Skip to contentSkip to conversation

Make IPA use PM_EM

PATCH Superseded Quentin Perret 2 participants 6 messages Aug 1, 2019 – Aug 7, 2019 Thermal

v6 of 7 Superseded Latest is v7 →

lore
Share & tools
lore
#0 lore raw

v6 cover Make IPA use PM_EM

Changes in v6
*************
- Added Daniel's and Viresh's Acked-by to all patches

Changes in v5:
**************
- Changed patch 02 to guard IPA-specific code in cpu_cooling.c with appropriate ifdefery (Daniel)
- Rebased on 5.2-rc2

Changes in v4:
**************
- Added Viresh's Acked-by to all 3 patches
- Improved commit message of patch 3/3 to explain how it has no functional impact on existing users (Eduardo)

Changes in v3:
**************
- Changed warning message for unordered tables to something more explicit (Viresh)
- Changed WARN() into a pr_err() for consistency

Changes in v2:
**************
- Fixed patch 01/03 to actually enable CONFIG_ENERGY_MODEL
- Added "depends on ENERGY_MODEL" to IPA (Daniel)
- Added check to bail out if the freq table is unsorted (Viresh)

Cover letter:
*************

The Intelligent Power Allocator (IPA) thermal governor uses an Energy Model (or EM) of the CPUs to re-distribute the power budget. To do so, it builds a table of <frequency, power> tuples where the power values are computed using the 'dynamic-power-coefficient' DT property. All of this is done in and only for the thermal subsystem, and more specifically for CPUs -- the power of other types of devices is obtained differently.

Recently, the CPU scheduler has seen the introduction of Energy Aware Scheduling (EAS) patches, which also rely on an EM of the CPUs. This EM, however, is managed by an independent framework, called PM_EM, aimed to be used by all kernel subsystems interested in the power consumed by CPUs, and not only the scheduler.

This patch series follows this logic and removes the (now redundant) thermal-specific EM computation code to migrate IPA to use PM_EM instead.

Doing so should have no visible functional impact for existing users of IPA since:

- during the 5.1 development cycle, a series of patches [1] introduced in PM_OPP some infrastructure (dev_pm_opp_of_register_em()) enabling the registration of EMs in PM_EM using the DT property used by IPA;

- the existing upstream cpufreq drivers marked with the 'CPUFREQ_IS_COOLING_DEV' flag all call dev_pm_opp_of_register_em(), which means they all support PM_EM (the only two exceptions are qoriq-cpufreq which doesn't in fact use an EM and scmi-cpufreq which already supports PM_EM without using the PM_OPP infrastructurei because it read power costs directly from firmware);

So, migrating IPA to using PM_EM should effectively be just plumbing since for the existing IPA users the PM_EM tables will contain the exact same power values that IPA used to compute on its own until now. The only new dependency is to compile in CONFIG_ENERGY_MODEL.

Why is this migration still a good thing ? For three main reasons.

1. it removes redundant code;

2. it introduces an abstraction layer between IPA and the EM computation. PM_EM offers to EAS and IPA (and potentially other clients) standardized EM tables and hides 'how' these tables have been obtained. PM_EM as of now supports power values either coming from the 'dynamic-power-coefficient' DT property or obtained directly from firmware using SCMI. The latter is a new feature for IPA and that comes 'for free' with the migration. This will also be true in the future every time PM_EM gets support for other ways of loading the EM. Moreover, PM_EM is documented and has a debugfs interface which should help adding support for new platforms.

3. it builds a consistent view of the EM of CPUs across kernel subsystems, which is a pre-requisite for any kind of future work aiming at a smarter power allocation using scheduler knowledge about the system for example.

[1] https://lore.kernel.org/lkml/[email protected]/

Quentin Perret (3):
arm64: defconfig: Enable CONFIG_ENERGY_MODEL
thermal: cpu_cooling: Make the power-related code depend on IPA
thermal: cpu_cooling: Migrate to using the EM framework

 arch/arm64/configs/defconfig  |   1 +
 drivers/thermal/Kconfig       |   1 +
 drivers/thermal/cpu_cooling.c | 428 ++++++++++++++--------------------
 3 files changed, 178 insertions(+), 252 deletions(-)
Signature · 1 line
2.22.0
#1 lore raw

PATCH v6 1/3 arm64: defconfig: Enable CONFIG_ENERGY_MODEL

The recently introduced Energy Model (EM) framework manages power cost tables for the CPUs of the system. Its only user right now is the scheduler, in the context of Energy Aware Scheduling (EAS).

However, the EM framework also offers a generic infrastructure that could replace subsystem-specific implementations of the same concepts, as this is the case in the thermal framework.

So, in order to prepare the migration of the thermal subsystem to use the EM framework, enable it in the default arm64 defconfig, which is the most commonly used architecture for IPA. This will also compile-in all of the EAS code, although it won't be enabled by default -- EAS requires to use the 'schedutil' CPUFreq governor while arm64 defaults to 'performance'.

---
 arch/arm64/configs/defconfig | 1 +
 1 file changed, 1 insertion(+)
arch/arm64/configs/defconfig +1 -0
index 0e58ef02880c..ad0e4944a71f 100644
@@ -71,6 +71,7 @@ CONFIG_COMPAT=y
 CONFIG_RANDOMIZE_BASE=y
 CONFIG_HIBERNATION=y
 CONFIG_WQ_POWER_EFFICIENT_DEFAULT=y
+CONFIG_ENERGY_MODEL=y
 CONFIG_ARM_CPUIDLE=y
 CONFIG_CPU_FREQ=y
 CONFIG_CPU_FREQ_STAT=y
Signature · 1 line
2.22.0
#2 lore raw

PATCH v6 2/3 thermal: cpu_cooling: Make the power-related code depend on IPA

In reply to Quentin Perret

The core CPU cooling infrastructure has power-related functions that have only one client: IPA. Since there can be no user of those functions if IPA is not compiled in, make sure to guard them with checks on CONFIG_THERMAL_GOV_POWER_ALLOCATOR to not waste space unnecessarily.

---
 drivers/thermal/cpu_cooling.c | 214 +++++++++++++++++-----------------
 1 file changed, 104 insertions(+), 110 deletions(-)
drivers/thermal/cpu_cooling.c +104 -110
index 4c5db59a619b..498f59ab64b2 100644
@@ -46,7 +46,9 @@
  */
 struct freq_table {
 	u32 frequency;
+#ifdef CONFIG_THERMAL_GOV_POWER_ALLOCATOR
 	u32 power;
+#endif
 };
 
 /**
@@ -96,28 +98,6 @@ static DEFINE_IDA(cpufreq_ida);
 static DEFINE_MUTEX(cooling_list_lock);
 static LIST_HEAD(cpufreq_cdev_list);
 
-/* Below code defines functions to be used for cpufreq as cooling device */
-
-/**
- * get_level: Find the level for a particular frequency
- * @cpufreq_cdev: cpufreq_cdev for which the property is required
- * @freq: Frequency
- *
- * Return: level corresponding to the frequency.
- */
-static unsigned long get_level(struct cpufreq_cooling_device *cpufreq_cdev,
-			       unsigned int freq)
-{
-	struct freq_table *freq_table = cpufreq_cdev->freq_table;
-	unsigned long level;
-
-	for (level = 1; level <= cpufreq_cdev->max_level; level++)
-		if (freq > freq_table[level].frequency)
-			break;
-
-	return level - 1;
-}
-
 /**
  * cpufreq_thermal_notifier - notifier callback for cpufreq policy change.
  * @nb:	struct notifier_block * with callback info.
@@ -171,6 +151,27 @@ static int cpufreq_thermal_notifier(struct notifier_block *nb,
 	return NOTIFY_OK;
 }
 
+#ifdef CONFIG_THERMAL_GOV_POWER_ALLOCATOR
+/**
+ * get_level: Find the level for a particular frequency
+ * @cpufreq_cdev: cpufreq_cdev for which the property is required
+ * @freq: Frequency
+ *
+ * Return: level corresponding to the frequency.
+ */
+static unsigned long get_level(struct cpufreq_cooling_device *cpufreq_cdev,
+			       unsigned int freq)
+{
+	struct freq_table *freq_table = cpufreq_cdev->freq_table;
+	unsigned long level;
+
+	for (level = 1; level <= cpufreq_cdev->max_level; level++)
+		if (freq > freq_table[level].frequency)
+			break;
+
+	return level - 1;
+}
+
 /**
  * update_freq_table() - Update the freq table with power numbers
  * @cpufreq_cdev:	the cpufreq cooling device in which to update the table
@@ -319,80 +320,6 @@ static u32 get_dynamic_power(struct cpufreq_cooling_device *cpufreq_cdev,
 	return (raw_cpu_power * cpufreq_cdev->last_load) / 100;
 }
 
-/* cpufreq cooling device callback functions are defined below */
-
-/**
- * cpufreq_get_max_state - callback function to get the max cooling state.
- * @cdev: thermal cooling device pointer.
- * @state: fill this variable with the max cooling state.
- *
- * Callback for the thermal cooling device to return the cpufreq
- * max cooling state.
- *
- * Return: 0 on success, an error code otherwise.
- */
-static int cpufreq_get_max_state(struct thermal_cooling_device *cdev,
-				 unsigned long *state)
-{
-	struct cpufreq_cooling_device *cpufreq_cdev = cdev->devdata;
-
-	*state = cpufreq_cdev->max_level;
-	return 0;
-}
-
-/**
- * cpufreq_get_cur_state - callback function to get the current cooling state.
- * @cdev: thermal cooling device pointer.
- * @state: fill this variable with the current cooling state.
- *
- * Callback for the thermal cooling device to return the cpufreq
- * current cooling state.
- *
- * Return: 0 on success, an error code otherwise.
- */
-static int cpufreq_get_cur_state(struct thermal_cooling_device *cdev,
-				 unsigned long *state)
-{
-	struct cpufreq_cooling_device *cpufreq_cdev = cdev->devdata;
-
-	*state = cpufreq_cdev->cpufreq_state;
-
-	return 0;
-}
-
-/**
- * cpufreq_set_cur_state - callback function to set the current cooling state.
- * @cdev: thermal cooling device pointer.
- * @state: set this variable to the current cooling state.
- *
- * Callback for the thermal cooling device to change the cpufreq
- * current cooling state.
- *
- * Return: 0 on success, an error code otherwise.
- */
-static int cpufreq_set_cur_state(struct thermal_cooling_device *cdev,
-				 unsigned long state)
-{
-	struct cpufreq_cooling_device *cpufreq_cdev = cdev->devdata;
-	unsigned int clip_freq;
-
-	/* Request state should be less than max_level */
-	if (WARN_ON(state > cpufreq_cdev->max_level))
-		return -EINVAL;
-
-	/* Check if the old cooling action is same as new cooling action */
-	if (cpufreq_cdev->cpufreq_state == state)
-		return 0;
-
-	clip_freq = cpufreq_cdev->freq_table[state].frequency;
-	cpufreq_cdev->cpufreq_state = state;
-	cpufreq_cdev->clipped_freq = clip_freq;
-
-	cpufreq_update_policy(cpufreq_cdev->policy->cpu);
-
-	return 0;
-}
-
 /**
  * cpufreq_get_requested_power() - get the current power
  * @cdev:	&thermal_cooling_device pointer
@@ -536,22 +463,88 @@ static int cpufreq_power2state(struct thermal_cooling_device *cdev,
 				      power);
 	return 0;
 }
+#endif /* CONFIG_THERMAL_GOV_POWER_ALLOCATOR */
+
+/* cpufreq cooling device callback functions are defined below */
+
+/**
+ * cpufreq_get_max_state - callback function to get the max cooling state.
+ * @cdev: thermal cooling device pointer.
+ * @state: fill this variable with the max cooling state.
+ *
+ * Callback for the thermal cooling device to return the cpufreq
+ * max cooling state.
+ *
+ * Return: 0 on success, an error code otherwise.
+ */
+static int cpufreq_get_max_state(struct thermal_cooling_device *cdev,
+				 unsigned long *state)
+{
+	struct cpufreq_cooling_device *cpufreq_cdev = cdev->devdata;
+
+	*state = cpufreq_cdev->max_level;
+	return 0;
+}
+
+/**
+ * cpufreq_get_cur_state - callback function to get the current cooling state.
+ * @cdev: thermal cooling device pointer.
+ * @state: fill this variable with the current cooling state.
+ *
+ * Callback for the thermal cooling device to return the cpufreq
+ * current cooling state.
+ *
+ * Return: 0 on success, an error code otherwise.
+ */
+static int cpufreq_get_cur_state(struct thermal_cooling_device *cdev,
+				 unsigned long *state)
+{
+	struct cpufreq_cooling_device *cpufreq_cdev = cdev->devdata;
+
+	*state = cpufreq_cdev->cpufreq_state;
+
+	return 0;
+}
+
+/**
+ * cpufreq_set_cur_state - callback function to set the current cooling state.
+ * @cdev: thermal cooling device pointer.
+ * @state: set this variable to the current cooling state.
+ *
+ * Callback for the thermal cooling device to change the cpufreq
+ * current cooling state.
+ *
+ * Return: 0 on success, an error code otherwise.
+ */
+static int cpufreq_set_cur_state(struct thermal_cooling_device *cdev,
+				 unsigned long state)
+{
+	struct cpufreq_cooling_device *cpufreq_cdev = cdev->devdata;
+	unsigned int clip_freq;
+
+	/* Request state should be less than max_level */
+	if (WARN_ON(state > cpufreq_cdev->max_level))
+		return -EINVAL;
+
+	/* Check if the old cooling action is same as new cooling action */
+	if (cpufreq_cdev->cpufreq_state == state)
+		return 0;
+
+	clip_freq = cpufreq_cdev->freq_table[state].frequency;
+	cpufreq_cdev->cpufreq_state = state;
+	cpufreq_cdev->clipped_freq = clip_freq;
+
+	cpufreq_update_policy(cpufreq_cdev->policy->cpu);
+
+	return 0;
+}
 
 /* Bind cpufreq callbacks to thermal cooling device ops */
 
 static struct thermal_cooling_device_ops cpufreq_cooling_ops = {
-	.get_max_state = cpufreq_get_max_state,
-	.get_cur_state = cpufreq_get_cur_state,
-	.set_cur_state = cpufreq_set_cur_state,
-};
-
-static struct thermal_cooling_device_ops cpufreq_power_cooling_ops = {
 	.get_max_state		= cpufreq_get_max_state,
 	.get_cur_state		= cpufreq_get_cur_state,
 	.set_cur_state		= cpufreq_set_cur_state,
-	.get_requested_power	= cpufreq_get_requested_power,
-	.state2power		= cpufreq_state2power,
-	.power2state		= cpufreq_power2state,
 };
 
 /* Notifier for cpufreq policy change */
@@ -659,18 +652,19 @@ __cpufreq_cooling_register(struct device_node *np,
 			pr_debug("%s: freq:%u KHz\n", __func__, freq);
 	}
 
+	cooling_ops = &cpufreq_cooling_ops;
+#ifdef CONFIG_THERMAL_GOV_POWER_ALLOCATOR
 	if (capacitance) {
 		ret = update_freq_table(cpufreq_cdev, capacitance);
 		if (ret) {
 			cdev = ERR_PTR(ret);
 			goto remove_ida;
 		}
-
-		cooling_ops = &cpufreq_power_cooling_ops;
-	} else {
-		cooling_ops = &cpufreq_cooling_ops;
+		cooling_ops->get_requested_power = cpufreq_get_requested_power;
+		cooling_ops->state2power = cpufreq_state2power;
+		cooling_ops->power2state = cpufreq_power2state;
 	}
-
+#endif
 	cdev = thermal_of_cooling_device_register(np, dev_name, cpufreq_cdev,
 						  cooling_ops);
 	if (IS_ERR(cdev))
Signature · 1 line
2.22.0
#3 lore raw

PATCH v6 3/3 thermal: cpu_cooling: Migrate to using the EM framework

In reply to Quentin Perret

The newly introduced Energy Model framework manages power cost tables in a generic way. Moreover, it supports several types of models since the tables can come from DT or firmware (through SCMI) for example. On the other hand, the cpu_cooling subsystem manages its own power cost tables using only DT data.

In order to avoid the duplication of data in the kernel, and in order to enable IPA with EMs coming from more than just DT, remove the private tables from cpu_cooling.c and migrate it to using the centralized EM framework. Doing so should have no visible functional impact for existing users of IPA since:

- recent extenstions to the the PM_OPP infrastructure enable the registration of EMs in PM_EM using the DT property used by IPA;

- the existing upstream cpufreq drivers marked with the 'CPUFREQ_IS_COOLING_DEV' flag all use the aforementioned PM_OPP infrastructure, which means they all support PM_EM. The only two exceptions are qoriq-cpufreq which doesn't in fact use an EM and scmi-cpufreq which doesn't use DT for power costs.

For existing users of cpu_cooling, PM_EM tables will contain the exact same power values that IPA used to compute on its own until now. The only new dependency for them is to compile in CONFIG_ENERGY_MODEL.

The case where the thermal subsystem is used without an Energy Model (cpufreq_cooling_ops) is handled by looking directly at CPUFreq's frequency table which is already a dependency for cpu_cooling.c anyway. Since the thermal framework expects the cooling states in a particular order, bail out whenever the CPUFreq table is unsorted, since that is fairly uncommon in general, and there are currently no users of cpu_cooling for this use-case.

---
 drivers/thermal/Kconfig       |   1 +
 drivers/thermal/cpu_cooling.c | 250 ++++++++++++----------------------
 2 files changed, 91 insertions(+), 160 deletions(-)
drivers/thermal/Kconfig +1 -0
index 9966364a6deb..340853a3ca48 100644
@@ -144,6 +144,7 @@ config THERMAL_GOV_USER_SPACE
 
 config THERMAL_GOV_POWER_ALLOCATOR
 	bool "Power allocator thermal governor"
+	depends on ENERGY_MODEL
 	help
 	  Enable this to manage platform thermals by dynamically
 	  allocating and limiting power to devices.
drivers/thermal/cpu_cooling.c +90 -160

406 lines not on this page · raw message

Signature · 1 line
2.22.0
#4 lore raw

Re: [PATCH v6 3/3] thermal: cpu_cooling: Migrate to using the EM framework

Hi all,

drivers/thermal/Kconfig config THERMAL_GOV_USER_SPACE · line 150

Quoted 12 lines · Quentin Perret
drivers/thermal/Kconfig +1 -0
index 9966364a6deb..340853a3ca48 100644
@@ -144,6 +144,7 @@ config THERMAL_GOV_USER_SPACE
 
 config THERMAL_GOV_POWER_ALLOCATOR
 	bool "Power allocator thermal governor"
+	depends on ENERGY_MODEL
 	help
 	  Enable this to manage platform thermals by dynamically
 	  allocating and limiting power to devices.

FYI, the kbuild bot just reported a randconfig build issue with this. THERMAL_DEFAULT_GOV_POWER_ALLOCATOR 'select' THERMAL_GOV_POWER_ALLOCATOR unconditionally. And I just learned 'select' will force the option ON and totally ignore its dependencies. That is, we can end up with IPA force-compiled in, and no PM_EM, which is broken.

So I guess the simplest fix is to do 'select ENERGY_MODEL' in this patch, instead of depending on it.

I'll send a v7 with this fixed shortly.

Thanks,
Quentin

#5 lore raw
In reply to Quentin Perret

Hi Rui,

is it possible to merge these patches? They are acked-by since May.

Thanks

-- Daniel

Quoted 101 lines · Quentin Perret

Changes in v6
*************
- Added Daniel's and Viresh's Acked-by to all patches

Changes in v5:
**************
- Changed patch 02 to guard IPA-specific code in cpu_cooling.c with appropriate ifdefery (Daniel)
- Rebased on 5.2-rc2

Changes in v4:
**************
- Added Viresh's Acked-by to all 3 patches
- Improved commit message of patch 3/3 to explain how it has no functional impact on existing users (Eduardo)

Changes in v3:
**************
- Changed warning message for unordered tables to something more explicit (Viresh)
- Changed WARN() into a pr_err() for consistency

Changes in v2:
**************
- Fixed patch 01/03 to actually enable CONFIG_ENERGY_MODEL
- Added "depends on ENERGY_MODEL" to IPA (Daniel)
- Added check to bail out if the freq table is unsorted (Viresh)

Cover letter:
*************

The Intelligent Power Allocator (IPA) thermal governor uses an Energy Model (or EM) of the CPUs to re-distribute the power budget. To do so, it builds a table of <frequency, power> tuples where the power values are computed using the 'dynamic-power-coefficient' DT property. All of this is done in and only for the thermal subsystem, and more specifically for CPUs -- the power of other types of devices is obtained differently.

Recently, the CPU scheduler has seen the introduction of Energy Aware Scheduling (EAS) patches, which also rely on an EM of the CPUs. This EM, however, is managed by an independent framework, called PM_EM, aimed to be used by all kernel subsystems interested in the power consumed by CPUs, and not only the scheduler.

This patch series follows this logic and removes the (now redundant) thermal-specific EM computation code to migrate IPA to use PM_EM instead.

Doing so should have no visible functional impact for existing users of IPA since:

- during the 5.1 development cycle, a series of patches [1] introduced in PM_OPP some infrastructure (dev_pm_opp_of_register_em()) enabling the registration of EMs in PM_EM using the DT property used by IPA;

- the existing upstream cpufreq drivers marked with the 'CPUFREQ_IS_COOLING_DEV' flag all call dev_pm_opp_of_register_em(), which means they all support PM_EM (the only two exceptions are qoriq-cpufreq which doesn't in fact use an EM and scmi-cpufreq which already supports PM_EM without using the PM_OPP infrastructurei because it read power costs directly from firmware);

So, migrating IPA to using PM_EM should effectively be just plumbing since for the existing IPA users the PM_EM tables will contain the exact same power values that IPA used to compute on its own until now. The only new dependency is to compile in CONFIG_ENERGY_MODEL.

Why is this migration still a good thing ? For three main reasons.

1. it removes redundant code;

2. it introduces an abstraction layer between IPA and the EM computation. PM_EM offers to EAS and IPA (and potentially other clients) standardized EM tables and hides 'how' these tables have been obtained. PM_EM as of now supports power values either coming from the 'dynamic-power-coefficient' DT property or obtained directly from firmware using SCMI. The latter is a new feature for IPA and that comes 'for free' with the migration. This will also be true in the future every time PM_EM gets support for other ways of loading the EM. Moreover, PM_EM is documented and has a debugfs interface which should help adding support for new platforms.

3. it builds a consistent view of the EM of CPUs across kernel subsystems, which is a pre-requisite for any kind of future work aiming at a smarter power allocation using scheduler knowledge about the system for example.

[1] https://lore.kernel.org/lkml/[email protected]/

Quentin Perret (3):
arm64: defconfig: Enable CONFIG_ENERGY_MODEL
thermal: cpu_cooling: Make the power-related code depend on IPA
thermal: cpu_cooling: Migrate to using the EM framework

 arch/arm64/configs/defconfig  |   1 +
 drivers/thermal/Kconfig       |   1 +
 drivers/thermal/cpu_cooling.c | 428 ++++++++++++++--------------------
 3 files changed, 178 insertions(+), 252 deletions(-)
Signature · 5 lines
 <http://www.linaro.org/> Linaro.org │ Open source software for ARM SoCs

Follow Linaro:  <http://www.facebook.com/pages/Linaro> Facebook |
<http://twitter.com/#!/linaroorg> Twitter |
<http://www.linaro.org/linaro-blog/> Blog