[mirror_ubuntu-bionic-kernel.git] / drivers / cpufreq / intel_pstate.c

/*
 * intel_pstate.c: Native P state management for Intel processors
 *
 * (C) Copyright 2012 Intel Corporation
 * Author: Dirk Brandewie <dirk.j.brandewie@intel.com>
 *
 * This program is free software; you can redistribute it and/or
 * modify it under the terms of the GNU General Public License
 * as published by the Free Software Foundation; version 2
 * of the License.
 */

#include <linux/kernel.h>
#include <linux/kernel_stat.h>
#include <linux/module.h>
#include <linux/ktime.h>
#include <linux/hrtimer.h>
#include <linux/tick.h>
#include <linux/slab.h>
#include <linux/sched.h>
#include <linux/list.h>
#include <linux/cpu.h>
#include <linux/cpufreq.h>
#include <linux/sysfs.h>
#include <linux/types.h>
#include <linux/fs.h>
#include <linux/debugfs.h>
#include <trace/events/power.h>

#include <asm/div64.h>
#include <asm/msr.h>
#include <asm/cpu_device_id.h>

#define SAMPLE_COUNT		3

#define FRAC_BITS 8
#define int_tofp(X) ((int64_t)(X) << FRAC_BITS)
#define fp_toint(X) ((X) >> FRAC_BITS)

static inline int32_t mul_fp(int32_t x, int32_t y)
{
	return ((int64_t)x * (int64_t)y) >> FRAC_BITS;
}

static inline int32_t div_fp(int32_t x, int32_t y)
{
	return div_s64((int64_t)x << FRAC_BITS, (int64_t)y);
}

struct sample {
	int core_pct_busy;
	u64 aperf;
	u64 mperf;
	int freq;
};

struct pstate_data {
	int	current_pstate;
	int	min_pstate;
	int	max_pstate;
	int	turbo_pstate;
};

struct _pid {
	int setpoint;
	int32_t integral;
	int32_t p_gain;
	int32_t i_gain;
	int32_t d_gain;
	int deadband;
	int last_err;
};

struct cpudata {
	int cpu;

	char name[64];

	struct timer_list timer;

	struct pstate_adjust_policy *pstate_policy;
	struct pstate_data pstate;
	struct _pid pid;

	int min_pstate_count;

	u64	prev_aperf;
	u64	prev_mperf;
	int	sample_ptr;
	struct sample samples[SAMPLE_COUNT];
};

static struct cpudata **all_cpu_data;
struct pstate_adjust_policy {
	int sample_rate_ms;
	int deadband;
	int setpoint;
	int p_gain_pct;
	int d_gain_pct;
	int i_gain_pct;
};

static struct pstate_adjust_policy default_policy = {
	.sample_rate_ms = 10,
	.deadband = 0,
	.setpoint = 97,
	.p_gain_pct = 20,
	.d_gain_pct = 0,
	.i_gain_pct = 0,
};

struct perf_limits {
	int no_turbo;
	int max_perf_pct;
	int min_perf_pct;
	int32_t max_perf;
	int32_t min_perf;
	int max_policy_pct;
	int max_sysfs_pct;
};

static struct perf_limits limits = {
	.no_turbo = 0,
	.max_perf_pct = 100,
	.max_perf = int_tofp(1),
	.min_perf_pct = 0,
	.min_perf = 0,
	.max_policy_pct = 100,
	.max_sysfs_pct = 100,
};

static inline void pid_reset(struct _pid *pid, int setpoint, int busy,
			int deadband, int integral) {
	pid->setpoint = setpoint;
	pid->deadband  = deadband;
	pid->integral  = int_tofp(integral);
	pid->last_err  = setpoint - busy;
}

static inline void pid_p_gain_set(struct _pid *pid, int percent)
{
	pid->p_gain = div_fp(int_tofp(percent), int_tofp(100));
}

static inline void pid_i_gain_set(struct _pid *pid, int percent)
{
	pid->i_gain = div_fp(int_tofp(percent), int_tofp(100));
}

static inline void pid_d_gain_set(struct _pid *pid, int percent)
{

	pid->d_gain = div_fp(int_tofp(percent), int_tofp(100));
}

static signed int pid_calc(struct _pid *pid, int busy)
{
	signed int err, result;
	int32_t pterm, dterm, fp_error;
	int32_t integral_limit;

	err = pid->setpoint - busy;
	fp_error = int_tofp(err);

	if (abs(err) <= pid->deadband)
		return 0;

	pterm = mul_fp(pid->p_gain, fp_error);

	pid->integral += fp_error;

	/* limit the integral term */
	integral_limit = int_tofp(30);
	if (pid->integral > integral_limit)
		pid->integral = integral_limit;
	if (pid->integral < -integral_limit)
		pid->integral = -integral_limit;

	dterm = mul_fp(pid->d_gain, (err - pid->last_err));
	pid->last_err = err;

	result = pterm + mul_fp(pid->integral, pid->i_gain) + dterm;

	return (signed int)fp_toint(result);
}

static inline void intel_pstate_busy_pid_reset(struct cpudata *cpu)
{
	pid_p_gain_set(&cpu->pid, cpu->pstate_policy->p_gain_pct);
	pid_d_gain_set(&cpu->pid, cpu->pstate_policy->d_gain_pct);
	pid_i_gain_set(&cpu->pid, cpu->pstate_policy->i_gain_pct);

	pid_reset(&cpu->pid,
		cpu->pstate_policy->setpoint,
		100,
		cpu->pstate_policy->deadband,
		0);
}

static inline void intel_pstate_reset_all_pid(void)
{
	unsigned int cpu;
	for_each_online_cpu(cpu) {
		if (all_cpu_data[cpu])
			intel_pstate_busy_pid_reset(all_cpu_data[cpu]);
	}
}

/************************** debugfs begin ************************/
static int pid_param_set(void *data, u64 val)
{
	*(u32 *)data = val;
	intel_pstate_reset_all_pid();
	return 0;
}
static int pid_param_get(void *data, u64 *val)
{
	*val = *(u32 *)data;
	return 0;
}
DEFINE_SIMPLE_ATTRIBUTE(fops_pid_param, pid_param_get,
			pid_param_set, "%llu\n");

struct pid_param {
	char *name;
	void *value;
};

static struct pid_param pid_files[] = {
	{"sample_rate_ms", &default_policy.sample_rate_ms},
	{"d_gain_pct", &default_policy.d_gain_pct},
	{"i_gain_pct", &default_policy.i_gain_pct},
	{"deadband", &default_policy.deadband},
	{"setpoint", &default_policy.setpoint},
	{"p_gain_pct", &default_policy.p_gain_pct},
	{NULL, NULL}
};

static struct dentry *debugfs_parent;
static void intel_pstate_debug_expose_params(void)
{
	int i = 0;

	debugfs_parent = debugfs_create_dir("pstate_snb", NULL);
	if (IS_ERR_OR_NULL(debugfs_parent))
		return;
	while (pid_files[i].name) {
		debugfs_create_file(pid_files[i].name, 0660,
				debugfs_parent, pid_files[i].value,
				&fops_pid_param);
		i++;
	}
}

/************************** debugfs end ************************/

/************************** sysfs begin ************************/
#define show_one(file_name, object)					\
	static ssize_t show_##file_name					\
	(struct kobject *kobj, struct attribute *attr, char *buf)	\
	{								\
		return sprintf(buf, "%u\n", limits.object);		\
	}

static ssize_t store_no_turbo(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;
	limits.no_turbo = clamp_t(int, input, 0 , 1);

	return count;
}

static ssize_t store_max_perf_pct(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;

	limits.max_sysfs_pct = clamp_t(int, input, 0 , 100);
	limits.max_perf_pct = min(limits.max_policy_pct, limits.max_sysfs_pct);
	limits.max_perf = div_fp(int_tofp(limits.max_perf_pct), int_tofp(100));
	return count;
}

static ssize_t store_min_perf_pct(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;
	limits.min_perf_pct = clamp_t(int, input, 0 , 100);
	limits.min_perf = div_fp(int_tofp(limits.min_perf_pct), int_tofp(100));

	return count;
}

show_one(no_turbo, no_turbo);
show_one(max_perf_pct, max_perf_pct);
show_one(min_perf_pct, min_perf_pct);

define_one_global_rw(no_turbo);
define_one_global_rw(max_perf_pct);
define_one_global_rw(min_perf_pct);

static struct attribute *intel_pstate_attributes[] = {
	&no_turbo.attr,
	&max_perf_pct.attr,
	&min_perf_pct.attr,
	NULL
};

static struct attribute_group intel_pstate_attr_group = {
	.attrs = intel_pstate_attributes,
};
static struct kobject *intel_pstate_kobject;

static void intel_pstate_sysfs_expose_params(void)
{
	int rc;

	intel_pstate_kobject = kobject_create_and_add("intel_pstate",
						&cpu_subsys.dev_root->kobj);
	BUG_ON(!intel_pstate_kobject);
	rc = sysfs_create_group(intel_pstate_kobject,
				&intel_pstate_attr_group);
	BUG_ON(rc);
}

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

static int intel_pstate_min_pstate(void)
{
	u64 value;
	rdmsrl(MSR_PLATFORM_INFO, value);
	return (value >> 40) & 0xFF;
}

static int intel_pstate_max_pstate(void)
{
	u64 value;
	rdmsrl(MSR_PLATFORM_INFO, value);
	return (value >> 8) & 0xFF;
}

static int intel_pstate_turbo_pstate(void)
{
	u64 value;
	int nont, ret;
	rdmsrl(MSR_NHM_TURBO_RATIO_LIMIT, value);
	nont = intel_pstate_max_pstate();
	ret = ((value) & 255);
	if (ret <= nont)
		ret = nont;
	return ret;
}

static void intel_pstate_get_min_max(struct cpudata *cpu, int *min, int *max)
{
	int max_perf = cpu->pstate.turbo_pstate;
	int min_perf;
	if (limits.no_turbo)
		max_perf = cpu->pstate.max_pstate;

	max_perf = fp_toint(mul_fp(int_tofp(max_perf), limits.max_perf));
	*max = clamp_t(int, max_perf,
			cpu->pstate.min_pstate, cpu->pstate.turbo_pstate);

	min_perf = fp_toint(mul_fp(int_tofp(max_perf), limits.min_perf));
	*min = clamp_t(int, min_perf,
			cpu->pstate.min_pstate, max_perf);
}

static void intel_pstate_set_pstate(struct cpudata *cpu, int pstate)
{
	int max_perf, min_perf;

	intel_pstate_get_min_max(cpu, &min_perf, &max_perf);

	pstate = clamp_t(int, pstate, min_perf, max_perf);

	if (pstate == cpu->pstate.current_pstate)
		return;

	trace_cpu_frequency(pstate * 100000, cpu->cpu);

	cpu->pstate.current_pstate = pstate;
	if (limits.no_turbo)
		wrmsrl(MSR_IA32_PERF_CTL, BIT(32) | (pstate << 8));
	else
		wrmsrl(MSR_IA32_PERF_CTL, pstate << 8);

}

static inline void intel_pstate_pstate_increase(struct cpudata *cpu, int steps)
{
	int target;
	target = cpu->pstate.current_pstate + steps;

	intel_pstate_set_pstate(cpu, target);
}

static inline void intel_pstate_pstate_decrease(struct cpudata *cpu, int steps)
{
	int target;
	target = cpu->pstate.current_pstate - steps;
	intel_pstate_set_pstate(cpu, target);
}

static void intel_pstate_get_cpu_pstates(struct cpudata *cpu)
{
	sprintf(cpu->name, "Intel 2nd generation core");

	cpu->pstate.min_pstate = intel_pstate_min_pstate();
	cpu->pstate.max_pstate = intel_pstate_max_pstate();
	cpu->pstate.turbo_pstate = intel_pstate_turbo_pstate();

	/*
	 * goto max pstate so we don't slow up boot if we are built-in if we are
	 * a module we will take care of it during normal operation
	 */
	intel_pstate_set_pstate(cpu, cpu->pstate.max_pstate);
}

static inline void intel_pstate_calc_busy(struct cpudata *cpu,
					struct sample *sample)
{
	u64 core_pct;
	core_pct = div64_u64(sample->aperf * 100, sample->mperf);
	sample->freq = cpu->pstate.max_pstate * core_pct * 1000;

	sample->core_pct_busy = core_pct;
}

static inline void intel_pstate_sample(struct cpudata *cpu)
{
	u64 aperf, mperf;

	rdmsrl(MSR_IA32_APERF, aperf);
	rdmsrl(MSR_IA32_MPERF, mperf);
	cpu->sample_ptr = (cpu->sample_ptr + 1) % SAMPLE_COUNT;
	cpu->samples[cpu->sample_ptr].aperf = aperf;
	cpu->samples[cpu->sample_ptr].mperf = mperf;
	cpu->samples[cpu->sample_ptr].aperf -= cpu->prev_aperf;
	cpu->samples[cpu->sample_ptr].mperf -= cpu->prev_mperf;

	intel_pstate_calc_busy(cpu, &cpu->samples[cpu->sample_ptr]);

	cpu->prev_aperf = aperf;
	cpu->prev_mperf = mperf;
}

static inline void intel_pstate_set_sample_time(struct cpudata *cpu)
{
	int sample_time, delay;

	sample_time = cpu->pstate_policy->sample_rate_ms;
	delay = msecs_to_jiffies(sample_time);
	mod_timer_pinned(&cpu->timer, jiffies + delay);
}

static inline int intel_pstate_get_scaled_busy(struct cpudata *cpu)
{
	int32_t busy_scaled;
	int32_t core_busy, max_pstate, current_pstate;

	core_busy = int_tofp(cpu->samples[cpu->sample_ptr].core_pct_busy);
	max_pstate = int_tofp(cpu->pstate.max_pstate);
	current_pstate = int_tofp(cpu->pstate.current_pstate);
	busy_scaled = mul_fp(core_busy, div_fp(max_pstate, current_pstate));

	return fp_toint(busy_scaled);
}

static inline void intel_pstate_adjust_busy_pstate(struct cpudata *cpu)
{
	int busy_scaled;
	struct _pid *pid;
	signed int ctl = 0;
	int steps;

	pid = &cpu->pid;
	busy_scaled = intel_pstate_get_scaled_busy(cpu);

	ctl = pid_calc(pid, busy_scaled);

	steps = abs(ctl);
	if (ctl < 0)
		intel_pstate_pstate_increase(cpu, steps);
	else
		intel_pstate_pstate_decrease(cpu, steps);
}

static void intel_pstate_timer_func(unsigned long __data)
{
	struct cpudata *cpu = (struct cpudata *) __data;

	intel_pstate_sample(cpu);
	intel_pstate_adjust_busy_pstate(cpu);

	if (cpu->pstate.current_pstate == cpu->pstate.min_pstate) {
		cpu->min_pstate_count++;
		if (!(cpu->min_pstate_count % 5)) {
			intel_pstate_set_pstate(cpu, cpu->pstate.max_pstate);
		}
	} else
		cpu->min_pstate_count = 0;

	intel_pstate_set_sample_time(cpu);
}

#define ICPU(model, policy) \
	{ X86_VENDOR_INTEL, 6, model, X86_FEATURE_ANY, (unsigned long)&policy }

static const struct x86_cpu_id intel_pstate_cpu_ids[] = {
	ICPU(0x2a, default_policy),
	ICPU(0x2d, default_policy),
	ICPU(0x3a, default_policy),
	ICPU(0x3c, default_policy),
	ICPU(0x3e, default_policy),
	ICPU(0x3f, default_policy),
	ICPU(0x45, default_policy),
	ICPU(0x46, default_policy),
	{}
};
MODULE_DEVICE_TABLE(x86cpu, intel_pstate_cpu_ids);

static int intel_pstate_init_cpu(unsigned int cpunum)
{

	const struct x86_cpu_id *id;
	struct cpudata *cpu;

	id = x86_match_cpu(intel_pstate_cpu_ids);
	if (!id)
		return -ENODEV;

	all_cpu_data[cpunum] = kzalloc(sizeof(struct cpudata), GFP_KERNEL);
	if (!all_cpu_data[cpunum])
		return -ENOMEM;

	cpu = all_cpu_data[cpunum];

	intel_pstate_get_cpu_pstates(cpu);

	cpu->cpu = cpunum;
	cpu->pstate_policy =
		(struct pstate_adjust_policy *)id->driver_data;
	init_timer_deferrable(&cpu->timer);
	cpu->timer.function = intel_pstate_timer_func;
	cpu->timer.data =
		(unsigned long)cpu;
	cpu->timer.expires = jiffies + HZ/100;
	intel_pstate_busy_pid_reset(cpu);
	intel_pstate_sample(cpu);
	intel_pstate_set_pstate(cpu, cpu->pstate.max_pstate);

	add_timer_on(&cpu->timer, cpunum);

	pr_info("Intel pstate controlling: cpu %d\n", cpunum);

	return 0;
}

static unsigned int intel_pstate_get(unsigned int cpu_num)
{
	struct sample *sample;
	struct cpudata *cpu;

	cpu = all_cpu_data[cpu_num];
	if (!cpu)
		return 0;
	sample = &cpu->samples[cpu->sample_ptr];
	return sample->freq;
}

static int intel_pstate_set_policy(struct cpufreq_policy *policy)
{
	struct cpudata *cpu;

	cpu = all_cpu_data[policy->cpu];

	if (!policy->cpuinfo.max_freq)
		return -ENODEV;

	if (policy->policy == CPUFREQ_POLICY_PERFORMANCE) {
		limits.min_perf_pct = 100;
		limits.min_perf = int_tofp(1);
		limits.max_perf_pct = 100;
		limits.max_perf = int_tofp(1);
		limits.no_turbo = 0;
		return 0;
	}
	limits.min_perf_pct = (policy->min * 100) / policy->cpuinfo.max_freq;
	limits.min_perf_pct = clamp_t(int, limits.min_perf_pct, 0 , 100);
	limits.min_perf = div_fp(int_tofp(limits.min_perf_pct), int_tofp(100));

	limits.max_policy_pct = policy->max * 100 / policy->cpuinfo.max_freq;
	limits.max_policy_pct = clamp_t(int, limits.max_policy_pct, 0 , 100);
	limits.max_perf_pct = min(limits.max_policy_pct, limits.max_sysfs_pct);
	limits.max_perf = div_fp(int_tofp(limits.max_perf_pct), int_tofp(100));

	return 0;
}

static int intel_pstate_verify_policy(struct cpufreq_policy *policy)
{
	cpufreq_verify_within_limits(policy,
				policy->cpuinfo.min_freq,
				policy->cpuinfo.max_freq);

	if ((policy->policy != CPUFREQ_POLICY_POWERSAVE) &&
		(policy->policy != CPUFREQ_POLICY_PERFORMANCE))
		return -EINVAL;

	return 0;
}

static int intel_pstate_cpu_exit(struct cpufreq_policy *policy)
{
	int cpu = policy->cpu;

	del_timer(&all_cpu_data[cpu]->timer);
	kfree(all_cpu_data[cpu]);
	all_cpu_data[cpu] = NULL;
	return 0;
}

static int intel_pstate_cpu_init(struct cpufreq_policy *policy)
{
	int rc, min_pstate, max_pstate;
	struct cpudata *cpu;

	rc = intel_pstate_init_cpu(policy->cpu);
	if (rc)
		return rc;

	cpu = all_cpu_data[policy->cpu];

	if (!limits.no_turbo &&
		limits.min_perf_pct == 100 && limits.max_perf_pct == 100)
		policy->policy = CPUFREQ_POLICY_PERFORMANCE;
	else
		policy->policy = CPUFREQ_POLICY_POWERSAVE;

	intel_pstate_get_min_max(cpu, &min_pstate, &max_pstate);
	policy->min = min_pstate * 100000;
	policy->max = max_pstate * 100000;

	/* cpuinfo and default policy values */
	policy->cpuinfo.min_freq = cpu->pstate.min_pstate * 100000;
	policy->cpuinfo.max_freq = cpu->pstate.turbo_pstate * 100000;
	policy->cpuinfo.transition_latency = CPUFREQ_ETERNAL;
	cpumask_set_cpu(policy->cpu, policy->cpus);

	return 0;
}

static struct cpufreq_driver intel_pstate_driver = {
	.flags		= CPUFREQ_CONST_LOOPS,
	.verify		= intel_pstate_verify_policy,
	.setpolicy	= intel_pstate_set_policy,
	.get		= intel_pstate_get,
	.init		= intel_pstate_cpu_init,
	.exit		= intel_pstate_cpu_exit,
	.name		= "intel_pstate",
};

static int __initdata no_load;

static int intel_pstate_msrs_not_valid(void)
{
	/* Check that all the msr's we are using are valid. */
	u64 aperf, mperf, tmp;

	rdmsrl(MSR_IA32_APERF, aperf);
	rdmsrl(MSR_IA32_MPERF, mperf);

	if (!intel_pstate_min_pstate() ||
		!intel_pstate_max_pstate() ||
		!intel_pstate_turbo_pstate())
		return -ENODEV;

	rdmsrl(MSR_IA32_APERF, tmp);
	if (!(tmp - aperf))
		return -ENODEV;

	rdmsrl(MSR_IA32_MPERF, tmp);
	if (!(tmp - mperf))
		return -ENODEV;

	return 0;
}
static int __init intel_pstate_init(void)
{
	int cpu, rc = 0;
	const struct x86_cpu_id *id;

	if (no_load)
		return -ENODEV;

	id = x86_match_cpu(intel_pstate_cpu_ids);
	if (!id)
		return -ENODEV;

	if (intel_pstate_msrs_not_valid())
		return -ENODEV;

	pr_info("Intel P-state driver initializing.\n");

	all_cpu_data = vzalloc(sizeof(void *) * num_possible_cpus());
	if (!all_cpu_data)
		return -ENOMEM;

	rc = cpufreq_register_driver(&intel_pstate_driver);
	if (rc)
		goto out;

	intel_pstate_debug_expose_params();
	intel_pstate_sysfs_expose_params();
	return rc;
out:
	get_online_cpus();
	for_each_online_cpu(cpu) {
		if (all_cpu_data[cpu]) {
			del_timer_sync(&all_cpu_data[cpu]->timer);
			kfree(all_cpu_data[cpu]);
		}
	}

	put_online_cpus();
	vfree(all_cpu_data);
	return -ENODEV;
}
device_initcall(intel_pstate_init);

static int __init intel_pstate_setup(char *str)
{
	if (!str)
		return -EINVAL;

	if (!strcmp(str, "disable"))
		no_load = 1;
	return 0;
}
early_param("intel_pstate", intel_pstate_setup);

MODULE_AUTHOR("Dirk Brandewie <dirk.j.brandewie@intel.com>");
MODULE_DESCRIPTION("'intel_pstate' - P state driver Intel Core processors");
MODULE_LICENSE("GPL");
Commit	Line	Data
93f0822d	1	/*
d1b68485	2	* intel_pstate.c: Native P state management for Intel processors
93f0822d DB	3	*
	4	* (C) Copyright 2012 Intel Corporation
	5	* Author: Dirk Brandewie <dirk.j.brandewie@intel.com>
	6	*
	7	* This program is free software; you can redistribute it and/or
	8	* modify it under the terms of the GNU General Public License
	9	* as published by the Free Software Foundation; version 2
	10	* of the License.
	11	*/
	12
	13	#include <linux/kernel.h>
	14	#include <linux/kernel_stat.h>
	15	#include <linux/module.h>
	16	#include <linux/ktime.h>
	17	#include <linux/hrtimer.h>
	18	#include <linux/tick.h>
	19	#include <linux/slab.h>
	20	#include <linux/sched.h>
	21	#include <linux/list.h>
	22	#include <linux/cpu.h>
	23	#include <linux/cpufreq.h>
	24	#include <linux/sysfs.h>
	25	#include <linux/types.h>
	26	#include <linux/fs.h>
	27	#include <linux/debugfs.h>
	28	#include <trace/events/power.h>
	29
	30	#include <asm/div64.h>
	31	#include <asm/msr.h>
	32	#include <asm/cpu_device_id.h>
	33
	34	#define SAMPLE_COUNT 3
	35
	36	#define FRAC_BITS 8
	37	#define int_tofp(X) ((int64_t)(X) << FRAC_BITS)
	38	#define fp_toint(X) ((X) >> FRAC_BITS)
	39
	40	static inline int32_t mul_fp(int32_t x, int32_t y)
	41	{
	42	return ((int64_t)x * (int64_t)y) >> FRAC_BITS;
	43	}
	44
	45	static inline int32_t div_fp(int32_t x, int32_t y)
	46	{
	47	return div_s64((int64_t)x << FRAC_BITS, (int64_t)y);
	48	}
	49
	50	struct sample {
93f0822d	51	int core_pct_busy;
93f0822d DB	52	u64 aperf;
	53	u64 mperf;
	54	int freq;
	55	};
	56
	57	struct pstate_data {
	58	int current_pstate;
	59	int min_pstate;
	60	int max_pstate;
	61	int turbo_pstate;
	62	};
	63
	64	struct _pid {
	65	int setpoint;
	66	int32_t integral;
	67	int32_t p_gain;
	68	int32_t i_gain;
	69	int32_t d_gain;
	70	int deadband;
	71	int last_err;
	72	};
	73
	74	struct cpudata {
	75	int cpu;
	76
	77	char name[64];
	78
	79	struct timer_list timer;
	80
	81	struct pstate_adjust_policy *pstate_policy;
	82	struct pstate_data pstate;
	83	struct _pid pid;
93f0822d DB	84
93f0822d DB	85	int min_pstate_count;
93f0822d	86
93f0822d DB	87	u64 prev_aperf;
	88	u64 prev_mperf;
	89	int sample_ptr;
	90	struct sample samples[SAMPLE_COUNT];
	91	};
	92
	93	static struct cpudata **all_cpu_data;
	94	struct pstate_adjust_policy {
	95	int sample_rate_ms;
	96	int deadband;
	97	int setpoint;
	98	int p_gain_pct;
	99	int d_gain_pct;
	100	int i_gain_pct;
	101	};
	102
	103	static struct pstate_adjust_policy default_policy = {
	104	.sample_rate_ms = 10,
	105	.deadband = 0,
2134ed4d DB	106	.setpoint = 97,
2134ed4d DB	107	.p_gain_pct = 20,
93f0822d	108	.d_gain_pct = 0,
2134ed4d	109	.i_gain_pct = 0,
93f0822d DB	110	};
	111
	112	struct perf_limits {
	113	int no_turbo;
	114	int max_perf_pct;
	115	int min_perf_pct;
	116	int32_t max_perf;
	117	int32_t min_perf;
d8f469e9 DB	118	int max_policy_pct;
d8f469e9 DB	119	int max_sysfs_pct;
93f0822d DB	120	};
	121
	122	static struct perf_limits limits = {
	123	.no_turbo = 0,
	124	.max_perf_pct = 100,
	125	.max_perf = int_tofp(1),
	126	.min_perf_pct = 0,
	127	.min_perf = 0,
d8f469e9 DB	128	.max_policy_pct = 100,
d8f469e9 DB	129	.max_sysfs_pct = 100,
93f0822d DB	130	};
	131
	132	static inline void pid_reset(struct _pid *pid, int setpoint, int busy,
	133	int deadband, int integral) {
	134	pid->setpoint = setpoint;
	135	pid->deadband = deadband;
	136	pid->integral = int_tofp(integral);
	137	pid->last_err = setpoint - busy;
	138	}
	139
	140	static inline void pid_p_gain_set(struct _pid *pid, int percent)
	141	{
	142	pid->p_gain = div_fp(int_tofp(percent), int_tofp(100));
	143	}
	144
	145	static inline void pid_i_gain_set(struct _pid *pid, int percent)
	146	{
	147	pid->i_gain = div_fp(int_tofp(percent), int_tofp(100));
	148	}
	149
	150	static inline void pid_d_gain_set(struct _pid *pid, int percent)
	151	{
	152
	153	pid->d_gain = div_fp(int_tofp(percent), int_tofp(100));
	154	}
	155
	156	static signed int pid_calc(struct _pid *pid, int busy)
	157	{
	158	signed int err, result;
	159	int32_t pterm, dterm, fp_error;
	160	int32_t integral_limit;
	161
	162	err = pid->setpoint - busy;
	163	fp_error = int_tofp(err);
	164
	165	if (abs(err) <= pid->deadband)
	166	return 0;
	167
	168	pterm = mul_fp(pid->p_gain, fp_error);
	169
	170	pid->integral += fp_error;
	171
	172	/* limit the integral term */
	173	integral_limit = int_tofp(30);
	174	if (pid->integral > integral_limit)
	175	pid->integral = integral_limit;
	176	if (pid->integral < -integral_limit)
	177	pid->integral = -integral_limit;
	178
	179	dterm = mul_fp(pid->d_gain, (err - pid->last_err));
	180	pid->last_err = err;
	181
	182	result = pterm + mul_fp(pid->integral, pid->i_gain) + dterm;
	183
	184	return (signed int)fp_toint(result);
	185	}
	186
	187	static inline void intel_pstate_busy_pid_reset(struct cpudata *cpu)
	188	{
	189	pid_p_gain_set(&cpu->pid, cpu->pstate_policy->p_gain_pct);
	190	pid_d_gain_set(&cpu->pid, cpu->pstate_policy->d_gain_pct);
	191	pid_i_gain_set(&cpu->pid, cpu->pstate_policy->i_gain_pct);
	192
	193	pid_reset(&cpu->pid,
194	cpu->pstate_policy->setpoint,
195	100,
196	cpu->pstate_policy->deadband,
197	0);
198	}
199
93f0822d DB	200	static inline void intel_pstate_reset_all_pid(void)
	201	{
	202	unsigned int cpu;
	203	for_each_online_cpu(cpu) {
	204	if (all_cpu_data[cpu])
	205	intel_pstate_busy_pid_reset(all_cpu_data[cpu]);
	206	}
	207	}
	208
	209	/************************ debugfs begin **********************/
	210	static int pid_param_set(void *data, u64 val)
	211	{
	212	(u32 )data = val;
	213	intel_pstate_reset_all_pid();
	214	return 0;
	215	}
	216	static int pid_param_get(void data, u64 val)
	217	{
	218	val = (u32 *)data;
	219	return 0;
	220	}
	221	DEFINE_SIMPLE_ATTRIBUTE(fops_pid_param, pid_param_get,
	222	pid_param_set, "%llu\n");
	223
	224	struct pid_param {
	225	char *name;
	226	void *value;
	227	};
	228
	229	static struct pid_param pid_files[] = {
	230	{"sample_rate_ms", &default_policy.sample_rate_ms},
	231	{"d_gain_pct", &default_policy.d_gain_pct},
	232	{"i_gain_pct", &default_policy.i_gain_pct},
	233	{"deadband", &default_policy.deadband},
	234	{"setpoint", &default_policy.setpoint},
	235	{"p_gain_pct", &default_policy.p_gain_pct},
	236	{NULL, NULL}
	237	};
	238
	239	static struct dentry *debugfs_parent;
	240	static void intel_pstate_debug_expose_params(void)
	241	{
	242	int i = 0;
	243
	244	debugfs_parent = debugfs_create_dir("pstate_snb", NULL);
	245	if (IS_ERR_OR_NULL(debugfs_parent))
	246	return;
	247	while (pid_files[i].name) {
	248	debugfs_create_file(pid_files[i].name, 0660,
	249	debugfs_parent, pid_files[i].value,
	250	&fops_pid_param);
	251	i++;
	252	}
	253	}
	254
	255	/************************ debugfs end **********************/
	256
	257	/************************ sysfs begin **********************/
	258	#define show_one(file_name, object) \
	259	static ssize_t show_##file_name \
	260	(struct kobject kobj, struct attribute attr, char *buf) \
	261	{ \
	262	return sprintf(buf, "%u\n", limits.object); \
	263	}
264
265	static ssize_t store_no_turbo(struct kobject a, struct attribute b,
266	const char *buf, size_t count)
267	{
268	unsigned int input;
269	int ret;
270	ret = sscanf(buf, "%u", &input);
271	if (ret != 1)
272	return -EINVAL;
273	limits.no_turbo = clamp_t(int, input, 0 , 1);
274
275	return count;
276	}
277
278	static ssize_t store_max_perf_pct(struct kobject a, struct attribute b,
279	const char *buf, size_t count)
280	{
281	unsigned int input;
282	int ret;
283	ret = sscanf(buf, "%u", &input);
284	if (ret != 1)
285	return -EINVAL;
286
d8f469e9 DB	287	limits.max_sysfs_pct = clamp_t(int, input, 0 , 100);
d8f469e9 DB	288	limits.max_perf_pct = min(limits.max_policy_pct, limits.max_sysfs_pct);
93f0822d DB	289	limits.max_perf = div_fp(int_tofp(limits.max_perf_pct), int_tofp(100));
	290	return count;
	291	}
	292
	293	static ssize_t store_min_perf_pct(struct kobject a, struct attribute b,
	294	const char *buf, size_t count)
	295	{
	296	unsigned int input;
	297	int ret;
	298	ret = sscanf(buf, "%u", &input);
	299	if (ret != 1)
	300	return -EINVAL;
	301	limits.min_perf_pct = clamp_t(int, input, 0 , 100);
	302	limits.min_perf = div_fp(int_tofp(limits.min_perf_pct), int_tofp(100));
	303
	304	return count;
	305	}
	306
	307	show_one(no_turbo, no_turbo);
	308	show_one(max_perf_pct, max_perf_pct);
	309	show_one(min_perf_pct, min_perf_pct);
	310
	311	define_one_global_rw(no_turbo);
	312	define_one_global_rw(max_perf_pct);
	313	define_one_global_rw(min_perf_pct);
	314
	315	static struct attribute *intel_pstate_attributes[] = {
	316	&no_turbo.attr,
	317	&max_perf_pct.attr,
	318	&min_perf_pct.attr,
	319	NULL
	320	};
	321
	322	static struct attribute_group intel_pstate_attr_group = {
	323	.attrs = intel_pstate_attributes,
	324	};
	325	static struct kobject *intel_pstate_kobject;
	326
	327	static void intel_pstate_sysfs_expose_params(void)
	328	{
	329	int rc;
	330
	331	intel_pstate_kobject = kobject_create_and_add("intel_pstate",
	332	&cpu_subsys.dev_root->kobj);
	333	BUG_ON(!intel_pstate_kobject);
	334	rc = sysfs_create_group(intel_pstate_kobject,
	335	&intel_pstate_attr_group);
	336	BUG_ON(rc);
	337	}
	338
	339	/************************ sysfs end **********************/
	340
	341	static int intel_pstate_min_pstate(void)
	342	{
	343	u64 value;
05e99c8c	344	rdmsrl(MSR_PLATFORM_INFO, value);
93f0822d DB	345	return (value >> 40) & 0xFF;
	346	}
	347
	348	static int intel_pstate_max_pstate(void)
	349	{
	350	u64 value;
05e99c8c	351	rdmsrl(MSR_PLATFORM_INFO, value);
93f0822d DB	352	return (value >> 8) & 0xFF;
	353	}
	354
	355	static int intel_pstate_turbo_pstate(void)
	356	{
	357	u64 value;
	358	int nont, ret;
05e99c8c	359	rdmsrl(MSR_NHM_TURBO_RATIO_LIMIT, value);
93f0822d DB	360	nont = intel_pstate_max_pstate();
	361	ret = ((value) & 255);
	362	if (ret <= nont)
	363	ret = nont;
	364	return ret;
	365	}
	366
	367	static void intel_pstate_get_min_max(struct cpudata cpu, int min, int *max)
	368	{
	369	int max_perf = cpu->pstate.turbo_pstate;
	370	int min_perf;
	371	if (limits.no_turbo)
	372	max_perf = cpu->pstate.max_pstate;
	373
	374	max_perf = fp_toint(mul_fp(int_tofp(max_perf), limits.max_perf));
	375	*max = clamp_t(int, max_perf,
	376	cpu->pstate.min_pstate, cpu->pstate.turbo_pstate);
	377
	378	min_perf = fp_toint(mul_fp(int_tofp(max_perf), limits.min_perf));
	379	*min = clamp_t(int, min_perf,
	380	cpu->pstate.min_pstate, max_perf);
	381	}
	382
	383	static void intel_pstate_set_pstate(struct cpudata *cpu, int pstate)
	384	{
	385	int max_perf, min_perf;
	386
	387	intel_pstate_get_min_max(cpu, &min_perf, &max_perf);
	388
	389	pstate = clamp_t(int, pstate, min_perf, max_perf);
	390
	391	if (pstate == cpu->pstate.current_pstate)
	392	return;
	393
93f0822d	394	trace_cpu_frequency(pstate * 100000, cpu->cpu);
35363e94	395
93f0822d	396	cpu->pstate.current_pstate = pstate;
1ccf7a1c SP	397	if (limits.no_turbo)
	398	wrmsrl(MSR_IA32_PERF_CTL, BIT(32) \| (pstate << 8));
	399	else
	400	wrmsrl(MSR_IA32_PERF_CTL, pstate << 8);
93f0822d DB	401
	402	}
	403
	404	static inline void intel_pstate_pstate_increase(struct cpudata *cpu, int steps)
	405	{
	406	int target;
	407	target = cpu->pstate.current_pstate + steps;
	408
	409	intel_pstate_set_pstate(cpu, target);
	410	}
	411
	412	static inline void intel_pstate_pstate_decrease(struct cpudata *cpu, int steps)
	413	{
	414	int target;
	415	target = cpu->pstate.current_pstate - steps;
	416	intel_pstate_set_pstate(cpu, target);
	417	}
	418
	419	static void intel_pstate_get_cpu_pstates(struct cpudata *cpu)
	420	{
	421	sprintf(cpu->name, "Intel 2nd generation core");
	422
	423	cpu->pstate.min_pstate = intel_pstate_min_pstate();
	424	cpu->pstate.max_pstate = intel_pstate_max_pstate();
	425	cpu->pstate.turbo_pstate = intel_pstate_turbo_pstate();
	426
	427	/*
	428	* goto max pstate so we don't slow up boot if we are built-in if we are
	429	* a module we will take care of it during normal operation
	430	*/
	431	intel_pstate_set_pstate(cpu, cpu->pstate.max_pstate);
	432	}
	433
	434	static inline void intel_pstate_calc_busy(struct cpudata *cpu,
	435	struct sample *sample)
	436	{
	437	u64 core_pct;
93f0822d	438	core_pct = div64_u64(sample->aperf * 100, sample->mperf);
e6f3eb29	439	sample->freq = cpu->pstate.max_pstate * core_pct * 1000;
93f0822d	440
1abc4b20	441	sample->core_pct_busy = core_pct;
93f0822d DB	442	}
	443
	444	static inline void intel_pstate_sample(struct cpudata *cpu)
	445	{
93f0822d DB	446	u64 aperf, mperf;
93f0822d DB	447
93f0822d DB	448	rdmsrl(MSR_IA32_APERF, aperf);
93f0822d DB	449	rdmsrl(MSR_IA32_MPERF, mperf);
1abc4b20 DB	450	cpu->sample_ptr = (cpu->sample_ptr + 1) % SAMPLE_COUNT;
	451	cpu->samples[cpu->sample_ptr].aperf = aperf;
	452	cpu->samples[cpu->sample_ptr].mperf = mperf;
	453	cpu->samples[cpu->sample_ptr].aperf -= cpu->prev_aperf;
	454	cpu->samples[cpu->sample_ptr].mperf -= cpu->prev_mperf;
	455
	456	intel_pstate_calc_busy(cpu, &cpu->samples[cpu->sample_ptr]);
93f0822d	457
93f0822d DB	458	cpu->prev_aperf = aperf;
	459	cpu->prev_mperf = mperf;
	460	}
	461
	462	static inline void intel_pstate_set_sample_time(struct cpudata *cpu)
	463	{
	464	int sample_time, delay;
	465
	466	sample_time = cpu->pstate_policy->sample_rate_ms;
	467	delay = msecs_to_jiffies(sample_time);
93f0822d DB	468	mod_timer_pinned(&cpu->timer, jiffies + delay);
	469	}
	470
93f0822d DB	471	static inline int intel_pstate_get_scaled_busy(struct cpudata *cpu)
	472	{
	473	int32_t busy_scaled;
2134ed4d	474	int32_t core_busy, max_pstate, current_pstate;
93f0822d DB	475
93f0822d DB	476	core_busy = int_tofp(cpu->samples[cpu->sample_ptr].core_pct_busy);
2134ed4d	477	max_pstate = int_tofp(cpu->pstate.max_pstate);
93f0822d	478	current_pstate = int_tofp(cpu->pstate.current_pstate);
2134ed4d	479	busy_scaled = mul_fp(core_busy, div_fp(max_pstate, current_pstate));
93f0822d DB	480
	481	return fp_toint(busy_scaled);
	482	}
	483
	484	static inline void intel_pstate_adjust_busy_pstate(struct cpudata *cpu)
	485	{
	486	int busy_scaled;
	487	struct _pid *pid;
	488	signed int ctl = 0;
	489	int steps;
	490
	491	pid = &cpu->pid;
	492	busy_scaled = intel_pstate_get_scaled_busy(cpu);
	493
	494	ctl = pid_calc(pid, busy_scaled);
	495
	496	steps = abs(ctl);
	497	if (ctl < 0)
	498	intel_pstate_pstate_increase(cpu, steps);
	499	else
	500	intel_pstate_pstate_decrease(cpu, steps);
	501	}
	502
93f0822d DB	503	static void intel_pstate_timer_func(unsigned long __data)
	504	{
	505	struct cpudata cpu = (struct cpudata ) __data;
	506
	507	intel_pstate_sample(cpu);
ca182aee	508	intel_pstate_adjust_busy_pstate(cpu);
93f0822d	509
93f0822d DB	510	if (cpu->pstate.current_pstate == cpu->pstate.min_pstate) {
	511	cpu->min_pstate_count++;
	512	if (!(cpu->min_pstate_count % 5)) {
	513	intel_pstate_set_pstate(cpu, cpu->pstate.max_pstate);
93f0822d DB	514	}
	515	} else
	516	cpu->min_pstate_count = 0;
ca182aee	517
93f0822d DB	518	intel_pstate_set_sample_time(cpu);
	519	}
	520
	521	#define ICPU(model, policy) \
	522	{ X86_VENDOR_INTEL, 6, model, X86_FEATURE_ANY, (unsigned long)&policy }
	523
	524	static const struct x86_cpu_id intel_pstate_cpu_ids[] = {
	525	ICPU(0x2a, default_policy),
	526	ICPU(0x2d, default_policy),
c96d53d6	527	ICPU(0x3a, default_policy),
6cdcdb79 NH	528	ICPU(0x3c, default_policy),
	529	ICPU(0x3e, default_policy),
	530	ICPU(0x3f, default_policy),
	531	ICPU(0x45, default_policy),
	532	ICPU(0x46, default_policy),
93f0822d DB	533	{}
	534	};
	535	MODULE_DEVICE_TABLE(x86cpu, intel_pstate_cpu_ids);
	536
	537	static int intel_pstate_init_cpu(unsigned int cpunum)
	538	{
	539
	540	const struct x86_cpu_id *id;
	541	struct cpudata *cpu;
	542
	543	id = x86_match_cpu(intel_pstate_cpu_ids);
	544	if (!id)
	545	return -ENODEV;
	546
	547	all_cpu_data[cpunum] = kzalloc(sizeof(struct cpudata), GFP_KERNEL);
	548	if (!all_cpu_data[cpunum])
	549	return -ENOMEM;
	550
	551	cpu = all_cpu_data[cpunum];
	552
	553	intel_pstate_get_cpu_pstates(cpu);
	554
	555	cpu->cpu = cpunum;
	556	cpu->pstate_policy =
	557	(struct pstate_adjust_policy *)id->driver_data;
	558	init_timer_deferrable(&cpu->timer);
	559	cpu->timer.function = intel_pstate_timer_func;
	560	cpu->timer.data =
	561	(unsigned long)cpu;
	562	cpu->timer.expires = jiffies + HZ/100;
	563	intel_pstate_busy_pid_reset(cpu);
93f0822d DB	564	intel_pstate_sample(cpu);
	565	intel_pstate_set_pstate(cpu, cpu->pstate.max_pstate);
	566
	567	add_timer_on(&cpu->timer, cpunum);
	568
	569	pr_info("Intel pstate controlling: cpu %d\n", cpunum);
	570
	571	return 0;
	572	}
	573
	574	static unsigned int intel_pstate_get(unsigned int cpu_num)
	575	{
	576	struct sample *sample;
	577	struct cpudata *cpu;
	578
	579	cpu = all_cpu_data[cpu_num];
	580	if (!cpu)
	581	return 0;
	582	sample = &cpu->samples[cpu->sample_ptr];
	583	return sample->freq;
	584	}
	585
	586	static int intel_pstate_set_policy(struct cpufreq_policy *policy)
	587	{
	588	struct cpudata *cpu;
93f0822d DB	589
	590	cpu = all_cpu_data[policy->cpu];
	591
d3929b83 DB	592	if (!policy->cpuinfo.max_freq)
	593	return -ENODEV;
	594
93f0822d DB	595	if (policy->policy == CPUFREQ_POLICY_PERFORMANCE) {
	596	limits.min_perf_pct = 100;
	597	limits.min_perf = int_tofp(1);
	598	limits.max_perf_pct = 100;
	599	limits.max_perf = int_tofp(1);
	600	limits.no_turbo = 0;
d1b68485	601	return 0;
93f0822d	602	}
d1b68485 SP	603	limits.min_perf_pct = (policy->min * 100) / policy->cpuinfo.max_freq;
	604	limits.min_perf_pct = clamp_t(int, limits.min_perf_pct, 0 , 100);
	605	limits.min_perf = div_fp(int_tofp(limits.min_perf_pct), int_tofp(100));
	606
d8f469e9 DB	607	limits.max_policy_pct = policy->max * 100 / policy->cpuinfo.max_freq;
	608	limits.max_policy_pct = clamp_t(int, limits.max_policy_pct, 0 , 100);
	609	limits.max_perf_pct = min(limits.max_policy_pct, limits.max_sysfs_pct);
d1b68485	610	limits.max_perf = div_fp(int_tofp(limits.max_perf_pct), int_tofp(100));
93f0822d DB	611
	612	return 0;
	613	}
	614
	615	static int intel_pstate_verify_policy(struct cpufreq_policy *policy)
	616	{
	617	cpufreq_verify_within_limits(policy,
	618	policy->cpuinfo.min_freq,
	619	policy->cpuinfo.max_freq);
	620
	621	if ((policy->policy != CPUFREQ_POLICY_POWERSAVE) &&
	622	(policy->policy != CPUFREQ_POLICY_PERFORMANCE))
	623	return -EINVAL;
	624
	625	return 0;
	626	}
	627
2760984f	628	static int intel_pstate_cpu_exit(struct cpufreq_policy *policy)
93f0822d DB	629	{
	630	int cpu = policy->cpu;
	631
	632	del_timer(&all_cpu_data[cpu]->timer);
	633	kfree(all_cpu_data[cpu]);
	634	all_cpu_data[cpu] = NULL;
	635	return 0;
	636	}
	637
2760984f	638	static int intel_pstate_cpu_init(struct cpufreq_policy *policy)
93f0822d DB	639	{
	640	int rc, min_pstate, max_pstate;
	641	struct cpudata *cpu;
	642
	643	rc = intel_pstate_init_cpu(policy->cpu);
	644	if (rc)
	645	return rc;
	646
	647	cpu = all_cpu_data[policy->cpu];
	648
	649	if (!limits.no_turbo &&
	650	limits.min_perf_pct == 100 && limits.max_perf_pct == 100)
	651	policy->policy = CPUFREQ_POLICY_PERFORMANCE;
	652	else
	653	policy->policy = CPUFREQ_POLICY_POWERSAVE;
	654
	655	intel_pstate_get_min_max(cpu, &min_pstate, &max_pstate);
	656	policy->min = min_pstate * 100000;
	657	policy->max = max_pstate * 100000;
	658
	659	/* cpuinfo and default policy values */
	660	policy->cpuinfo.min_freq = cpu->pstate.min_pstate * 100000;
	661	policy->cpuinfo.max_freq = cpu->pstate.turbo_pstate * 100000;
	662	policy->cpuinfo.transition_latency = CPUFREQ_ETERNAL;
	663	cpumask_set_cpu(policy->cpu, policy->cpus);
	664
	665	return 0;
	666	}
	667
	668	static struct cpufreq_driver intel_pstate_driver = {
	669	.flags = CPUFREQ_CONST_LOOPS,
	670	.verify = intel_pstate_verify_policy,
	671	.setpolicy = intel_pstate_set_policy,
	672	.get = intel_pstate_get,
	673	.init = intel_pstate_cpu_init,
	674	.exit = intel_pstate_cpu_exit,
	675	.name = "intel_pstate",
93f0822d DB	676	};
93f0822d DB	677
6be26498 DB	678	static int __initdata no_load;
6be26498 DB	679
b563b4e3 DB	680	static int intel_pstate_msrs_not_valid(void)
	681	{
	682	/* Check that all the msr's we are using are valid. */
	683	u64 aperf, mperf, tmp;
	684
	685	rdmsrl(MSR_IA32_APERF, aperf);
	686	rdmsrl(MSR_IA32_MPERF, mperf);
	687
	688	if (!intel_pstate_min_pstate() \|\|
	689	!intel_pstate_max_pstate() \|\|
	690	!intel_pstate_turbo_pstate())
	691	return -ENODEV;
	692
	693	rdmsrl(MSR_IA32_APERF, tmp);
	694	if (!(tmp - aperf))
	695	return -ENODEV;
	696
	697	rdmsrl(MSR_IA32_MPERF, tmp);
	698	if (!(tmp - mperf))
	699	return -ENODEV;
	700
	701	return 0;
	702	}
93f0822d DB	703	static int __init intel_pstate_init(void)
93f0822d DB	704	{
907cc908	705	int cpu, rc = 0;
93f0822d DB	706	const struct x86_cpu_id *id;
93f0822d DB	707
6be26498 DB	708	if (no_load)
	709	return -ENODEV;
	710
93f0822d DB	711	id = x86_match_cpu(intel_pstate_cpu_ids);
	712	if (!id)
	713	return -ENODEV;
	714
b563b4e3 DB	715	if (intel_pstate_msrs_not_valid())
	716	return -ENODEV;
	717
93f0822d DB	718	pr_info("Intel P-state driver initializing.\n");
93f0822d DB	719
b57ffac5	720	all_cpu_data = vzalloc(sizeof(void ) num_possible_cpus());
93f0822d DB	721	if (!all_cpu_data)
93f0822d DB	722	return -ENOMEM;
93f0822d DB	723
	724	rc = cpufreq_register_driver(&intel_pstate_driver);
	725	if (rc)
	726	goto out;
	727
	728	intel_pstate_debug_expose_params();
	729	intel_pstate_sysfs_expose_params();
	730	return rc;
	731	out:
907cc908 DB	732	get_online_cpus();
	733	for_each_online_cpu(cpu) {
	734	if (all_cpu_data[cpu]) {
	735	del_timer_sync(&all_cpu_data[cpu]->timer);
	736	kfree(all_cpu_data[cpu]);
	737	}
	738	}
	739
	740	put_online_cpus();
	741	vfree(all_cpu_data);
93f0822d DB	742	return -ENODEV;
	743	}
	744	device_initcall(intel_pstate_init);
	745
6be26498 DB	746	static int __init intel_pstate_setup(char *str)
	747	{
	748	if (!str)
	749	return -EINVAL;
	750
	751	if (!strcmp(str, "disable"))
	752	no_load = 1;
	753	return 0;
	754	}
	755	early_param("intel_pstate", intel_pstate_setup);
	756
93f0822d DB	757	MODULE_AUTHOR("Dirk Brandewie <dirk.j.brandewie@intel.com>");
	758	MODULE_DESCRIPTION("'intel_pstate' - P state driver Intel Core processors");
	759	MODULE_LICENSE("GPL");