down_factor(struct kobject *a,
					  struct attribute *b,
					  const char *buf, size_t count)
{
	unsigned int input;
	int ret;
	ret = sscanf(buf, "%u", &input);

	if (ret != 1 || input > MAX_SAMPLING_DOWN_FACTOR || input < 1)
		return -EINVAL;

	cs_tuners.sampling_down_factor = input;
	return count;
}

static ssize_t store_sampling_rate(struct kobject *a, struct attribute *b,
				   const char *buf, size_t count)
{
	unsigned int input;
	int ret;
	ret = sscanf(buf, "%u", &input);

	if (ret != 1)
		return -EINVAL;

	cs_tuners.sampling_rate = max(input, cs_dbs_data.min_sampling_rate);
	return count;
}

static ssize_t store_up_threshold(struct kobject *a, struct attribute *b,
				  const char *buf, size_t count)
{
	unsigned int input;
	int ret;
	ret = sscanf(buf, "%u", &input);

	if (ret != 1 || input > 100 || input <= cs_tuners.down_threshold)
		return -EINVAL;

	cs_tuners.up_threshold = input;
	return count;
}

static ssize_t store_down_threshold(struct kobject *a, struct attribute *b,
				    const char *buf, size_t count)
{
	unsigned int input;
	int ret;
	ret = sscanf(buf, "%u", &input);

	/* cannot be lower than 11 otherwise freq will not fall */
	if (ret != 1 || input < 11 || input > 100 ||
			input >= cs_tuners.up_threshold)
		return -EINVAL;

	cs_tuners.down_threshold = input;
	return count;
}

static ssize_t store_ignore_nice_load(struct kobject *a, struct attribute *b,
				      const char *buf, size_t count)
{
	unsigned int input, j;
	int ret;

	ret = sscanf(buf, "%u", &input);
	if (ret != 1)
		return -EINVAL;

	if (input > 1)
		input = 1;

	if (input == cs_tuners.ignore_nice) /* nothing to do */
		return count;

	cs_tuners.ignore_nice = input;

	/* we need to re-evaluate prev_cpu_idle */
	for_each_online_cpu(j) {
		struct cs_cpu_dbs_info_s *dbs_info;
		dbs_info = &per_cpu(cs_cpu_dbs_info, j);
		dbs_info->cdbs.prev_cpu_idle = get_cpu_idle_time(j,
						&dbs_info->cdbs.prev_cpu_wall);
		if (cs_tuners.ignore_nice)
			dbs_info->cdbs.prev_cpu_nice =
				kcpustat_cpu(j).cpustat[CPUTIME_NICE];
	}
	return count;
}

static ssize_t store_freq_step(struct kobject *a, struct attribute *b,
			       const char *buf, size_t count)
{
	unsigned int input;
	int ret;
	ret = sscanf(buf, "%u", &input);

	if (ret != 1)
		return -EINVAL;

	if (input > 100)
		input = 100;

	/*
	 * no need to test here if freq_step is zero as the user might actually
	 * want this, they would be crazy though :)
	 */
	cs_tuners.freq_step = input;
	return count;
}

show_one(cs, sampling_rate, sampling_rate);
show_one(cs, sampling_down_factor, sampling_down_factor);
show_one(cs, up_threshold, up_threshold);
show_one(cs, down_threshold, down_threshold);
show_one(cs, ignore_nice_load, ignore_nice);
show_one(cs, freq_step, freq_step);

define_one_global_rw(sampling_rate);
define_one_global_rw(sampling_down_factor);
define_one_global_rw(up_threshold);
define_one_global_rw(down_threshold);
define_one_global_rw(ignore_nice_load);
define_one_global_rw(freq_step);
define_one_global_ro(sampling_rate_min);

static struct attribute *dbs_attributes[] = {
	&sampling_rate_min.attr,
	&sampling_rate.attr,
	&sampling_down_factor.attr,
	&up_threshold.attr,
	&down_threshold.attr,
	&ignore_nice_load.attr,
	&freq_step.attr,
	NULL
};

static struct attribute_group cs_attr_group = {
	.attrs = dbs_attributes,
	.name = "conservative",
};

/************************** sysfs end ************************/

define_get_cpu_dbs_routines(cs_cpu_dbs_info);

static struct notifier_block cs_cpufreq_notifier_block = {
	.notifier_call = dbs_cpufreq_notifier,
};

static struct cs_ops cs_ops = {
	.notifier_block = &cs_cpufreq_notifier_block,
};

static struct dbs_data cs_dbs_data = {
	.governor = GOV_CONSERVATIVE,
	.attr_group = &cs_attr_group,
	.tuners = &cs_tuners,
	.get_cpu_cdbs = get_cpu_cdbs,
	.get_cpu_dbs_info_s = get_cpu_dbs_info_s,
	.gov_dbs_timer = cs_dbs_timer,
	.gov_check_cpu = cs_check_cpu,
	.gov_ops = &cs_ops,
};

static int cs_cpufreq_governor_dbs(struct cpufreq_policy *policy,
				   unsigned int event)
{
	return cpufreq_governor_dbs(&cs_dbs_data, policy, event);
}

#ifndef CONFIG_CPU_FREQ_DEFAULT_GOV_CONSERVATIVE
static
#endif
struct cpufreq_governor cpufreq_gov_conservative = {
	.name			= "conservative",
	.governor		= cs_cpufreq_governor_dbs,
	.max_transition_latency	= TRANSITION_LATENCY_LIMIT,
	.owner			= THIS_MODULE,
};

static int __init cpufreq_gov_dbs_init(void)
{
	mutex_init(&cs_dbs_data.mutex);
	return cpufreq_register_governor(&cpufreq_gov_conservative);
}

static void __exit cpufreq_gov_dbs_exit(void)
{
	cpufreq_unregister_governor(&cpufreq_gov_conservative);
}

MODULE_AUTHOR("Alexander Clouter <alex@digriz.org.uk>");
MODULE_DESCRIPTION("'cpufreq_conservative' - A dynamic cpufreq governor for "
		"Low Latency Frequency Transition capable processors "
		"optimised for use in a battery environment");
MODULE_LICENSE("GPL");

#ifdef CONFIG_CPU_FREQ_DEFAULT_GOV_CONSERVATIVE
fs_initcall(cpufreq_gov_dbs_init);
#else
module_init(cpufreq_gov_dbs_init);
#endif
module_exit(cpufreq_gov_dbs_exit);
OpenPOWER on IntegriCloud