[linux-2.6-block.git] / drivers / cpufreq / cpufreq_stats.c

/*
 *  drivers/cpufreq/cpufreq_stats.c
 *
 *  Copyright (C) 2003-2004 Venkatesh Pallipadi <venkatesh.pallipadi@intel.com>.
 *  (C) 2004 Zou Nan hai <nanhai.zou@intel.com>.
 *
 * 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/kernel.h>
#include <linux/slab.h>
#include <linux/cpu.h>
#include <linux/sysfs.h>
#include <linux/cpufreq.h>
#include <linux/module.h>
#include <linux/jiffies.h>
#include <linux/percpu.h>
#include <linux/kobject.h>
#include <linux/spinlock.h>
#include <linux/notifier.h>
#include <asm/cputime.h>

static spinlock_t cpufreq_stats_lock;

#define CPUFREQ_STATDEVICE_ATTR(_name, _mode, _show) \
static struct freq_attr _attr_##_name = {\
	.attr = {.name = __stringify(_name), .mode = _mode, }, \
	.show = _show,\
};

struct cpufreq_stats {
	unsigned int cpu;
	unsigned int total_trans;
	unsigned long long  last_time;
	unsigned int max_state;
	unsigned int state_num;
	unsigned int last_index;
	u64 *time_in_state;
	unsigned int *freq_table;
#ifdef CONFIG_CPU_FREQ_STAT_DETAILS
	unsigned int *trans_table;
#endif
};

static DEFINE_PER_CPU(struct cpufreq_stats *, cpufreq_stats_table);

struct cpufreq_stats_attribute {
	struct attribute attr;
	ssize_t(*show) (struct cpufreq_stats *, char *);
};

static int cpufreq_stats_update(unsigned int cpu)
{
	struct cpufreq_stats *stat;
	unsigned long long cur_time;

	cur_time = get_jiffies_64();
	spin_lock(&cpufreq_stats_lock);
	stat = per_cpu(cpufreq_stats_table, cpu);
	if (stat->time_in_state)
		stat->time_in_state[stat->last_index] +=
			cur_time - stat->last_time;
	stat->last_time = cur_time;
	spin_unlock(&cpufreq_stats_lock);
	return 0;
}

static ssize_t show_total_trans(struct cpufreq_policy *policy, char *buf)
{
	struct cpufreq_stats *stat = per_cpu(cpufreq_stats_table, policy->cpu);
	if (!stat)
		return 0;
	return sprintf(buf, "%d\n",
			per_cpu(cpufreq_stats_table, stat->cpu)->total_trans);
}

static ssize_t show_time_in_state(struct cpufreq_policy *policy, char *buf)
{
	ssize_t len = 0;
	int i;
	struct cpufreq_stats *stat = per_cpu(cpufreq_stats_table, policy->cpu);
	if (!stat)
		return 0;
	cpufreq_stats_update(stat->cpu);
	for (i = 0; i < stat->state_num; i++) {
		len += sprintf(buf + len, "%u %llu\n", stat->freq_table[i],
			(unsigned long long)
			cputime64_to_clock_t(stat->time_in_state[i]));
	}
	return len;
}

#ifdef CONFIG_CPU_FREQ_STAT_DETAILS
static ssize_t show_trans_table(struct cpufreq_policy *policy, char *buf)
{
	ssize_t len = 0;
	int i, j;

	struct cpufreq_stats *stat = per_cpu(cpufreq_stats_table, policy->cpu);
	if (!stat)
		return 0;
	cpufreq_stats_update(stat->cpu);
	len += snprintf(buf + len, PAGE_SIZE - len, "   From  :    To\n");
	len += snprintf(buf + len, PAGE_SIZE - len, "         : ");
	for (i = 0; i < stat->state_num; i++) {
		if (len >= PAGE_SIZE)
			break;
		len += snprintf(buf + len, PAGE_SIZE - len, "%9u ",
				stat->freq_table[i]);
	}
	if (len >= PAGE_SIZE)
		return PAGE_SIZE;

	len += snprintf(buf + len, PAGE_SIZE - len, "\n");

	for (i = 0; i < stat->state_num; i++) {
		if (len >= PAGE_SIZE)
			break;

		len += snprintf(buf + len, PAGE_SIZE - len, "%9u: ",
				stat->freq_table[i]);

		for (j = 0; j < stat->state_num; j++)   {
			if (len >= PAGE_SIZE)
				break;
			len += snprintf(buf + len, PAGE_SIZE - len, "%9u ",
					stat->trans_table[i*stat->max_state+j]);
		}
		if (len >= PAGE_SIZE)
			break;
		len += snprintf(buf + len, PAGE_SIZE - len, "\n");
	}
	if (len >= PAGE_SIZE)
		return PAGE_SIZE;
	return len;
}
CPUFREQ_STATDEVICE_ATTR(trans_table, 0444, show_trans_table);
#endif

CPUFREQ_STATDEVICE_ATTR(total_trans, 0444, show_total_trans);
CPUFREQ_STATDEVICE_ATTR(time_in_state, 0444, show_time_in_state);

static struct attribute *default_attrs[] = {
	&_attr_total_trans.attr,
	&_attr_time_in_state.attr,
#ifdef CONFIG_CPU_FREQ_STAT_DETAILS
	&_attr_trans_table.attr,
#endif
	NULL
};
static struct attribute_group stats_attr_group = {
	.attrs = default_attrs,
	.name = "stats"
};

static int freq_table_get_index(struct cpufreq_stats *stat, unsigned int freq)
{
	int index;
	for (index = 0; index < stat->max_state; index++)
		if (stat->freq_table[index] == freq)
			return index;
	return -1;
}

/* should be called late in the CPU removal sequence so that the stats
 * memory is still available in case someone tries to use it.
 */
static void cpufreq_stats_free_table(unsigned int cpu)
{
	struct cpufreq_stats *stat = per_cpu(cpufreq_stats_table, cpu);

	if (stat) {
		pr_debug("%s: Free stat table\n", __func__);
		kfree(stat->time_in_state);
		kfree(stat);
		per_cpu(cpufreq_stats_table, cpu) = NULL;
	}
}

/* must be called early in the CPU removal sequence (before
 * cpufreq_remove_dev) so that policy is still valid.
 */
static void cpufreq_stats_free_sysfs(unsigned int cpu)
{
	struct cpufreq_policy *policy = cpufreq_cpu_get(cpu);
	if (policy && !policy_is_shared(policy)) {
		pr_debug("%s: Free sysfs stat\n", __func__);
		sysfs_remove_group(&policy->kobj, &stats_attr_group);
	}
	if (policy)
		cpufreq_cpu_put(policy);
}

static int cpufreq_stats_create_table(struct cpufreq_policy *policy,
		struct cpufreq_frequency_table *table)
{
	unsigned int i, j, count = 0, ret = 0;
	struct cpufreq_stats *stat;
	struct cpufreq_policy *data;
	unsigned int alloc_size;
	unsigned int cpu = policy->cpu;
	if (per_cpu(cpufreq_stats_table, cpu))
		return -EBUSY;
	stat = kzalloc(sizeof(struct cpufreq_stats), GFP_KERNEL);
	if ((stat) == NULL)
		return -ENOMEM;

	data = cpufreq_cpu_get(cpu);
	if (data == NULL) {
		ret = -EINVAL;
		goto error_get_fail;
	}

	ret = sysfs_create_group(&data->kobj, &stats_attr_group);
	if (ret)
		goto error_out;

	stat->cpu = cpu;
	per_cpu(cpufreq_stats_table, cpu) = stat;

	for (i = 0; table[i].frequency != CPUFREQ_TABLE_END; i++) {
		unsigned int freq = table[i].frequency;
		if (freq == CPUFREQ_ENTRY_INVALID)
			continue;
		count++;
	}

	alloc_size = count * sizeof(int) + count * sizeof(u64);

#ifdef CONFIG_CPU_FREQ_STAT_DETAILS
	alloc_size += count * count * sizeof(int);
#endif
	stat->max_state = count;
	stat->time_in_state = kzalloc(alloc_size, GFP_KERNEL);
	if (!stat->time_in_state) {
		ret = -ENOMEM;
		goto error_out;
	}
	stat->freq_table = (unsigned int *)(stat->time_in_state + count);

#ifdef CONFIG_CPU_FREQ_STAT_DETAILS
	stat->trans_table = stat->freq_table + count;
#endif
	j = 0;
	for (i = 0; table[i].frequency != CPUFREQ_TABLE_END; i++) {
		unsigned int freq = table[i].frequency;
		if (freq == CPUFREQ_ENTRY_INVALID)
			continue;
		if (freq_table_get_index(stat, freq) == -1)
			stat->freq_table[j++] = freq;
	}
	stat->state_num = j;
	spin_lock(&cpufreq_stats_lock);
	stat->last_time = get_jiffies_64();
	stat->last_index = freq_table_get_index(stat, policy->cur);
	spin_unlock(&cpufreq_stats_lock);
	cpufreq_cpu_put(data);
	return 0;
error_out:
	cpufreq_cpu_put(data);
error_get_fail:
	kfree(stat);
	per_cpu(cpufreq_stats_table, cpu) = NULL;
	return ret;
}

static void cpufreq_stats_update_policy_cpu(struct cpufreq_policy *policy)
{
	struct cpufreq_stats *stat = per_cpu(cpufreq_stats_table,
			policy->last_cpu);

	pr_debug("Updating stats_table for new_cpu %u from last_cpu %u\n",
			policy->cpu, policy->last_cpu);
	per_cpu(cpufreq_stats_table, policy->cpu) = per_cpu(cpufreq_stats_table,
			policy->last_cpu);
	per_cpu(cpufreq_stats_table, policy->last_cpu) = NULL;
	stat->cpu = policy->cpu;
}

static int cpufreq_stat_notifier_policy(struct notifier_block *nb,
		unsigned long val, void *data)
{
	int ret;
	struct cpufreq_policy *policy = data;
	struct cpufreq_frequency_table *table;
	unsigned int cpu = policy->cpu;

	if (val == CPUFREQ_UPDATE_POLICY_CPU) {
		cpufreq_stats_update_policy_cpu(policy);
		return 0;
	}

	if (val != CPUFREQ_NOTIFY)
		return 0;
	table = cpufreq_frequency_get_table(cpu);
	if (!table)
		return 0;
	ret = cpufreq_stats_create_table(policy, table);
	if (ret)
		return ret;
	return 0;
}

static int cpufreq_stat_notifier_trans(struct notifier_block *nb,
		unsigned long val, void *data)
{
	struct cpufreq_freqs *freq = data;
	struct cpufreq_stats *stat;
	int old_index, new_index;

	if (val != CPUFREQ_POSTCHANGE)
		return 0;

	stat = per_cpu(cpufreq_stats_table, freq->cpu);
	if (!stat)
		return 0;

	old_index = stat->last_index;
	new_index = freq_table_get_index(stat, freq->new);

	/* We can't do stat->time_in_state[-1]= .. */
	if (old_index == -1 || new_index == -1)
		return 0;

	cpufreq_stats_update(freq->cpu);

	if (old_index == new_index)
		return 0;

	spin_lock(&cpufreq_stats_lock);
	stat->last_index = new_index;
#ifdef CONFIG_CPU_FREQ_STAT_DETAILS
	stat->trans_table[old_index * stat->max_state + new_index]++;
#endif
	stat->total_trans++;
	spin_unlock(&cpufreq_stats_lock);
	return 0;
}

static int __cpuinit cpufreq_stat_cpu_callback(struct notifier_block *nfb,
					       unsigned long action,
					       void *hcpu)
{
	unsigned int cpu = (unsigned long)hcpu;

	switch (action) {
	case CPU_ONLINE:
	case CPU_ONLINE_FROZEN:
		cpufreq_update_policy(cpu);
		break;
	case CPU_DOWN_PREPARE:
	case CPU_DOWN_PREPARE_FROZEN:
		cpufreq_stats_free_sysfs(cpu);
		break;
	case CPU_DEAD:
	case CPU_DEAD_FROZEN:
		cpufreq_stats_free_table(cpu);
		break;
	}
	return NOTIFY_OK;
}

/* priority=1 so this will get called before cpufreq_remove_dev */
static struct notifier_block cpufreq_stat_cpu_notifier __refdata = {
	.notifier_call = cpufreq_stat_cpu_callback,
	.priority = 1,
};

static struct notifier_block notifier_policy_block = {
	.notifier_call = cpufreq_stat_notifier_policy
};

static struct notifier_block notifier_trans_block = {
	.notifier_call = cpufreq_stat_notifier_trans
};

static int __init cpufreq_stats_init(void)
{
	int ret;
	unsigned int cpu;

	spin_lock_init(&cpufreq_stats_lock);
	ret = cpufreq_register_notifier(&notifier_policy_block,
				CPUFREQ_POLICY_NOTIFIER);
	if (ret)
		return ret;

	register_hotcpu_notifier(&cpufreq_stat_cpu_notifier);
	for_each_online_cpu(cpu)
		cpufreq_update_policy(cpu);

	ret = cpufreq_register_notifier(&notifier_trans_block,
				CPUFREQ_TRANSITION_NOTIFIER);
	if (ret) {
		cpufreq_unregister_notifier(&notifier_policy_block,
				CPUFREQ_POLICY_NOTIFIER);
		unregister_hotcpu_notifier(&cpufreq_stat_cpu_notifier);
		for_each_online_cpu(cpu)
			cpufreq_stats_free_table(cpu);
		return ret;
	}

	return 0;
}
static void __exit cpufreq_stats_exit(void)
{
	unsigned int cpu;

	cpufreq_unregister_notifier(&notifier_policy_block,
			CPUFREQ_POLICY_NOTIFIER);
	cpufreq_unregister_notifier(&notifier_trans_block,
			CPUFREQ_TRANSITION_NOTIFIER);
	unregister_hotcpu_notifier(&cpufreq_stat_cpu_notifier);
	for_each_online_cpu(cpu) {
		cpufreq_stats_free_table(cpu);
		cpufreq_stats_free_sysfs(cpu);
	}
}

MODULE_AUTHOR("Zou Nan hai <nanhai.zou@intel.com>");
MODULE_DESCRIPTION("'cpufreq_stats' - A driver to export cpufreq stats "
				"through sysfs filesystem");
MODULE_LICENSE("GPL");

module_init(cpufreq_stats_init);
module_exit(cpufreq_stats_exit);
Commit	Line	Data
1da177e4 LT	1	/*
	2	* drivers/cpufreq/cpufreq_stats.c
	3	*
	4	* Copyright (C) 2003-2004 Venkatesh Pallipadi <venkatesh.pallipadi@intel.com>.
0a829c5a	5	* (C) 2004 Zou Nan hai <nanhai.zou@intel.com>.
1da177e4 LT	6	*
	7	* This program is free software; you can redistribute it and/or modify
	8	* it under the terms of the GNU General Public License version 2 as
	9	* published by the Free Software Foundation.
	10	*/
	11
1da177e4	12	#include <linux/kernel.h>
5a0e3ad6	13	#include <linux/slab.h>
1da177e4 LT	14	#include <linux/cpu.h>
	15	#include <linux/sysfs.h>
	16	#include <linux/cpufreq.h>
5c720d37	17	#include <linux/module.h>
1da177e4 LT	18	#include <linux/jiffies.h>
	19	#include <linux/percpu.h>
	20	#include <linux/kobject.h>
	21	#include <linux/spinlock.h>
c32b6b8e	22	#include <linux/notifier.h>
58f1df25	23	#include <asm/cputime.h>
1da177e4 LT	24
	25	static spinlock_t cpufreq_stats_lock;
	26
0a829c5a	27	#define CPUFREQ_STATDEVICE_ATTR(_name, _mode, _show) \
1da177e4	28	static struct freq_attr _attr_##_name = {\
7b595756	29	.attr = {.name = __stringify(_name), .mode = _mode, }, \
1da177e4 LT	30	.show = _show,\
	31	};
	32
1da177e4 LT	33	struct cpufreq_stats {
	34	unsigned int cpu;
	35	unsigned int total_trans;
58f1df25	36	unsigned long long last_time;
1da177e4 LT	37	unsigned int max_state;
	38	unsigned int state_num;
	39	unsigned int last_index;
1e7586a1	40	u64 *time_in_state;
1da177e4 LT	41	unsigned int *freq_table;
	42	#ifdef CONFIG_CPU_FREQ_STAT_DETAILS
	43	unsigned int *trans_table;
	44	#endif
	45	};
	46
7a6aedfa	47	static DEFINE_PER_CPU(struct cpufreq_stats *, cpufreq_stats_table);
1da177e4 LT	48
	49	struct cpufreq_stats_attribute {
	50	struct attribute attr;
	51	ssize_t(show) (struct cpufreq_stats , char *);
	52	};
	53
0a829c5a	54	static int cpufreq_stats_update(unsigned int cpu)
1da177e4 LT	55	{
1da177e4 LT	56	struct cpufreq_stats *stat;
58f1df25 VP	57	unsigned long long cur_time;
	58
	59	cur_time = get_jiffies_64();
1da177e4	60	spin_lock(&cpufreq_stats_lock);
7a6aedfa	61	stat = per_cpu(cpufreq_stats_table, cpu);
1da177e4	62	if (stat->time_in_state)
64861634 MS	63	stat->time_in_state[stat->last_index] +=
64861634 MS	64	cur_time - stat->last_time;
58f1df25	65	stat->last_time = cur_time;
1da177e4 LT	66	spin_unlock(&cpufreq_stats_lock);
	67	return 0;
	68	}
	69
0a829c5a	70	static ssize_t show_total_trans(struct cpufreq_policy policy, char buf)
1da177e4	71	{
7a6aedfa	72	struct cpufreq_stats *stat = per_cpu(cpufreq_stats_table, policy->cpu);
511e9ee1	73	if (!stat)
1da177e4 LT	74	return 0;
1da177e4 LT	75	return sprintf(buf, "%d\n",
7a6aedfa	76	per_cpu(cpufreq_stats_table, stat->cpu)->total_trans);
1da177e4 LT	77	}
1da177e4 LT	78
0a829c5a	79	static ssize_t show_time_in_state(struct cpufreq_policy policy, char buf)
1da177e4 LT	80	{
	81	ssize_t len = 0;
	82	int i;
7a6aedfa	83	struct cpufreq_stats *stat = per_cpu(cpufreq_stats_table, policy->cpu);
511e9ee1	84	if (!stat)
1da177e4 LT	85	return 0;
	86	cpufreq_stats_update(stat->cpu);
	87	for (i = 0; i < stat->state_num; i++) {
32ee8c3e	88	len += sprintf(buf + len, "%u %llu\n", stat->freq_table[i],
0a829c5a DJ	89	(unsigned long long)
0a829c5a DJ	90	cputime64_to_clock_t(stat->time_in_state[i]));
1da177e4 LT	91	}
	92	return len;
	93	}
	94
	95	#ifdef CONFIG_CPU_FREQ_STAT_DETAILS
0a829c5a	96	static ssize_t show_trans_table(struct cpufreq_policy policy, char buf)
1da177e4 LT	97	{
	98	ssize_t len = 0;
	99	int i, j;
	100
7a6aedfa	101	struct cpufreq_stats *stat = per_cpu(cpufreq_stats_table, policy->cpu);
511e9ee1	102	if (!stat)
1da177e4 LT	103	return 0;
1da177e4 LT	104	cpufreq_stats_update(stat->cpu);
58f1df25 VP	105	len += snprintf(buf + len, PAGE_SIZE - len, " From : To\n");
	106	len += snprintf(buf + len, PAGE_SIZE - len, " : ");
	107	for (i = 0; i < stat->state_num; i++) {
	108	if (len >= PAGE_SIZE)
	109	break;
	110	len += snprintf(buf + len, PAGE_SIZE - len, "%9u ",
	111	stat->freq_table[i]);
	112	}
	113	if (len >= PAGE_SIZE)
25aca347	114	return PAGE_SIZE;
58f1df25 VP	115
	116	len += snprintf(buf + len, PAGE_SIZE - len, "\n");
	117
1da177e4 LT	118	for (i = 0; i < stat->state_num; i++) {
	119	if (len >= PAGE_SIZE)
	120	break;
58f1df25 VP	121
58f1df25 VP	122	len += snprintf(buf + len, PAGE_SIZE - len, "%9u: ",
1da177e4 LT	123	stat->freq_table[i]);
	124
	125	for (j = 0; j < stat->state_num; j++) {
	126	if (len >= PAGE_SIZE)
	127	break;
58f1df25	128	len += snprintf(buf + len, PAGE_SIZE - len, "%9u ",
1da177e4 LT	129	stat->trans_table[i*stat->max_state+j]);
1da177e4 LT	130	}
25aca347 CEB	131	if (len >= PAGE_SIZE)
25aca347 CEB	132	break;
1da177e4 LT	133	len += snprintf(buf + len, PAGE_SIZE - len, "\n");
1da177e4 LT	134	}
25aca347 CEB	135	if (len >= PAGE_SIZE)
25aca347 CEB	136	return PAGE_SIZE;
1da177e4 LT	137	return len;
1da177e4 LT	138	}
0a829c5a	139	CPUFREQ_STATDEVICE_ATTR(trans_table, 0444, show_trans_table);
1da177e4 LT	140	#endif
1da177e4 LT	141
0a829c5a DJ	142	CPUFREQ_STATDEVICE_ATTR(total_trans, 0444, show_total_trans);
0a829c5a DJ	143	CPUFREQ_STATDEVICE_ATTR(time_in_state, 0444, show_time_in_state);
1da177e4 LT	144
	145	static struct attribute *default_attrs[] = {
	146	&_attr_total_trans.attr,
	147	&_attr_time_in_state.attr,
	148	#ifdef CONFIG_CPU_FREQ_STAT_DETAILS
	149	&_attr_trans_table.attr,
	150	#endif
	151	NULL
	152	};
	153	static struct attribute_group stats_attr_group = {
	154	.attrs = default_attrs,
	155	.name = "stats"
	156	};
	157
0a829c5a	158	static int freq_table_get_index(struct cpufreq_stats *stat, unsigned int freq)
1da177e4 LT	159	{
	160	int index;
	161	for (index = 0; index < stat->max_state; index++)
	162	if (stat->freq_table[index] == freq)
	163	return index;
	164	return -1;
	165	}
	166
98586ed8	167	/* should be called late in the CPU removal sequence so that the stats
	168	* memory is still available in case someone tries to use it.
	169	*/
a3323473	170	static void cpufreq_stats_free_table(unsigned int cpu)
1da177e4	171	{
7a6aedfa	172	struct cpufreq_stats *stat = per_cpu(cpufreq_stats_table, cpu);
b8eed8af	173
1da177e4	174	if (stat) {
b8eed8af	175	pr_debug("%s: Free stat table\n", __func__);
1da177e4 LT	176	kfree(stat->time_in_state);
1da177e4 LT	177	kfree(stat);
b8eed8af	178	per_cpu(cpufreq_stats_table, cpu) = NULL;
1da177e4	179	}
98586ed8	180	}
	181
	182	/* must be called early in the CPU removal sequence (before
	183	* cpufreq_remove_dev) so that policy is still valid.
	184	*/
	185	static void cpufreq_stats_free_sysfs(unsigned int cpu)
	186	{
	187	struct cpufreq_policy *policy = cpufreq_cpu_get(cpu);
2624f90c	188	if (policy && !policy_is_shared(policy)) {
b8eed8af	189	pr_debug("%s: Free sysfs stat\n", __func__);
98586ed8	190	sysfs_remove_group(&policy->kobj, &stats_attr_group);
b8eed8af	191	}
1da177e4 LT	192	if (policy)
	193	cpufreq_cpu_put(policy);
	194	}
	195
0a829c5a	196	static int cpufreq_stats_create_table(struct cpufreq_policy *policy,
1da177e4 LT	197	struct cpufreq_frequency_table *table)
	198	{
	199	unsigned int i, j, count = 0, ret = 0;
	200	struct cpufreq_stats *stat;
	201	struct cpufreq_policy *data;
	202	unsigned int alloc_size;
	203	unsigned int cpu = policy->cpu;
7a6aedfa	204	if (per_cpu(cpufreq_stats_table, cpu))
1da177e4	205	return -EBUSY;
0a829c5a DJ	206	stat = kzalloc(sizeof(struct cpufreq_stats), GFP_KERNEL);
0a829c5a DJ	207	if ((stat) == NULL)
1da177e4	208	return -ENOMEM;
1da177e4 LT	209
1da177e4 LT	210	data = cpufreq_cpu_get(cpu);
bc7b26fd DJ	211	if (data == NULL) {
	212	ret = -EINVAL;
	213	goto error_get_fail;
	214	}
	215
0a829c5a DJ	216	ret = sysfs_create_group(&data->kobj, &stats_attr_group);
0a829c5a DJ	217	if (ret)
1da177e4 LT	218	goto error_out;
	219
	220	stat->cpu = cpu;
7a6aedfa	221	per_cpu(cpufreq_stats_table, cpu) = stat;
1da177e4	222
0a829c5a	223	for (i = 0; table[i].frequency != CPUFREQ_TABLE_END; i++) {
1da177e4 LT	224	unsigned int freq = table[i].frequency;
	225	if (freq == CPUFREQ_ENTRY_INVALID)
	226	continue;
	227	count++;
	228	}
	229
1e7586a1	230	alloc_size = count * sizeof(int) + count * sizeof(u64);
1da177e4 LT	231
	232	#ifdef CONFIG_CPU_FREQ_STAT_DETAILS
	233	alloc_size += count * count * sizeof(int);
	234	#endif
	235	stat->max_state = count;
e98df50c	236	stat->time_in_state = kzalloc(alloc_size, GFP_KERNEL);
1da177e4 LT	237	if (!stat->time_in_state) {
	238	ret = -ENOMEM;
	239	goto error_out;
	240	}
1da177e4 LT	241	stat->freq_table = (unsigned int *)(stat->time_in_state + count);
	242
	243	#ifdef CONFIG_CPU_FREQ_STAT_DETAILS
	244	stat->trans_table = stat->freq_table + count;
	245	#endif
	246	j = 0;
	247	for (i = 0; table[i].frequency != CPUFREQ_TABLE_END; i++) {
	248	unsigned int freq = table[i].frequency;
	249	if (freq == CPUFREQ_ENTRY_INVALID)
	250	continue;
	251	if (freq_table_get_index(stat, freq) == -1)
	252	stat->freq_table[j++] = freq;
	253	}
	254	stat->state_num = j;
	255	spin_lock(&cpufreq_stats_lock);
58f1df25	256	stat->last_time = get_jiffies_64();
1da177e4 LT	257	stat->last_index = freq_table_get_index(stat, policy->cur);
	258	spin_unlock(&cpufreq_stats_lock);
	259	cpufreq_cpu_put(data);
	260	return 0;
	261	error_out:
	262	cpufreq_cpu_put(data);
b7fb358c	263	error_get_fail:
1da177e4	264	kfree(stat);
7a6aedfa	265	per_cpu(cpufreq_stats_table, cpu) = NULL;
1da177e4 LT	266	return ret;
	267	}
	268
b8eed8af VK	269	static void cpufreq_stats_update_policy_cpu(struct cpufreq_policy *policy)
	270	{
	271	struct cpufreq_stats *stat = per_cpu(cpufreq_stats_table,
	272	policy->last_cpu);
	273
	274	pr_debug("Updating stats_table for new_cpu %u from last_cpu %u\n",
	275	policy->cpu, policy->last_cpu);
	276	per_cpu(cpufreq_stats_table, policy->cpu) = per_cpu(cpufreq_stats_table,
	277	policy->last_cpu);
	278	per_cpu(cpufreq_stats_table, policy->last_cpu) = NULL;
	279	stat->cpu = policy->cpu;
	280	}
	281
0a829c5a DJ	282	static int cpufreq_stat_notifier_policy(struct notifier_block *nb,
0a829c5a DJ	283	unsigned long val, void *data)
1da177e4 LT	284	{
	285	int ret;
	286	struct cpufreq_policy *policy = data;
	287	struct cpufreq_frequency_table *table;
	288	unsigned int cpu = policy->cpu;
b8eed8af VK	289
	290	if (val == CPUFREQ_UPDATE_POLICY_CPU) {
	291	cpufreq_stats_update_policy_cpu(policy);
	292	return 0;
	293	}
	294
1da177e4 LT	295	if (val != CPUFREQ_NOTIFY)
	296	return 0;
	297	table = cpufreq_frequency_get_table(cpu);
	298	if (!table)
	299	return 0;
0a829c5a DJ	300	ret = cpufreq_stats_create_table(policy, table);
0a829c5a DJ	301	if (ret)
1da177e4 LT	302	return ret;
	303	return 0;
	304	}
	305
0a829c5a DJ	306	static int cpufreq_stat_notifier_trans(struct notifier_block *nb,
0a829c5a DJ	307	unsigned long val, void *data)
1da177e4 LT	308	{
	309	struct cpufreq_freqs *freq = data;
	310	struct cpufreq_stats *stat;
	311	int old_index, new_index;
	312
	313	if (val != CPUFREQ_POSTCHANGE)
	314	return 0;
	315
7a6aedfa	316	stat = per_cpu(cpufreq_stats_table, freq->cpu);
1da177e4 LT	317	if (!stat)
1da177e4 LT	318	return 0;
8edc59d9	319
6501faf8	320	old_index = stat->last_index;
1da177e4 LT	321	new_index = freq_table_get_index(stat, freq->new);
1da177e4 LT	322
46a310b8 KRW	323	/* We can't do stat->time_in_state[-1]= .. */
46a310b8 KRW	324	if (old_index == -1 \|\| new_index == -1)
1da177e4 LT	325	return 0;
1da177e4 LT	326
46a310b8 KRW	327	cpufreq_stats_update(freq->cpu);
	328
	329	if (old_index == new_index)
8edc59d9 VP	330	return 0;
8edc59d9 VP	331
1da177e4 LT	332	spin_lock(&cpufreq_stats_lock);
	333	stat->last_index = new_index;
	334	#ifdef CONFIG_CPU_FREQ_STAT_DETAILS
	335	stat->trans_table[old_index * stat->max_state + new_index]++;
	336	#endif
	337	stat->total_trans++;
	338	spin_unlock(&cpufreq_stats_lock);
	339	return 0;
	340	}
	341
55395ae7 SS	342	static int __cpuinit cpufreq_stat_cpu_callback(struct notifier_block *nfb,
	343	unsigned long action,
	344	void *hcpu)
c32b6b8e AR	345	{
	346	unsigned int cpu = (unsigned long)hcpu;
	347
	348	switch (action) {
	349	case CPU_ONLINE:
8bb78442	350	case CPU_ONLINE_FROZEN:
c32b6b8e AR	351	cpufreq_update_policy(cpu);
c32b6b8e AR	352	break;
98586ed8	353	case CPU_DOWN_PREPARE:
e3773677	354	case CPU_DOWN_PREPARE_FROZEN:
98586ed8	355	cpufreq_stats_free_sysfs(cpu);
98586ed8	356	break;
c32b6b8e	357	case CPU_DEAD:
8bb78442	358	case CPU_DEAD_FROZEN:
c32b6b8e AR	359	cpufreq_stats_free_table(cpu);
	360	break;
	361	}
	362	return NOTIFY_OK;
	363	}
	364
98586ed8	365	/* priority=1 so this will get called before cpufreq_remove_dev */
469057d5	366	static struct notifier_block cpufreq_stat_cpu_notifier __refdata = {
c32b6b8e	367	.notifier_call = cpufreq_stat_cpu_callback,
98586ed8	368	.priority = 1,
c32b6b8e AR	369	};
c32b6b8e AR	370
1da177e4 LT	371	static struct notifier_block notifier_policy_block = {
	372	.notifier_call = cpufreq_stat_notifier_policy
	373	};
	374
	375	static struct notifier_block notifier_trans_block = {
	376	.notifier_call = cpufreq_stat_notifier_trans
	377	};
	378
0a829c5a	379	static int __init cpufreq_stats_init(void)
1da177e4 LT	380	{
	381	int ret;
	382	unsigned int cpu;
c32b6b8e	383
1da177e4	384	spin_lock_init(&cpufreq_stats_lock);
0a829c5a DJ	385	ret = cpufreq_register_notifier(&notifier_policy_block,
	386	CPUFREQ_POLICY_NOTIFIER);
	387	if (ret)
1da177e4 LT	388	return ret;
1da177e4 LT	389
56836fb4 KK	390	register_hotcpu_notifier(&cpufreq_stat_cpu_notifier);
	391	for_each_online_cpu(cpu)
	392	cpufreq_update_policy(cpu);
	393
0a829c5a DJ	394	ret = cpufreq_register_notifier(&notifier_trans_block,
	395	CPUFREQ_TRANSITION_NOTIFIER);
	396	if (ret) {
1da177e4 LT	397	cpufreq_unregister_notifier(&notifier_policy_block,
1da177e4 LT	398	CPUFREQ_POLICY_NOTIFIER);
56836fb4 KK	399	unregister_hotcpu_notifier(&cpufreq_stat_cpu_notifier);
	400	for_each_online_cpu(cpu)
	401	cpufreq_stats_free_table(cpu);
1da177e4 LT	402	return ret;
	403	}
	404
1da177e4 LT	405	return 0;
1da177e4 LT	406	}
0a829c5a	407	static void __exit cpufreq_stats_exit(void)
1da177e4 LT	408	{
1da177e4 LT	409	unsigned int cpu;
c32b6b8e	410
1da177e4 LT	411	cpufreq_unregister_notifier(&notifier_policy_block,
	412	CPUFREQ_POLICY_NOTIFIER);
	413	cpufreq_unregister_notifier(&notifier_trans_block,
	414	CPUFREQ_TRANSITION_NOTIFIER);
65edc68c	415	unregister_hotcpu_notifier(&cpufreq_stat_cpu_notifier);
c32b6b8e	416	for_each_online_cpu(cpu) {
55395ae7	417	cpufreq_stats_free_table(cpu);
13f06753	418	cpufreq_stats_free_sysfs(cpu);
c32b6b8e	419	}
1da177e4 LT	420	}
1da177e4 LT	421
0a829c5a DJ	422	MODULE_AUTHOR("Zou Nan hai <nanhai.zou@intel.com>");
0a829c5a DJ	423	MODULE_DESCRIPTION("'cpufreq_stats' - A driver to export cpufreq stats "
e08f5f5b	424	"through sysfs filesystem");
0a829c5a	425	MODULE_LICENSE("GPL");
1da177e4 LT	426
	427	module_init(cpufreq_stats_init);
	428	module_exit(cpufreq_stats_exit);