Merge branch 'pm-cpufreq'

* pm-cpufreq: (46 commits)
  intel_pstate: provide option to only use intel_pstate with HWP
  cpufreq-dt: Drop unnecessary check before cpufreq_cooling_unregister() invocation
  cpufreq: Create for_each_governor()
  cpufreq: Create for_each_policy()
  cpufreq: Drop cpufreq_disabled() check from cpufreq_cpu_{get|put}()
  cpufreq: Set cpufreq_cpu_data to NULL before putting kobject
  intel_pstate: honor user space min_perf_pct override on resume
  intel_pstate: respect cpufreq policy request
  intel_pstate: Add num_pstates to sysfs
  intel_pstate: expose turbo range to sysfs
  intel_pstate: Add support for SkyLake
  cpufreq: stats: drop unnecessary locking
  cpufreq: stats: don't update stats on false notifiers
  cpufreq: stats: don't update stats from show_trans_table()
  cpufreq: stats: time_in_state can't be NULL in cpufreq_stats_update()
  cpufreq: stats: create sysfs group once we are ready
  cpufreq: remove CPUFREQ_UPDATE_POLICY_CPU notifications
  cpufreq: stats: drop 'cpu' field of struct cpufreq_stats
  cpufreq: Remove (now) unused 'last_cpu' from struct cpufreq_policy
  cpufreq: stats: rename 'struct cpufreq_stats' objects as 'stats'
  ...

Documentation/cpu-freq/intel-pstate.txt

@@ -37,6 +37,14 @@ controlling P state selection. These files have been added to
       no_turbo: limits the driver to selecting P states below the turbo
       frequency range.
 
+      turbo_pct: displays the percentage of the total performance that
+      is supported by hardware that is in the turbo range.  This number
+      is independent of whether turbo has been disabled or not.
+
+      num_pstates: displays the number of pstates that are supported
+      by hardware.  This number is independent of whether turbo has
+      been disabled or not.
+
 For contemporary Intel processors, the frequency is controlled by the
 processor itself and the P-states exposed to software are related to
 performance levels.  The idea that frequency can be set to a single

Documentation/kernel-parameters.txt

@@ -1470,6 +1470,9 @@ bytes respectively. Such letter suffixes can also be entirely omitted.
 		       no_hwp
 		         Do not enable hardware P state control (HWP)
 			 if available.
+		hwp_only
+			Only load intel_pstate on systems which support
+			hardware P state control (HWP) if available.
 
 	intremap=	[X86-64, Intel-IOMMU]
 			on	enable Interrupt Remapping (default)

arch/x86/include/uapi/asm/msr-index.h

@@ -358,6 +358,7 @@
 
 #define MSR_IA32_PERF_STATUS		0x00000198
 #define MSR_IA32_PERF_CTL		0x00000199
+#define INTEL_PERF_CTL_MASK		0xffff
 #define MSR_AMD_PSTATE_DEF_BASE		0xc0010064
 #define MSR_AMD_PERF_STATUS		0xc0010063
 #define MSR_AMD_PERF_CTL		0xc0010062

drivers/cpufreq/Kconfig.x86

@@ -57,6 +57,16 @@ config X86_ACPI_CPUFREQ_CPB
 	  By enabling this option the acpi_cpufreq driver provides the old
 	  entry in addition to the new boost ones, for compatibility reasons.
 
+config X86_SFI_CPUFREQ
+	tristate "SFI Performance-States driver"
+	depends on X86_INTEL_MID && SFI
+	help
+	  This adds a CPUFreq driver for some Silvermont based Intel Atom
+	  architectures like Z34xx and Z35xx which enumerate processor
+	  performance states through SFI.
+
+	  If in doubt, say N.
+
 config ELAN_CPUFREQ
 	tristate "AMD Elan SC400 and SC410"
 	depends on MELAN

drivers/cpufreq/Makefile

@@ -41,6 +41,7 @@ obj-$(CONFIG_X86_P4_CLOCKMOD)		+= p4-clockmod.o
 obj-$(CONFIG_X86_CPUFREQ_NFORCE2)	+= cpufreq-nforce2.o
 obj-$(CONFIG_X86_INTEL_PSTATE)		+= intel_pstate.o
 obj-$(CONFIG_X86_AMD_FREQ_SENSITIVITY)	+= amd_freq_sensitivity.o
+obj-$(CONFIG_X86_SFI_CPUFREQ)		+= sfi-cpufreq.o
 
 ##################################################################################
 # ARM SoC drivers

drivers/cpufreq/cpufreq-dt.c

@@ -320,8 +320,7 @@ static int cpufreq_exit(struct cpufreq_policy *policy)
 {
 	struct private_data *priv = policy->driver_data;
 
-	if (priv->cdev)
-		cpufreq_cooling_unregister(priv->cdev);
+	cpufreq_cooling_unregister(priv->cdev);
 	dev_pm_opp_free_cpufreq_table(priv->cpu_dev, &policy->freq_table);
 	of_free_opp_table(priv->cpu_dev);
 	clk_put(policy->clk);

drivers/cpufreq/intel_pstate.c

@@ -148,6 +148,8 @@ struct perf_limits {
 	int32_t min_perf;
 	int max_policy_pct;
 	int max_sysfs_pct;
+	int min_policy_pct;
+	int min_sysfs_pct;
 };
 
 static struct perf_limits limits = {
@@ -159,6 +161,8 @@ static struct perf_limits limits = {
 	.min_perf = 0,
 	.max_policy_pct = 100,
 	.max_sysfs_pct = 100,
+	.min_policy_pct = 0,
+	.min_sysfs_pct = 0,
 };
 
 static inline void pid_reset(struct _pid *pid, int setpoint, int busy,
@@ -338,6 +342,33 @@ static void __init intel_pstate_debug_expose_params(void)
 		return sprintf(buf, "%u\n", limits.object);		\
 	}
 
+static ssize_t show_turbo_pct(struct kobject *kobj,
+				struct attribute *attr, char *buf)
+{
+	struct cpudata *cpu;
+	int total, no_turbo, turbo_pct;
+	uint32_t turbo_fp;
+
+	cpu = all_cpu_data[0];
+
+	total = cpu->pstate.turbo_pstate - cpu->pstate.min_pstate + 1;
+	no_turbo = cpu->pstate.max_pstate - cpu->pstate.min_pstate + 1;
+	turbo_fp = div_fp(int_tofp(no_turbo), int_tofp(total));
+	turbo_pct = 100 - fp_toint(mul_fp(turbo_fp, int_tofp(100)));
+	return sprintf(buf, "%u\n", turbo_pct);
+}
+
+static ssize_t show_num_pstates(struct kobject *kobj,
+				struct attribute *attr, char *buf)
+{
+	struct cpudata *cpu;
+	int total;
+
+	cpu = all_cpu_data[0];
+	total = cpu->pstate.turbo_pstate - cpu->pstate.min_pstate + 1;
+	return sprintf(buf, "%u\n", total);
+}
+
 static ssize_t show_no_turbo(struct kobject *kobj,
 			     struct attribute *attr, char *buf)
 {
@@ -404,7 +435,9 @@ static ssize_t store_min_perf_pct(struct kobject *a, struct attribute *b,
 	ret = sscanf(buf, "%u", &input);
 	if (ret != 1)
 		return -EINVAL;
-	limits.min_perf_pct = clamp_t(int, input, 0 , 100);
+
+	limits.min_sysfs_pct = clamp_t(int, input, 0 , 100);
+	limits.min_perf_pct = max(limits.min_policy_pct, limits.min_sysfs_pct);
 	limits.min_perf = div_fp(int_tofp(limits.min_perf_pct), int_tofp(100));
 
 	if (hwp_active)
@@ -418,11 +451,15 @@ 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);
+define_one_global_ro(turbo_pct);
+define_one_global_ro(num_pstates);
 
 static struct attribute *intel_pstate_attributes[] = {
 	&no_turbo.attr,
 	&max_perf_pct.attr,
 	&min_perf_pct.attr,
+	&turbo_pct.attr,
+	&num_pstates.attr,
 	NULL
 };
 
@@ -825,6 +862,7 @@ static const struct x86_cpu_id intel_pstate_cpu_ids[] = {
 	ICPU(0x46, core_params),
 	ICPU(0x47, core_params),
 	ICPU(0x4c, byt_params),
+	ICPU(0x4e, core_params),
 	ICPU(0x4f, core_params),
 	ICPU(0x56, core_params),
 	{}
@@ -887,7 +925,9 @@ static int intel_pstate_set_policy(struct cpufreq_policy *policy)
 	if (!policy->cpuinfo.max_freq)
 		return -ENODEV;
 
-	if (policy->policy == CPUFREQ_POLICY_PERFORMANCE) {
+	if (policy->policy == CPUFREQ_POLICY_PERFORMANCE &&
+	    policy->max >= policy->cpuinfo.max_freq) {
+		limits.min_policy_pct = 100;
 		limits.min_perf_pct = 100;
 		limits.min_perf = int_tofp(1);
 		limits.max_policy_pct = 100;
@@ -897,8 +937,9 @@ static int intel_pstate_set_policy(struct cpufreq_policy *policy)
 		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_policy_pct = (policy->min * 100) / policy->cpuinfo.max_freq;
+	limits.min_policy_pct = clamp_t(int, limits.min_policy_pct, 0 , 100);
+	limits.min_perf_pct = max(limits.min_policy_pct, limits.min_sysfs_pct);
 	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;
@@ -978,6 +1019,7 @@ static struct cpufreq_driver intel_pstate_driver = {
 
 static int __initdata no_load;
 static int __initdata no_hwp;
+static int __initdata hwp_only;
 static unsigned int force_load;
 
 static int intel_pstate_msrs_not_valid(void)
@@ -1175,6 +1217,9 @@ static int __init intel_pstate_init(void)
 	if (cpu_has(c,X86_FEATURE_HWP) && !no_hwp)
 		intel_pstate_hwp_enable();
 
+	if (!hwp_active && hwp_only)
+		goto out;
+
 	rc = cpufreq_register_driver(&intel_pstate_driver);
 	if (rc)
 		goto out;
@@ -1209,6 +1254,8 @@ static int __init intel_pstate_setup(char *str)
 		no_hwp = 1;
 	if (!strcmp(str, "force"))
 		force_load = 1;
+	if (!strcmp(str, "hwp_only"))
+		hwp_only = 1;
 	return 0;
 }
 early_param("intel_pstate", intel_pstate_setup);

drivers/cpufreq/ls1x-cpufreq.c

@@ -210,7 +210,6 @@ static int ls1x_cpufreq_probe(struct platform_device *pdev)
 static struct platform_driver ls1x_cpufreq_platdrv = {
 	.driver = {
 		.name	= "ls1x-cpufreq",
-		.owner	= THIS_MODULE,
 	},
 	.probe		= ls1x_cpufreq_probe,
 	.remove		= ls1x_cpufreq_remove,

drivers/cpufreq/sfi-cpufreq.c

+/*
+ *  SFI Performance States Driver
+ *
+ *  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; either version 2 of the License.
+ *
+ *  This program is distributed in the hope that it will be useful, but
+ *  WITHOUT ANY WARRANTY; without even the implied warranty of
+ *  MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+ *  General Public License for more details.
+ *
+ *  Author: Vishwesh M Rudramuni <vishwesh.m.rudramuni@intel.com>
+ *  Author: Srinidhi Kasagar <srinidhi.kasagar@intel.com>
+ */
+
+#include <linux/cpufreq.h>
+#include <linux/init.h>
+#include <linux/kernel.h>
+#include <linux/module.h>
+#include <linux/sfi.h>
+#include <linux/slab.h>
+#include <linux/smp.h>
+
+#include <asm/msr.h>
+
+struct cpufreq_frequency_table *freq_table;
+static struct sfi_freq_table_entry *sfi_cpufreq_array;
+static int num_freq_table_entries;
+
+static int sfi_parse_freq(struct sfi_table_header *table)
+{
+	struct sfi_table_simple *sb;
+	struct sfi_freq_table_entry *pentry;
+	int totallen;
+
+	sb = (struct sfi_table_simple *)table;
+	num_freq_table_entries = SFI_GET_NUM_ENTRIES(sb,
+			struct sfi_freq_table_entry);
+	if (num_freq_table_entries <= 1) {
+		pr_err("No p-states discovered\n");
+		return -ENODEV;
+	}
+
+	pentry = (struct sfi_freq_table_entry *)sb->pentry;
+	totallen = num_freq_table_entries * sizeof(*pentry);
+
+	sfi_cpufreq_array = kzalloc(totallen, GFP_KERNEL);
+	if (!sfi_cpufreq_array)
+		return -ENOMEM;
+
+	memcpy(sfi_cpufreq_array, pentry, totallen);
+
+	return 0;
+}
+
+static int sfi_cpufreq_target(struct cpufreq_policy *policy, unsigned int index)
+{
+	unsigned int next_perf_state = 0; /* Index into perf table */
+	u32 lo, hi;
+
+	next_perf_state = policy->freq_table[index].driver_data;
+
+	rdmsr_on_cpu(policy->cpu, MSR_IA32_PERF_CTL, &lo, &hi);
+	lo = (lo & ~INTEL_PERF_CTL_MASK) |
+		((u32) sfi_cpufreq_array[next_perf_state].ctrl_val &
+		INTEL_PERF_CTL_MASK);
+	wrmsr_on_cpu(policy->cpu, MSR_IA32_PERF_CTL, lo, hi);
+
+	return 0;
+}
+
+static int sfi_cpufreq_cpu_init(struct cpufreq_policy *policy)
+{
+	policy->shared_type = CPUFREQ_SHARED_TYPE_HW;
+	policy->cpuinfo.transition_latency = 100000;	/* 100us */
+
+	return cpufreq_table_validate_and_show(policy, freq_table);
+}
+
+static struct cpufreq_driver sfi_cpufreq_driver = {
+	.flags		= CPUFREQ_CONST_LOOPS,
+	.verify		= cpufreq_generic_frequency_table_verify,
+	.target_index	= sfi_cpufreq_target,
+	.init		= sfi_cpufreq_cpu_init,
+	.name		= "sfi-cpufreq",
+	.attr		= cpufreq_generic_attr,
+};
+
+static int __init sfi_cpufreq_init(void)
+{
+	int ret, i;
+
+	/* parse the freq table from SFI */
+	ret = sfi_table_parse(SFI_SIG_FREQ, NULL, NULL, sfi_parse_freq);
+	if (ret)
+		return ret;
+
+	freq_table = kzalloc(sizeof(*freq_table) *
+			(num_freq_table_entries + 1), GFP_KERNEL);
+	if (!freq_table) {
+		ret = -ENOMEM;
+		goto err_free_array;
+	}
+
+	for (i = 0; i < num_freq_table_entries; i++) {
+		freq_table[i].driver_data = i;
+		freq_table[i].frequency = sfi_cpufreq_array[i].freq_mhz * 1000;
+	}
+	freq_table[i].frequency = CPUFREQ_TABLE_END;
+
+	ret = cpufreq_register_driver(&sfi_cpufreq_driver);
+	if (ret)
+		goto err_free_tbl;
+
+	return ret;
+
+err_free_tbl:
+	kfree(freq_table);
+err_free_array:
+	kfree(sfi_cpufreq_array);
+	return ret;
+}
+late_initcall(sfi_cpufreq_init);
+
+static void __exit sfi_cpufreq_exit(void)
+{
+	cpufreq_unregister_driver(&sfi_cpufreq_driver);
+	kfree(freq_table);
+	kfree(sfi_cpufreq_array);
+}
+module_exit(sfi_cpufreq_exit);
+
+MODULE_AUTHOR("Vishwesh M Rudramuni <vishwesh.m.rudramuni@intel.com>");
+MODULE_DESCRIPTION("SFI Performance-States Driver");
+MODULE_LICENSE("GPL");

drivers/sfi/sfi_core.c

@@ -161,7 +161,7 @@ static int sfi_verify_table(struct sfi_table_header *table)
  * Check for common case that we can re-use mapping to SYST,
  * which requires syst_pa, syst_va to be initialized.
  */
-struct sfi_table_header *sfi_map_table(u64 pa)
+static struct sfi_table_header *sfi_map_table(u64 pa)
 {
 	struct sfi_table_header *th;
 	u32 length;
@@ -189,7 +189,7 @@ struct sfi_table_header *sfi_map_table(u64 pa)
  * Undoes effect of sfi_map_table() by unmapping table
  * if it did not completely fit on same page as SYST.
  */
-void sfi_unmap_table(struct sfi_table_header *th)
+static void sfi_unmap_table(struct sfi_table_header *th)
 {
 	if (!TABLE_ON_PAGE(syst_va, th, th->len))
 		sfi_unmap_memory(th, TABLE_ON_PAGE(th, th, th->len) ?

include/linux/cpufreq.h

@@ -66,8 +66,6 @@ struct cpufreq_policy {
 	unsigned int		shared_type; /* ACPI: ANY or ALL affected CPUs
 						should set cpufreq */
 	unsigned int		cpu;    /* cpu nr of CPU managing this policy */
-	unsigned int		last_cpu; /* cpu nr of previous CPU that managed
-					   * this policy */
 	struct clk		*clk;
 	struct cpufreq_cpuinfo	cpuinfo;/* see above */
 
@@ -113,6 +111,9 @@ struct cpufreq_policy {
 	wait_queue_head_t	transition_wait;
 	struct task_struct	*transition_task; /* Task which is doing the transition */
 
+	/* cpufreq-stats */
+	struct cpufreq_stats	*stats;
+
 	/* For cpufreq driver's internal use */
 	void			*driver_data;
 };
@@ -367,9 +368,8 @@ static inline void cpufreq_resume(void) {}
 #define CPUFREQ_INCOMPATIBLE		(1)
 #define CPUFREQ_NOTIFY			(2)
 #define CPUFREQ_START			(3)
-#define CPUFREQ_UPDATE_POLICY_CPU	(4)
-#define CPUFREQ_CREATE_POLICY		(5)
-#define CPUFREQ_REMOVE_POLICY		(6)
+#define CPUFREQ_CREATE_POLICY		(4)
+#define CPUFREQ_REMOVE_POLICY		(5)
 
 #ifdef CONFIG_CPU_FREQ
 int cpufreq_register_notifier(struct notifier_block *nb, unsigned int list);