Merge branch 'next' of git://git.kernel.org/pub/scm/linux/kernel/git/davej/cpufreq

* 'next' of git://git.kernel.org/pub/scm/linux/kernel/git/davej/cpufreq:
  [CPUFREQ] Fix BUG: using smp_processor_id() in preemptible code
  [CPUFREQ] Don't export governors for default governor
  [CPUFREQ][6/6] cpufreq: Add idle microaccounting in ondemand governor
  [CPUFREQ][5/6] cpufreq: Changes to get_cpu_idle_time_us(), used by ondemand governor
  [CPUFREQ][4/6] cpufreq_ondemand: Parameterize down differential
  [CPUFREQ][3/6] cpufreq: get_cpu_idle_time() changes in ondemand for idle-microaccounting
  [CPUFREQ][2/6] cpufreq: Change load calculation in ondemand for software coordination
  [CPUFREQ][1/6] cpufreq: Add cpu number parameter to __cpufreq_driver_getavg()
  [CPUFREQ] use deferrable delayed work init in conservative governor
  [CPUFREQ] drivers/cpufreq/cpufreq.c: Adjust error handling code involving cpufreq_cpu_put
  [CPUFREQ] add error handling for cpufreq_register_governor() error
  [CPUFREQ] acpi-cpufreq: add error handling for cpufreq_register_driver() error
  [CPUFREQ] Coding style fixes to arch/x86/kernel/cpu/cpufreq/powernow-k6.c
  [CPUFREQ] Coding style fixes to arch/x86/kernel/cpu/cpufreq/elanfreq.c

arch/x86/kernel/cpu/cpufreq/acpi-cpufreq.c

@@ -256,7 +256,8 @@ static u32 get_cur_val(const cpumask_t *mask)
  * Only IA32_APERF/IA32_MPERF ratio is architecturally defined and
  * no meaning should be associated with absolute values of these MSRs.
  */
-static unsigned int get_measured_perf(unsigned int cpu)
+static unsigned int get_measured_perf(struct cpufreq_policy *policy,
+				      unsigned int cpu)
 {
 	union {
 		struct {
@@ -326,7 +327,7 @@ static unsigned int get_measured_perf(unsigned int cpu)
 
 #endif
 
-	retval = per_cpu(drv_data, cpu)->max_freq * perf_percent / 100;
+	retval = per_cpu(drv_data, policy->cpu)->max_freq * perf_percent / 100;
 
 	put_cpu();
 	set_cpus_allowed_ptr(current, &saved_mask);
@@ -785,7 +786,11 @@ static int __init acpi_cpufreq_init(void)
 	if (ret)
 		return ret;
 
-	return cpufreq_register_driver(&acpi_cpufreq_driver);
+	ret = cpufreq_register_driver(&acpi_cpufreq_driver);
+	if (ret)
+		free_percpu(acpi_perf_data);
+
+	return ret;
 }
 
 static void __exit acpi_cpufreq_exit(void)
@@ -795,8 +800,6 @@ static void __exit acpi_cpufreq_exit(void)
 	cpufreq_unregister_driver(&acpi_cpufreq_driver);
 
 	free_percpu(acpi_perf_data);
-
-	return;
 }
 
 module_param(acpi_pstate_strict, uint, 0644);

arch/x86/kernel/cpu/cpufreq/elanfreq.c

@@ -25,8 +25,8 @@
 #include <linux/cpufreq.h>
 
 #include <asm/msr.h>
-#include <asm/timex.h>
-#include <asm/io.h>
+#include <linux/timex.h>
+#include <linux/io.h>
 
 #define REG_CSCIR 0x22		/* Chip Setup and Control Index Register    */
 #define REG_CSCDR 0x23		/* Chip Setup and Control Data  Register    */
@@ -82,7 +82,7 @@ static unsigned int elanfreq_get_cpu_frequency(unsigned int cpu)
 	u8 clockspeed_reg;    /* Clock Speed Register */
 
 	local_irq_disable();
-	outb_p(0x80,REG_CSCIR);
+	outb_p(0x80, REG_CSCIR);
 	clockspeed_reg = inb_p(REG_CSCDR);
 	local_irq_enable();
 
@@ -98,10 +98,10 @@ static unsigned int elanfreq_get_cpu_frequency(unsigned int cpu)
 	}
 
 	/* 33 MHz is not 32 MHz... */
-	if ((clockspeed_reg & 0xE0)==0xA0)
+	if ((clockspeed_reg & 0xE0) == 0xA0)
 		return 33000;
 
-	return ((1<<((clockspeed_reg & 0xE0) >> 5)) * 1000);
+	return (1<<((clockspeed_reg & 0xE0) >> 5)) * 1000;
 }
 
 
@@ -117,7 +117,7 @@ static unsigned int elanfreq_get_cpu_frequency(unsigned int cpu)
  *	There is no return value.
  */
 
-static void elanfreq_set_cpu_state (unsigned int state)
+static void elanfreq_set_cpu_state(unsigned int state)
 {
 	struct cpufreq_freqs    freqs;
 
@@ -144,20 +144,20 @@ static void elanfreq_set_cpu_state (unsigned int state)
 	 */
 
 	local_irq_disable();
-	outb_p(0x40,REG_CSCIR);		/* Disable hyperspeed mode */
-	outb_p(0x00,REG_CSCDR);
+	outb_p(0x40, REG_CSCIR);		/* Disable hyperspeed mode */
+	outb_p(0x00, REG_CSCDR);
 	local_irq_enable();		/* wait till internal pipelines and */
 	udelay(1000);			/* buffers have cleaned up          */
 
 	local_irq_disable();
 
 	/* now, set the CPU clock speed register (0x80) */
-	outb_p(0x80,REG_CSCIR);
-	outb_p(elan_multiplier[state].val80h,REG_CSCDR);
+	outb_p(0x80, REG_CSCIR);
+	outb_p(elan_multiplier[state].val80h, REG_CSCDR);
 
 	/* now, the hyperspeed bit in PMU Force Mode Register (0x40) */
-	outb_p(0x40,REG_CSCIR);
-	outb_p(elan_multiplier[state].val40h,REG_CSCDR);
+	outb_p(0x40, REG_CSCIR);
+	outb_p(elan_multiplier[state].val40h, REG_CSCDR);
 	udelay(10000);
 	local_irq_enable();
 
@@ -173,12 +173,12 @@ static void elanfreq_set_cpu_state (unsigned int state)
  *	for the hardware supported by the driver.
  */
 
-static int elanfreq_verify (struct cpufreq_policy *policy)
+static int elanfreq_verify(struct cpufreq_policy *policy)
 {
 	return cpufreq_frequency_table_verify(policy, &elanfreq_table[0]);
 }
 
-static int elanfreq_target (struct cpufreq_policy *policy,
+static int elanfreq_target(struct cpufreq_policy *policy,
 			    unsigned int target_freq,
 			    unsigned int relation)
 {
@@ -205,7 +205,7 @@ static int elanfreq_cpu_init(struct cpufreq_policy *policy)
 
 	/* capability check */
 	if ((c->x86_vendor != X86_VENDOR_AMD) ||
-	    (c->x86 != 4) || (c->x86_model!=10))
+	    (c->x86 != 4) || (c->x86_model != 10))
 		return -ENODEV;
 
 	/* max freq */
@@ -213,7 +213,7 @@ static int elanfreq_cpu_init(struct cpufreq_policy *policy)
 		max_freq = elanfreq_get_cpu_frequency(0);
 
 	/* table init */
-	for (i=0; (elanfreq_table[i].frequency != CPUFREQ_TABLE_END); i++) {
+	for (i = 0; (elanfreq_table[i].frequency != CPUFREQ_TABLE_END); i++) {
 		if (elanfreq_table[i].frequency > max_freq)
 			elanfreq_table[i].frequency = CPUFREQ_ENTRY_INVALID;
 	}
@@ -224,7 +224,7 @@ static int elanfreq_cpu_init(struct cpufreq_policy *policy)
 
 	result = cpufreq_frequency_table_cpuinfo(policy, elanfreq_table);
 	if (result)
-		return (result);
+		return result;
 
 	cpufreq_frequency_table_get_attr(elanfreq_table, policy->cpu);
 	return 0;
@@ -260,7 +260,7 @@ __setup("elanfreq=", elanfreq_setup);
 #endif
 
 
-static struct freq_attr* elanfreq_attr[] = {
+static struct freq_attr *elanfreq_attr[] = {
 	&cpufreq_freq_attr_scaling_available_freqs,
 	NULL,
 };
@@ -284,9 +284,9 @@ static int __init elanfreq_init(void)
 
 	/* Test if we have the right hardware */
 	if ((c->x86_vendor != X86_VENDOR_AMD) ||
-		(c->x86 != 4) || (c->x86_model!=10)) {
+		(c->x86 != 4) || (c->x86_model != 10)) {
 		printk(KERN_INFO "elanfreq: error: no Elan processor found!\n");
-                return -ENODEV;
+		return -ENODEV;
 	}
 	return cpufreq_register_driver(&elanfreq_driver);
 }
@@ -298,7 +298,7 @@ static void __exit elanfreq_exit(void)
 }
 
 
-module_param (max_freq, int, 0444);
+module_param(max_freq, int, 0444);
 
 MODULE_LICENSE("GPL");
 MODULE_AUTHOR("Robert Schwebel <r.schwebel@pengutronix.de>, Sven Geggus <sven@geggus.net>");

arch/x86/kernel/cpu/cpufreq/powernow-k6.c

@@ -15,12 +15,11 @@
 #include <linux/slab.h>
 
 #include <asm/msr.h>
-#include <asm/timex.h>
-#include <asm/io.h>
+#include <linux/timex.h>
+#include <linux/io.h>
 
-
-#define POWERNOW_IOPORT 0xfff0         /* it doesn't matter where, as long
-					  as it is unused */
+#define POWERNOW_IOPORT 0xfff0          /* it doesn't matter where, as long
+					   as it is unused */
 
 static unsigned int                     busfreq;   /* FSB, in 10 kHz */
 static unsigned int                     max_multiplier;
@@ -53,7 +52,7 @@ static int powernow_k6_get_cpu_multiplier(void)
 
 	msrval = POWERNOW_IOPORT + 0x1;
 	wrmsr(MSR_K6_EPMR, msrval, 0); /* enable the PowerNow port */
-	invalue=inl(POWERNOW_IOPORT + 0x8);
+	invalue = inl(POWERNOW_IOPORT + 0x8);
 	msrval = POWERNOW_IOPORT + 0x0;
 	wrmsr(MSR_K6_EPMR, msrval, 0); /* disable it again */
 
@@ -67,9 +66,9 @@ static int powernow_k6_get_cpu_multiplier(void)
  *
  *   Tries to change the PowerNow! multiplier
  */
-static void powernow_k6_set_state (unsigned int best_i)
+static void powernow_k6_set_state(unsigned int best_i)
 {
-	unsigned long           outvalue=0, invalue=0;
+	unsigned long           outvalue = 0, invalue = 0;
 	unsigned long           msrval;
 	struct cpufreq_freqs    freqs;
 
@@ -90,10 +89,10 @@ static void powernow_k6_set_state (unsigned int best_i)
 
 	msrval = POWERNOW_IOPORT + 0x1;
 	wrmsr(MSR_K6_EPMR, msrval, 0); /* enable the PowerNow port */
-	invalue=inl(POWERNOW_IOPORT + 0x8);
+	invalue = inl(POWERNOW_IOPORT + 0x8);
 	invalue = invalue & 0xf;
 	outvalue = outvalue | invalue;
-	outl(outvalue ,(POWERNOW_IOPORT + 0x8));
+	outl(outvalue , (POWERNOW_IOPORT + 0x8));
 	msrval = POWERNOW_IOPORT + 0x0;
 	wrmsr(MSR_K6_EPMR, msrval, 0); /* disable it again */
 
@@ -124,7 +123,7 @@ static int powernow_k6_verify(struct cpufreq_policy *policy)
  *
  * sets a new CPUFreq policy
  */
-static int powernow_k6_target (struct cpufreq_policy *policy,
+static int powernow_k6_target(struct cpufreq_policy *policy,
 			       unsigned int target_freq,
 			       unsigned int relation)
 {
@@ -152,7 +151,7 @@ static int powernow_k6_cpu_init(struct cpufreq_policy *policy)
 	busfreq = cpu_khz / max_multiplier;
 
 	/* table init */
-	for (i=0; (clock_ratio[i].frequency != CPUFREQ_TABLE_END); i++) {
+	for (i = 0; (clock_ratio[i].frequency != CPUFREQ_TABLE_END); i++) {
 		if (clock_ratio[i].index > max_multiplier)
 			clock_ratio[i].frequency = CPUFREQ_ENTRY_INVALID;
 		else
@@ -165,7 +164,7 @@ static int powernow_k6_cpu_init(struct cpufreq_policy *policy)
 
 	result = cpufreq_frequency_table_cpuinfo(policy, clock_ratio);
 	if (result)
-		return (result);
+		return result;
 
 	cpufreq_frequency_table_get_attr(clock_ratio, policy->cpu);
 
@@ -176,8 +175,8 @@ static int powernow_k6_cpu_init(struct cpufreq_policy *policy)
 static int powernow_k6_cpu_exit(struct cpufreq_policy *policy)
 {
 	unsigned int i;
-	for (i=0; i<8; i++) {
-		if (i==max_multiplier)
+	for (i = 0; i < 8; i++) {
+		if (i == max_multiplier)
 			powernow_k6_set_state(i);
 	}
 	cpufreq_frequency_table_put_attr(policy->cpu);
@@ -189,7 +188,7 @@ static unsigned int powernow_k6_get(unsigned int cpu)
 	return busfreq * powernow_k6_get_cpu_multiplier();
 }
 
-static struct freq_attr* powernow_k6_attr[] = {
+static struct freq_attr *powernow_k6_attr[] = {
 	&cpufreq_freq_attr_scaling_available_freqs,
 	NULL,
 };
@@ -227,7 +226,7 @@ static int __init powernow_k6_init(void)
 	}
 
 	if (cpufreq_register_driver(&powernow_k6_driver)) {
-		release_region (POWERNOW_IOPORT, 16);
+		release_region(POWERNOW_IOPORT, 16);
 		return -EINVAL;
 	}
 
@@ -243,13 +242,13 @@ static int __init powernow_k6_init(void)
 static void __exit powernow_k6_exit(void)
 {
 	cpufreq_unregister_driver(&powernow_k6_driver);
-	release_region (POWERNOW_IOPORT, 16);
+	release_region(POWERNOW_IOPORT, 16);
 }
 
 
-MODULE_AUTHOR ("Arjan van de Ven <arjanv@redhat.com>, Dave Jones <davej@codemonkey.org.uk>, Dominik Brodowski <linux@brodo.de>");
-MODULE_DESCRIPTION ("PowerNow! driver for AMD K6-2+ / K6-3+ processors.");
-MODULE_LICENSE ("GPL");
+MODULE_AUTHOR("Arjan van de Ven <arjanv@redhat.com>, Dave Jones <davej@codemonkey.org.uk>, Dominik Brodowski <linux@brodo.de>");
+MODULE_DESCRIPTION("PowerNow! driver for AMD K6-2+ / K6-3+ processors.");
+MODULE_LICENSE("GPL");
 
 module_init(powernow_k6_init);
 module_exit(powernow_k6_exit);

drivers/cpufreq/cpufreq.c

@@ -1467,25 +1467,27 @@ int cpufreq_driver_target(struct cpufreq_policy *policy,
 			  unsigned int target_freq,
 			  unsigned int relation)
 {
-	int ret;
+	int ret = -EINVAL;
 
 	policy = cpufreq_cpu_get(policy->cpu);
 	if (!policy)
-		return -EINVAL;
+		goto no_policy;
 
 	if (unlikely(lock_policy_rwsem_write(policy->cpu)))
-		return -EINVAL;
+		goto fail;
 
 	ret = __cpufreq_driver_target(policy, target_freq, relation);
 
 	unlock_policy_rwsem_write(policy->cpu);
 
+fail:
 	cpufreq_cpu_put(policy);
+no_policy:
 	return ret;
 }
 EXPORT_SYMBOL_GPL(cpufreq_driver_target);
 
-int __cpufreq_driver_getavg(struct cpufreq_policy *policy)
+int __cpufreq_driver_getavg(struct cpufreq_policy *policy, unsigned int cpu)
 {
 	int ret = 0;
 
@@ -1493,8 +1495,8 @@ int __cpufreq_driver_getavg(struct cpufreq_policy *policy)
 	if (!policy)
 		return -EINVAL;
 
-	if (cpu_online(policy->cpu) && cpufreq_driver->getavg)
-		ret = cpufreq_driver->getavg(policy->cpu);
+	if (cpu_online(cpu) && cpufreq_driver->getavg)
+		ret = cpufreq_driver->getavg(policy, cpu);
 
 	cpufreq_cpu_put(policy);
 	return ret;
@@ -1717,13 +1719,17 @@ int cpufreq_update_policy(unsigned int cpu)
 {
 	struct cpufreq_policy *data = cpufreq_cpu_get(cpu);
 	struct cpufreq_policy policy;
-	int ret = 0;
+	int ret;
 
-	if (!data)
-		return -ENODEV;
+	if (!data) {
+		ret = -ENODEV;
+		goto no_policy;
+	}
 
-	if (unlikely(lock_policy_rwsem_write(cpu)))
-		return -EINVAL;
+	if (unlikely(lock_policy_rwsem_write(cpu))) {
+		ret = -EINVAL;
+		goto fail;
+	}
 
 	dprintk("updating policy for CPU %u\n", cpu);
 	memcpy(&policy, data, sizeof(struct cpufreq_policy));
@@ -1750,7 +1756,9 @@ int cpufreq_update_policy(unsigned int cpu)
 
 	unlock_policy_rwsem_write(cpu);
 
+fail:
 	cpufreq_cpu_put(data);
+no_policy:
 	return ret;
 }
 EXPORT_SYMBOL(cpufreq_update_policy);

drivers/cpufreq/cpufreq_conservative.c

@@ -460,6 +460,7 @@ static void do_dbs_timer(struct work_struct *work)
 
 static inline void dbs_timer_init(void)
 {
+	init_timer_deferrable(&dbs_work.timer);
 	schedule_delayed_work(&dbs_work,
 			usecs_to_jiffies(dbs_tuners_ins.sampling_rate));
 	return;
@@ -575,13 +576,15 @@ static int cpufreq_governor_dbs(struct cpufreq_policy *policy,
 	return 0;
 }
 
+#ifndef CONFIG_CPU_FREQ_DEFAULT_GOV_CONSERVATIVE
+static
+#endif
 struct cpufreq_governor cpufreq_gov_conservative = {
 	.name			= "conservative",
 	.governor		= cpufreq_governor_dbs,
 	.max_transition_latency	= TRANSITION_LATENCY_LIMIT,
 	.owner			= THIS_MODULE,
 };
-EXPORT_SYMBOL(cpufreq_gov_conservative);
 
 static int __init cpufreq_gov_dbs_init(void)
 {

drivers/cpufreq/cpufreq_ondemand.c

@@ -18,13 +18,19 @@
 #include <linux/jiffies.h>
 #include <linux/kernel_stat.h>
 #include <linux/mutex.h>
+#include <linux/hrtimer.h>
+#include <linux/tick.h>
+#include <linux/ktime.h>
 
 /*
  * dbs is used in this file as a shortform for demandbased switching
  * It helps to keep variable names smaller, simpler
  */
 
+#define DEF_FREQUENCY_DOWN_DIFFERENTIAL		(10)
 #define DEF_FREQUENCY_UP_THRESHOLD		(80)
+#define MICRO_FREQUENCY_DOWN_DIFFERENTIAL	(3)
+#define MICRO_FREQUENCY_UP_THRESHOLD		(95)
 #define MIN_FREQUENCY_UP_THRESHOLD		(11)
 #define MAX_FREQUENCY_UP_THRESHOLD		(100)
 
@@ -57,6 +63,7 @@ enum {DBS_NORMAL_SAMPLE, DBS_SUB_SAMPLE};
 struct cpu_dbs_info_s {
 	cputime64_t prev_cpu_idle;
 	cputime64_t prev_cpu_wall;
+	cputime64_t prev_cpu_nice;
 	struct cpufreq_policy *cur_policy;
  	struct delayed_work work;
 	struct cpufreq_frequency_table *freq_table;
@@ -86,21 +93,24 @@ static struct workqueue_struct	*kondemand_wq;
 static struct dbs_tuners {
 	unsigned int sampling_rate;
 	unsigned int up_threshold;
+	unsigned int down_differential;
 	unsigned int ignore_nice;
 	unsigned int powersave_bias;
 } dbs_tuners_ins = {
 	.up_threshold = DEF_FREQUENCY_UP_THRESHOLD,
+	.down_differential = DEF_FREQUENCY_DOWN_DIFFERENTIAL,
 	.ignore_nice = 0,
 	.powersave_bias = 0,
 };
 
-static inline cputime64_t get_cpu_idle_time(unsigned int cpu)
+static inline cputime64_t get_cpu_idle_time_jiffy(unsigned int cpu,
+							cputime64_t *wall)
 {
 	cputime64_t idle_time;
-	cputime64_t cur_jiffies;
+	cputime64_t cur_wall_time;
 	cputime64_t busy_time;
 
-	cur_jiffies = jiffies64_to_cputime64(get_jiffies_64());
+	cur_wall_time = jiffies64_to_cputime64(get_jiffies_64());
 	busy_time = cputime64_add(kstat_cpu(cpu).cpustat.user,
 			kstat_cpu(cpu).cpustat.system);
 
@@ -113,7 +123,37 @@ static inline cputime64_t get_cpu_idle_time(unsigned int cpu)
 				kstat_cpu(cpu).cpustat.nice);
 	}
 
-	idle_time = cputime64_sub(cur_jiffies, busy_time);
+	idle_time = cputime64_sub(cur_wall_time, busy_time);
+	if (wall)
+		*wall = cur_wall_time;
+
+	return idle_time;
+}
+
+static inline cputime64_t get_cpu_idle_time(unsigned int cpu, cputime64_t *wall)
+{
+	u64 idle_time = get_cpu_idle_time_us(cpu, wall);
+
+	if (idle_time == -1ULL)
+		return get_cpu_idle_time_jiffy(cpu, wall);
+
+	if (dbs_tuners_ins.ignore_nice) {
+		cputime64_t cur_nice;
+		unsigned long cur_nice_jiffies;
+		struct cpu_dbs_info_s *dbs_info;
+
+		dbs_info = &per_cpu(cpu_dbs_info, cpu);
+		cur_nice = cputime64_sub(kstat_cpu(cpu).cpustat.nice,
+					 dbs_info->prev_cpu_nice);
+		/*
+		 * Assumption: nice time between sampling periods will be
+		 * less than 2^32 jiffies for 32 bit sys
+		 */
+		cur_nice_jiffies = (unsigned long)
+					cputime64_to_jiffies64(cur_nice);
+		dbs_info->prev_cpu_nice = kstat_cpu(cpu).cpustat.nice;
+		return idle_time + jiffies_to_usecs(cur_nice_jiffies);
+	}
 	return idle_time;
 }
 
@@ -277,8 +317,8 @@ static ssize_t store_ignore_nice_load(struct cpufreq_policy *policy,
 	for_each_online_cpu(j) {
 		struct cpu_dbs_info_s *dbs_info;
 		dbs_info = &per_cpu(cpu_dbs_info, j);
-		dbs_info->prev_cpu_idle = get_cpu_idle_time(j);
-		dbs_info->prev_cpu_wall = get_jiffies_64();
+		dbs_info->prev_cpu_idle = get_cpu_idle_time(j,
+						&dbs_info->prev_cpu_wall);
 	}
 	mutex_unlock(&dbs_mutex);
 
@@ -334,9 +374,7 @@ static struct attribute_group dbs_attr_group = {
 
 static void dbs_check_cpu(struct cpu_dbs_info_s *this_dbs_info)
 {
-	unsigned int idle_ticks, total_ticks;
-	unsigned int load = 0;
-	cputime64_t cur_jiffies;
+	unsigned int max_load_freq;
 
 	struct cpufreq_policy *policy;
 	unsigned int j;
@@ -346,13 +384,7 @@ static void dbs_check_cpu(struct cpu_dbs_info_s *this_dbs_info)
 
 	this_dbs_info->freq_lo = 0;
 	policy = this_dbs_info->cur_policy;
-	cur_jiffies = jiffies64_to_cputime64(get_jiffies_64());
-	total_ticks = (unsigned int) cputime64_sub(cur_jiffies,
-			this_dbs_info->prev_cpu_wall);
-	this_dbs_info->prev_cpu_wall = get_jiffies_64();
 
-	if (!total_ticks)
-		return;
 	/*
 	 * Every sampling_rate, we check, if current idle time is less
 	 * than 20% (default), then we try to increase frequency
@@ -365,27 +397,44 @@ static void dbs_check_cpu(struct cpu_dbs_info_s *this_dbs_info)
 	 * 5% (default) of current frequency
 	 */
 
-	/* Get Idle Time */
-	idle_ticks = UINT_MAX;
+	/* Get Absolute Load - in terms of freq */
+	max_load_freq = 0;
+
 	for_each_cpu_mask_nr(j, policy->cpus) {
-		cputime64_t total_idle_ticks;
-		unsigned int tmp_idle_ticks;
 		struct cpu_dbs_info_s *j_dbs_info;
+		cputime64_t cur_wall_time, cur_idle_time;
+		unsigned int idle_time, wall_time;
+		unsigned int load, load_freq;
+		int freq_avg;
 
 		j_dbs_info = &per_cpu(cpu_dbs_info, j);
-		total_idle_ticks = get_cpu_idle_time(j);
-		tmp_idle_ticks = (unsigned int) cputime64_sub(total_idle_ticks,
+
+		cur_idle_time = get_cpu_idle_time(j, &cur_wall_time);
+
+		wall_time = (unsigned int) cputime64_sub(cur_wall_time,
+				j_dbs_info->prev_cpu_wall);
+		j_dbs_info->prev_cpu_wall = cur_wall_time;
+
+		idle_time = (unsigned int) cputime64_sub(cur_idle_time,
 				j_dbs_info->prev_cpu_idle);
-		j_dbs_info->prev_cpu_idle = total_idle_ticks;
+		j_dbs_info->prev_cpu_idle = cur_idle_time;
+
+		if (unlikely(!wall_time || wall_time < idle_time))
+			continue;
+
+		load = 100 * (wall_time - idle_time) / wall_time;
+
+		freq_avg = __cpufreq_driver_getavg(policy, j);
+		if (freq_avg <= 0)
+			freq_avg = policy->cur;
 
-		if (tmp_idle_ticks < idle_ticks)
-			idle_ticks = tmp_idle_ticks;
+		load_freq = load * freq_avg;
+		if (load_freq > max_load_freq)
+			max_load_freq = load_freq;
 	}
-	if (likely(total_ticks > idle_ticks))
-		load = (100 * (total_ticks - idle_ticks)) / total_ticks;
 
 	/* Check for frequency increase */
-	if (load > dbs_tuners_ins.up_threshold) {
+	if (max_load_freq > dbs_tuners_ins.up_threshold * policy->cur) {
 		/* if we are already at full speed then break out early */
 		if (!dbs_tuners_ins.powersave_bias) {
 			if (policy->cur == policy->max)
@@ -412,15 +461,13 @@ static void dbs_check_cpu(struct cpu_dbs_info_s *this_dbs_info)
 	 * can support the current CPU usage without triggering the up
 	 * policy. To be safe, we focus 10 points under the threshold.
 	 */
-	if (load < (dbs_tuners_ins.up_threshold - 10)) {
-		unsigned int freq_next, freq_cur;
-
-		freq_cur = __cpufreq_driver_getavg(policy);
-		if (!freq_cur)
-			freq_cur = policy->cur;
-
-		freq_next = (freq_cur * load) /
-			(dbs_tuners_ins.up_threshold - 10);
+	if (max_load_freq <
+	    (dbs_tuners_ins.up_threshold - dbs_tuners_ins.down_differential) *
+	     policy->cur) {
+		unsigned int freq_next;
+		freq_next = max_load_freq /
+				(dbs_tuners_ins.up_threshold -
+				 dbs_tuners_ins.down_differential);
 
 		if (!dbs_tuners_ins.powersave_bias) {
 			__cpufreq_driver_target(policy, freq_next,
@@ -526,8 +573,8 @@ static int cpufreq_governor_dbs(struct cpufreq_policy *policy,
 			j_dbs_info = &per_cpu(cpu_dbs_info, j);
 			j_dbs_info->cur_policy = policy;
 
-			j_dbs_info->prev_cpu_idle = get_cpu_idle_time(j);
-			j_dbs_info->prev_cpu_wall = get_jiffies_64();
+			j_dbs_info->prev_cpu_idle = get_cpu_idle_time(j,
+						&j_dbs_info->prev_cpu_wall);
 		}
 		this_dbs_info->cpu = cpu;
 		/*
@@ -579,22 +626,42 @@ static int cpufreq_governor_dbs(struct cpufreq_policy *policy,
 	return 0;
 }
 
+#ifndef CONFIG_CPU_FREQ_DEFAULT_GOV_ONDEMAND
+static
+#endif
 struct cpufreq_governor cpufreq_gov_ondemand = {
 	.name			= "ondemand",
 	.governor		= cpufreq_governor_dbs,
 	.max_transition_latency = TRANSITION_LATENCY_LIMIT,
 	.owner			= THIS_MODULE,
 };
-EXPORT_SYMBOL(cpufreq_gov_ondemand);
 
 static int __init cpufreq_gov_dbs_init(void)
 {
+	int err;
+	cputime64_t wall;
+	u64 idle_time;
+	int cpu = get_cpu();
+
+	idle_time = get_cpu_idle_time_us(cpu, &wall);
+	put_cpu();
+	if (idle_time != -1ULL) {
+		/* Idle micro accounting is supported. Use finer thresholds */
+		dbs_tuners_ins.up_threshold = MICRO_FREQUENCY_UP_THRESHOLD;
+		dbs_tuners_ins.down_differential =
+					MICRO_FREQUENCY_DOWN_DIFFERENTIAL;
+	}
+
 	kondemand_wq = create_workqueue("kondemand");
 	if (!kondemand_wq) {
 		printk(KERN_ERR "Creation of kondemand failed\n");
 		return -EFAULT;
 	}
-	return cpufreq_register_governor(&cpufreq_gov_ondemand);
+	err = cpufreq_register_governor(&cpufreq_gov_ondemand);
+	if (err)
+		destroy_workqueue(kondemand_wq);
+
+	return err;
 }
 
 static void __exit cpufreq_gov_dbs_exit(void)

drivers/cpufreq/cpufreq_performance.c

@@ -36,12 +36,14 @@ static int cpufreq_governor_performance(struct cpufreq_policy *policy,
 	return 0;
 }
 
+#ifdef CONFIG_CPU_FREQ_GOV_PERFORMANCE_MODULE
+static
+#endif
 struct cpufreq_governor cpufreq_gov_performance = {
 	.name		= "performance",
 	.governor	= cpufreq_governor_performance,
 	.owner		= THIS_MODULE,
 };
-EXPORT_SYMBOL(cpufreq_gov_performance);
 
 
 static int __init cpufreq_gov_performance_init(void)

drivers/cpufreq/cpufreq_powersave.c

@@ -35,12 +35,14 @@ static int cpufreq_governor_powersave(struct cpufreq_policy *policy,
 	return 0;
 }
 
+#ifndef CONFIG_CPU_FREQ_DEFAULT_GOV_POWERSAVE
+static
+#endif
 struct cpufreq_governor cpufreq_gov_powersave = {
 	.name		= "powersave",
 	.governor	= cpufreq_governor_powersave,
 	.owner		= THIS_MODULE,
 };
-EXPORT_SYMBOL(cpufreq_gov_powersave);
 
 static int __init cpufreq_gov_powersave_init(void)
 {

drivers/cpufreq/cpufreq_userspace.c

@@ -187,6 +187,9 @@ static int cpufreq_governor_userspace(struct cpufreq_policy *policy,
 }
 
 
+#ifndef CONFIG_CPU_FREQ_DEFAULT_GOV_USERSPACE
+static
+#endif
 struct cpufreq_governor cpufreq_gov_userspace = {
 	.name		= "userspace",
 	.governor	= cpufreq_governor_userspace,
@@ -194,7 +197,6 @@ struct cpufreq_governor cpufreq_gov_userspace = {
 	.show_setspeed	= show_speed,
 	.owner		= THIS_MODULE,
 };
-EXPORT_SYMBOL(cpufreq_gov_userspace);
 
 static int __init cpufreq_gov_userspace_init(void)
 {

include/linux/cpufreq.h

@@ -187,7 +187,8 @@ extern int __cpufreq_driver_target(struct cpufreq_policy *policy,
 				   unsigned int relation);
 
 
-extern int __cpufreq_driver_getavg(struct cpufreq_policy *policy);
+extern int __cpufreq_driver_getavg(struct cpufreq_policy *policy,
+				   unsigned int cpu);
 
 int cpufreq_register_governor(struct cpufreq_governor *governor);
 void cpufreq_unregister_governor(struct cpufreq_governor *governor);
@@ -226,7 +227,9 @@ struct cpufreq_driver {
 	unsigned int	(*get)	(unsigned int cpu);
 
 	/* optional */
-	unsigned int (*getavg)	(unsigned int cpu);
+	unsigned int (*getavg)	(struct cpufreq_policy *policy,
+				 unsigned int cpu);
+
 	int	(*exit)		(struct cpufreq_policy *policy);
 	int	(*suspend)	(struct cpufreq_policy *policy, pm_message_t pmsg);
 	int	(*resume)	(struct cpufreq_policy *policy);

include/linux/tick.h

@@ -126,7 +126,7 @@ static inline ktime_t tick_nohz_get_sleep_length(void)
 	return len;
 }
 static inline void tick_nohz_stop_idle(int cpu) { }
-static inline u64 get_cpu_idle_time_us(int cpu, u64 *unused) { return 0; }
+static inline u64 get_cpu_idle_time_us(int cpu, u64 *unused) { return -1; }
 # endif /* !NO_HZ */
 
 #endif

kernel/time/tick-sched.c

@@ -20,6 +20,7 @@
 #include <linux/profile.h>
 #include <linux/sched.h>
 #include <linux/tick.h>
+#include <linux/module.h>
 
 #include <asm/irq_regs.h>
 
@@ -190,9 +191,17 @@ u64 get_cpu_idle_time_us(int cpu, u64 *last_update_time)
 {
 	struct tick_sched *ts = &per_cpu(tick_cpu_sched, cpu);
 
-	*last_update_time = ktime_to_us(ts->idle_lastupdate);
+	if (!tick_nohz_enabled)
+		return -1;
+
+	if (ts->idle_active)
+		*last_update_time = ktime_to_us(ts->idle_lastupdate);
+	else
+		*last_update_time = ktime_to_us(ktime_get());
+
 	return ktime_to_us(ts->idle_sleeptime);
 }
+EXPORT_SYMBOL_GPL(get_cpu_idle_time_us);
 
 /**
  * tick_nohz_stop_sched_tick - stop the idle tick from the idle task