Merge tag 'pm+acpi-3.19-rc3' of git://git.kernel.org/pub/scm/linux/kernel/git/rafael/linux-pm

Pull power management and ACPI material from Rafael J Wysocki: "These are fixes (operating performance points library, cpufreq-dt driver, cpufreq core, ACPI backlight, cpupower tool), cleanups (cpuidle), new processor IDs for the RAPL (Running Average Power Limit) power capping driver, and a modification of the generic power domains framework allowing modular drivers to call one of its helper functions. Specifics: - Fix for a potential NULL pointer dereference in the cpufreq core due to an initialization race condition (Ethan Zhao). - Fixes for abuse of the OPP (Operating Performance Points) API related to RCU and other minor issues in the OPP library and the cpufreq-dt driver (Dmitry Torokhov). - cpuidle governors cleanup making them measure idle duration in a better way without using the CPUIDLE_FLAG_TIME_INVALID flag which allows that flag to be dropped from the ACPI cpuidle driver and from the core too (Len Brown). - New ACPI backlight blacklist entries for Samsung machines without a working native backlight interface that need to use the ACPI backlight instead (Aaron Lu). - New CPU IDs of future Intel Xeon CPUs for the Intel RAPL power capping driver (Jacob Pan). - Generic power domains framework modification to export the of_genpd_get_from_provider() function to modular drivers that will allow future driver modifications to be based on the mainline (Amit Daniel Kachhap). - Two fixes for the cpupower tool (Michal Privoznik, Prarit Bhargava)" * tag 'pm+acpi-3.19-rc3' of git://git.kernel.org/pub/scm/linux/kernel/git/rafael/linux-pm: ACPI / video: Add some Samsung models to disable_native_backlight list tools / cpupower: Fix no idle state information return value tools / cpupower: Correctly detect if running as root cpufreq: fix a NULL pointer dereference in __cpufreq_governor() cpufreq-dt: defer probing if OPP table is not ready PM / OPP: take RCU lock in dev_pm_opp_get_opp_count PM / OPP: fix warning in of_free_opp_table() PM / OPP: add some lockdep annotations powercap / RAPL: add IDs for future Xeon CPUs PM / Domains: Export of_genpd_get_from_provider function cpuidle / ACPI: remove unused CPUIDLE_FLAG_TIME_INVALID cpuidle: ladder: Better idle duration measurement without using CPUIDLE_FLAG_TIME_INVALID cpuidle: menu: Better idle duration measurement without using CPUIDLE_FLAG_TIME_INVALID

Merge tag 'pm+acpi-3.19-rc3' of git://git.kernel.org/pub/scm/linux/kernel/git/rafael/linux-pm
Pull power management and ACPI material from Rafael J Wysocki: "These are fixes (operating performance points library, cpufreq-dt driver, cpufreq core, ACPI backlight, cpupower tool), cleanups (cpuidle), new processor IDs for the RAPL (Running Average Power Limit) power capping driver, and a modification of the generic power domains framework allowing modular drivers to call one of its helper functions. Specifics: - Fix for a potential NULL pointer dereference in the cpufreq core due to an initialization race condition (Ethan Zhao). - Fixes for abuse of the OPP (Operating Performance Points) API related to RCU and other minor issues in the OPP library and the cpufreq-dt driver (Dmitry Torokhov). - cpuidle governors cleanup making them measure idle duration in a better way without using the CPUIDLE_FLAG_TIME_INVALID flag which allows that flag to be dropped from the ACPI cpuidle driver and from the core too (Len Brown). - New ACPI backlight blacklist entries for Samsung machines without a working native backlight interface that need to use the ACPI backlight instead (Aaron Lu). - New CPU IDs of future Intel Xeon CPUs for the Intel RAPL power capping driver (Jacob Pan). - Generic power domains framework modification to export the of_genpd_get_from_provider() function to modular drivers that will allow future driver modifications to be based on the mainline (Amit Daniel Kachhap). - Two fixes for the cpupower tool (Michal Privoznik, Prarit Bhargava)" * tag 'pm+acpi-3.19-rc3' of git://git.kernel.org/pub/scm/linux/kernel/git/rafael/linux-pm: ACPI / video: Add some Samsung models to disable_native_backlight list tools / cpupower: Fix no idle state information return value tools / cpupower: Correctly detect if running as root cpufreq: fix a NULL pointer dereference in __cpufreq_governor() cpufreq-dt: defer probing if OPP table is not ready PM / OPP: take RCU lock in dev_pm_opp_get_opp_count PM / OPP: fix warning in of_free_opp_table() PM / OPP: add some lockdep annotations powercap / RAPL: add IDs for future Xeon CPUs PM / Domains: Export of_genpd_get_from_provider function cpuidle / ACPI: remove unused CPUIDLE_FLAG_TIME_INVALID cpuidle: ladder: Better idle duration measurement without using CPUIDLE_FLAG_TIME_INVALID cpuidle: menu: Better idle duration measurement without using CPUIDLE_FLAG_TIME_INVALID
Linus Torvalds
2 parents 346eef2fc0 651c3293bf
Showing 13 changed files Side-by-side Diff
drivers/acpi/processor_idle.c
drivers/acpi/video.c
drivers/base/power/domain.c
drivers/base/power/opp.c
drivers/cpufreq/cpufreq-dt.c
drivers/cpufreq/cpufreq.c
drivers/cpuidle/governors/ladder.c
drivers/cpuidle/governors/menu.c
drivers/powercap/intel_rapl.c
include/linux/cpuidle.h
include/linux/pm_domain.h
tools/power/cpupower/utils/cpupower.c
tools/power/cpupower/utils/helpers/sysfs.c
@@ -985,8 +985,6 @@
 		state->flags = 0;
 		switch (cx->type) {
 			case ACPI_STATE_C1:
-			if (cx->entry_method != ACPI_CSTATE_FFH)
-				state->flags |= CPUIDLE_FLAG_TIME_INVALID;
  
 			state->enter = acpi_idle_enter_c1;
 			state->enter_dead = acpi_idle_play_dead;
@@ -505,6 +505,23 @@
 		DMI_MATCH(DMI_PRODUCT_NAME, "HP ENVY 15 Notebook PC"),
 		},
 	},
+
+	{
+	 .callback = video_disable_native_backlight,
+	 .ident = "SAMSUNG 870Z5E/880Z5E/680Z5E",
+	 .matches = {
+		DMI_MATCH(DMI_SYS_VENDOR, "SAMSUNG ELECTRONICS CO., LTD."),
+		DMI_MATCH(DMI_PRODUCT_NAME, "870Z5E/880Z5E/680Z5E"),
+		},
+	},
+	{
+	 .callback = video_disable_native_backlight,
+	 .ident = "SAMSUNG 370R4E/370R4V/370R5E/3570RE/370R5V",
+	 .matches = {
+		DMI_MATCH(DMI_SYS_VENDOR, "SAMSUNG ELECTRONICS CO., LTD."),
+		DMI_MATCH(DMI_PRODUCT_NAME, "370R4E/370R4V/370R5E/3570RE/370R5V"),
+		},
+	},
 	{}
 };
  
@@ -2088,7 +2088,7 @@
  * Returns a valid pointer to struct generic_pm_domain on success or ERR_PTR()
  * on failure.
  */
-static struct generic_pm_domain *of_genpd_get_from_provider(
+struct generic_pm_domain *of_genpd_get_from_provider(
 					struct of_phandle_args *genpdspec)
 {
 	struct generic_pm_domain *genpd = ERR_PTR(-ENOENT);
@@ -2108,6 +2108,7 @@
  
 	return genpd;
 }
+EXPORT_SYMBOL_GPL(of_genpd_get_from_provider);
  
 /**
  * genpd_dev_pm_detach - Detach a device from its PM domain.
@@ -108,6 +108,14 @@
 /* Lock to allow exclusive modification to the device and opp lists */
 static DEFINE_MUTEX(dev_opp_list_lock);
  
+#define opp_rcu_lockdep_assert()					\
+do {									\
+	rcu_lockdep_assert(rcu_read_lock_held() ||			\
+				lockdep_is_held(&dev_opp_list_lock),	\
+			   "Missing rcu_read_lock() or "		\
+			   "dev_opp_list_lock protection");		\
+} while (0)
+
 /**
  * find_device_opp() - find device_opp struct using device pointer
  * @dev:	device pointer used to lookup device OPPs
@@ -208,9 +216,7 @@
  * This function returns the number of available opps if there are any,
  * else returns 0 if none or the corresponding error value.
  *
- * Locking: This function must be called under rcu_read_lock(). This function
- * internally references two RCU protected structures: device_opp and opp which
- * are safe as long as we are under a common RCU locked section.
+ * Locking: This function takes rcu_read_lock().
  */
 int dev_pm_opp_get_opp_count(struct device *dev)
 {
  
@@ -218,11 +224,14 @@
 	struct dev_pm_opp *temp_opp;
 	int count = 0;
  
+	rcu_read_lock();
+
 	dev_opp = find_device_opp(dev);
 	if (IS_ERR(dev_opp)) {
-		int r = PTR_ERR(dev_opp);
-		dev_err(dev, "%s: device OPP not found (%d)\n", __func__, r);
-		return r;
+		count = PTR_ERR(dev_opp);
+		dev_err(dev, "%s: device OPP not found (%d)\n",
+			__func__, count);
+		goto out_unlock;
 	}
  
 	list_for_each_entry_rcu(temp_opp, &dev_opp->opp_list, node) {
@@ -230,6 +239,8 @@
 			count++;
 	}
  
+out_unlock:
+	rcu_read_unlock();
 	return count;
 }
 EXPORT_SYMBOL_GPL(dev_pm_opp_get_opp_count);
@@ -267,6 +278,8 @@
 	struct device_opp *dev_opp;
 	struct dev_pm_opp *temp_opp, *opp = ERR_PTR(-ERANGE);
  
+	opp_rcu_lockdep_assert();
+
 	dev_opp = find_device_opp(dev);
 	if (IS_ERR(dev_opp)) {
 		int r = PTR_ERR(dev_opp);
@@ -313,6 +326,8 @@
 	struct device_opp *dev_opp;
 	struct dev_pm_opp *temp_opp, *opp = ERR_PTR(-ERANGE);
  
+	opp_rcu_lockdep_assert();
+
 	if (!dev || !freq) {
 		dev_err(dev, "%s: Invalid argument freq=%p\n", __func__, freq);
 		return ERR_PTR(-EINVAL);
@@ -361,6 +376,8 @@
 	struct device_opp *dev_opp;
 	struct dev_pm_opp *temp_opp, *opp = ERR_PTR(-ERANGE);
  
+	opp_rcu_lockdep_assert();
+
 	if (!dev || !freq) {
 		dev_err(dev, "%s: Invalid argument freq=%p\n", __func__, freq);
 		return ERR_PTR(-EINVAL);
  
@@ -783,9 +800,15 @@
  
 	/* Check for existing list for 'dev' */
 	dev_opp = find_device_opp(dev);
-	if (WARN(IS_ERR(dev_opp), "%s: dev_opp: %ld\n", dev_name(dev),
-		 PTR_ERR(dev_opp)))
+	if (IS_ERR(dev_opp)) {
+		int error = PTR_ERR(dev_opp);
+		if (error != -ENODEV)
+			WARN(1, "%s: dev_opp: %d\n",
+			     IS_ERR_OR_NULL(dev) ?
+					"Invalid device" : dev_name(dev),
+			     error);
 		return;
+	}
  
 	/* Hold our list modification lock here */
 	mutex_lock(&dev_opp_list_lock);
@@ -211,6 +211,17 @@
 	/* OPPs might be populated at runtime, don't check for error here */
 	of_init_opp_table(cpu_dev);
  
+	/*
+	 * But we need OPP table to function so if it is not there let's
+	 * give platform code chance to provide it for us.
+	 */
+	ret = dev_pm_opp_get_opp_count(cpu_dev);
+	if (ret <= 0) {
+		pr_debug("OPP table is not ready, deferring probe\n");
+		ret = -EPROBE_DEFER;
+		goto out_free_opp;
+	}
+
 	priv = kzalloc(sizeof(*priv), GFP_KERNEL);
 	if (!priv) {
 		ret = -ENOMEM;
@@ -2028,6 +2028,12 @@
 	/* Don't start any governor operations if we are entering suspend */
 	if (cpufreq_suspended)
 		return 0;
+	/*
+	 * Governor might not be initiated here if ACPI _PPC changed
+	 * notification happened, so check it.
+	 */
+	if (!policy->governor)
+		return -EINVAL;
  
 	if (policy->governor->max_transition_latency &&
 	    policy->cpuinfo.transition_latency >
@@ -79,12 +79,7 @@
  
 	last_state = &ldev->states[last_idx];
  
-	if (!(drv->states[last_idx].flags & CPUIDLE_FLAG_TIME_INVALID)) {
-		last_residency = cpuidle_get_last_residency(dev) - \
-					 drv->states[last_idx].exit_latency;
-	}
-	else
-		last_residency = last_state->threshold.promotion_time + 1;
+	last_residency = cpuidle_get_last_residency(dev) - drv->states[last_idx].exit_latency;
  
 	/* consider promotion */
 	if (last_idx < drv->state_count - 1 &&
@@ -396,8 +396,8 @@
 	 * power state and occurrence of the wakeup event.
 	 *
 	 * If the entered idle state didn't support residency measurements,
-	 * we are basically lost in the dark how much time passed.
-	 * As a compromise, assume we slept for the whole expected time.
+	 * we use them anyway if they are short, and if long,
+	 * truncate to the whole expected time.
 	 *
 	 * Any measured amount of time will include the exit latency.
 	 * Since we are interested in when the wakeup begun, not when it
  
  
  
@@ -405,22 +405,17 @@
 	 * the measured amount of time is less than the exit latency,
 	 * assume the state was never reached and the exit latency is 0.
 	 */
-	if (unlikely(target->flags & CPUIDLE_FLAG_TIME_INVALID)) {
-		/* Use timer value as is */
-		measured_us = data->next_timer_us;
  
-	} else {
-		/* Use measured value */
-		measured_us = cpuidle_get_last_residency(dev);
+	/* measured value */
+	measured_us = cpuidle_get_last_residency(dev);
  
-		/* Deduct exit latency */
-		if (measured_us > target->exit_latency)
-			measured_us -= target->exit_latency;
+	/* Deduct exit latency */
+	if (measured_us > target->exit_latency)
+		measured_us -= target->exit_latency;
  
-		/* Make sure our coefficients do not exceed unity */
-		if (measured_us > data->next_timer_us)
-			measured_us = data->next_timer_us;
-	}
+	/* Make sure our coefficients do not exceed unity */
+	if (measured_us > data->next_timer_us)
+		measured_us = data->next_timer_us;
  
 	/* Update our correction ratio */
 	new_factor = data->correction_factor[data->bucket];
@@ -1041,6 +1041,7 @@
 	RAPL_CPU(0x45, rapl_defaults_core),/* Haswell ULT */
 	RAPL_CPU(0x4C, rapl_defaults_atom),/* Braswell */
 	RAPL_CPU(0x4A, rapl_defaults_atom),/* Tangier */
+	RAPL_CPU(0x56, rapl_defaults_core),/* Future Xeon */
 	RAPL_CPU(0x5A, rapl_defaults_atom),/* Annidale */
 	{}
 };
@@ -53,7 +53,6 @@
 };
  
 /* Idle State Flags */
-#define CPUIDLE_FLAG_TIME_INVALID	(0x01) /* is residency time measurable? */
 #define CPUIDLE_FLAG_COUPLED	(0x02) /* state applies to multiple cpus */
 #define CPUIDLE_FLAG_TIMER_STOP (0x04)  /* timer is stopped on this state */
  
@@ -89,8 +88,6 @@
 /**
  * cpuidle_get_last_residency - retrieves the last state's residency time
  * @dev: the target CPU
- *
- * NOTE: this value is invalid if CPUIDLE_FLAG_TIME_INVALID is set
  */
 static inline int cpuidle_get_last_residency(struct cpuidle_device *dev)
 {
@@ -271,6 +271,8 @@
 int __of_genpd_add_provider(struct device_node *np, genpd_xlate_t xlate,
 			void *data);
 void of_genpd_del_provider(struct device_node *np);
+struct generic_pm_domain *of_genpd_get_from_provider(
+			struct of_phandle_args *genpdspec);
  
 struct generic_pm_domain *__of_genpd_xlate_simple(
 					struct of_phandle_args *genpdspec,
@@ -287,6 +289,12 @@
 	return 0;
 }
 static inline void of_genpd_del_provider(struct device_node *np) {}
+
+static inline struct generic_pm_domain *of_genpd_get_from_provider(
+			struct of_phandle_args *genpdspec)
+{
+	return NULL;
+}
  
 #define __of_genpd_xlate_simple		NULL
 #define __of_genpd_xlate_onecell	NULL
@@ -199,7 +199,7 @@
 	}
  
 	get_cpu_info(0, &cpupower_cpu_info);
-	run_as_root = !getuid();
+	run_as_root = !geteuid();
 	if (run_as_root) {
 		ret = uname(&uts);
 		if (!ret && !strcmp(uts.machine, "x86_64") &&
@@ -361,7 +361,7 @@
  
 	snprintf(file, SYSFS_PATH_MAX, PATH_TO_CPU "cpuidle");
 	if (stat(file, &statbuf) != 0 || !S_ISDIR(statbuf.st_mode))
-		return -ENODEV;
+		return 0;
  
 	snprintf(file, SYSFS_PATH_MAX, PATH_TO_CPU "cpu%u/cpuidle/state0", cpu);
 	if (stat(file, &statbuf) != 0 || !S_ISDIR(statbuf.st_mode))
...	...	@@ -985,8 +985,6 @@
985	985	state->flags = 0;
986	986	switch (cx->type) {
987	987	case ACPI_STATE_C1:
988		- if (cx->entry_method != ACPI_CSTATE_FFH)
989		- state->flags \|= CPUIDLE_FLAG_TIME_INVALID;
990	988
991	989	state->enter = acpi_idle_enter_c1;
992	990	state->enter_dead = acpi_idle_play_dead;
...	...	@@ -505,6 +505,23 @@
505	505	DMI_MATCH(DMI_PRODUCT_NAME, "HP ENVY 15 Notebook PC"),
506	506	},
507	507	},
	508	+
	509	+ {
	510	+ .callback = video_disable_native_backlight,
	511	+ .ident = "SAMSUNG 870Z5E/880Z5E/680Z5E",
	512	+ .matches = {
	513	+ DMI_MATCH(DMI_SYS_VENDOR, "SAMSUNG ELECTRONICS CO., LTD."),
	514	+ DMI_MATCH(DMI_PRODUCT_NAME, "870Z5E/880Z5E/680Z5E"),
	515	+ },
	516	+ },
	517	+ {
	518	+ .callback = video_disable_native_backlight,
	519	+ .ident = "SAMSUNG 370R4E/370R4V/370R5E/3570RE/370R5V",
	520	+ .matches = {
	521	+ DMI_MATCH(DMI_SYS_VENDOR, "SAMSUNG ELECTRONICS CO., LTD."),
	522	+ DMI_MATCH(DMI_PRODUCT_NAME, "370R4E/370R4V/370R5E/3570RE/370R5V"),
	523	+ },
	524	+ },
508	525	{}
509	526	};
510	527
...	...	@@ -2088,7 +2088,7 @@
2088	2088	* Returns a valid pointer to struct generic_pm_domain on success or ERR_PTR()
2089	2089	* on failure.
2090	2090	*/
2091		-static struct generic_pm_domain *of_genpd_get_from_provider(
	2091	+struct generic_pm_domain *of_genpd_get_from_provider(
2092	2092	struct of_phandle_args *genpdspec)
2093	2093	{
2094	2094	struct generic_pm_domain *genpd = ERR_PTR(-ENOENT);
...	...	@@ -2108,6 +2108,7 @@
2108	2108
2109	2109	return genpd;
2110	2110	}
	2111	+EXPORT_SYMBOL_GPL(of_genpd_get_from_provider);
2111	2112
2112	2113	/**
2113	2114	* genpd_dev_pm_detach - Detach a device from its PM domain.
...	...	@@ -108,6 +108,14 @@
108	108	/* Lock to allow exclusive modification to the device and opp lists */
109	109	static DEFINE_MUTEX(dev_opp_list_lock);
110	110
	111	+#define opp_rcu_lockdep_assert() \
	112	+do { \
	113	+ rcu_lockdep_assert(rcu_read_lock_held() \|\| \
	114	+ lockdep_is_held(&dev_opp_list_lock), \
	115	+ "Missing rcu_read_lock() or " \
	116	+ "dev_opp_list_lock protection"); \
	117	+} while (0)
	118	+
111	119	/**
112	120	* find_device_opp() - find device_opp struct using device pointer
113	121	* @dev: device pointer used to lookup device OPPs
...	...	@@ -208,9 +216,7 @@
208	216	* This function returns the number of available opps if there are any,
209	217	* else returns 0 if none or the corresponding error value.
210	218	*
211		- * Locking: This function must be called under rcu_read_lock(). This function
212		- * internally references two RCU protected structures: device_opp and opp which
213		- * are safe as long as we are under a common RCU locked section.
	219	+ * Locking: This function takes rcu_read_lock().
214	220	*/
215	221	int dev_pm_opp_get_opp_count(struct device *dev)
216	222	{
217	223
...	...	@@ -218,11 +224,14 @@
218	224	struct dev_pm_opp *temp_opp;
219	225	int count = 0;
220	226
	227	+ rcu_read_lock();
	228	+
221	229	dev_opp = find_device_opp(dev);
222	230	if (IS_ERR(dev_opp)) {
223		- int r = PTR_ERR(dev_opp);
224		- dev_err(dev, "%s: device OPP not found (%d)\n", __func__, r);
225		- return r;
	231	+ count = PTR_ERR(dev_opp);
	232	+ dev_err(dev, "%s: device OPP not found (%d)\n",
	233	+ __func__, count);
	234	+ goto out_unlock;
226	235	}
227	236
228	237	list_for_each_entry_rcu(temp_opp, &dev_opp->opp_list, node) {
...	...	@@ -230,6 +239,8 @@
230	239	count++;
231	240	}
232	241
	242	+out_unlock:
	243	+ rcu_read_unlock();
233	244	return count;
234	245	}
235	246	EXPORT_SYMBOL_GPL(dev_pm_opp_get_opp_count);
...	...	@@ -267,6 +278,8 @@
267	278	struct device_opp *dev_opp;
268	279	struct dev_pm_opp temp_opp, opp = ERR_PTR(-ERANGE);
269	280
	281	+ opp_rcu_lockdep_assert();
	282	+
270	283	dev_opp = find_device_opp(dev);
271	284	if (IS_ERR(dev_opp)) {
272	285	int r = PTR_ERR(dev_opp);
...	...	@@ -313,6 +326,8 @@
313	326	struct device_opp *dev_opp;
314	327	struct dev_pm_opp temp_opp, opp = ERR_PTR(-ERANGE);
315	328
	329	+ opp_rcu_lockdep_assert();
	330	+
316	331	if (!dev \|\| !freq) {
317	332	dev_err(dev, "%s: Invalid argument freq=%p\n", __func__, freq);
318	333	return ERR_PTR(-EINVAL);
...	...	@@ -361,6 +376,8 @@
361	376	struct device_opp *dev_opp;
362	377	struct dev_pm_opp temp_opp, opp = ERR_PTR(-ERANGE);
363	378
	379	+ opp_rcu_lockdep_assert();
	380	+
364	381	if (!dev \|\| !freq) {
365	382	dev_err(dev, "%s: Invalid argument freq=%p\n", __func__, freq);
366	383	return ERR_PTR(-EINVAL);
367	384
...	...	@@ -783,9 +800,15 @@
783	800
784	801	/* Check for existing list for 'dev' */
785	802	dev_opp = find_device_opp(dev);
786		- if (WARN(IS_ERR(dev_opp), "%s: dev_opp: %ld\n", dev_name(dev),
787		- PTR_ERR(dev_opp)))
	803	+ if (IS_ERR(dev_opp)) {
	804	+ int error = PTR_ERR(dev_opp);
	805	+ if (error != -ENODEV)
	806	+ WARN(1, "%s: dev_opp: %d\n",
	807	+ IS_ERR_OR_NULL(dev) ?
	808	+ "Invalid device" : dev_name(dev),
	809	+ error);
788	810	return;
	811	+ }
789	812
790	813	/* Hold our list modification lock here */
791	814	mutex_lock(&dev_opp_list_lock);
...	...	@@ -211,6 +211,17 @@
211	211	/* OPPs might be populated at runtime, don't check for error here */
212	212	of_init_opp_table(cpu_dev);
213	213
	214	+ /*
	215	+ * But we need OPP table to function so if it is not there let's
	216	+ * give platform code chance to provide it for us.
	217	+ */
	218	+ ret = dev_pm_opp_get_opp_count(cpu_dev);
	219	+ if (ret <= 0) {
	220	+ pr_debug("OPP table is not ready, deferring probe\n");
	221	+ ret = -EPROBE_DEFER;
	222	+ goto out_free_opp;
	223	+ }
	224	+
214	225	priv = kzalloc(sizeof(*priv), GFP_KERNEL);
215	226	if (!priv) {
216	227	ret = -ENOMEM;
...	...	@@ -2028,6 +2028,12 @@
2028	2028	/* Don't start any governor operations if we are entering suspend */
2029	2029	if (cpufreq_suspended)
2030	2030	return 0;
	2031	+ /*
	2032	+ * Governor might not be initiated here if ACPI _PPC changed
	2033	+ * notification happened, so check it.
	2034	+ */
	2035	+ if (!policy->governor)
	2036	+ return -EINVAL;
2031	2037
2032	2038	if (policy->governor->max_transition_latency &&
2033	2039	policy->cpuinfo.transition_latency >
...	...	@@ -79,12 +79,7 @@
79	79
80	80	last_state = &ldev->states[last_idx];
81	81
82		- if (!(drv->states[last_idx].flags & CPUIDLE_FLAG_TIME_INVALID)) {
83		- last_residency = cpuidle_get_last_residency(dev) - \
84		- drv->states[last_idx].exit_latency;
85		- }
86		- else
87		- last_residency = last_state->threshold.promotion_time + 1;
	82	+ last_residency = cpuidle_get_last_residency(dev) - drv->states[last_idx].exit_latency;
88	83
89	84	/* consider promotion */
90	85	if (last_idx < drv->state_count - 1 &&
...	...	@@ -396,8 +396,8 @@
396	396	* power state and occurrence of the wakeup event.
397	397	*
398	398	* If the entered idle state didn't support residency measurements,
399		- * we are basically lost in the dark how much time passed.
400		- * As a compromise, assume we slept for the whole expected time.
	399	+ * we use them anyway if they are short, and if long,
	400	+ * truncate to the whole expected time.
401	401	*
402	402	* Any measured amount of time will include the exit latency.
403	403	* Since we are interested in when the wakeup begun, not when it
404	404
405	405
406	406
...	...	@@ -405,22 +405,17 @@
405	405	* the measured amount of time is less than the exit latency,
406	406	* assume the state was never reached and the exit latency is 0.
407	407	*/
408		- if (unlikely(target->flags & CPUIDLE_FLAG_TIME_INVALID)) {
409		- /* Use timer value as is */
410		- measured_us = data->next_timer_us;
411	408
412		- } else {
413		- /* Use measured value */
414		- measured_us = cpuidle_get_last_residency(dev);
	409	+ /* measured value */
	410	+ measured_us = cpuidle_get_last_residency(dev);
415	411
416		- /* Deduct exit latency */
417		- if (measured_us > target->exit_latency)
418		- measured_us -= target->exit_latency;
	412	+ /* Deduct exit latency */
	413	+ if (measured_us > target->exit_latency)
	414	+ measured_us -= target->exit_latency;
419	415
420		- /* Make sure our coefficients do not exceed unity */
421		- if (measured_us > data->next_timer_us)
422		- measured_us = data->next_timer_us;
423		- }
	416	+ /* Make sure our coefficients do not exceed unity */
	417	+ if (measured_us > data->next_timer_us)
	418	+ measured_us = data->next_timer_us;
424	419
425	420	/* Update our correction ratio */
426	421	new_factor = data->correction_factor[data->bucket];
...	...	@@ -1041,6 +1041,7 @@
1041	1041	RAPL_CPU(0x45, rapl_defaults_core),/* Haswell ULT */
1042	1042	RAPL_CPU(0x4C, rapl_defaults_atom),/* Braswell */
1043	1043	RAPL_CPU(0x4A, rapl_defaults_atom),/* Tangier */
	1044	+ RAPL_CPU(0x56, rapl_defaults_core),/* Future Xeon */
1044	1045	RAPL_CPU(0x5A, rapl_defaults_atom),/* Annidale */
1045	1046	{}
1046	1047	};
...	...	@@ -53,7 +53,6 @@
53	53	};
54	54
55	55	/* Idle State Flags */
56		-#define CPUIDLE_FLAG_TIME_INVALID (0x01) /* is residency time measurable? */
57	56	#define CPUIDLE_FLAG_COUPLED (0x02) /* state applies to multiple cpus */
58	57	#define CPUIDLE_FLAG_TIMER_STOP (0x04) /* timer is stopped on this state */
59	58
...	...	@@ -89,8 +88,6 @@
89	88	/**
90	89	* cpuidle_get_last_residency - retrieves the last state's residency time
91	90	* @dev: the target CPU
92		- *
93		- * NOTE: this value is invalid if CPUIDLE_FLAG_TIME_INVALID is set
94	91	*/
95	92	static inline int cpuidle_get_last_residency(struct cpuidle_device *dev)
96	93	{