2019-05-29 17:57:24 -06:00
|
|
|
// SPDX-License-Identifier: GPL-2.0-only
|
PowerCap: Introduce Intel RAPL power capping driver
The Intel Running Average Power Limit (RAPL) technology provides platform
software with the ability to monitor, control, and get notifications on
power usage.
This feature is present in all Sandy Bridge and later Intel processors.
Newer models allow more fine grained controls to be applied. In RAPL,
power control is divided into domains, which include package, DRAM
controller, CPU core (Power Plane 0), graphics uncore (power plane 1), etc.
The purpose of this driver is to expose the RAPL settings to userspace.
Overall, RAPL fits in the new powercap class driver in that platform
level power capping controls are exposed via this generic interface.
This driver is based on an earlier patch from Zhang Rui.
However, while the previous work was mainly focused on thermal monitoring
the focus here is on the usability from user space perspective.
References: https://lkml.org/lkml/2011/5/26/93
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Signed-off-by: Jacob Pan <jacob.jun.pan@linux.intel.com>
Reviewed-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
2013-10-17 11:28:35 -06:00
|
|
|
/*
|
2019-07-10 07:44:30 -06:00
|
|
|
* Common code for Intel Running Average Power Limit (RAPL) support.
|
|
|
|
* Copyright (c) 2019, Intel Corporation.
|
PowerCap: Introduce Intel RAPL power capping driver
The Intel Running Average Power Limit (RAPL) technology provides platform
software with the ability to monitor, control, and get notifications on
power usage.
This feature is present in all Sandy Bridge and later Intel processors.
Newer models allow more fine grained controls to be applied. In RAPL,
power control is divided into domains, which include package, DRAM
controller, CPU core (Power Plane 0), graphics uncore (power plane 1), etc.
The purpose of this driver is to expose the RAPL settings to userspace.
Overall, RAPL fits in the new powercap class driver in that platform
level power capping controls are exposed via this generic interface.
This driver is based on an earlier patch from Zhang Rui.
However, while the previous work was mainly focused on thermal monitoring
the focus here is on the usability from user space perspective.
References: https://lkml.org/lkml/2011/5/26/93
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Signed-off-by: Jacob Pan <jacob.jun.pan@linux.intel.com>
Reviewed-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
2013-10-17 11:28:35 -06:00
|
|
|
*/
|
|
|
|
#define pr_fmt(fmt) KBUILD_MODNAME ": " fmt
|
|
|
|
|
|
|
|
#include <linux/kernel.h>
|
|
|
|
#include <linux/module.h>
|
|
|
|
#include <linux/list.h>
|
|
|
|
#include <linux/types.h>
|
|
|
|
#include <linux/device.h>
|
|
|
|
#include <linux/slab.h>
|
|
|
|
#include <linux/log2.h>
|
|
|
|
#include <linux/bitmap.h>
|
|
|
|
#include <linux/delay.h>
|
|
|
|
#include <linux/sysfs.h>
|
|
|
|
#include <linux/cpu.h>
|
|
|
|
#include <linux/powercap.h>
|
2018-01-09 17:38:23 -07:00
|
|
|
#include <linux/suspend.h>
|
2019-07-10 07:44:24 -06:00
|
|
|
#include <linux/intel_rapl.h>
|
2019-07-10 07:44:30 -06:00
|
|
|
#include <linux/processor.h>
|
2019-07-10 07:44:34 -06:00
|
|
|
#include <linux/platform_device.h>
|
|
|
|
|
|
|
|
#include <asm/iosf_mbi.h>
|
PowerCap: Introduce Intel RAPL power capping driver
The Intel Running Average Power Limit (RAPL) technology provides platform
software with the ability to monitor, control, and get notifications on
power usage.
This feature is present in all Sandy Bridge and later Intel processors.
Newer models allow more fine grained controls to be applied. In RAPL,
power control is divided into domains, which include package, DRAM
controller, CPU core (Power Plane 0), graphics uncore (power plane 1), etc.
The purpose of this driver is to expose the RAPL settings to userspace.
Overall, RAPL fits in the new powercap class driver in that platform
level power capping controls are exposed via this generic interface.
This driver is based on an earlier patch from Zhang Rui.
However, while the previous work was mainly focused on thermal monitoring
the focus here is on the usability from user space perspective.
References: https://lkml.org/lkml/2011/5/26/93
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Signed-off-by: Jacob Pan <jacob.jun.pan@linux.intel.com>
Reviewed-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
2013-10-17 11:28:35 -06:00
|
|
|
#include <asm/cpu_device_id.h>
|
2016-06-02 18:19:36 -06:00
|
|
|
#include <asm/intel-family.h>
|
PowerCap: Introduce Intel RAPL power capping driver
The Intel Running Average Power Limit (RAPL) technology provides platform
software with the ability to monitor, control, and get notifications on
power usage.
This feature is present in all Sandy Bridge and later Intel processors.
Newer models allow more fine grained controls to be applied. In RAPL,
power control is divided into domains, which include package, DRAM
controller, CPU core (Power Plane 0), graphics uncore (power plane 1), etc.
The purpose of this driver is to expose the RAPL settings to userspace.
Overall, RAPL fits in the new powercap class driver in that platform
level power capping controls are exposed via this generic interface.
This driver is based on an earlier patch from Zhang Rui.
However, while the previous work was mainly focused on thermal monitoring
the focus here is on the usability from user space perspective.
References: https://lkml.org/lkml/2011/5/26/93
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Signed-off-by: Jacob Pan <jacob.jun.pan@linux.intel.com>
Reviewed-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
2013-10-17 11:28:35 -06:00
|
|
|
|
2016-04-17 16:03:01 -06:00
|
|
|
/* Local defines */
|
|
|
|
#define MSR_PLATFORM_POWER_LIMIT 0x0000065C
|
|
|
|
|
PowerCap: Introduce Intel RAPL power capping driver
The Intel Running Average Power Limit (RAPL) technology provides platform
software with the ability to monitor, control, and get notifications on
power usage.
This feature is present in all Sandy Bridge and later Intel processors.
Newer models allow more fine grained controls to be applied. In RAPL,
power control is divided into domains, which include package, DRAM
controller, CPU core (Power Plane 0), graphics uncore (power plane 1), etc.
The purpose of this driver is to expose the RAPL settings to userspace.
Overall, RAPL fits in the new powercap class driver in that platform
level power capping controls are exposed via this generic interface.
This driver is based on an earlier patch from Zhang Rui.
However, while the previous work was mainly focused on thermal monitoring
the focus here is on the usability from user space perspective.
References: https://lkml.org/lkml/2011/5/26/93
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Signed-off-by: Jacob Pan <jacob.jun.pan@linux.intel.com>
Reviewed-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
2013-10-17 11:28:35 -06:00
|
|
|
/* bitmasks for RAPL MSRs, used by primitive access functions */
|
|
|
|
#define ENERGY_STATUS_MASK 0xffffffff
|
|
|
|
|
|
|
|
#define POWER_LIMIT1_MASK 0x7FFF
|
|
|
|
#define POWER_LIMIT1_ENABLE BIT(15)
|
|
|
|
#define POWER_LIMIT1_CLAMP BIT(16)
|
|
|
|
|
|
|
|
#define POWER_LIMIT2_MASK (0x7FFFULL<<32)
|
|
|
|
#define POWER_LIMIT2_ENABLE BIT_ULL(47)
|
|
|
|
#define POWER_LIMIT2_CLAMP BIT_ULL(48)
|
2019-07-10 07:44:32 -06:00
|
|
|
#define POWER_HIGH_LOCK BIT_ULL(63)
|
|
|
|
#define POWER_LOW_LOCK BIT(31)
|
PowerCap: Introduce Intel RAPL power capping driver
The Intel Running Average Power Limit (RAPL) technology provides platform
software with the ability to monitor, control, and get notifications on
power usage.
This feature is present in all Sandy Bridge and later Intel processors.
Newer models allow more fine grained controls to be applied. In RAPL,
power control is divided into domains, which include package, DRAM
controller, CPU core (Power Plane 0), graphics uncore (power plane 1), etc.
The purpose of this driver is to expose the RAPL settings to userspace.
Overall, RAPL fits in the new powercap class driver in that platform
level power capping controls are exposed via this generic interface.
This driver is based on an earlier patch from Zhang Rui.
However, while the previous work was mainly focused on thermal monitoring
the focus here is on the usability from user space perspective.
References: https://lkml.org/lkml/2011/5/26/93
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Signed-off-by: Jacob Pan <jacob.jun.pan@linux.intel.com>
Reviewed-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
2013-10-17 11:28:35 -06:00
|
|
|
|
|
|
|
#define TIME_WINDOW1_MASK (0x7FULL<<17)
|
|
|
|
#define TIME_WINDOW2_MASK (0x7FULL<<49)
|
|
|
|
|
|
|
|
#define POWER_UNIT_OFFSET 0
|
|
|
|
#define POWER_UNIT_MASK 0x0F
|
|
|
|
|
|
|
|
#define ENERGY_UNIT_OFFSET 0x08
|
|
|
|
#define ENERGY_UNIT_MASK 0x1F00
|
|
|
|
|
|
|
|
#define TIME_UNIT_OFFSET 0x10
|
|
|
|
#define TIME_UNIT_MASK 0xF0000
|
|
|
|
|
|
|
|
#define POWER_INFO_MAX_MASK (0x7fffULL<<32)
|
|
|
|
#define POWER_INFO_MIN_MASK (0x7fffULL<<16)
|
|
|
|
#define POWER_INFO_MAX_TIME_WIN_MASK (0x3fULL<<48)
|
|
|
|
#define POWER_INFO_THERMAL_SPEC_MASK 0x7fff
|
|
|
|
|
|
|
|
#define PERF_STATUS_THROTTLE_TIME_MASK 0xffffffff
|
|
|
|
#define PP_POLICY_MASK 0x1F
|
|
|
|
|
|
|
|
/* Non HW constants */
|
2019-07-10 07:44:30 -06:00
|
|
|
#define RAPL_PRIMITIVE_DERIVED BIT(1) /* not from raw data */
|
PowerCap: Introduce Intel RAPL power capping driver
The Intel Running Average Power Limit (RAPL) technology provides platform
software with the ability to monitor, control, and get notifications on
power usage.
This feature is present in all Sandy Bridge and later Intel processors.
Newer models allow more fine grained controls to be applied. In RAPL,
power control is divided into domains, which include package, DRAM
controller, CPU core (Power Plane 0), graphics uncore (power plane 1), etc.
The purpose of this driver is to expose the RAPL settings to userspace.
Overall, RAPL fits in the new powercap class driver in that platform
level power capping controls are exposed via this generic interface.
This driver is based on an earlier patch from Zhang Rui.
However, while the previous work was mainly focused on thermal monitoring
the focus here is on the usability from user space perspective.
References: https://lkml.org/lkml/2011/5/26/93
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Signed-off-by: Jacob Pan <jacob.jun.pan@linux.intel.com>
Reviewed-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
2013-10-17 11:28:35 -06:00
|
|
|
#define RAPL_PRIMITIVE_DUMMY BIT(2)
|
|
|
|
|
|
|
|
#define TIME_WINDOW_MAX_MSEC 40000
|
|
|
|
#define TIME_WINDOW_MIN_MSEC 250
|
2019-07-10 07:44:30 -06:00
|
|
|
#define ENERGY_UNIT_SCALE 1000 /* scale from driver unit to powercap unit */
|
PowerCap: Introduce Intel RAPL power capping driver
The Intel Running Average Power Limit (RAPL) technology provides platform
software with the ability to monitor, control, and get notifications on
power usage.
This feature is present in all Sandy Bridge and later Intel processors.
Newer models allow more fine grained controls to be applied. In RAPL,
power control is divided into domains, which include package, DRAM
controller, CPU core (Power Plane 0), graphics uncore (power plane 1), etc.
The purpose of this driver is to expose the RAPL settings to userspace.
Overall, RAPL fits in the new powercap class driver in that platform
level power capping controls are exposed via this generic interface.
This driver is based on an earlier patch from Zhang Rui.
However, while the previous work was mainly focused on thermal monitoring
the focus here is on the usability from user space perspective.
References: https://lkml.org/lkml/2011/5/26/93
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Signed-off-by: Jacob Pan <jacob.jun.pan@linux.intel.com>
Reviewed-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
2013-10-17 11:28:35 -06:00
|
|
|
enum unit_type {
|
2019-07-10 07:44:30 -06:00
|
|
|
ARBITRARY_UNIT, /* no translation */
|
PowerCap: Introduce Intel RAPL power capping driver
The Intel Running Average Power Limit (RAPL) technology provides platform
software with the ability to monitor, control, and get notifications on
power usage.
This feature is present in all Sandy Bridge and later Intel processors.
Newer models allow more fine grained controls to be applied. In RAPL,
power control is divided into domains, which include package, DRAM
controller, CPU core (Power Plane 0), graphics uncore (power plane 1), etc.
The purpose of this driver is to expose the RAPL settings to userspace.
Overall, RAPL fits in the new powercap class driver in that platform
level power capping controls are exposed via this generic interface.
This driver is based on an earlier patch from Zhang Rui.
However, while the previous work was mainly focused on thermal monitoring
the focus here is on the usability from user space perspective.
References: https://lkml.org/lkml/2011/5/26/93
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Signed-off-by: Jacob Pan <jacob.jun.pan@linux.intel.com>
Reviewed-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
2013-10-17 11:28:35 -06:00
|
|
|
POWER_UNIT,
|
|
|
|
ENERGY_UNIT,
|
|
|
|
TIME_UNIT,
|
|
|
|
};
|
|
|
|
|
|
|
|
/* per domain data, some are optional */
|
|
|
|
#define NR_RAW_PRIMITIVES (NR_RAPL_PRIMITIVES - 2)
|
|
|
|
|
|
|
|
#define DOMAIN_STATE_INACTIVE BIT(0)
|
|
|
|
#define DOMAIN_STATE_POWER_LIMIT_SET BIT(1)
|
|
|
|
#define DOMAIN_STATE_BIOS_LOCKED BIT(2)
|
|
|
|
|
|
|
|
static const char pl1_name[] = "long_term";
|
|
|
|
static const char pl2_name[] = "short_term";
|
|
|
|
|
|
|
|
#define power_zone_to_rapl_domain(_zone) \
|
|
|
|
container_of(_zone, struct rapl_domain, power_zone)
|
|
|
|
|
2014-11-07 10:29:25 -07:00
|
|
|
struct rapl_defaults {
|
2015-04-29 14:13:23 -06:00
|
|
|
u8 floor_freq_reg_addr;
|
2014-11-07 10:29:25 -07:00
|
|
|
int (*check_unit)(struct rapl_package *rp, int cpu);
|
|
|
|
void (*set_floor_freq)(struct rapl_domain *rd, bool mode);
|
|
|
|
u64 (*compute_time_window)(struct rapl_package *rp, u64 val,
|
2019-07-10 07:44:30 -06:00
|
|
|
bool to_raw);
|
2015-03-13 04:48:56 -06:00
|
|
|
unsigned int dram_domain_energy_unit;
|
2014-11-07 10:29:25 -07:00
|
|
|
};
|
|
|
|
static struct rapl_defaults *rapl_defaults;
|
|
|
|
|
2014-11-07 10:29:26 -07:00
|
|
|
/* Sideband MBI registers */
|
2015-04-29 14:13:23 -06:00
|
|
|
#define IOSF_CPU_POWER_BUDGET_CTL_BYT (0x2)
|
|
|
|
#define IOSF_CPU_POWER_BUDGET_CTL_TNG (0xdf)
|
2014-11-07 10:29:26 -07:00
|
|
|
|
PowerCap: Introduce Intel RAPL power capping driver
The Intel Running Average Power Limit (RAPL) technology provides platform
software with the ability to monitor, control, and get notifications on
power usage.
This feature is present in all Sandy Bridge and later Intel processors.
Newer models allow more fine grained controls to be applied. In RAPL,
power control is divided into domains, which include package, DRAM
controller, CPU core (Power Plane 0), graphics uncore (power plane 1), etc.
The purpose of this driver is to expose the RAPL settings to userspace.
Overall, RAPL fits in the new powercap class driver in that platform
level power capping controls are exposed via this generic interface.
This driver is based on an earlier patch from Zhang Rui.
However, while the previous work was mainly focused on thermal monitoring
the focus here is on the usability from user space perspective.
References: https://lkml.org/lkml/2011/5/26/93
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Signed-off-by: Jacob Pan <jacob.jun.pan@linux.intel.com>
Reviewed-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
2013-10-17 11:28:35 -06:00
|
|
|
#define PACKAGE_PLN_INT_SAVED BIT(0)
|
|
|
|
#define MAX_PRIM_NAME (32)
|
|
|
|
|
|
|
|
/* per domain data. used to describe individual knobs such that access function
|
|
|
|
* can be consolidated into one instead of many inline functions.
|
|
|
|
*/
|
|
|
|
struct rapl_primitive_info {
|
|
|
|
const char *name;
|
|
|
|
u64 mask;
|
|
|
|
int shift;
|
2019-07-10 07:44:22 -06:00
|
|
|
enum rapl_domain_reg_id id;
|
PowerCap: Introduce Intel RAPL power capping driver
The Intel Running Average Power Limit (RAPL) technology provides platform
software with the ability to monitor, control, and get notifications on
power usage.
This feature is present in all Sandy Bridge and later Intel processors.
Newer models allow more fine grained controls to be applied. In RAPL,
power control is divided into domains, which include package, DRAM
controller, CPU core (Power Plane 0), graphics uncore (power plane 1), etc.
The purpose of this driver is to expose the RAPL settings to userspace.
Overall, RAPL fits in the new powercap class driver in that platform
level power capping controls are exposed via this generic interface.
This driver is based on an earlier patch from Zhang Rui.
However, while the previous work was mainly focused on thermal monitoring
the focus here is on the usability from user space perspective.
References: https://lkml.org/lkml/2011/5/26/93
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Signed-off-by: Jacob Pan <jacob.jun.pan@linux.intel.com>
Reviewed-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
2013-10-17 11:28:35 -06:00
|
|
|
enum unit_type unit;
|
|
|
|
u32 flag;
|
|
|
|
};
|
|
|
|
|
|
|
|
#define PRIMITIVE_INFO_INIT(p, m, s, i, u, f) { \
|
|
|
|
.name = #p, \
|
|
|
|
.mask = m, \
|
|
|
|
.shift = s, \
|
|
|
|
.id = i, \
|
|
|
|
.unit = u, \
|
|
|
|
.flag = f \
|
|
|
|
}
|
|
|
|
|
|
|
|
static void rapl_init_domains(struct rapl_package *rp);
|
|
|
|
static int rapl_read_data_raw(struct rapl_domain *rd,
|
2019-07-10 07:44:30 -06:00
|
|
|
enum rapl_primitives prim,
|
|
|
|
bool xlate, u64 *data);
|
PowerCap: Introduce Intel RAPL power capping driver
The Intel Running Average Power Limit (RAPL) technology provides platform
software with the ability to monitor, control, and get notifications on
power usage.
This feature is present in all Sandy Bridge and later Intel processors.
Newer models allow more fine grained controls to be applied. In RAPL,
power control is divided into domains, which include package, DRAM
controller, CPU core (Power Plane 0), graphics uncore (power plane 1), etc.
The purpose of this driver is to expose the RAPL settings to userspace.
Overall, RAPL fits in the new powercap class driver in that platform
level power capping controls are exposed via this generic interface.
This driver is based on an earlier patch from Zhang Rui.
However, while the previous work was mainly focused on thermal monitoring
the focus here is on the usability from user space perspective.
References: https://lkml.org/lkml/2011/5/26/93
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Signed-off-by: Jacob Pan <jacob.jun.pan@linux.intel.com>
Reviewed-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
2013-10-17 11:28:35 -06:00
|
|
|
static int rapl_write_data_raw(struct rapl_domain *rd,
|
2019-07-10 07:44:30 -06:00
|
|
|
enum rapl_primitives prim,
|
|
|
|
unsigned long long value);
|
2016-02-24 14:31:37 -07:00
|
|
|
static u64 rapl_unit_xlate(struct rapl_domain *rd,
|
2019-07-10 07:44:30 -06:00
|
|
|
enum unit_type type, u64 value, int to_raw);
|
2016-02-24 14:31:37 -07:00
|
|
|
static void package_power_limit_irq_save(struct rapl_package *rp);
|
PowerCap: Introduce Intel RAPL power capping driver
The Intel Running Average Power Limit (RAPL) technology provides platform
software with the ability to monitor, control, and get notifications on
power usage.
This feature is present in all Sandy Bridge and later Intel processors.
Newer models allow more fine grained controls to be applied. In RAPL,
power control is divided into domains, which include package, DRAM
controller, CPU core (Power Plane 0), graphics uncore (power plane 1), etc.
The purpose of this driver is to expose the RAPL settings to userspace.
Overall, RAPL fits in the new powercap class driver in that platform
level power capping controls are exposed via this generic interface.
This driver is based on an earlier patch from Zhang Rui.
However, while the previous work was mainly focused on thermal monitoring
the focus here is on the usability from user space perspective.
References: https://lkml.org/lkml/2011/5/26/93
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Signed-off-by: Jacob Pan <jacob.jun.pan@linux.intel.com>
Reviewed-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
2013-10-17 11:28:35 -06:00
|
|
|
|
2019-07-10 07:44:30 -06:00
|
|
|
static LIST_HEAD(rapl_packages); /* guarded by CPU hotplug lock */
|
PowerCap: Introduce Intel RAPL power capping driver
The Intel Running Average Power Limit (RAPL) technology provides platform
software with the ability to monitor, control, and get notifications on
power usage.
This feature is present in all Sandy Bridge and later Intel processors.
Newer models allow more fine grained controls to be applied. In RAPL,
power control is divided into domains, which include package, DRAM
controller, CPU core (Power Plane 0), graphics uncore (power plane 1), etc.
The purpose of this driver is to expose the RAPL settings to userspace.
Overall, RAPL fits in the new powercap class driver in that platform
level power capping controls are exposed via this generic interface.
This driver is based on an earlier patch from Zhang Rui.
However, while the previous work was mainly focused on thermal monitoring
the focus here is on the usability from user space perspective.
References: https://lkml.org/lkml/2011/5/26/93
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Signed-off-by: Jacob Pan <jacob.jun.pan@linux.intel.com>
Reviewed-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
2013-10-17 11:28:35 -06:00
|
|
|
|
2019-07-10 07:44:30 -06:00
|
|
|
static const char *const rapl_domain_names[] = {
|
PowerCap: Introduce Intel RAPL power capping driver
The Intel Running Average Power Limit (RAPL) technology provides platform
software with the ability to monitor, control, and get notifications on
power usage.
This feature is present in all Sandy Bridge and later Intel processors.
Newer models allow more fine grained controls to be applied. In RAPL,
power control is divided into domains, which include package, DRAM
controller, CPU core (Power Plane 0), graphics uncore (power plane 1), etc.
The purpose of this driver is to expose the RAPL settings to userspace.
Overall, RAPL fits in the new powercap class driver in that platform
level power capping controls are exposed via this generic interface.
This driver is based on an earlier patch from Zhang Rui.
However, while the previous work was mainly focused on thermal monitoring
the focus here is on the usability from user space perspective.
References: https://lkml.org/lkml/2011/5/26/93
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Signed-off-by: Jacob Pan <jacob.jun.pan@linux.intel.com>
Reviewed-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
2013-10-17 11:28:35 -06:00
|
|
|
"package",
|
|
|
|
"core",
|
|
|
|
"uncore",
|
|
|
|
"dram",
|
2016-04-17 16:03:01 -06:00
|
|
|
"psys",
|
PowerCap: Introduce Intel RAPL power capping driver
The Intel Running Average Power Limit (RAPL) technology provides platform
software with the ability to monitor, control, and get notifications on
power usage.
This feature is present in all Sandy Bridge and later Intel processors.
Newer models allow more fine grained controls to be applied. In RAPL,
power control is divided into domains, which include package, DRAM
controller, CPU core (Power Plane 0), graphics uncore (power plane 1), etc.
The purpose of this driver is to expose the RAPL settings to userspace.
Overall, RAPL fits in the new powercap class driver in that platform
level power capping controls are exposed via this generic interface.
This driver is based on an earlier patch from Zhang Rui.
However, while the previous work was mainly focused on thermal monitoring
the focus here is on the usability from user space perspective.
References: https://lkml.org/lkml/2011/5/26/93
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Signed-off-by: Jacob Pan <jacob.jun.pan@linux.intel.com>
Reviewed-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
2013-10-17 11:28:35 -06:00
|
|
|
};
|
|
|
|
|
2019-07-10 07:44:30 -06:00
|
|
|
static int get_energy_counter(struct powercap_zone *power_zone,
|
|
|
|
u64 *energy_raw)
|
PowerCap: Introduce Intel RAPL power capping driver
The Intel Running Average Power Limit (RAPL) technology provides platform
software with the ability to monitor, control, and get notifications on
power usage.
This feature is present in all Sandy Bridge and later Intel processors.
Newer models allow more fine grained controls to be applied. In RAPL,
power control is divided into domains, which include package, DRAM
controller, CPU core (Power Plane 0), graphics uncore (power plane 1), etc.
The purpose of this driver is to expose the RAPL settings to userspace.
Overall, RAPL fits in the new powercap class driver in that platform
level power capping controls are exposed via this generic interface.
This driver is based on an earlier patch from Zhang Rui.
However, while the previous work was mainly focused on thermal monitoring
the focus here is on the usability from user space perspective.
References: https://lkml.org/lkml/2011/5/26/93
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Signed-off-by: Jacob Pan <jacob.jun.pan@linux.intel.com>
Reviewed-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
2013-10-17 11:28:35 -06:00
|
|
|
{
|
|
|
|
struct rapl_domain *rd;
|
|
|
|
u64 energy_now;
|
|
|
|
|
|
|
|
/* prevent CPU hotplug, make sure the RAPL domain does not go
|
|
|
|
* away while reading the counter.
|
|
|
|
*/
|
|
|
|
get_online_cpus();
|
|
|
|
rd = power_zone_to_rapl_domain(power_zone);
|
|
|
|
|
|
|
|
if (!rapl_read_data_raw(rd, ENERGY_COUNTER, true, &energy_now)) {
|
|
|
|
*energy_raw = energy_now;
|
|
|
|
put_online_cpus();
|
|
|
|
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
put_online_cpus();
|
|
|
|
|
|
|
|
return -EIO;
|
|
|
|
}
|
|
|
|
|
|
|
|
static int get_max_energy_counter(struct powercap_zone *pcd_dev, u64 *energy)
|
|
|
|
{
|
2015-03-13 04:48:56 -06:00
|
|
|
struct rapl_domain *rd = power_zone_to_rapl_domain(pcd_dev);
|
|
|
|
|
2016-02-24 14:31:37 -07:00
|
|
|
*energy = rapl_unit_xlate(rd, ENERGY_UNIT, ENERGY_STATUS_MASK, 0);
|
PowerCap: Introduce Intel RAPL power capping driver
The Intel Running Average Power Limit (RAPL) technology provides platform
software with the ability to monitor, control, and get notifications on
power usage.
This feature is present in all Sandy Bridge and later Intel processors.
Newer models allow more fine grained controls to be applied. In RAPL,
power control is divided into domains, which include package, DRAM
controller, CPU core (Power Plane 0), graphics uncore (power plane 1), etc.
The purpose of this driver is to expose the RAPL settings to userspace.
Overall, RAPL fits in the new powercap class driver in that platform
level power capping controls are exposed via this generic interface.
This driver is based on an earlier patch from Zhang Rui.
However, while the previous work was mainly focused on thermal monitoring
the focus here is on the usability from user space perspective.
References: https://lkml.org/lkml/2011/5/26/93
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Signed-off-by: Jacob Pan <jacob.jun.pan@linux.intel.com>
Reviewed-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
2013-10-17 11:28:35 -06:00
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
static int release_zone(struct powercap_zone *power_zone)
|
|
|
|
{
|
|
|
|
struct rapl_domain *rd = power_zone_to_rapl_domain(power_zone);
|
2016-02-24 14:31:37 -07:00
|
|
|
struct rapl_package *rp = rd->rp;
|
PowerCap: Introduce Intel RAPL power capping driver
The Intel Running Average Power Limit (RAPL) technology provides platform
software with the ability to monitor, control, and get notifications on
power usage.
This feature is present in all Sandy Bridge and later Intel processors.
Newer models allow more fine grained controls to be applied. In RAPL,
power control is divided into domains, which include package, DRAM
controller, CPU core (Power Plane 0), graphics uncore (power plane 1), etc.
The purpose of this driver is to expose the RAPL settings to userspace.
Overall, RAPL fits in the new powercap class driver in that platform
level power capping controls are exposed via this generic interface.
This driver is based on an earlier patch from Zhang Rui.
However, while the previous work was mainly focused on thermal monitoring
the focus here is on the usability from user space perspective.
References: https://lkml.org/lkml/2011/5/26/93
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Signed-off-by: Jacob Pan <jacob.jun.pan@linux.intel.com>
Reviewed-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
2013-10-17 11:28:35 -06:00
|
|
|
|
|
|
|
/* package zone is the last zone of a package, we can free
|
|
|
|
* memory here since all children has been unregistered.
|
|
|
|
*/
|
|
|
|
if (rd->id == RAPL_DOMAIN_PACKAGE) {
|
|
|
|
kfree(rd);
|
|
|
|
rp->domains = NULL;
|
|
|
|
}
|
|
|
|
|
|
|
|
return 0;
|
|
|
|
|
|
|
|
}
|
|
|
|
|
|
|
|
static int find_nr_power_limit(struct rapl_domain *rd)
|
|
|
|
{
|
2016-05-31 14:41:29 -06:00
|
|
|
int i, nr_pl = 0;
|
PowerCap: Introduce Intel RAPL power capping driver
The Intel Running Average Power Limit (RAPL) technology provides platform
software with the ability to monitor, control, and get notifications on
power usage.
This feature is present in all Sandy Bridge and later Intel processors.
Newer models allow more fine grained controls to be applied. In RAPL,
power control is divided into domains, which include package, DRAM
controller, CPU core (Power Plane 0), graphics uncore (power plane 1), etc.
The purpose of this driver is to expose the RAPL settings to userspace.
Overall, RAPL fits in the new powercap class driver in that platform
level power capping controls are exposed via this generic interface.
This driver is based on an earlier patch from Zhang Rui.
However, while the previous work was mainly focused on thermal monitoring
the focus here is on the usability from user space perspective.
References: https://lkml.org/lkml/2011/5/26/93
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Signed-off-by: Jacob Pan <jacob.jun.pan@linux.intel.com>
Reviewed-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
2013-10-17 11:28:35 -06:00
|
|
|
|
|
|
|
for (i = 0; i < NR_POWER_LIMITS; i++) {
|
2016-05-31 14:41:29 -06:00
|
|
|
if (rd->rpl[i].name)
|
|
|
|
nr_pl++;
|
PowerCap: Introduce Intel RAPL power capping driver
The Intel Running Average Power Limit (RAPL) technology provides platform
software with the ability to monitor, control, and get notifications on
power usage.
This feature is present in all Sandy Bridge and later Intel processors.
Newer models allow more fine grained controls to be applied. In RAPL,
power control is divided into domains, which include package, DRAM
controller, CPU core (Power Plane 0), graphics uncore (power plane 1), etc.
The purpose of this driver is to expose the RAPL settings to userspace.
Overall, RAPL fits in the new powercap class driver in that platform
level power capping controls are exposed via this generic interface.
This driver is based on an earlier patch from Zhang Rui.
However, while the previous work was mainly focused on thermal monitoring
the focus here is on the usability from user space perspective.
References: https://lkml.org/lkml/2011/5/26/93
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Signed-off-by: Jacob Pan <jacob.jun.pan@linux.intel.com>
Reviewed-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
2013-10-17 11:28:35 -06:00
|
|
|
}
|
|
|
|
|
2016-05-31 14:41:29 -06:00
|
|
|
return nr_pl;
|
PowerCap: Introduce Intel RAPL power capping driver
The Intel Running Average Power Limit (RAPL) technology provides platform
software with the ability to monitor, control, and get notifications on
power usage.
This feature is present in all Sandy Bridge and later Intel processors.
Newer models allow more fine grained controls to be applied. In RAPL,
power control is divided into domains, which include package, DRAM
controller, CPU core (Power Plane 0), graphics uncore (power plane 1), etc.
The purpose of this driver is to expose the RAPL settings to userspace.
Overall, RAPL fits in the new powercap class driver in that platform
level power capping controls are exposed via this generic interface.
This driver is based on an earlier patch from Zhang Rui.
However, while the previous work was mainly focused on thermal monitoring
the focus here is on the usability from user space perspective.
References: https://lkml.org/lkml/2011/5/26/93
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Signed-off-by: Jacob Pan <jacob.jun.pan@linux.intel.com>
Reviewed-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
2013-10-17 11:28:35 -06:00
|
|
|
}
|
|
|
|
|
|
|
|
static int set_domain_enable(struct powercap_zone *power_zone, bool mode)
|
|
|
|
{
|
|
|
|
struct rapl_domain *rd = power_zone_to_rapl_domain(power_zone);
|
|
|
|
|
|
|
|
if (rd->state & DOMAIN_STATE_BIOS_LOCKED)
|
|
|
|
return -EACCES;
|
2014-11-07 10:29:26 -07:00
|
|
|
|
PowerCap: Introduce Intel RAPL power capping driver
The Intel Running Average Power Limit (RAPL) technology provides platform
software with the ability to monitor, control, and get notifications on
power usage.
This feature is present in all Sandy Bridge and later Intel processors.
Newer models allow more fine grained controls to be applied. In RAPL,
power control is divided into domains, which include package, DRAM
controller, CPU core (Power Plane 0), graphics uncore (power plane 1), etc.
The purpose of this driver is to expose the RAPL settings to userspace.
Overall, RAPL fits in the new powercap class driver in that platform
level power capping controls are exposed via this generic interface.
This driver is based on an earlier patch from Zhang Rui.
However, while the previous work was mainly focused on thermal monitoring
the focus here is on the usability from user space perspective.
References: https://lkml.org/lkml/2011/5/26/93
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Signed-off-by: Jacob Pan <jacob.jun.pan@linux.intel.com>
Reviewed-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
2013-10-17 11:28:35 -06:00
|
|
|
get_online_cpus();
|
|
|
|
rapl_write_data_raw(rd, PL1_ENABLE, mode);
|
2015-04-29 14:13:23 -06:00
|
|
|
if (rapl_defaults->set_floor_freq)
|
|
|
|
rapl_defaults->set_floor_freq(rd, mode);
|
PowerCap: Introduce Intel RAPL power capping driver
The Intel Running Average Power Limit (RAPL) technology provides platform
software with the ability to monitor, control, and get notifications on
power usage.
This feature is present in all Sandy Bridge and later Intel processors.
Newer models allow more fine grained controls to be applied. In RAPL,
power control is divided into domains, which include package, DRAM
controller, CPU core (Power Plane 0), graphics uncore (power plane 1), etc.
The purpose of this driver is to expose the RAPL settings to userspace.
Overall, RAPL fits in the new powercap class driver in that platform
level power capping controls are exposed via this generic interface.
This driver is based on an earlier patch from Zhang Rui.
However, while the previous work was mainly focused on thermal monitoring
the focus here is on the usability from user space perspective.
References: https://lkml.org/lkml/2011/5/26/93
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Signed-off-by: Jacob Pan <jacob.jun.pan@linux.intel.com>
Reviewed-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
2013-10-17 11:28:35 -06:00
|
|
|
put_online_cpus();
|
|
|
|
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
static int get_domain_enable(struct powercap_zone *power_zone, bool *mode)
|
|
|
|
{
|
|
|
|
struct rapl_domain *rd = power_zone_to_rapl_domain(power_zone);
|
|
|
|
u64 val;
|
|
|
|
|
|
|
|
if (rd->state & DOMAIN_STATE_BIOS_LOCKED) {
|
|
|
|
*mode = false;
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
get_online_cpus();
|
|
|
|
if (rapl_read_data_raw(rd, PL1_ENABLE, true, &val)) {
|
|
|
|
put_online_cpus();
|
|
|
|
return -EIO;
|
|
|
|
}
|
|
|
|
*mode = val;
|
|
|
|
put_online_cpus();
|
|
|
|
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
/* per RAPL domain ops, in the order of rapl_domain_type */
|
2015-12-23 14:59:55 -07:00
|
|
|
static const struct powercap_zone_ops zone_ops[] = {
|
PowerCap: Introduce Intel RAPL power capping driver
The Intel Running Average Power Limit (RAPL) technology provides platform
software with the ability to monitor, control, and get notifications on
power usage.
This feature is present in all Sandy Bridge and later Intel processors.
Newer models allow more fine grained controls to be applied. In RAPL,
power control is divided into domains, which include package, DRAM
controller, CPU core (Power Plane 0), graphics uncore (power plane 1), etc.
The purpose of this driver is to expose the RAPL settings to userspace.
Overall, RAPL fits in the new powercap class driver in that platform
level power capping controls are exposed via this generic interface.
This driver is based on an earlier patch from Zhang Rui.
However, while the previous work was mainly focused on thermal monitoring
the focus here is on the usability from user space perspective.
References: https://lkml.org/lkml/2011/5/26/93
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Signed-off-by: Jacob Pan <jacob.jun.pan@linux.intel.com>
Reviewed-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
2013-10-17 11:28:35 -06:00
|
|
|
/* RAPL_DOMAIN_PACKAGE */
|
|
|
|
{
|
2019-07-10 07:44:30 -06:00
|
|
|
.get_energy_uj = get_energy_counter,
|
|
|
|
.get_max_energy_range_uj = get_max_energy_counter,
|
|
|
|
.release = release_zone,
|
|
|
|
.set_enable = set_domain_enable,
|
|
|
|
.get_enable = get_domain_enable,
|
|
|
|
},
|
PowerCap: Introduce Intel RAPL power capping driver
The Intel Running Average Power Limit (RAPL) technology provides platform
software with the ability to monitor, control, and get notifications on
power usage.
This feature is present in all Sandy Bridge and later Intel processors.
Newer models allow more fine grained controls to be applied. In RAPL,
power control is divided into domains, which include package, DRAM
controller, CPU core (Power Plane 0), graphics uncore (power plane 1), etc.
The purpose of this driver is to expose the RAPL settings to userspace.
Overall, RAPL fits in the new powercap class driver in that platform
level power capping controls are exposed via this generic interface.
This driver is based on an earlier patch from Zhang Rui.
However, while the previous work was mainly focused on thermal monitoring
the focus here is on the usability from user space perspective.
References: https://lkml.org/lkml/2011/5/26/93
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Signed-off-by: Jacob Pan <jacob.jun.pan@linux.intel.com>
Reviewed-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
2013-10-17 11:28:35 -06:00
|
|
|
/* RAPL_DOMAIN_PP0 */
|
|
|
|
{
|
2019-07-10 07:44:30 -06:00
|
|
|
.get_energy_uj = get_energy_counter,
|
|
|
|
.get_max_energy_range_uj = get_max_energy_counter,
|
|
|
|
.release = release_zone,
|
|
|
|
.set_enable = set_domain_enable,
|
|
|
|
.get_enable = get_domain_enable,
|
|
|
|
},
|
PowerCap: Introduce Intel RAPL power capping driver
The Intel Running Average Power Limit (RAPL) technology provides platform
software with the ability to monitor, control, and get notifications on
power usage.
This feature is present in all Sandy Bridge and later Intel processors.
Newer models allow more fine grained controls to be applied. In RAPL,
power control is divided into domains, which include package, DRAM
controller, CPU core (Power Plane 0), graphics uncore (power plane 1), etc.
The purpose of this driver is to expose the RAPL settings to userspace.
Overall, RAPL fits in the new powercap class driver in that platform
level power capping controls are exposed via this generic interface.
This driver is based on an earlier patch from Zhang Rui.
However, while the previous work was mainly focused on thermal monitoring
the focus here is on the usability from user space perspective.
References: https://lkml.org/lkml/2011/5/26/93
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Signed-off-by: Jacob Pan <jacob.jun.pan@linux.intel.com>
Reviewed-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
2013-10-17 11:28:35 -06:00
|
|
|
/* RAPL_DOMAIN_PP1 */
|
|
|
|
{
|
2019-07-10 07:44:30 -06:00
|
|
|
.get_energy_uj = get_energy_counter,
|
|
|
|
.get_max_energy_range_uj = get_max_energy_counter,
|
|
|
|
.release = release_zone,
|
|
|
|
.set_enable = set_domain_enable,
|
|
|
|
.get_enable = get_domain_enable,
|
|
|
|
},
|
PowerCap: Introduce Intel RAPL power capping driver
The Intel Running Average Power Limit (RAPL) technology provides platform
software with the ability to monitor, control, and get notifications on
power usage.
This feature is present in all Sandy Bridge and later Intel processors.
Newer models allow more fine grained controls to be applied. In RAPL,
power control is divided into domains, which include package, DRAM
controller, CPU core (Power Plane 0), graphics uncore (power plane 1), etc.
The purpose of this driver is to expose the RAPL settings to userspace.
Overall, RAPL fits in the new powercap class driver in that platform
level power capping controls are exposed via this generic interface.
This driver is based on an earlier patch from Zhang Rui.
However, while the previous work was mainly focused on thermal monitoring
the focus here is on the usability from user space perspective.
References: https://lkml.org/lkml/2011/5/26/93
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Signed-off-by: Jacob Pan <jacob.jun.pan@linux.intel.com>
Reviewed-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
2013-10-17 11:28:35 -06:00
|
|
|
/* RAPL_DOMAIN_DRAM */
|
|
|
|
{
|
2019-07-10 07:44:30 -06:00
|
|
|
.get_energy_uj = get_energy_counter,
|
|
|
|
.get_max_energy_range_uj = get_max_energy_counter,
|
|
|
|
.release = release_zone,
|
|
|
|
.set_enable = set_domain_enable,
|
|
|
|
.get_enable = get_domain_enable,
|
|
|
|
},
|
2016-04-17 16:03:01 -06:00
|
|
|
/* RAPL_DOMAIN_PLATFORM */
|
|
|
|
{
|
2019-07-10 07:44:30 -06:00
|
|
|
.get_energy_uj = get_energy_counter,
|
|
|
|
.get_max_energy_range_uj = get_max_energy_counter,
|
|
|
|
.release = release_zone,
|
|
|
|
.set_enable = set_domain_enable,
|
|
|
|
.get_enable = get_domain_enable,
|
|
|
|
},
|
PowerCap: Introduce Intel RAPL power capping driver
The Intel Running Average Power Limit (RAPL) technology provides platform
software with the ability to monitor, control, and get notifications on
power usage.
This feature is present in all Sandy Bridge and later Intel processors.
Newer models allow more fine grained controls to be applied. In RAPL,
power control is divided into domains, which include package, DRAM
controller, CPU core (Power Plane 0), graphics uncore (power plane 1), etc.
The purpose of this driver is to expose the RAPL settings to userspace.
Overall, RAPL fits in the new powercap class driver in that platform
level power capping controls are exposed via this generic interface.
This driver is based on an earlier patch from Zhang Rui.
However, while the previous work was mainly focused on thermal monitoring
the focus here is on the usability from user space perspective.
References: https://lkml.org/lkml/2011/5/26/93
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Signed-off-by: Jacob Pan <jacob.jun.pan@linux.intel.com>
Reviewed-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
2013-10-17 11:28:35 -06:00
|
|
|
};
|
|
|
|
|
2016-05-31 14:41:29 -06:00
|
|
|
/*
|
|
|
|
* Constraint index used by powercap can be different than power limit (PL)
|
2019-07-10 07:44:30 -06:00
|
|
|
* index in that some PLs maybe missing due to non-existent MSRs. So we
|
2016-05-31 14:41:29 -06:00
|
|
|
* need to convert here by finding the valid PLs only (name populated).
|
|
|
|
*/
|
|
|
|
static int contraint_to_pl(struct rapl_domain *rd, int cid)
|
|
|
|
{
|
|
|
|
int i, j;
|
|
|
|
|
|
|
|
for (i = 0, j = 0; i < NR_POWER_LIMITS; i++) {
|
|
|
|
if ((rd->rpl[i].name) && j++ == cid) {
|
|
|
|
pr_debug("%s: index %d\n", __func__, i);
|
|
|
|
return i;
|
|
|
|
}
|
|
|
|
}
|
2016-11-28 14:53:11 -07:00
|
|
|
pr_err("Cannot find matching power limit for constraint %d\n", cid);
|
2016-05-31 14:41:29 -06:00
|
|
|
|
|
|
|
return -EINVAL;
|
|
|
|
}
|
|
|
|
|
|
|
|
static int set_power_limit(struct powercap_zone *power_zone, int cid,
|
2019-07-10 07:44:30 -06:00
|
|
|
u64 power_limit)
|
PowerCap: Introduce Intel RAPL power capping driver
The Intel Running Average Power Limit (RAPL) technology provides platform
software with the ability to monitor, control, and get notifications on
power usage.
This feature is present in all Sandy Bridge and later Intel processors.
Newer models allow more fine grained controls to be applied. In RAPL,
power control is divided into domains, which include package, DRAM
controller, CPU core (Power Plane 0), graphics uncore (power plane 1), etc.
The purpose of this driver is to expose the RAPL settings to userspace.
Overall, RAPL fits in the new powercap class driver in that platform
level power capping controls are exposed via this generic interface.
This driver is based on an earlier patch from Zhang Rui.
However, while the previous work was mainly focused on thermal monitoring
the focus here is on the usability from user space perspective.
References: https://lkml.org/lkml/2011/5/26/93
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Signed-off-by: Jacob Pan <jacob.jun.pan@linux.intel.com>
Reviewed-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
2013-10-17 11:28:35 -06:00
|
|
|
{
|
|
|
|
struct rapl_domain *rd;
|
|
|
|
struct rapl_package *rp;
|
|
|
|
int ret = 0;
|
2016-05-31 14:41:29 -06:00
|
|
|
int id;
|
PowerCap: Introduce Intel RAPL power capping driver
The Intel Running Average Power Limit (RAPL) technology provides platform
software with the ability to monitor, control, and get notifications on
power usage.
This feature is present in all Sandy Bridge and later Intel processors.
Newer models allow more fine grained controls to be applied. In RAPL,
power control is divided into domains, which include package, DRAM
controller, CPU core (Power Plane 0), graphics uncore (power plane 1), etc.
The purpose of this driver is to expose the RAPL settings to userspace.
Overall, RAPL fits in the new powercap class driver in that platform
level power capping controls are exposed via this generic interface.
This driver is based on an earlier patch from Zhang Rui.
However, while the previous work was mainly focused on thermal monitoring
the focus here is on the usability from user space perspective.
References: https://lkml.org/lkml/2011/5/26/93
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Signed-off-by: Jacob Pan <jacob.jun.pan@linux.intel.com>
Reviewed-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
2013-10-17 11:28:35 -06:00
|
|
|
|
|
|
|
get_online_cpus();
|
|
|
|
rd = power_zone_to_rapl_domain(power_zone);
|
2016-05-31 14:41:29 -06:00
|
|
|
id = contraint_to_pl(rd, cid);
|
2016-11-28 14:53:11 -07:00
|
|
|
if (id < 0) {
|
|
|
|
ret = id;
|
|
|
|
goto set_exit;
|
|
|
|
}
|
2016-05-31 14:41:29 -06:00
|
|
|
|
2016-02-24 14:31:37 -07:00
|
|
|
rp = rd->rp;
|
PowerCap: Introduce Intel RAPL power capping driver
The Intel Running Average Power Limit (RAPL) technology provides platform
software with the ability to monitor, control, and get notifications on
power usage.
This feature is present in all Sandy Bridge and later Intel processors.
Newer models allow more fine grained controls to be applied. In RAPL,
power control is divided into domains, which include package, DRAM
controller, CPU core (Power Plane 0), graphics uncore (power plane 1), etc.
The purpose of this driver is to expose the RAPL settings to userspace.
Overall, RAPL fits in the new powercap class driver in that platform
level power capping controls are exposed via this generic interface.
This driver is based on an earlier patch from Zhang Rui.
However, while the previous work was mainly focused on thermal monitoring
the focus here is on the usability from user space perspective.
References: https://lkml.org/lkml/2011/5/26/93
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Signed-off-by: Jacob Pan <jacob.jun.pan@linux.intel.com>
Reviewed-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
2013-10-17 11:28:35 -06:00
|
|
|
|
|
|
|
if (rd->state & DOMAIN_STATE_BIOS_LOCKED) {
|
2019-07-10 07:44:30 -06:00
|
|
|
dev_warn(&power_zone->dev,
|
|
|
|
"%s locked by BIOS, monitoring only\n", rd->name);
|
PowerCap: Introduce Intel RAPL power capping driver
The Intel Running Average Power Limit (RAPL) technology provides platform
software with the ability to monitor, control, and get notifications on
power usage.
This feature is present in all Sandy Bridge and later Intel processors.
Newer models allow more fine grained controls to be applied. In RAPL,
power control is divided into domains, which include package, DRAM
controller, CPU core (Power Plane 0), graphics uncore (power plane 1), etc.
The purpose of this driver is to expose the RAPL settings to userspace.
Overall, RAPL fits in the new powercap class driver in that platform
level power capping controls are exposed via this generic interface.
This driver is based on an earlier patch from Zhang Rui.
However, while the previous work was mainly focused on thermal monitoring
the focus here is on the usability from user space perspective.
References: https://lkml.org/lkml/2011/5/26/93
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Signed-off-by: Jacob Pan <jacob.jun.pan@linux.intel.com>
Reviewed-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
2013-10-17 11:28:35 -06:00
|
|
|
ret = -EACCES;
|
|
|
|
goto set_exit;
|
|
|
|
}
|
|
|
|
|
|
|
|
switch (rd->rpl[id].prim_id) {
|
|
|
|
case PL1_ENABLE:
|
|
|
|
rapl_write_data_raw(rd, POWER_LIMIT1, power_limit);
|
|
|
|
break;
|
|
|
|
case PL2_ENABLE:
|
|
|
|
rapl_write_data_raw(rd, POWER_LIMIT2, power_limit);
|
|
|
|
break;
|
|
|
|
default:
|
|
|
|
ret = -EINVAL;
|
|
|
|
}
|
|
|
|
if (!ret)
|
2016-02-24 14:31:37 -07:00
|
|
|
package_power_limit_irq_save(rp);
|
PowerCap: Introduce Intel RAPL power capping driver
The Intel Running Average Power Limit (RAPL) technology provides platform
software with the ability to monitor, control, and get notifications on
power usage.
This feature is present in all Sandy Bridge and later Intel processors.
Newer models allow more fine grained controls to be applied. In RAPL,
power control is divided into domains, which include package, DRAM
controller, CPU core (Power Plane 0), graphics uncore (power plane 1), etc.
The purpose of this driver is to expose the RAPL settings to userspace.
Overall, RAPL fits in the new powercap class driver in that platform
level power capping controls are exposed via this generic interface.
This driver is based on an earlier patch from Zhang Rui.
However, while the previous work was mainly focused on thermal monitoring
the focus here is on the usability from user space perspective.
References: https://lkml.org/lkml/2011/5/26/93
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Signed-off-by: Jacob Pan <jacob.jun.pan@linux.intel.com>
Reviewed-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
2013-10-17 11:28:35 -06:00
|
|
|
set_exit:
|
|
|
|
put_online_cpus();
|
|
|
|
return ret;
|
|
|
|
}
|
|
|
|
|
2016-05-31 14:41:29 -06:00
|
|
|
static int get_current_power_limit(struct powercap_zone *power_zone, int cid,
|
2019-07-10 07:44:30 -06:00
|
|
|
u64 *data)
|
PowerCap: Introduce Intel RAPL power capping driver
The Intel Running Average Power Limit (RAPL) technology provides platform
software with the ability to monitor, control, and get notifications on
power usage.
This feature is present in all Sandy Bridge and later Intel processors.
Newer models allow more fine grained controls to be applied. In RAPL,
power control is divided into domains, which include package, DRAM
controller, CPU core (Power Plane 0), graphics uncore (power plane 1), etc.
The purpose of this driver is to expose the RAPL settings to userspace.
Overall, RAPL fits in the new powercap class driver in that platform
level power capping controls are exposed via this generic interface.
This driver is based on an earlier patch from Zhang Rui.
However, while the previous work was mainly focused on thermal monitoring
the focus here is on the usability from user space perspective.
References: https://lkml.org/lkml/2011/5/26/93
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Signed-off-by: Jacob Pan <jacob.jun.pan@linux.intel.com>
Reviewed-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
2013-10-17 11:28:35 -06:00
|
|
|
{
|
|
|
|
struct rapl_domain *rd;
|
|
|
|
u64 val;
|
|
|
|
int prim;
|
|
|
|
int ret = 0;
|
2016-05-31 14:41:29 -06:00
|
|
|
int id;
|
PowerCap: Introduce Intel RAPL power capping driver
The Intel Running Average Power Limit (RAPL) technology provides platform
software with the ability to monitor, control, and get notifications on
power usage.
This feature is present in all Sandy Bridge and later Intel processors.
Newer models allow more fine grained controls to be applied. In RAPL,
power control is divided into domains, which include package, DRAM
controller, CPU core (Power Plane 0), graphics uncore (power plane 1), etc.
The purpose of this driver is to expose the RAPL settings to userspace.
Overall, RAPL fits in the new powercap class driver in that platform
level power capping controls are exposed via this generic interface.
This driver is based on an earlier patch from Zhang Rui.
However, while the previous work was mainly focused on thermal monitoring
the focus here is on the usability from user space perspective.
References: https://lkml.org/lkml/2011/5/26/93
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Signed-off-by: Jacob Pan <jacob.jun.pan@linux.intel.com>
Reviewed-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
2013-10-17 11:28:35 -06:00
|
|
|
|
|
|
|
get_online_cpus();
|
|
|
|
rd = power_zone_to_rapl_domain(power_zone);
|
2016-05-31 14:41:29 -06:00
|
|
|
id = contraint_to_pl(rd, cid);
|
2016-11-28 14:53:11 -07:00
|
|
|
if (id < 0) {
|
|
|
|
ret = id;
|
|
|
|
goto get_exit;
|
|
|
|
}
|
|
|
|
|
PowerCap: Introduce Intel RAPL power capping driver
The Intel Running Average Power Limit (RAPL) technology provides platform
software with the ability to monitor, control, and get notifications on
power usage.
This feature is present in all Sandy Bridge and later Intel processors.
Newer models allow more fine grained controls to be applied. In RAPL,
power control is divided into domains, which include package, DRAM
controller, CPU core (Power Plane 0), graphics uncore (power plane 1), etc.
The purpose of this driver is to expose the RAPL settings to userspace.
Overall, RAPL fits in the new powercap class driver in that platform
level power capping controls are exposed via this generic interface.
This driver is based on an earlier patch from Zhang Rui.
However, while the previous work was mainly focused on thermal monitoring
the focus here is on the usability from user space perspective.
References: https://lkml.org/lkml/2011/5/26/93
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Signed-off-by: Jacob Pan <jacob.jun.pan@linux.intel.com>
Reviewed-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
2013-10-17 11:28:35 -06:00
|
|
|
switch (rd->rpl[id].prim_id) {
|
|
|
|
case PL1_ENABLE:
|
|
|
|
prim = POWER_LIMIT1;
|
|
|
|
break;
|
|
|
|
case PL2_ENABLE:
|
|
|
|
prim = POWER_LIMIT2;
|
|
|
|
break;
|
|
|
|
default:
|
|
|
|
put_online_cpus();
|
|
|
|
return -EINVAL;
|
|
|
|
}
|
|
|
|
if (rapl_read_data_raw(rd, prim, true, &val))
|
|
|
|
ret = -EIO;
|
|
|
|
else
|
|
|
|
*data = val;
|
|
|
|
|
2016-11-28 14:53:11 -07:00
|
|
|
get_exit:
|
PowerCap: Introduce Intel RAPL power capping driver
The Intel Running Average Power Limit (RAPL) technology provides platform
software with the ability to monitor, control, and get notifications on
power usage.
This feature is present in all Sandy Bridge and later Intel processors.
Newer models allow more fine grained controls to be applied. In RAPL,
power control is divided into domains, which include package, DRAM
controller, CPU core (Power Plane 0), graphics uncore (power plane 1), etc.
The purpose of this driver is to expose the RAPL settings to userspace.
Overall, RAPL fits in the new powercap class driver in that platform
level power capping controls are exposed via this generic interface.
This driver is based on an earlier patch from Zhang Rui.
However, while the previous work was mainly focused on thermal monitoring
the focus here is on the usability from user space perspective.
References: https://lkml.org/lkml/2011/5/26/93
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Signed-off-by: Jacob Pan <jacob.jun.pan@linux.intel.com>
Reviewed-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
2013-10-17 11:28:35 -06:00
|
|
|
put_online_cpus();
|
|
|
|
|
|
|
|
return ret;
|
|
|
|
}
|
|
|
|
|
2016-05-31 14:41:29 -06:00
|
|
|
static int set_time_window(struct powercap_zone *power_zone, int cid,
|
2019-07-10 07:44:30 -06:00
|
|
|
u64 window)
|
PowerCap: Introduce Intel RAPL power capping driver
The Intel Running Average Power Limit (RAPL) technology provides platform
software with the ability to monitor, control, and get notifications on
power usage.
This feature is present in all Sandy Bridge and later Intel processors.
Newer models allow more fine grained controls to be applied. In RAPL,
power control is divided into domains, which include package, DRAM
controller, CPU core (Power Plane 0), graphics uncore (power plane 1), etc.
The purpose of this driver is to expose the RAPL settings to userspace.
Overall, RAPL fits in the new powercap class driver in that platform
level power capping controls are exposed via this generic interface.
This driver is based on an earlier patch from Zhang Rui.
However, while the previous work was mainly focused on thermal monitoring
the focus here is on the usability from user space perspective.
References: https://lkml.org/lkml/2011/5/26/93
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Signed-off-by: Jacob Pan <jacob.jun.pan@linux.intel.com>
Reviewed-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
2013-10-17 11:28:35 -06:00
|
|
|
{
|
|
|
|
struct rapl_domain *rd;
|
|
|
|
int ret = 0;
|
2016-05-31 14:41:29 -06:00
|
|
|
int id;
|
PowerCap: Introduce Intel RAPL power capping driver
The Intel Running Average Power Limit (RAPL) technology provides platform
software with the ability to monitor, control, and get notifications on
power usage.
This feature is present in all Sandy Bridge and later Intel processors.
Newer models allow more fine grained controls to be applied. In RAPL,
power control is divided into domains, which include package, DRAM
controller, CPU core (Power Plane 0), graphics uncore (power plane 1), etc.
The purpose of this driver is to expose the RAPL settings to userspace.
Overall, RAPL fits in the new powercap class driver in that platform
level power capping controls are exposed via this generic interface.
This driver is based on an earlier patch from Zhang Rui.
However, while the previous work was mainly focused on thermal monitoring
the focus here is on the usability from user space perspective.
References: https://lkml.org/lkml/2011/5/26/93
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Signed-off-by: Jacob Pan <jacob.jun.pan@linux.intel.com>
Reviewed-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
2013-10-17 11:28:35 -06:00
|
|
|
|
|
|
|
get_online_cpus();
|
|
|
|
rd = power_zone_to_rapl_domain(power_zone);
|
2016-05-31 14:41:29 -06:00
|
|
|
id = contraint_to_pl(rd, cid);
|
2016-11-28 14:53:11 -07:00
|
|
|
if (id < 0) {
|
|
|
|
ret = id;
|
|
|
|
goto set_time_exit;
|
|
|
|
}
|
2016-05-31 14:41:29 -06:00
|
|
|
|
PowerCap: Introduce Intel RAPL power capping driver
The Intel Running Average Power Limit (RAPL) technology provides platform
software with the ability to monitor, control, and get notifications on
power usage.
This feature is present in all Sandy Bridge and later Intel processors.
Newer models allow more fine grained controls to be applied. In RAPL,
power control is divided into domains, which include package, DRAM
controller, CPU core (Power Plane 0), graphics uncore (power plane 1), etc.
The purpose of this driver is to expose the RAPL settings to userspace.
Overall, RAPL fits in the new powercap class driver in that platform
level power capping controls are exposed via this generic interface.
This driver is based on an earlier patch from Zhang Rui.
However, while the previous work was mainly focused on thermal monitoring
the focus here is on the usability from user space perspective.
References: https://lkml.org/lkml/2011/5/26/93
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Signed-off-by: Jacob Pan <jacob.jun.pan@linux.intel.com>
Reviewed-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
2013-10-17 11:28:35 -06:00
|
|
|
switch (rd->rpl[id].prim_id) {
|
|
|
|
case PL1_ENABLE:
|
|
|
|
rapl_write_data_raw(rd, TIME_WINDOW1, window);
|
|
|
|
break;
|
|
|
|
case PL2_ENABLE:
|
|
|
|
rapl_write_data_raw(rd, TIME_WINDOW2, window);
|
|
|
|
break;
|
|
|
|
default:
|
|
|
|
ret = -EINVAL;
|
|
|
|
}
|
2016-11-28 14:53:11 -07:00
|
|
|
|
|
|
|
set_time_exit:
|
PowerCap: Introduce Intel RAPL power capping driver
The Intel Running Average Power Limit (RAPL) technology provides platform
software with the ability to monitor, control, and get notifications on
power usage.
This feature is present in all Sandy Bridge and later Intel processors.
Newer models allow more fine grained controls to be applied. In RAPL,
power control is divided into domains, which include package, DRAM
controller, CPU core (Power Plane 0), graphics uncore (power plane 1), etc.
The purpose of this driver is to expose the RAPL settings to userspace.
Overall, RAPL fits in the new powercap class driver in that platform
level power capping controls are exposed via this generic interface.
This driver is based on an earlier patch from Zhang Rui.
However, while the previous work was mainly focused on thermal monitoring
the focus here is on the usability from user space perspective.
References: https://lkml.org/lkml/2011/5/26/93
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Signed-off-by: Jacob Pan <jacob.jun.pan@linux.intel.com>
Reviewed-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
2013-10-17 11:28:35 -06:00
|
|
|
put_online_cpus();
|
|
|
|
return ret;
|
|
|
|
}
|
|
|
|
|
2019-07-10 07:44:30 -06:00
|
|
|
static int get_time_window(struct powercap_zone *power_zone, int cid,
|
|
|
|
u64 *data)
|
PowerCap: Introduce Intel RAPL power capping driver
The Intel Running Average Power Limit (RAPL) technology provides platform
software with the ability to monitor, control, and get notifications on
power usage.
This feature is present in all Sandy Bridge and later Intel processors.
Newer models allow more fine grained controls to be applied. In RAPL,
power control is divided into domains, which include package, DRAM
controller, CPU core (Power Plane 0), graphics uncore (power plane 1), etc.
The purpose of this driver is to expose the RAPL settings to userspace.
Overall, RAPL fits in the new powercap class driver in that platform
level power capping controls are exposed via this generic interface.
This driver is based on an earlier patch from Zhang Rui.
However, while the previous work was mainly focused on thermal monitoring
the focus here is on the usability from user space perspective.
References: https://lkml.org/lkml/2011/5/26/93
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Signed-off-by: Jacob Pan <jacob.jun.pan@linux.intel.com>
Reviewed-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
2013-10-17 11:28:35 -06:00
|
|
|
{
|
|
|
|
struct rapl_domain *rd;
|
|
|
|
u64 val;
|
|
|
|
int ret = 0;
|
2016-05-31 14:41:29 -06:00
|
|
|
int id;
|
PowerCap: Introduce Intel RAPL power capping driver
The Intel Running Average Power Limit (RAPL) technology provides platform
software with the ability to monitor, control, and get notifications on
power usage.
This feature is present in all Sandy Bridge and later Intel processors.
Newer models allow more fine grained controls to be applied. In RAPL,
power control is divided into domains, which include package, DRAM
controller, CPU core (Power Plane 0), graphics uncore (power plane 1), etc.
The purpose of this driver is to expose the RAPL settings to userspace.
Overall, RAPL fits in the new powercap class driver in that platform
level power capping controls are exposed via this generic interface.
This driver is based on an earlier patch from Zhang Rui.
However, while the previous work was mainly focused on thermal monitoring
the focus here is on the usability from user space perspective.
References: https://lkml.org/lkml/2011/5/26/93
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Signed-off-by: Jacob Pan <jacob.jun.pan@linux.intel.com>
Reviewed-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
2013-10-17 11:28:35 -06:00
|
|
|
|
|
|
|
get_online_cpus();
|
|
|
|
rd = power_zone_to_rapl_domain(power_zone);
|
2016-05-31 14:41:29 -06:00
|
|
|
id = contraint_to_pl(rd, cid);
|
2016-11-28 14:53:11 -07:00
|
|
|
if (id < 0) {
|
|
|
|
ret = id;
|
|
|
|
goto get_time_exit;
|
|
|
|
}
|
2016-05-31 14:41:29 -06:00
|
|
|
|
PowerCap: Introduce Intel RAPL power capping driver
The Intel Running Average Power Limit (RAPL) technology provides platform
software with the ability to monitor, control, and get notifications on
power usage.
This feature is present in all Sandy Bridge and later Intel processors.
Newer models allow more fine grained controls to be applied. In RAPL,
power control is divided into domains, which include package, DRAM
controller, CPU core (Power Plane 0), graphics uncore (power plane 1), etc.
The purpose of this driver is to expose the RAPL settings to userspace.
Overall, RAPL fits in the new powercap class driver in that platform
level power capping controls are exposed via this generic interface.
This driver is based on an earlier patch from Zhang Rui.
However, while the previous work was mainly focused on thermal monitoring
the focus here is on the usability from user space perspective.
References: https://lkml.org/lkml/2011/5/26/93
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Signed-off-by: Jacob Pan <jacob.jun.pan@linux.intel.com>
Reviewed-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
2013-10-17 11:28:35 -06:00
|
|
|
switch (rd->rpl[id].prim_id) {
|
|
|
|
case PL1_ENABLE:
|
|
|
|
ret = rapl_read_data_raw(rd, TIME_WINDOW1, true, &val);
|
|
|
|
break;
|
|
|
|
case PL2_ENABLE:
|
|
|
|
ret = rapl_read_data_raw(rd, TIME_WINDOW2, true, &val);
|
|
|
|
break;
|
|
|
|
default:
|
|
|
|
put_online_cpus();
|
|
|
|
return -EINVAL;
|
|
|
|
}
|
|
|
|
if (!ret)
|
|
|
|
*data = val;
|
2016-11-28 14:53:11 -07:00
|
|
|
|
|
|
|
get_time_exit:
|
PowerCap: Introduce Intel RAPL power capping driver
The Intel Running Average Power Limit (RAPL) technology provides platform
software with the ability to monitor, control, and get notifications on
power usage.
This feature is present in all Sandy Bridge and later Intel processors.
Newer models allow more fine grained controls to be applied. In RAPL,
power control is divided into domains, which include package, DRAM
controller, CPU core (Power Plane 0), graphics uncore (power plane 1), etc.
The purpose of this driver is to expose the RAPL settings to userspace.
Overall, RAPL fits in the new powercap class driver in that platform
level power capping controls are exposed via this generic interface.
This driver is based on an earlier patch from Zhang Rui.
However, while the previous work was mainly focused on thermal monitoring
the focus here is on the usability from user space perspective.
References: https://lkml.org/lkml/2011/5/26/93
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Signed-off-by: Jacob Pan <jacob.jun.pan@linux.intel.com>
Reviewed-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
2013-10-17 11:28:35 -06:00
|
|
|
put_online_cpus();
|
|
|
|
|
|
|
|
return ret;
|
|
|
|
}
|
|
|
|
|
2019-07-10 07:44:30 -06:00
|
|
|
static const char *get_constraint_name(struct powercap_zone *power_zone,
|
|
|
|
int cid)
|
PowerCap: Introduce Intel RAPL power capping driver
The Intel Running Average Power Limit (RAPL) technology provides platform
software with the ability to monitor, control, and get notifications on
power usage.
This feature is present in all Sandy Bridge and later Intel processors.
Newer models allow more fine grained controls to be applied. In RAPL,
power control is divided into domains, which include package, DRAM
controller, CPU core (Power Plane 0), graphics uncore (power plane 1), etc.
The purpose of this driver is to expose the RAPL settings to userspace.
Overall, RAPL fits in the new powercap class driver in that platform
level power capping controls are exposed via this generic interface.
This driver is based on an earlier patch from Zhang Rui.
However, while the previous work was mainly focused on thermal monitoring
the focus here is on the usability from user space perspective.
References: https://lkml.org/lkml/2011/5/26/93
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Signed-off-by: Jacob Pan <jacob.jun.pan@linux.intel.com>
Reviewed-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
2013-10-17 11:28:35 -06:00
|
|
|
{
|
|
|
|
struct rapl_domain *rd;
|
2016-05-31 14:41:29 -06:00
|
|
|
int id;
|
PowerCap: Introduce Intel RAPL power capping driver
The Intel Running Average Power Limit (RAPL) technology provides platform
software with the ability to monitor, control, and get notifications on
power usage.
This feature is present in all Sandy Bridge and later Intel processors.
Newer models allow more fine grained controls to be applied. In RAPL,
power control is divided into domains, which include package, DRAM
controller, CPU core (Power Plane 0), graphics uncore (power plane 1), etc.
The purpose of this driver is to expose the RAPL settings to userspace.
Overall, RAPL fits in the new powercap class driver in that platform
level power capping controls are exposed via this generic interface.
This driver is based on an earlier patch from Zhang Rui.
However, while the previous work was mainly focused on thermal monitoring
the focus here is on the usability from user space perspective.
References: https://lkml.org/lkml/2011/5/26/93
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Signed-off-by: Jacob Pan <jacob.jun.pan@linux.intel.com>
Reviewed-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
2013-10-17 11:28:35 -06:00
|
|
|
|
|
|
|
rd = power_zone_to_rapl_domain(power_zone);
|
2016-05-31 14:41:29 -06:00
|
|
|
id = contraint_to_pl(rd, cid);
|
|
|
|
if (id >= 0)
|
|
|
|
return rd->rpl[id].name;
|
PowerCap: Introduce Intel RAPL power capping driver
The Intel Running Average Power Limit (RAPL) technology provides platform
software with the ability to monitor, control, and get notifications on
power usage.
This feature is present in all Sandy Bridge and later Intel processors.
Newer models allow more fine grained controls to be applied. In RAPL,
power control is divided into domains, which include package, DRAM
controller, CPU core (Power Plane 0), graphics uncore (power plane 1), etc.
The purpose of this driver is to expose the RAPL settings to userspace.
Overall, RAPL fits in the new powercap class driver in that platform
level power capping controls are exposed via this generic interface.
This driver is based on an earlier patch from Zhang Rui.
However, while the previous work was mainly focused on thermal monitoring
the focus here is on the usability from user space perspective.
References: https://lkml.org/lkml/2011/5/26/93
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Signed-off-by: Jacob Pan <jacob.jun.pan@linux.intel.com>
Reviewed-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
2013-10-17 11:28:35 -06:00
|
|
|
|
2016-05-31 14:41:29 -06:00
|
|
|
return NULL;
|
PowerCap: Introduce Intel RAPL power capping driver
The Intel Running Average Power Limit (RAPL) technology provides platform
software with the ability to monitor, control, and get notifications on
power usage.
This feature is present in all Sandy Bridge and later Intel processors.
Newer models allow more fine grained controls to be applied. In RAPL,
power control is divided into domains, which include package, DRAM
controller, CPU core (Power Plane 0), graphics uncore (power plane 1), etc.
The purpose of this driver is to expose the RAPL settings to userspace.
Overall, RAPL fits in the new powercap class driver in that platform
level power capping controls are exposed via this generic interface.
This driver is based on an earlier patch from Zhang Rui.
However, while the previous work was mainly focused on thermal monitoring
the focus here is on the usability from user space perspective.
References: https://lkml.org/lkml/2011/5/26/93
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Signed-off-by: Jacob Pan <jacob.jun.pan@linux.intel.com>
Reviewed-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
2013-10-17 11:28:35 -06:00
|
|
|
}
|
|
|
|
|
2019-07-10 07:44:30 -06:00
|
|
|
static int get_max_power(struct powercap_zone *power_zone, int id, u64 *data)
|
PowerCap: Introduce Intel RAPL power capping driver
The Intel Running Average Power Limit (RAPL) technology provides platform
software with the ability to monitor, control, and get notifications on
power usage.
This feature is present in all Sandy Bridge and later Intel processors.
Newer models allow more fine grained controls to be applied. In RAPL,
power control is divided into domains, which include package, DRAM
controller, CPU core (Power Plane 0), graphics uncore (power plane 1), etc.
The purpose of this driver is to expose the RAPL settings to userspace.
Overall, RAPL fits in the new powercap class driver in that platform
level power capping controls are exposed via this generic interface.
This driver is based on an earlier patch from Zhang Rui.
However, while the previous work was mainly focused on thermal monitoring
the focus here is on the usability from user space perspective.
References: https://lkml.org/lkml/2011/5/26/93
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Signed-off-by: Jacob Pan <jacob.jun.pan@linux.intel.com>
Reviewed-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
2013-10-17 11:28:35 -06:00
|
|
|
{
|
|
|
|
struct rapl_domain *rd;
|
|
|
|
u64 val;
|
|
|
|
int prim;
|
|
|
|
int ret = 0;
|
|
|
|
|
|
|
|
get_online_cpus();
|
|
|
|
rd = power_zone_to_rapl_domain(power_zone);
|
|
|
|
switch (rd->rpl[id].prim_id) {
|
|
|
|
case PL1_ENABLE:
|
|
|
|
prim = THERMAL_SPEC_POWER;
|
|
|
|
break;
|
|
|
|
case PL2_ENABLE:
|
|
|
|
prim = MAX_POWER;
|
|
|
|
break;
|
|
|
|
default:
|
|
|
|
put_online_cpus();
|
|
|
|
return -EINVAL;
|
|
|
|
}
|
|
|
|
if (rapl_read_data_raw(rd, prim, true, &val))
|
|
|
|
ret = -EIO;
|
|
|
|
else
|
|
|
|
*data = val;
|
|
|
|
|
|
|
|
put_online_cpus();
|
|
|
|
|
|
|
|
return ret;
|
|
|
|
}
|
|
|
|
|
2015-12-23 14:59:55 -07:00
|
|
|
static const struct powercap_zone_constraint_ops constraint_ops = {
|
PowerCap: Introduce Intel RAPL power capping driver
The Intel Running Average Power Limit (RAPL) technology provides platform
software with the ability to monitor, control, and get notifications on
power usage.
This feature is present in all Sandy Bridge and later Intel processors.
Newer models allow more fine grained controls to be applied. In RAPL,
power control is divided into domains, which include package, DRAM
controller, CPU core (Power Plane 0), graphics uncore (power plane 1), etc.
The purpose of this driver is to expose the RAPL settings to userspace.
Overall, RAPL fits in the new powercap class driver in that platform
level power capping controls are exposed via this generic interface.
This driver is based on an earlier patch from Zhang Rui.
However, while the previous work was mainly focused on thermal monitoring
the focus here is on the usability from user space perspective.
References: https://lkml.org/lkml/2011/5/26/93
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Signed-off-by: Jacob Pan <jacob.jun.pan@linux.intel.com>
Reviewed-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
2013-10-17 11:28:35 -06:00
|
|
|
.set_power_limit_uw = set_power_limit,
|
|
|
|
.get_power_limit_uw = get_current_power_limit,
|
|
|
|
.set_time_window_us = set_time_window,
|
|
|
|
.get_time_window_us = get_time_window,
|
|
|
|
.get_max_power_uw = get_max_power,
|
|
|
|
.get_name = get_constraint_name,
|
|
|
|
};
|
|
|
|
|
|
|
|
/* called after domain detection and package level data are set */
|
|
|
|
static void rapl_init_domains(struct rapl_package *rp)
|
|
|
|
{
|
2019-07-10 07:44:32 -06:00
|
|
|
enum rapl_domain_type i;
|
|
|
|
enum rapl_domain_reg_id j;
|
PowerCap: Introduce Intel RAPL power capping driver
The Intel Running Average Power Limit (RAPL) technology provides platform
software with the ability to monitor, control, and get notifications on
power usage.
This feature is present in all Sandy Bridge and later Intel processors.
Newer models allow more fine grained controls to be applied. In RAPL,
power control is divided into domains, which include package, DRAM
controller, CPU core (Power Plane 0), graphics uncore (power plane 1), etc.
The purpose of this driver is to expose the RAPL settings to userspace.
Overall, RAPL fits in the new powercap class driver in that platform
level power capping controls are exposed via this generic interface.
This driver is based on an earlier patch from Zhang Rui.
However, while the previous work was mainly focused on thermal monitoring
the focus here is on the usability from user space perspective.
References: https://lkml.org/lkml/2011/5/26/93
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Signed-off-by: Jacob Pan <jacob.jun.pan@linux.intel.com>
Reviewed-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
2013-10-17 11:28:35 -06:00
|
|
|
struct rapl_domain *rd = rp->domains;
|
|
|
|
|
|
|
|
for (i = 0; i < RAPL_DOMAIN_MAX; i++) {
|
|
|
|
unsigned int mask = rp->domain_map & (1 << i);
|
2019-07-10 07:44:26 -06:00
|
|
|
|
2019-07-10 07:44:32 -06:00
|
|
|
if (!mask)
|
|
|
|
continue;
|
|
|
|
|
|
|
|
rd->rp = rp;
|
|
|
|
rd->name = rapl_domain_names[i];
|
|
|
|
rd->id = i;
|
|
|
|
rd->rpl[0].prim_id = PL1_ENABLE;
|
|
|
|
rd->rpl[0].name = pl1_name;
|
|
|
|
/* some domain may support two power limits */
|
|
|
|
if (rp->priv->limits[i] == 2) {
|
PowerCap: Introduce Intel RAPL power capping driver
The Intel Running Average Power Limit (RAPL) technology provides platform
software with the ability to monitor, control, and get notifications on
power usage.
This feature is present in all Sandy Bridge and later Intel processors.
Newer models allow more fine grained controls to be applied. In RAPL,
power control is divided into domains, which include package, DRAM
controller, CPU core (Power Plane 0), graphics uncore (power plane 1), etc.
The purpose of this driver is to expose the RAPL settings to userspace.
Overall, RAPL fits in the new powercap class driver in that platform
level power capping controls are exposed via this generic interface.
This driver is based on an earlier patch from Zhang Rui.
However, while the previous work was mainly focused on thermal monitoring
the focus here is on the usability from user space perspective.
References: https://lkml.org/lkml/2011/5/26/93
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Signed-off-by: Jacob Pan <jacob.jun.pan@linux.intel.com>
Reviewed-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
2013-10-17 11:28:35 -06:00
|
|
|
rd->rpl[1].prim_id = PL2_ENABLE;
|
|
|
|
rd->rpl[1].name = pl2_name;
|
2019-07-10 07:44:32 -06:00
|
|
|
}
|
|
|
|
|
|
|
|
for (j = 0; j < RAPL_DOMAIN_REG_MAX; j++)
|
|
|
|
rd->regs[j] = rp->priv->regs[i][j];
|
|
|
|
|
|
|
|
if (i == RAPL_DOMAIN_DRAM) {
|
2015-03-13 04:48:56 -06:00
|
|
|
rd->domain_energy_unit =
|
2019-07-10 07:44:30 -06:00
|
|
|
rapl_defaults->dram_domain_energy_unit;
|
2015-03-13 04:48:56 -06:00
|
|
|
if (rd->domain_energy_unit)
|
|
|
|
pr_info("DRAM domain energy unit %dpj\n",
|
|
|
|
rd->domain_energy_unit);
|
PowerCap: Introduce Intel RAPL power capping driver
The Intel Running Average Power Limit (RAPL) technology provides platform
software with the ability to monitor, control, and get notifications on
power usage.
This feature is present in all Sandy Bridge and later Intel processors.
Newer models allow more fine grained controls to be applied. In RAPL,
power control is divided into domains, which include package, DRAM
controller, CPU core (Power Plane 0), graphics uncore (power plane 1), etc.
The purpose of this driver is to expose the RAPL settings to userspace.
Overall, RAPL fits in the new powercap class driver in that platform
level power capping controls are exposed via this generic interface.
This driver is based on an earlier patch from Zhang Rui.
However, while the previous work was mainly focused on thermal monitoring
the focus here is on the usability from user space perspective.
References: https://lkml.org/lkml/2011/5/26/93
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Signed-off-by: Jacob Pan <jacob.jun.pan@linux.intel.com>
Reviewed-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
2013-10-17 11:28:35 -06:00
|
|
|
}
|
2019-07-10 07:44:32 -06:00
|
|
|
rd++;
|
PowerCap: Introduce Intel RAPL power capping driver
The Intel Running Average Power Limit (RAPL) technology provides platform
software with the ability to monitor, control, and get notifications on
power usage.
This feature is present in all Sandy Bridge and later Intel processors.
Newer models allow more fine grained controls to be applied. In RAPL,
power control is divided into domains, which include package, DRAM
controller, CPU core (Power Plane 0), graphics uncore (power plane 1), etc.
The purpose of this driver is to expose the RAPL settings to userspace.
Overall, RAPL fits in the new powercap class driver in that platform
level power capping controls are exposed via this generic interface.
This driver is based on an earlier patch from Zhang Rui.
However, while the previous work was mainly focused on thermal monitoring
the focus here is on the usability from user space perspective.
References: https://lkml.org/lkml/2011/5/26/93
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Signed-off-by: Jacob Pan <jacob.jun.pan@linux.intel.com>
Reviewed-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
2013-10-17 11:28:35 -06:00
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2016-02-24 14:31:37 -07:00
|
|
|
static u64 rapl_unit_xlate(struct rapl_domain *rd, enum unit_type type,
|
2019-07-10 07:44:30 -06:00
|
|
|
u64 value, int to_raw)
|
PowerCap: Introduce Intel RAPL power capping driver
The Intel Running Average Power Limit (RAPL) technology provides platform
software with the ability to monitor, control, and get notifications on
power usage.
This feature is present in all Sandy Bridge and later Intel processors.
Newer models allow more fine grained controls to be applied. In RAPL,
power control is divided into domains, which include package, DRAM
controller, CPU core (Power Plane 0), graphics uncore (power plane 1), etc.
The purpose of this driver is to expose the RAPL settings to userspace.
Overall, RAPL fits in the new powercap class driver in that platform
level power capping controls are exposed via this generic interface.
This driver is based on an earlier patch from Zhang Rui.
However, while the previous work was mainly focused on thermal monitoring
the focus here is on the usability from user space perspective.
References: https://lkml.org/lkml/2011/5/26/93
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Signed-off-by: Jacob Pan <jacob.jun.pan@linux.intel.com>
Reviewed-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
2013-10-17 11:28:35 -06:00
|
|
|
{
|
2014-11-07 10:29:26 -07:00
|
|
|
u64 units = 1;
|
2016-02-24 14:31:37 -07:00
|
|
|
struct rapl_package *rp = rd->rp;
|
2015-03-13 04:48:56 -06:00
|
|
|
u64 scale = 1;
|
PowerCap: Introduce Intel RAPL power capping driver
The Intel Running Average Power Limit (RAPL) technology provides platform
software with the ability to monitor, control, and get notifications on
power usage.
This feature is present in all Sandy Bridge and later Intel processors.
Newer models allow more fine grained controls to be applied. In RAPL,
power control is divided into domains, which include package, DRAM
controller, CPU core (Power Plane 0), graphics uncore (power plane 1), etc.
The purpose of this driver is to expose the RAPL settings to userspace.
Overall, RAPL fits in the new powercap class driver in that platform
level power capping controls are exposed via this generic interface.
This driver is based on an earlier patch from Zhang Rui.
However, while the previous work was mainly focused on thermal monitoring
the focus here is on the usability from user space perspective.
References: https://lkml.org/lkml/2011/5/26/93
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Signed-off-by: Jacob Pan <jacob.jun.pan@linux.intel.com>
Reviewed-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
2013-10-17 11:28:35 -06:00
|
|
|
|
|
|
|
switch (type) {
|
|
|
|
case POWER_UNIT:
|
2014-11-07 10:29:26 -07:00
|
|
|
units = rp->power_unit;
|
PowerCap: Introduce Intel RAPL power capping driver
The Intel Running Average Power Limit (RAPL) technology provides platform
software with the ability to monitor, control, and get notifications on
power usage.
This feature is present in all Sandy Bridge and later Intel processors.
Newer models allow more fine grained controls to be applied. In RAPL,
power control is divided into domains, which include package, DRAM
controller, CPU core (Power Plane 0), graphics uncore (power plane 1), etc.
The purpose of this driver is to expose the RAPL settings to userspace.
Overall, RAPL fits in the new powercap class driver in that platform
level power capping controls are exposed via this generic interface.
This driver is based on an earlier patch from Zhang Rui.
However, while the previous work was mainly focused on thermal monitoring
the focus here is on the usability from user space perspective.
References: https://lkml.org/lkml/2011/5/26/93
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Signed-off-by: Jacob Pan <jacob.jun.pan@linux.intel.com>
Reviewed-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
2013-10-17 11:28:35 -06:00
|
|
|
break;
|
|
|
|
case ENERGY_UNIT:
|
2015-03-13 04:48:56 -06:00
|
|
|
scale = ENERGY_UNIT_SCALE;
|
|
|
|
/* per domain unit takes precedence */
|
2016-11-28 14:53:11 -07:00
|
|
|
if (rd->domain_energy_unit)
|
2015-03-13 04:48:56 -06:00
|
|
|
units = rd->domain_energy_unit;
|
|
|
|
else
|
|
|
|
units = rp->energy_unit;
|
PowerCap: Introduce Intel RAPL power capping driver
The Intel Running Average Power Limit (RAPL) technology provides platform
software with the ability to monitor, control, and get notifications on
power usage.
This feature is present in all Sandy Bridge and later Intel processors.
Newer models allow more fine grained controls to be applied. In RAPL,
power control is divided into domains, which include package, DRAM
controller, CPU core (Power Plane 0), graphics uncore (power plane 1), etc.
The purpose of this driver is to expose the RAPL settings to userspace.
Overall, RAPL fits in the new powercap class driver in that platform
level power capping controls are exposed via this generic interface.
This driver is based on an earlier patch from Zhang Rui.
However, while the previous work was mainly focused on thermal monitoring
the focus here is on the usability from user space perspective.
References: https://lkml.org/lkml/2011/5/26/93
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Signed-off-by: Jacob Pan <jacob.jun.pan@linux.intel.com>
Reviewed-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
2013-10-17 11:28:35 -06:00
|
|
|
break;
|
|
|
|
case TIME_UNIT:
|
2014-11-07 10:29:26 -07:00
|
|
|
return rapl_defaults->compute_time_window(rp, value, to_raw);
|
PowerCap: Introduce Intel RAPL power capping driver
The Intel Running Average Power Limit (RAPL) technology provides platform
software with the ability to monitor, control, and get notifications on
power usage.
This feature is present in all Sandy Bridge and later Intel processors.
Newer models allow more fine grained controls to be applied. In RAPL,
power control is divided into domains, which include package, DRAM
controller, CPU core (Power Plane 0), graphics uncore (power plane 1), etc.
The purpose of this driver is to expose the RAPL settings to userspace.
Overall, RAPL fits in the new powercap class driver in that platform
level power capping controls are exposed via this generic interface.
This driver is based on an earlier patch from Zhang Rui.
However, while the previous work was mainly focused on thermal monitoring
the focus here is on the usability from user space perspective.
References: https://lkml.org/lkml/2011/5/26/93
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Signed-off-by: Jacob Pan <jacob.jun.pan@linux.intel.com>
Reviewed-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
2013-10-17 11:28:35 -06:00
|
|
|
case ARBITRARY_UNIT:
|
|
|
|
default:
|
|
|
|
return value;
|
|
|
|
};
|
|
|
|
|
|
|
|
if (to_raw)
|
2015-03-13 04:48:56 -06:00
|
|
|
return div64_u64(value, units) * scale;
|
2014-11-07 10:29:26 -07:00
|
|
|
|
|
|
|
value *= units;
|
|
|
|
|
2015-03-13 04:48:56 -06:00
|
|
|
return div64_u64(value, scale);
|
PowerCap: Introduce Intel RAPL power capping driver
The Intel Running Average Power Limit (RAPL) technology provides platform
software with the ability to monitor, control, and get notifications on
power usage.
This feature is present in all Sandy Bridge and later Intel processors.
Newer models allow more fine grained controls to be applied. In RAPL,
power control is divided into domains, which include package, DRAM
controller, CPU core (Power Plane 0), graphics uncore (power plane 1), etc.
The purpose of this driver is to expose the RAPL settings to userspace.
Overall, RAPL fits in the new powercap class driver in that platform
level power capping controls are exposed via this generic interface.
This driver is based on an earlier patch from Zhang Rui.
However, while the previous work was mainly focused on thermal monitoring
the focus here is on the usability from user space perspective.
References: https://lkml.org/lkml/2011/5/26/93
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Signed-off-by: Jacob Pan <jacob.jun.pan@linux.intel.com>
Reviewed-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
2013-10-17 11:28:35 -06:00
|
|
|
}
|
|
|
|
|
|
|
|
/* in the order of enum rapl_primitives */
|
|
|
|
static struct rapl_primitive_info rpi[] = {
|
|
|
|
/* name, mask, shift, msr index, unit divisor */
|
|
|
|
PRIMITIVE_INFO_INIT(ENERGY_COUNTER, ENERGY_STATUS_MASK, 0,
|
2019-07-10 07:44:30 -06:00
|
|
|
RAPL_DOMAIN_REG_STATUS, ENERGY_UNIT, 0),
|
PowerCap: Introduce Intel RAPL power capping driver
The Intel Running Average Power Limit (RAPL) technology provides platform
software with the ability to monitor, control, and get notifications on
power usage.
This feature is present in all Sandy Bridge and later Intel processors.
Newer models allow more fine grained controls to be applied. In RAPL,
power control is divided into domains, which include package, DRAM
controller, CPU core (Power Plane 0), graphics uncore (power plane 1), etc.
The purpose of this driver is to expose the RAPL settings to userspace.
Overall, RAPL fits in the new powercap class driver in that platform
level power capping controls are exposed via this generic interface.
This driver is based on an earlier patch from Zhang Rui.
However, while the previous work was mainly focused on thermal monitoring
the focus here is on the usability from user space perspective.
References: https://lkml.org/lkml/2011/5/26/93
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Signed-off-by: Jacob Pan <jacob.jun.pan@linux.intel.com>
Reviewed-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
2013-10-17 11:28:35 -06:00
|
|
|
PRIMITIVE_INFO_INIT(POWER_LIMIT1, POWER_LIMIT1_MASK, 0,
|
2019-07-10 07:44:30 -06:00
|
|
|
RAPL_DOMAIN_REG_LIMIT, POWER_UNIT, 0),
|
PowerCap: Introduce Intel RAPL power capping driver
The Intel Running Average Power Limit (RAPL) technology provides platform
software with the ability to monitor, control, and get notifications on
power usage.
This feature is present in all Sandy Bridge and later Intel processors.
Newer models allow more fine grained controls to be applied. In RAPL,
power control is divided into domains, which include package, DRAM
controller, CPU core (Power Plane 0), graphics uncore (power plane 1), etc.
The purpose of this driver is to expose the RAPL settings to userspace.
Overall, RAPL fits in the new powercap class driver in that platform
level power capping controls are exposed via this generic interface.
This driver is based on an earlier patch from Zhang Rui.
However, while the previous work was mainly focused on thermal monitoring
the focus here is on the usability from user space perspective.
References: https://lkml.org/lkml/2011/5/26/93
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Signed-off-by: Jacob Pan <jacob.jun.pan@linux.intel.com>
Reviewed-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
2013-10-17 11:28:35 -06:00
|
|
|
PRIMITIVE_INFO_INIT(POWER_LIMIT2, POWER_LIMIT2_MASK, 32,
|
2019-07-10 07:44:30 -06:00
|
|
|
RAPL_DOMAIN_REG_LIMIT, POWER_UNIT, 0),
|
2019-07-10 07:44:32 -06:00
|
|
|
PRIMITIVE_INFO_INIT(FW_LOCK, POWER_LOW_LOCK, 31,
|
2019-07-10 07:44:30 -06:00
|
|
|
RAPL_DOMAIN_REG_LIMIT, ARBITRARY_UNIT, 0),
|
PowerCap: Introduce Intel RAPL power capping driver
The Intel Running Average Power Limit (RAPL) technology provides platform
software with the ability to monitor, control, and get notifications on
power usage.
This feature is present in all Sandy Bridge and later Intel processors.
Newer models allow more fine grained controls to be applied. In RAPL,
power control is divided into domains, which include package, DRAM
controller, CPU core (Power Plane 0), graphics uncore (power plane 1), etc.
The purpose of this driver is to expose the RAPL settings to userspace.
Overall, RAPL fits in the new powercap class driver in that platform
level power capping controls are exposed via this generic interface.
This driver is based on an earlier patch from Zhang Rui.
However, while the previous work was mainly focused on thermal monitoring
the focus here is on the usability from user space perspective.
References: https://lkml.org/lkml/2011/5/26/93
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Signed-off-by: Jacob Pan <jacob.jun.pan@linux.intel.com>
Reviewed-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
2013-10-17 11:28:35 -06:00
|
|
|
PRIMITIVE_INFO_INIT(PL1_ENABLE, POWER_LIMIT1_ENABLE, 15,
|
2019-07-10 07:44:30 -06:00
|
|
|
RAPL_DOMAIN_REG_LIMIT, ARBITRARY_UNIT, 0),
|
PowerCap: Introduce Intel RAPL power capping driver
The Intel Running Average Power Limit (RAPL) technology provides platform
software with the ability to monitor, control, and get notifications on
power usage.
This feature is present in all Sandy Bridge and later Intel processors.
Newer models allow more fine grained controls to be applied. In RAPL,
power control is divided into domains, which include package, DRAM
controller, CPU core (Power Plane 0), graphics uncore (power plane 1), etc.
The purpose of this driver is to expose the RAPL settings to userspace.
Overall, RAPL fits in the new powercap class driver in that platform
level power capping controls are exposed via this generic interface.
This driver is based on an earlier patch from Zhang Rui.
However, while the previous work was mainly focused on thermal monitoring
the focus here is on the usability from user space perspective.
References: https://lkml.org/lkml/2011/5/26/93
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Signed-off-by: Jacob Pan <jacob.jun.pan@linux.intel.com>
Reviewed-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
2013-10-17 11:28:35 -06:00
|
|
|
PRIMITIVE_INFO_INIT(PL1_CLAMP, POWER_LIMIT1_CLAMP, 16,
|
2019-07-10 07:44:30 -06:00
|
|
|
RAPL_DOMAIN_REG_LIMIT, ARBITRARY_UNIT, 0),
|
PowerCap: Introduce Intel RAPL power capping driver
The Intel Running Average Power Limit (RAPL) technology provides platform
software with the ability to monitor, control, and get notifications on
power usage.
This feature is present in all Sandy Bridge and later Intel processors.
Newer models allow more fine grained controls to be applied. In RAPL,
power control is divided into domains, which include package, DRAM
controller, CPU core (Power Plane 0), graphics uncore (power plane 1), etc.
The purpose of this driver is to expose the RAPL settings to userspace.
Overall, RAPL fits in the new powercap class driver in that platform
level power capping controls are exposed via this generic interface.
This driver is based on an earlier patch from Zhang Rui.
However, while the previous work was mainly focused on thermal monitoring
the focus here is on the usability from user space perspective.
References: https://lkml.org/lkml/2011/5/26/93
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Signed-off-by: Jacob Pan <jacob.jun.pan@linux.intel.com>
Reviewed-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
2013-10-17 11:28:35 -06:00
|
|
|
PRIMITIVE_INFO_INIT(PL2_ENABLE, POWER_LIMIT2_ENABLE, 47,
|
2019-07-10 07:44:30 -06:00
|
|
|
RAPL_DOMAIN_REG_LIMIT, ARBITRARY_UNIT, 0),
|
PowerCap: Introduce Intel RAPL power capping driver
The Intel Running Average Power Limit (RAPL) technology provides platform
software with the ability to monitor, control, and get notifications on
power usage.
This feature is present in all Sandy Bridge and later Intel processors.
Newer models allow more fine grained controls to be applied. In RAPL,
power control is divided into domains, which include package, DRAM
controller, CPU core (Power Plane 0), graphics uncore (power plane 1), etc.
The purpose of this driver is to expose the RAPL settings to userspace.
Overall, RAPL fits in the new powercap class driver in that platform
level power capping controls are exposed via this generic interface.
This driver is based on an earlier patch from Zhang Rui.
However, while the previous work was mainly focused on thermal monitoring
the focus here is on the usability from user space perspective.
References: https://lkml.org/lkml/2011/5/26/93
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Signed-off-by: Jacob Pan <jacob.jun.pan@linux.intel.com>
Reviewed-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
2013-10-17 11:28:35 -06:00
|
|
|
PRIMITIVE_INFO_INIT(PL2_CLAMP, POWER_LIMIT2_CLAMP, 48,
|
2019-07-10 07:44:30 -06:00
|
|
|
RAPL_DOMAIN_REG_LIMIT, ARBITRARY_UNIT, 0),
|
PowerCap: Introduce Intel RAPL power capping driver
The Intel Running Average Power Limit (RAPL) technology provides platform
software with the ability to monitor, control, and get notifications on
power usage.
This feature is present in all Sandy Bridge and later Intel processors.
Newer models allow more fine grained controls to be applied. In RAPL,
power control is divided into domains, which include package, DRAM
controller, CPU core (Power Plane 0), graphics uncore (power plane 1), etc.
The purpose of this driver is to expose the RAPL settings to userspace.
Overall, RAPL fits in the new powercap class driver in that platform
level power capping controls are exposed via this generic interface.
This driver is based on an earlier patch from Zhang Rui.
However, while the previous work was mainly focused on thermal monitoring
the focus here is on the usability from user space perspective.
References: https://lkml.org/lkml/2011/5/26/93
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Signed-off-by: Jacob Pan <jacob.jun.pan@linux.intel.com>
Reviewed-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
2013-10-17 11:28:35 -06:00
|
|
|
PRIMITIVE_INFO_INIT(TIME_WINDOW1, TIME_WINDOW1_MASK, 17,
|
2019-07-10 07:44:30 -06:00
|
|
|
RAPL_DOMAIN_REG_LIMIT, TIME_UNIT, 0),
|
PowerCap: Introduce Intel RAPL power capping driver
The Intel Running Average Power Limit (RAPL) technology provides platform
software with the ability to monitor, control, and get notifications on
power usage.
This feature is present in all Sandy Bridge and later Intel processors.
Newer models allow more fine grained controls to be applied. In RAPL,
power control is divided into domains, which include package, DRAM
controller, CPU core (Power Plane 0), graphics uncore (power plane 1), etc.
The purpose of this driver is to expose the RAPL settings to userspace.
Overall, RAPL fits in the new powercap class driver in that platform
level power capping controls are exposed via this generic interface.
This driver is based on an earlier patch from Zhang Rui.
However, while the previous work was mainly focused on thermal monitoring
the focus here is on the usability from user space perspective.
References: https://lkml.org/lkml/2011/5/26/93
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Signed-off-by: Jacob Pan <jacob.jun.pan@linux.intel.com>
Reviewed-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
2013-10-17 11:28:35 -06:00
|
|
|
PRIMITIVE_INFO_INIT(TIME_WINDOW2, TIME_WINDOW2_MASK, 49,
|
2019-07-10 07:44:30 -06:00
|
|
|
RAPL_DOMAIN_REG_LIMIT, TIME_UNIT, 0),
|
PowerCap: Introduce Intel RAPL power capping driver
The Intel Running Average Power Limit (RAPL) technology provides platform
software with the ability to monitor, control, and get notifications on
power usage.
This feature is present in all Sandy Bridge and later Intel processors.
Newer models allow more fine grained controls to be applied. In RAPL,
power control is divided into domains, which include package, DRAM
controller, CPU core (Power Plane 0), graphics uncore (power plane 1), etc.
The purpose of this driver is to expose the RAPL settings to userspace.
Overall, RAPL fits in the new powercap class driver in that platform
level power capping controls are exposed via this generic interface.
This driver is based on an earlier patch from Zhang Rui.
However, while the previous work was mainly focused on thermal monitoring
the focus here is on the usability from user space perspective.
References: https://lkml.org/lkml/2011/5/26/93
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Signed-off-by: Jacob Pan <jacob.jun.pan@linux.intel.com>
Reviewed-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
2013-10-17 11:28:35 -06:00
|
|
|
PRIMITIVE_INFO_INIT(THERMAL_SPEC_POWER, POWER_INFO_THERMAL_SPEC_MASK,
|
2019-07-10 07:44:30 -06:00
|
|
|
0, RAPL_DOMAIN_REG_INFO, POWER_UNIT, 0),
|
PowerCap: Introduce Intel RAPL power capping driver
The Intel Running Average Power Limit (RAPL) technology provides platform
software with the ability to monitor, control, and get notifications on
power usage.
This feature is present in all Sandy Bridge and later Intel processors.
Newer models allow more fine grained controls to be applied. In RAPL,
power control is divided into domains, which include package, DRAM
controller, CPU core (Power Plane 0), graphics uncore (power plane 1), etc.
The purpose of this driver is to expose the RAPL settings to userspace.
Overall, RAPL fits in the new powercap class driver in that platform
level power capping controls are exposed via this generic interface.
This driver is based on an earlier patch from Zhang Rui.
However, while the previous work was mainly focused on thermal monitoring
the focus here is on the usability from user space perspective.
References: https://lkml.org/lkml/2011/5/26/93
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Signed-off-by: Jacob Pan <jacob.jun.pan@linux.intel.com>
Reviewed-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
2013-10-17 11:28:35 -06:00
|
|
|
PRIMITIVE_INFO_INIT(MAX_POWER, POWER_INFO_MAX_MASK, 32,
|
2019-07-10 07:44:30 -06:00
|
|
|
RAPL_DOMAIN_REG_INFO, POWER_UNIT, 0),
|
PowerCap: Introduce Intel RAPL power capping driver
The Intel Running Average Power Limit (RAPL) technology provides platform
software with the ability to monitor, control, and get notifications on
power usage.
This feature is present in all Sandy Bridge and later Intel processors.
Newer models allow more fine grained controls to be applied. In RAPL,
power control is divided into domains, which include package, DRAM
controller, CPU core (Power Plane 0), graphics uncore (power plane 1), etc.
The purpose of this driver is to expose the RAPL settings to userspace.
Overall, RAPL fits in the new powercap class driver in that platform
level power capping controls are exposed via this generic interface.
This driver is based on an earlier patch from Zhang Rui.
However, while the previous work was mainly focused on thermal monitoring
the focus here is on the usability from user space perspective.
References: https://lkml.org/lkml/2011/5/26/93
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Signed-off-by: Jacob Pan <jacob.jun.pan@linux.intel.com>
Reviewed-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
2013-10-17 11:28:35 -06:00
|
|
|
PRIMITIVE_INFO_INIT(MIN_POWER, POWER_INFO_MIN_MASK, 16,
|
2019-07-10 07:44:30 -06:00
|
|
|
RAPL_DOMAIN_REG_INFO, POWER_UNIT, 0),
|
PowerCap: Introduce Intel RAPL power capping driver
The Intel Running Average Power Limit (RAPL) technology provides platform
software with the ability to monitor, control, and get notifications on
power usage.
This feature is present in all Sandy Bridge and later Intel processors.
Newer models allow more fine grained controls to be applied. In RAPL,
power control is divided into domains, which include package, DRAM
controller, CPU core (Power Plane 0), graphics uncore (power plane 1), etc.
The purpose of this driver is to expose the RAPL settings to userspace.
Overall, RAPL fits in the new powercap class driver in that platform
level power capping controls are exposed via this generic interface.
This driver is based on an earlier patch from Zhang Rui.
However, while the previous work was mainly focused on thermal monitoring
the focus here is on the usability from user space perspective.
References: https://lkml.org/lkml/2011/5/26/93
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Signed-off-by: Jacob Pan <jacob.jun.pan@linux.intel.com>
Reviewed-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
2013-10-17 11:28:35 -06:00
|
|
|
PRIMITIVE_INFO_INIT(MAX_TIME_WINDOW, POWER_INFO_MAX_TIME_WIN_MASK, 48,
|
2019-07-10 07:44:30 -06:00
|
|
|
RAPL_DOMAIN_REG_INFO, TIME_UNIT, 0),
|
PowerCap: Introduce Intel RAPL power capping driver
The Intel Running Average Power Limit (RAPL) technology provides platform
software with the ability to monitor, control, and get notifications on
power usage.
This feature is present in all Sandy Bridge and later Intel processors.
Newer models allow more fine grained controls to be applied. In RAPL,
power control is divided into domains, which include package, DRAM
controller, CPU core (Power Plane 0), graphics uncore (power plane 1), etc.
The purpose of this driver is to expose the RAPL settings to userspace.
Overall, RAPL fits in the new powercap class driver in that platform
level power capping controls are exposed via this generic interface.
This driver is based on an earlier patch from Zhang Rui.
However, while the previous work was mainly focused on thermal monitoring
the focus here is on the usability from user space perspective.
References: https://lkml.org/lkml/2011/5/26/93
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Signed-off-by: Jacob Pan <jacob.jun.pan@linux.intel.com>
Reviewed-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
2013-10-17 11:28:35 -06:00
|
|
|
PRIMITIVE_INFO_INIT(THROTTLED_TIME, PERF_STATUS_THROTTLE_TIME_MASK, 0,
|
2019-07-10 07:44:30 -06:00
|
|
|
RAPL_DOMAIN_REG_PERF, TIME_UNIT, 0),
|
PowerCap: Introduce Intel RAPL power capping driver
The Intel Running Average Power Limit (RAPL) technology provides platform
software with the ability to monitor, control, and get notifications on
power usage.
This feature is present in all Sandy Bridge and later Intel processors.
Newer models allow more fine grained controls to be applied. In RAPL,
power control is divided into domains, which include package, DRAM
controller, CPU core (Power Plane 0), graphics uncore (power plane 1), etc.
The purpose of this driver is to expose the RAPL settings to userspace.
Overall, RAPL fits in the new powercap class driver in that platform
level power capping controls are exposed via this generic interface.
This driver is based on an earlier patch from Zhang Rui.
However, while the previous work was mainly focused on thermal monitoring
the focus here is on the usability from user space perspective.
References: https://lkml.org/lkml/2011/5/26/93
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Signed-off-by: Jacob Pan <jacob.jun.pan@linux.intel.com>
Reviewed-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
2013-10-17 11:28:35 -06:00
|
|
|
PRIMITIVE_INFO_INIT(PRIORITY_LEVEL, PP_POLICY_MASK, 0,
|
2019-07-10 07:44:30 -06:00
|
|
|
RAPL_DOMAIN_REG_POLICY, ARBITRARY_UNIT, 0),
|
PowerCap: Introduce Intel RAPL power capping driver
The Intel Running Average Power Limit (RAPL) technology provides platform
software with the ability to monitor, control, and get notifications on
power usage.
This feature is present in all Sandy Bridge and later Intel processors.
Newer models allow more fine grained controls to be applied. In RAPL,
power control is divided into domains, which include package, DRAM
controller, CPU core (Power Plane 0), graphics uncore (power plane 1), etc.
The purpose of this driver is to expose the RAPL settings to userspace.
Overall, RAPL fits in the new powercap class driver in that platform
level power capping controls are exposed via this generic interface.
This driver is based on an earlier patch from Zhang Rui.
However, while the previous work was mainly focused on thermal monitoring
the focus here is on the usability from user space perspective.
References: https://lkml.org/lkml/2011/5/26/93
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Signed-off-by: Jacob Pan <jacob.jun.pan@linux.intel.com>
Reviewed-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
2013-10-17 11:28:35 -06:00
|
|
|
/* non-hardware */
|
|
|
|
PRIMITIVE_INFO_INIT(AVERAGE_POWER, 0, 0, 0, POWER_UNIT,
|
2019-07-10 07:44:30 -06:00
|
|
|
RAPL_PRIMITIVE_DERIVED),
|
PowerCap: Introduce Intel RAPL power capping driver
The Intel Running Average Power Limit (RAPL) technology provides platform
software with the ability to monitor, control, and get notifications on
power usage.
This feature is present in all Sandy Bridge and later Intel processors.
Newer models allow more fine grained controls to be applied. In RAPL,
power control is divided into domains, which include package, DRAM
controller, CPU core (Power Plane 0), graphics uncore (power plane 1), etc.
The purpose of this driver is to expose the RAPL settings to userspace.
Overall, RAPL fits in the new powercap class driver in that platform
level power capping controls are exposed via this generic interface.
This driver is based on an earlier patch from Zhang Rui.
However, while the previous work was mainly focused on thermal monitoring
the focus here is on the usability from user space perspective.
References: https://lkml.org/lkml/2011/5/26/93
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Signed-off-by: Jacob Pan <jacob.jun.pan@linux.intel.com>
Reviewed-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
2013-10-17 11:28:35 -06:00
|
|
|
{NULL, 0, 0, 0},
|
|
|
|
};
|
|
|
|
|
|
|
|
/* Read primitive data based on its related struct rapl_primitive_info.
|
|
|
|
* if xlate flag is set, return translated data based on data units, i.e.
|
|
|
|
* time, energy, and power.
|
|
|
|
* RAPL MSRs are non-architectual and are laid out not consistently across
|
|
|
|
* domains. Here we use primitive info to allow writing consolidated access
|
|
|
|
* functions.
|
|
|
|
* For a given primitive, it is processed by MSR mask and shift. Unit conversion
|
|
|
|
* is pre-assigned based on RAPL unit MSRs read at init time.
|
|
|
|
* 63-------------------------- 31--------------------------- 0
|
|
|
|
* | xxxxx (mask) |
|
|
|
|
* | |<- shift ----------------|
|
|
|
|
* 63-------------------------- 31--------------------------- 0
|
|
|
|
*/
|
|
|
|
static int rapl_read_data_raw(struct rapl_domain *rd,
|
2019-07-10 07:44:30 -06:00
|
|
|
enum rapl_primitives prim, bool xlate, u64 *data)
|
PowerCap: Introduce Intel RAPL power capping driver
The Intel Running Average Power Limit (RAPL) technology provides platform
software with the ability to monitor, control, and get notifications on
power usage.
This feature is present in all Sandy Bridge and later Intel processors.
Newer models allow more fine grained controls to be applied. In RAPL,
power control is divided into domains, which include package, DRAM
controller, CPU core (Power Plane 0), graphics uncore (power plane 1), etc.
The purpose of this driver is to expose the RAPL settings to userspace.
Overall, RAPL fits in the new powercap class driver in that platform
level power capping controls are exposed via this generic interface.
This driver is based on an earlier patch from Zhang Rui.
However, while the previous work was mainly focused on thermal monitoring
the focus here is on the usability from user space perspective.
References: https://lkml.org/lkml/2011/5/26/93
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Signed-off-by: Jacob Pan <jacob.jun.pan@linux.intel.com>
Reviewed-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
2013-10-17 11:28:35 -06:00
|
|
|
{
|
2019-07-10 07:44:27 -06:00
|
|
|
u64 value;
|
PowerCap: Introduce Intel RAPL power capping driver
The Intel Running Average Power Limit (RAPL) technology provides platform
software with the ability to monitor, control, and get notifications on
power usage.
This feature is present in all Sandy Bridge and later Intel processors.
Newer models allow more fine grained controls to be applied. In RAPL,
power control is divided into domains, which include package, DRAM
controller, CPU core (Power Plane 0), graphics uncore (power plane 1), etc.
The purpose of this driver is to expose the RAPL settings to userspace.
Overall, RAPL fits in the new powercap class driver in that platform
level power capping controls are exposed via this generic interface.
This driver is based on an earlier patch from Zhang Rui.
However, while the previous work was mainly focused on thermal monitoring
the focus here is on the usability from user space perspective.
References: https://lkml.org/lkml/2011/5/26/93
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Signed-off-by: Jacob Pan <jacob.jun.pan@linux.intel.com>
Reviewed-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
2013-10-17 11:28:35 -06:00
|
|
|
struct rapl_primitive_info *rp = &rpi[prim];
|
2019-07-10 07:44:27 -06:00
|
|
|
struct reg_action ra;
|
PowerCap: Introduce Intel RAPL power capping driver
The Intel Running Average Power Limit (RAPL) technology provides platform
software with the ability to monitor, control, and get notifications on
power usage.
This feature is present in all Sandy Bridge and later Intel processors.
Newer models allow more fine grained controls to be applied. In RAPL,
power control is divided into domains, which include package, DRAM
controller, CPU core (Power Plane 0), graphics uncore (power plane 1), etc.
The purpose of this driver is to expose the RAPL settings to userspace.
Overall, RAPL fits in the new powercap class driver in that platform
level power capping controls are exposed via this generic interface.
This driver is based on an earlier patch from Zhang Rui.
However, while the previous work was mainly focused on thermal monitoring
the focus here is on the usability from user space perspective.
References: https://lkml.org/lkml/2011/5/26/93
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Signed-off-by: Jacob Pan <jacob.jun.pan@linux.intel.com>
Reviewed-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
2013-10-17 11:28:35 -06:00
|
|
|
int cpu;
|
|
|
|
|
|
|
|
if (!rp->name || rp->flag & RAPL_PRIMITIVE_DUMMY)
|
|
|
|
return -EINVAL;
|
|
|
|
|
2019-07-10 07:44:27 -06:00
|
|
|
ra.reg = rd->regs[rp->id];
|
|
|
|
if (!ra.reg)
|
PowerCap: Introduce Intel RAPL power capping driver
The Intel Running Average Power Limit (RAPL) technology provides platform
software with the ability to monitor, control, and get notifications on
power usage.
This feature is present in all Sandy Bridge and later Intel processors.
Newer models allow more fine grained controls to be applied. In RAPL,
power control is divided into domains, which include package, DRAM
controller, CPU core (Power Plane 0), graphics uncore (power plane 1), etc.
The purpose of this driver is to expose the RAPL settings to userspace.
Overall, RAPL fits in the new powercap class driver in that platform
level power capping controls are exposed via this generic interface.
This driver is based on an earlier patch from Zhang Rui.
However, while the previous work was mainly focused on thermal monitoring
the focus here is on the usability from user space perspective.
References: https://lkml.org/lkml/2011/5/26/93
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Signed-off-by: Jacob Pan <jacob.jun.pan@linux.intel.com>
Reviewed-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
2013-10-17 11:28:35 -06:00
|
|
|
return -EINVAL;
|
2016-02-24 14:31:38 -07:00
|
|
|
|
|
|
|
cpu = rd->rp->lead_cpu;
|
PowerCap: Introduce Intel RAPL power capping driver
The Intel Running Average Power Limit (RAPL) technology provides platform
software with the ability to monitor, control, and get notifications on
power usage.
This feature is present in all Sandy Bridge and later Intel processors.
Newer models allow more fine grained controls to be applied. In RAPL,
power control is divided into domains, which include package, DRAM
controller, CPU core (Power Plane 0), graphics uncore (power plane 1), etc.
The purpose of this driver is to expose the RAPL settings to userspace.
Overall, RAPL fits in the new powercap class driver in that platform
level power capping controls are exposed via this generic interface.
This driver is based on an earlier patch from Zhang Rui.
However, while the previous work was mainly focused on thermal monitoring
the focus here is on the usability from user space perspective.
References: https://lkml.org/lkml/2011/5/26/93
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Signed-off-by: Jacob Pan <jacob.jun.pan@linux.intel.com>
Reviewed-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
2013-10-17 11:28:35 -06:00
|
|
|
|
2019-07-10 07:44:32 -06:00
|
|
|
/* domain with 2 limits has different bit */
|
|
|
|
if (prim == FW_LOCK && rd->rp->priv->limits[rd->id] == 2) {
|
|
|
|
rp->mask = POWER_HIGH_LOCK;
|
PowerCap: Introduce Intel RAPL power capping driver
The Intel Running Average Power Limit (RAPL) technology provides platform
software with the ability to monitor, control, and get notifications on
power usage.
This feature is present in all Sandy Bridge and later Intel processors.
Newer models allow more fine grained controls to be applied. In RAPL,
power control is divided into domains, which include package, DRAM
controller, CPU core (Power Plane 0), graphics uncore (power plane 1), etc.
The purpose of this driver is to expose the RAPL settings to userspace.
Overall, RAPL fits in the new powercap class driver in that platform
level power capping controls are exposed via this generic interface.
This driver is based on an earlier patch from Zhang Rui.
However, while the previous work was mainly focused on thermal monitoring
the focus here is on the usability from user space perspective.
References: https://lkml.org/lkml/2011/5/26/93
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Signed-off-by: Jacob Pan <jacob.jun.pan@linux.intel.com>
Reviewed-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
2013-10-17 11:28:35 -06:00
|
|
|
rp->shift = 63;
|
|
|
|
}
|
|
|
|
/* non-hardware data are collected by the polling thread */
|
|
|
|
if (rp->flag & RAPL_PRIMITIVE_DERIVED) {
|
|
|
|
*data = rd->rdd.primitives[prim];
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
2019-07-10 07:44:27 -06:00
|
|
|
ra.mask = rp->mask;
|
|
|
|
|
|
|
|
if (rd->rp->priv->read_raw(cpu, &ra)) {
|
2019-07-10 07:44:31 -06:00
|
|
|
pr_debug("failed to read reg 0x%llx on cpu %d\n", ra.reg, cpu);
|
PowerCap: Introduce Intel RAPL power capping driver
The Intel Running Average Power Limit (RAPL) technology provides platform
software with the ability to monitor, control, and get notifications on
power usage.
This feature is present in all Sandy Bridge and later Intel processors.
Newer models allow more fine grained controls to be applied. In RAPL,
power control is divided into domains, which include package, DRAM
controller, CPU core (Power Plane 0), graphics uncore (power plane 1), etc.
The purpose of this driver is to expose the RAPL settings to userspace.
Overall, RAPL fits in the new powercap class driver in that platform
level power capping controls are exposed via this generic interface.
This driver is based on an earlier patch from Zhang Rui.
However, while the previous work was mainly focused on thermal monitoring
the focus here is on the usability from user space perspective.
References: https://lkml.org/lkml/2011/5/26/93
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Signed-off-by: Jacob Pan <jacob.jun.pan@linux.intel.com>
Reviewed-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
2013-10-17 11:28:35 -06:00
|
|
|
return -EIO;
|
|
|
|
}
|
|
|
|
|
2019-07-10 07:44:27 -06:00
|
|
|
value = ra.value >> rp->shift;
|
|
|
|
|
PowerCap: Introduce Intel RAPL power capping driver
The Intel Running Average Power Limit (RAPL) technology provides platform
software with the ability to monitor, control, and get notifications on
power usage.
This feature is present in all Sandy Bridge and later Intel processors.
Newer models allow more fine grained controls to be applied. In RAPL,
power control is divided into domains, which include package, DRAM
controller, CPU core (Power Plane 0), graphics uncore (power plane 1), etc.
The purpose of this driver is to expose the RAPL settings to userspace.
Overall, RAPL fits in the new powercap class driver in that platform
level power capping controls are exposed via this generic interface.
This driver is based on an earlier patch from Zhang Rui.
However, while the previous work was mainly focused on thermal monitoring
the focus here is on the usability from user space perspective.
References: https://lkml.org/lkml/2011/5/26/93
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Signed-off-by: Jacob Pan <jacob.jun.pan@linux.intel.com>
Reviewed-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
2013-10-17 11:28:35 -06:00
|
|
|
if (xlate)
|
2019-07-10 07:44:27 -06:00
|
|
|
*data = rapl_unit_xlate(rd, rp->unit, value, 0);
|
PowerCap: Introduce Intel RAPL power capping driver
The Intel Running Average Power Limit (RAPL) technology provides platform
software with the ability to monitor, control, and get notifications on
power usage.
This feature is present in all Sandy Bridge and later Intel processors.
Newer models allow more fine grained controls to be applied. In RAPL,
power control is divided into domains, which include package, DRAM
controller, CPU core (Power Plane 0), graphics uncore (power plane 1), etc.
The purpose of this driver is to expose the RAPL settings to userspace.
Overall, RAPL fits in the new powercap class driver in that platform
level power capping controls are exposed via this generic interface.
This driver is based on an earlier patch from Zhang Rui.
However, while the previous work was mainly focused on thermal monitoring
the focus here is on the usability from user space perspective.
References: https://lkml.org/lkml/2011/5/26/93
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Signed-off-by: Jacob Pan <jacob.jun.pan@linux.intel.com>
Reviewed-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
2013-10-17 11:28:35 -06:00
|
|
|
else
|
2019-07-10 07:44:27 -06:00
|
|
|
*data = value;
|
PowerCap: Introduce Intel RAPL power capping driver
The Intel Running Average Power Limit (RAPL) technology provides platform
software with the ability to monitor, control, and get notifications on
power usage.
This feature is present in all Sandy Bridge and later Intel processors.
Newer models allow more fine grained controls to be applied. In RAPL,
power control is divided into domains, which include package, DRAM
controller, CPU core (Power Plane 0), graphics uncore (power plane 1), etc.
The purpose of this driver is to expose the RAPL settings to userspace.
Overall, RAPL fits in the new powercap class driver in that platform
level power capping controls are exposed via this generic interface.
This driver is based on an earlier patch from Zhang Rui.
However, while the previous work was mainly focused on thermal monitoring
the focus here is on the usability from user space perspective.
References: https://lkml.org/lkml/2011/5/26/93
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Signed-off-by: Jacob Pan <jacob.jun.pan@linux.intel.com>
Reviewed-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
2013-10-17 11:28:35 -06:00
|
|
|
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
/* Similar use of primitive info in the read counterpart */
|
|
|
|
static int rapl_write_data_raw(struct rapl_domain *rd,
|
2019-07-10 07:44:30 -06:00
|
|
|
enum rapl_primitives prim,
|
|
|
|
unsigned long long value)
|
PowerCap: Introduce Intel RAPL power capping driver
The Intel Running Average Power Limit (RAPL) technology provides platform
software with the ability to monitor, control, and get notifications on
power usage.
This feature is present in all Sandy Bridge and later Intel processors.
Newer models allow more fine grained controls to be applied. In RAPL,
power control is divided into domains, which include package, DRAM
controller, CPU core (Power Plane 0), graphics uncore (power plane 1), etc.
The purpose of this driver is to expose the RAPL settings to userspace.
Overall, RAPL fits in the new powercap class driver in that platform
level power capping controls are exposed via this generic interface.
This driver is based on an earlier patch from Zhang Rui.
However, while the previous work was mainly focused on thermal monitoring
the focus here is on the usability from user space perspective.
References: https://lkml.org/lkml/2011/5/26/93
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Signed-off-by: Jacob Pan <jacob.jun.pan@linux.intel.com>
Reviewed-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
2013-10-17 11:28:35 -06:00
|
|
|
{
|
|
|
|
struct rapl_primitive_info *rp = &rpi[prim];
|
|
|
|
int cpu;
|
2016-02-24 14:31:36 -07:00
|
|
|
u64 bits;
|
2019-07-10 07:44:27 -06:00
|
|
|
struct reg_action ra;
|
2016-02-24 14:31:36 -07:00
|
|
|
int ret;
|
PowerCap: Introduce Intel RAPL power capping driver
The Intel Running Average Power Limit (RAPL) technology provides platform
software with the ability to monitor, control, and get notifications on
power usage.
This feature is present in all Sandy Bridge and later Intel processors.
Newer models allow more fine grained controls to be applied. In RAPL,
power control is divided into domains, which include package, DRAM
controller, CPU core (Power Plane 0), graphics uncore (power plane 1), etc.
The purpose of this driver is to expose the RAPL settings to userspace.
Overall, RAPL fits in the new powercap class driver in that platform
level power capping controls are exposed via this generic interface.
This driver is based on an earlier patch from Zhang Rui.
However, while the previous work was mainly focused on thermal monitoring
the focus here is on the usability from user space perspective.
References: https://lkml.org/lkml/2011/5/26/93
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Signed-off-by: Jacob Pan <jacob.jun.pan@linux.intel.com>
Reviewed-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
2013-10-17 11:28:35 -06:00
|
|
|
|
2016-02-24 14:31:38 -07:00
|
|
|
cpu = rd->rp->lead_cpu;
|
2016-02-24 14:31:37 -07:00
|
|
|
bits = rapl_unit_xlate(rd, rp->unit, value, 1);
|
2017-06-01 03:21:50 -06:00
|
|
|
bits <<= rp->shift;
|
|
|
|
bits &= rp->mask;
|
|
|
|
|
2019-07-10 07:44:27 -06:00
|
|
|
memset(&ra, 0, sizeof(ra));
|
2016-02-24 14:31:36 -07:00
|
|
|
|
2019-07-10 07:44:27 -06:00
|
|
|
ra.reg = rd->regs[rp->id];
|
|
|
|
ra.mask = rp->mask;
|
|
|
|
ra.value = bits;
|
2016-02-24 14:31:36 -07:00
|
|
|
|
2019-07-10 07:44:27 -06:00
|
|
|
ret = rd->rp->priv->write_raw(cpu, &ra);
|
2016-02-24 14:31:36 -07:00
|
|
|
|
|
|
|
return ret;
|
PowerCap: Introduce Intel RAPL power capping driver
The Intel Running Average Power Limit (RAPL) technology provides platform
software with the ability to monitor, control, and get notifications on
power usage.
This feature is present in all Sandy Bridge and later Intel processors.
Newer models allow more fine grained controls to be applied. In RAPL,
power control is divided into domains, which include package, DRAM
controller, CPU core (Power Plane 0), graphics uncore (power plane 1), etc.
The purpose of this driver is to expose the RAPL settings to userspace.
Overall, RAPL fits in the new powercap class driver in that platform
level power capping controls are exposed via this generic interface.
This driver is based on an earlier patch from Zhang Rui.
However, while the previous work was mainly focused on thermal monitoring
the focus here is on the usability from user space perspective.
References: https://lkml.org/lkml/2011/5/26/93
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Signed-off-by: Jacob Pan <jacob.jun.pan@linux.intel.com>
Reviewed-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
2013-10-17 11:28:35 -06:00
|
|
|
}
|
|
|
|
|
2014-11-07 10:29:26 -07:00
|
|
|
/*
|
|
|
|
* Raw RAPL data stored in MSRs are in certain scales. We need to
|
|
|
|
* convert them into standard units based on the units reported in
|
|
|
|
* the RAPL unit MSRs. This is specific to CPUs as the method to
|
|
|
|
* calculate units differ on different CPUs.
|
|
|
|
* We convert the units to below format based on CPUs.
|
|
|
|
* i.e.
|
2015-03-13 04:48:56 -06:00
|
|
|
* energy unit: picoJoules : Represented in picoJoules by default
|
2014-11-07 10:29:26 -07:00
|
|
|
* power unit : microWatts : Represented in milliWatts by default
|
|
|
|
* time unit : microseconds: Represented in seconds by default
|
|
|
|
*/
|
|
|
|
static int rapl_check_unit_core(struct rapl_package *rp, int cpu)
|
PowerCap: Introduce Intel RAPL power capping driver
The Intel Running Average Power Limit (RAPL) technology provides platform
software with the ability to monitor, control, and get notifications on
power usage.
This feature is present in all Sandy Bridge and later Intel processors.
Newer models allow more fine grained controls to be applied. In RAPL,
power control is divided into domains, which include package, DRAM
controller, CPU core (Power Plane 0), graphics uncore (power plane 1), etc.
The purpose of this driver is to expose the RAPL settings to userspace.
Overall, RAPL fits in the new powercap class driver in that platform
level power capping controls are exposed via this generic interface.
This driver is based on an earlier patch from Zhang Rui.
However, while the previous work was mainly focused on thermal monitoring
the focus here is on the usability from user space perspective.
References: https://lkml.org/lkml/2011/5/26/93
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Signed-off-by: Jacob Pan <jacob.jun.pan@linux.intel.com>
Reviewed-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
2013-10-17 11:28:35 -06:00
|
|
|
{
|
2019-07-10 07:44:29 -06:00
|
|
|
struct reg_action ra;
|
PowerCap: Introduce Intel RAPL power capping driver
The Intel Running Average Power Limit (RAPL) technology provides platform
software with the ability to monitor, control, and get notifications on
power usage.
This feature is present in all Sandy Bridge and later Intel processors.
Newer models allow more fine grained controls to be applied. In RAPL,
power control is divided into domains, which include package, DRAM
controller, CPU core (Power Plane 0), graphics uncore (power plane 1), etc.
The purpose of this driver is to expose the RAPL settings to userspace.
Overall, RAPL fits in the new powercap class driver in that platform
level power capping controls are exposed via this generic interface.
This driver is based on an earlier patch from Zhang Rui.
However, while the previous work was mainly focused on thermal monitoring
the focus here is on the usability from user space perspective.
References: https://lkml.org/lkml/2011/5/26/93
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Signed-off-by: Jacob Pan <jacob.jun.pan@linux.intel.com>
Reviewed-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
2013-10-17 11:28:35 -06:00
|
|
|
u32 value;
|
|
|
|
|
2019-07-10 07:44:29 -06:00
|
|
|
ra.reg = rp->priv->reg_unit;
|
|
|
|
ra.mask = ~0;
|
|
|
|
if (rp->priv->read_raw(cpu, &ra)) {
|
2019-07-10 07:44:31 -06:00
|
|
|
pr_err("Failed to read power unit REG 0x%llx on CPU %d, exit.\n",
|
2019-07-10 07:44:30 -06:00
|
|
|
rp->priv->reg_unit, cpu);
|
PowerCap: Introduce Intel RAPL power capping driver
The Intel Running Average Power Limit (RAPL) technology provides platform
software with the ability to monitor, control, and get notifications on
power usage.
This feature is present in all Sandy Bridge and later Intel processors.
Newer models allow more fine grained controls to be applied. In RAPL,
power control is divided into domains, which include package, DRAM
controller, CPU core (Power Plane 0), graphics uncore (power plane 1), etc.
The purpose of this driver is to expose the RAPL settings to userspace.
Overall, RAPL fits in the new powercap class driver in that platform
level power capping controls are exposed via this generic interface.
This driver is based on an earlier patch from Zhang Rui.
However, while the previous work was mainly focused on thermal monitoring
the focus here is on the usability from user space perspective.
References: https://lkml.org/lkml/2011/5/26/93
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Signed-off-by: Jacob Pan <jacob.jun.pan@linux.intel.com>
Reviewed-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
2013-10-17 11:28:35 -06:00
|
|
|
return -ENODEV;
|
|
|
|
}
|
|
|
|
|
2019-07-10 07:44:29 -06:00
|
|
|
value = (ra.value & ENERGY_UNIT_MASK) >> ENERGY_UNIT_OFFSET;
|
2015-03-13 04:48:56 -06:00
|
|
|
rp->energy_unit = ENERGY_UNIT_SCALE * 1000000 / (1 << value);
|
PowerCap: Introduce Intel RAPL power capping driver
The Intel Running Average Power Limit (RAPL) technology provides platform
software with the ability to monitor, control, and get notifications on
power usage.
This feature is present in all Sandy Bridge and later Intel processors.
Newer models allow more fine grained controls to be applied. In RAPL,
power control is divided into domains, which include package, DRAM
controller, CPU core (Power Plane 0), graphics uncore (power plane 1), etc.
The purpose of this driver is to expose the RAPL settings to userspace.
Overall, RAPL fits in the new powercap class driver in that platform
level power capping controls are exposed via this generic interface.
This driver is based on an earlier patch from Zhang Rui.
However, while the previous work was mainly focused on thermal monitoring
the focus here is on the usability from user space perspective.
References: https://lkml.org/lkml/2011/5/26/93
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Signed-off-by: Jacob Pan <jacob.jun.pan@linux.intel.com>
Reviewed-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
2013-10-17 11:28:35 -06:00
|
|
|
|
2019-07-10 07:44:29 -06:00
|
|
|
value = (ra.value & POWER_UNIT_MASK) >> POWER_UNIT_OFFSET;
|
2014-11-07 10:29:26 -07:00
|
|
|
rp->power_unit = 1000000 / (1 << value);
|
PowerCap: Introduce Intel RAPL power capping driver
The Intel Running Average Power Limit (RAPL) technology provides platform
software with the ability to monitor, control, and get notifications on
power usage.
This feature is present in all Sandy Bridge and later Intel processors.
Newer models allow more fine grained controls to be applied. In RAPL,
power control is divided into domains, which include package, DRAM
controller, CPU core (Power Plane 0), graphics uncore (power plane 1), etc.
The purpose of this driver is to expose the RAPL settings to userspace.
Overall, RAPL fits in the new powercap class driver in that platform
level power capping controls are exposed via this generic interface.
This driver is based on an earlier patch from Zhang Rui.
However, while the previous work was mainly focused on thermal monitoring
the focus here is on the usability from user space perspective.
References: https://lkml.org/lkml/2011/5/26/93
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Signed-off-by: Jacob Pan <jacob.jun.pan@linux.intel.com>
Reviewed-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
2013-10-17 11:28:35 -06:00
|
|
|
|
2019-07-10 07:44:29 -06:00
|
|
|
value = (ra.value & TIME_UNIT_MASK) >> TIME_UNIT_OFFSET;
|
2014-11-07 10:29:26 -07:00
|
|
|
rp->time_unit = 1000000 / (1 << value);
|
PowerCap: Introduce Intel RAPL power capping driver
The Intel Running Average Power Limit (RAPL) technology provides platform
software with the ability to monitor, control, and get notifications on
power usage.
This feature is present in all Sandy Bridge and later Intel processors.
Newer models allow more fine grained controls to be applied. In RAPL,
power control is divided into domains, which include package, DRAM
controller, CPU core (Power Plane 0), graphics uncore (power plane 1), etc.
The purpose of this driver is to expose the RAPL settings to userspace.
Overall, RAPL fits in the new powercap class driver in that platform
level power capping controls are exposed via this generic interface.
This driver is based on an earlier patch from Zhang Rui.
However, while the previous work was mainly focused on thermal monitoring
the focus here is on the usability from user space perspective.
References: https://lkml.org/lkml/2011/5/26/93
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Signed-off-by: Jacob Pan <jacob.jun.pan@linux.intel.com>
Reviewed-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
2013-10-17 11:28:35 -06:00
|
|
|
|
2019-05-13 11:58:53 -06:00
|
|
|
pr_debug("Core CPU %s energy=%dpJ, time=%dus, power=%duW\n",
|
2019-07-10 07:44:30 -06:00
|
|
|
rp->name, rp->energy_unit, rp->time_unit, rp->power_unit);
|
PowerCap: Introduce Intel RAPL power capping driver
The Intel Running Average Power Limit (RAPL) technology provides platform
software with the ability to monitor, control, and get notifications on
power usage.
This feature is present in all Sandy Bridge and later Intel processors.
Newer models allow more fine grained controls to be applied. In RAPL,
power control is divided into domains, which include package, DRAM
controller, CPU core (Power Plane 0), graphics uncore (power plane 1), etc.
The purpose of this driver is to expose the RAPL settings to userspace.
Overall, RAPL fits in the new powercap class driver in that platform
level power capping controls are exposed via this generic interface.
This driver is based on an earlier patch from Zhang Rui.
However, while the previous work was mainly focused on thermal monitoring
the focus here is on the usability from user space perspective.
References: https://lkml.org/lkml/2011/5/26/93
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Signed-off-by: Jacob Pan <jacob.jun.pan@linux.intel.com>
Reviewed-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
2013-10-17 11:28:35 -06:00
|
|
|
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
2014-11-07 10:29:26 -07:00
|
|
|
static int rapl_check_unit_atom(struct rapl_package *rp, int cpu)
|
|
|
|
{
|
2019-07-10 07:44:29 -06:00
|
|
|
struct reg_action ra;
|
2014-11-07 10:29:26 -07:00
|
|
|
u32 value;
|
|
|
|
|
2019-07-10 07:44:29 -06:00
|
|
|
ra.reg = rp->priv->reg_unit;
|
|
|
|
ra.mask = ~0;
|
|
|
|
if (rp->priv->read_raw(cpu, &ra)) {
|
2019-07-10 07:44:31 -06:00
|
|
|
pr_err("Failed to read power unit REG 0x%llx on CPU %d, exit.\n",
|
2019-07-10 07:44:30 -06:00
|
|
|
rp->priv->reg_unit, cpu);
|
2014-11-07 10:29:26 -07:00
|
|
|
return -ENODEV;
|
|
|
|
}
|
2019-07-10 07:44:29 -06:00
|
|
|
|
|
|
|
value = (ra.value & ENERGY_UNIT_MASK) >> ENERGY_UNIT_OFFSET;
|
2015-03-13 04:48:56 -06:00
|
|
|
rp->energy_unit = ENERGY_UNIT_SCALE * 1 << value;
|
2014-11-07 10:29:26 -07:00
|
|
|
|
2019-07-10 07:44:29 -06:00
|
|
|
value = (ra.value & POWER_UNIT_MASK) >> POWER_UNIT_OFFSET;
|
2014-11-07 10:29:26 -07:00
|
|
|
rp->power_unit = (1 << value) * 1000;
|
|
|
|
|
2019-07-10 07:44:29 -06:00
|
|
|
value = (ra.value & TIME_UNIT_MASK) >> TIME_UNIT_OFFSET;
|
2014-11-07 10:29:26 -07:00
|
|
|
rp->time_unit = 1000000 / (1 << value);
|
|
|
|
|
2019-05-13 11:58:53 -06:00
|
|
|
pr_debug("Atom %s energy=%dpJ, time=%dus, power=%duW\n",
|
2019-07-10 07:44:30 -06:00
|
|
|
rp->name, rp->energy_unit, rp->time_unit, rp->power_unit);
|
2014-11-07 10:29:26 -07:00
|
|
|
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
2016-02-24 14:31:36 -07:00
|
|
|
static void power_limit_irq_save_cpu(void *info)
|
|
|
|
{
|
|
|
|
u32 l, h = 0;
|
|
|
|
struct rapl_package *rp = (struct rapl_package *)info;
|
|
|
|
|
|
|
|
/* save the state of PLN irq mask bit before disabling it */
|
|
|
|
rdmsr_safe(MSR_IA32_PACKAGE_THERM_INTERRUPT, &l, &h);
|
|
|
|
if (!(rp->power_limit_irq & PACKAGE_PLN_INT_SAVED)) {
|
|
|
|
rp->power_limit_irq = l & PACKAGE_THERM_INT_PLN_ENABLE;
|
|
|
|
rp->power_limit_irq |= PACKAGE_PLN_INT_SAVED;
|
|
|
|
}
|
|
|
|
l &= ~PACKAGE_THERM_INT_PLN_ENABLE;
|
|
|
|
wrmsr_safe(MSR_IA32_PACKAGE_THERM_INTERRUPT, l, h);
|
|
|
|
}
|
|
|
|
|
PowerCap: Introduce Intel RAPL power capping driver
The Intel Running Average Power Limit (RAPL) technology provides platform
software with the ability to monitor, control, and get notifications on
power usage.
This feature is present in all Sandy Bridge and later Intel processors.
Newer models allow more fine grained controls to be applied. In RAPL,
power control is divided into domains, which include package, DRAM
controller, CPU core (Power Plane 0), graphics uncore (power plane 1), etc.
The purpose of this driver is to expose the RAPL settings to userspace.
Overall, RAPL fits in the new powercap class driver in that platform
level power capping controls are exposed via this generic interface.
This driver is based on an earlier patch from Zhang Rui.
However, while the previous work was mainly focused on thermal monitoring
the focus here is on the usability from user space perspective.
References: https://lkml.org/lkml/2011/5/26/93
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Signed-off-by: Jacob Pan <jacob.jun.pan@linux.intel.com>
Reviewed-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
2013-10-17 11:28:35 -06:00
|
|
|
/* REVISIT:
|
|
|
|
* When package power limit is set artificially low by RAPL, LVT
|
|
|
|
* thermal interrupt for package power limit should be ignored
|
|
|
|
* since we are not really exceeding the real limit. The intention
|
|
|
|
* is to avoid excessive interrupts while we are trying to save power.
|
|
|
|
* A useful feature might be routing the package_power_limit interrupt
|
|
|
|
* to userspace via eventfd. once we have a usecase, this is simple
|
|
|
|
* to do by adding an atomic notifier.
|
|
|
|
*/
|
|
|
|
|
2016-02-24 14:31:37 -07:00
|
|
|
static void package_power_limit_irq_save(struct rapl_package *rp)
|
PowerCap: Introduce Intel RAPL power capping driver
The Intel Running Average Power Limit (RAPL) technology provides platform
software with the ability to monitor, control, and get notifications on
power usage.
This feature is present in all Sandy Bridge and later Intel processors.
Newer models allow more fine grained controls to be applied. In RAPL,
power control is divided into domains, which include package, DRAM
controller, CPU core (Power Plane 0), graphics uncore (power plane 1), etc.
The purpose of this driver is to expose the RAPL settings to userspace.
Overall, RAPL fits in the new powercap class driver in that platform
level power capping controls are exposed via this generic interface.
This driver is based on an earlier patch from Zhang Rui.
However, while the previous work was mainly focused on thermal monitoring
the focus here is on the usability from user space perspective.
References: https://lkml.org/lkml/2011/5/26/93
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Signed-off-by: Jacob Pan <jacob.jun.pan@linux.intel.com>
Reviewed-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
2013-10-17 11:28:35 -06:00
|
|
|
{
|
2016-02-24 14:31:36 -07:00
|
|
|
if (!boot_cpu_has(X86_FEATURE_PTS) || !boot_cpu_has(X86_FEATURE_PLN))
|
|
|
|
return;
|
|
|
|
|
2016-02-24 14:31:38 -07:00
|
|
|
smp_call_function_single(rp->lead_cpu, power_limit_irq_save_cpu, rp, 1);
|
2016-02-24 14:31:36 -07:00
|
|
|
}
|
|
|
|
|
2016-11-22 14:16:02 -07:00
|
|
|
/*
|
|
|
|
* Restore per package power limit interrupt enable state. Called from cpu
|
|
|
|
* hotplug code on package removal.
|
|
|
|
*/
|
|
|
|
static void package_power_limit_irq_restore(struct rapl_package *rp)
|
2016-02-24 14:31:36 -07:00
|
|
|
{
|
2016-11-22 14:16:02 -07:00
|
|
|
u32 l, h;
|
|
|
|
|
|
|
|
if (!boot_cpu_has(X86_FEATURE_PTS) || !boot_cpu_has(X86_FEATURE_PLN))
|
|
|
|
return;
|
|
|
|
|
|
|
|
/* irq enable state not saved, nothing to restore */
|
|
|
|
if (!(rp->power_limit_irq & PACKAGE_PLN_INT_SAVED))
|
|
|
|
return;
|
2016-02-24 14:31:36 -07:00
|
|
|
|
|
|
|
rdmsr_safe(MSR_IA32_PACKAGE_THERM_INTERRUPT, &l, &h);
|
|
|
|
|
|
|
|
if (rp->power_limit_irq & PACKAGE_THERM_INT_PLN_ENABLE)
|
|
|
|
l |= PACKAGE_THERM_INT_PLN_ENABLE;
|
|
|
|
else
|
|
|
|
l &= ~PACKAGE_THERM_INT_PLN_ENABLE;
|
|
|
|
|
|
|
|
wrmsr_safe(MSR_IA32_PACKAGE_THERM_INTERRUPT, l, h);
|
PowerCap: Introduce Intel RAPL power capping driver
The Intel Running Average Power Limit (RAPL) technology provides platform
software with the ability to monitor, control, and get notifications on
power usage.
This feature is present in all Sandy Bridge and later Intel processors.
Newer models allow more fine grained controls to be applied. In RAPL,
power control is divided into domains, which include package, DRAM
controller, CPU core (Power Plane 0), graphics uncore (power plane 1), etc.
The purpose of this driver is to expose the RAPL settings to userspace.
Overall, RAPL fits in the new powercap class driver in that platform
level power capping controls are exposed via this generic interface.
This driver is based on an earlier patch from Zhang Rui.
However, while the previous work was mainly focused on thermal monitoring
the focus here is on the usability from user space perspective.
References: https://lkml.org/lkml/2011/5/26/93
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Signed-off-by: Jacob Pan <jacob.jun.pan@linux.intel.com>
Reviewed-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
2013-10-17 11:28:35 -06:00
|
|
|
}
|
|
|
|
|
2014-11-07 10:29:26 -07:00
|
|
|
static void set_floor_freq_default(struct rapl_domain *rd, bool mode)
|
|
|
|
{
|
|
|
|
int nr_powerlimit = find_nr_power_limit(rd);
|
|
|
|
|
|
|
|
/* always enable clamp such that p-state can go below OS requested
|
|
|
|
* range. power capping priority over guranteed frequency.
|
|
|
|
*/
|
|
|
|
rapl_write_data_raw(rd, PL1_CLAMP, mode);
|
|
|
|
|
|
|
|
/* some domains have pl2 */
|
|
|
|
if (nr_powerlimit > 1) {
|
|
|
|
rapl_write_data_raw(rd, PL2_ENABLE, mode);
|
|
|
|
rapl_write_data_raw(rd, PL2_CLAMP, mode);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
static void set_floor_freq_atom(struct rapl_domain *rd, bool enable)
|
|
|
|
{
|
|
|
|
static u32 power_ctrl_orig_val;
|
|
|
|
u32 mdata;
|
|
|
|
|
2015-04-29 14:13:23 -06:00
|
|
|
if (!rapl_defaults->floor_freq_reg_addr) {
|
|
|
|
pr_err("Invalid floor frequency config register\n");
|
|
|
|
return;
|
|
|
|
}
|
|
|
|
|
2014-11-07 10:29:26 -07:00
|
|
|
if (!power_ctrl_orig_val)
|
2015-11-11 10:59:29 -07:00
|
|
|
iosf_mbi_read(BT_MBI_UNIT_PMC, MBI_CR_READ,
|
|
|
|
rapl_defaults->floor_freq_reg_addr,
|
|
|
|
&power_ctrl_orig_val);
|
2014-11-07 10:29:26 -07:00
|
|
|
mdata = power_ctrl_orig_val;
|
|
|
|
if (enable) {
|
|
|
|
mdata &= ~(0x7f << 8);
|
|
|
|
mdata |= 1 << 8;
|
|
|
|
}
|
2015-11-11 10:59:29 -07:00
|
|
|
iosf_mbi_write(BT_MBI_UNIT_PMC, MBI_CR_WRITE,
|
|
|
|
rapl_defaults->floor_freq_reg_addr, mdata);
|
2014-11-07 10:29:26 -07:00
|
|
|
}
|
|
|
|
|
|
|
|
static u64 rapl_compute_time_window_core(struct rapl_package *rp, u64 value,
|
2019-07-10 07:44:30 -06:00
|
|
|
bool to_raw)
|
2014-11-07 10:29:26 -07:00
|
|
|
{
|
2019-07-10 07:44:30 -06:00
|
|
|
u64 f, y; /* fraction and exp. used for time unit */
|
2014-11-07 10:29:26 -07:00
|
|
|
|
|
|
|
/*
|
|
|
|
* Special processing based on 2^Y*(1+F/4), refer
|
|
|
|
* to Intel Software Developer's manual Vol.3B: CH 14.9.3.
|
|
|
|
*/
|
|
|
|
if (!to_raw) {
|
|
|
|
f = (value & 0x60) >> 5;
|
|
|
|
y = value & 0x1f;
|
|
|
|
value = (1 << y) * (4 + f) * rp->time_unit / 4;
|
|
|
|
} else {
|
|
|
|
do_div(value, rp->time_unit);
|
|
|
|
y = ilog2(value);
|
|
|
|
f = div64_u64(4 * (value - (1 << y)), 1 << y);
|
|
|
|
value = (y & 0x1f) | ((f & 0x3) << 5);
|
|
|
|
}
|
|
|
|
return value;
|
|
|
|
}
|
|
|
|
|
|
|
|
static u64 rapl_compute_time_window_atom(struct rapl_package *rp, u64 value,
|
2019-07-10 07:44:30 -06:00
|
|
|
bool to_raw)
|
2014-11-07 10:29:26 -07:00
|
|
|
{
|
|
|
|
/*
|
|
|
|
* Atom time unit encoding is straight forward val * time_unit,
|
|
|
|
* where time_unit is default to 1 sec. Never 0.
|
|
|
|
*/
|
|
|
|
if (!to_raw)
|
|
|
|
return (value) ? value *= rp->time_unit : rp->time_unit;
|
2019-07-10 07:44:30 -06:00
|
|
|
|
|
|
|
value = div64_u64(value, rp->time_unit);
|
2014-11-07 10:29:26 -07:00
|
|
|
|
|
|
|
return value;
|
|
|
|
}
|
|
|
|
|
2014-11-07 10:29:25 -07:00
|
|
|
static const struct rapl_defaults rapl_defaults_core = {
|
2015-04-29 14:13:23 -06:00
|
|
|
.floor_freq_reg_addr = 0,
|
2014-11-07 10:29:26 -07:00
|
|
|
.check_unit = rapl_check_unit_core,
|
|
|
|
.set_floor_freq = set_floor_freq_default,
|
|
|
|
.compute_time_window = rapl_compute_time_window_core,
|
2014-11-07 10:29:25 -07:00
|
|
|
};
|
|
|
|
|
2015-03-13 04:48:56 -06:00
|
|
|
static const struct rapl_defaults rapl_defaults_hsw_server = {
|
|
|
|
.check_unit = rapl_check_unit_core,
|
|
|
|
.set_floor_freq = set_floor_freq_default,
|
|
|
|
.compute_time_window = rapl_compute_time_window_core,
|
|
|
|
.dram_domain_energy_unit = 15300,
|
|
|
|
};
|
|
|
|
|
2015-04-29 14:13:23 -06:00
|
|
|
static const struct rapl_defaults rapl_defaults_byt = {
|
|
|
|
.floor_freq_reg_addr = IOSF_CPU_POWER_BUDGET_CTL_BYT,
|
|
|
|
.check_unit = rapl_check_unit_atom,
|
|
|
|
.set_floor_freq = set_floor_freq_atom,
|
|
|
|
.compute_time_window = rapl_compute_time_window_atom,
|
|
|
|
};
|
|
|
|
|
|
|
|
static const struct rapl_defaults rapl_defaults_tng = {
|
|
|
|
.floor_freq_reg_addr = IOSF_CPU_POWER_BUDGET_CTL_TNG,
|
2014-11-07 10:29:26 -07:00
|
|
|
.check_unit = rapl_check_unit_atom,
|
|
|
|
.set_floor_freq = set_floor_freq_atom,
|
|
|
|
.compute_time_window = rapl_compute_time_window_atom,
|
2014-11-07 10:29:25 -07:00
|
|
|
};
|
|
|
|
|
2015-04-29 14:13:23 -06:00
|
|
|
static const struct rapl_defaults rapl_defaults_ann = {
|
|
|
|
.floor_freq_reg_addr = 0,
|
|
|
|
.check_unit = rapl_check_unit_atom,
|
|
|
|
.set_floor_freq = NULL,
|
|
|
|
.compute_time_window = rapl_compute_time_window_atom,
|
|
|
|
};
|
|
|
|
|
|
|
|
static const struct rapl_defaults rapl_defaults_cht = {
|
|
|
|
.floor_freq_reg_addr = 0,
|
|
|
|
.check_unit = rapl_check_unit_atom,
|
|
|
|
.set_floor_freq = NULL,
|
|
|
|
.compute_time_window = rapl_compute_time_window_atom,
|
|
|
|
};
|
|
|
|
|
2015-03-25 15:15:52 -06:00
|
|
|
static const struct x86_cpu_id rapl_ids[] __initconst = {
|
2019-07-10 07:44:30 -06:00
|
|
|
INTEL_CPU_FAM6(SANDYBRIDGE, rapl_defaults_core),
|
|
|
|
INTEL_CPU_FAM6(SANDYBRIDGE_X, rapl_defaults_core),
|
|
|
|
|
|
|
|
INTEL_CPU_FAM6(IVYBRIDGE, rapl_defaults_core),
|
|
|
|
INTEL_CPU_FAM6(IVYBRIDGE_X, rapl_defaults_core),
|
|
|
|
|
|
|
|
INTEL_CPU_FAM6(HASWELL_CORE, rapl_defaults_core),
|
|
|
|
INTEL_CPU_FAM6(HASWELL_ULT, rapl_defaults_core),
|
|
|
|
INTEL_CPU_FAM6(HASWELL_GT3E, rapl_defaults_core),
|
|
|
|
INTEL_CPU_FAM6(HASWELL_X, rapl_defaults_hsw_server),
|
|
|
|
|
|
|
|
INTEL_CPU_FAM6(BROADWELL_CORE, rapl_defaults_core),
|
|
|
|
INTEL_CPU_FAM6(BROADWELL_GT3E, rapl_defaults_core),
|
|
|
|
INTEL_CPU_FAM6(BROADWELL_XEON_D, rapl_defaults_core),
|
|
|
|
INTEL_CPU_FAM6(BROADWELL_X, rapl_defaults_hsw_server),
|
|
|
|
|
|
|
|
INTEL_CPU_FAM6(SKYLAKE_DESKTOP, rapl_defaults_core),
|
|
|
|
INTEL_CPU_FAM6(SKYLAKE_MOBILE, rapl_defaults_core),
|
|
|
|
INTEL_CPU_FAM6(SKYLAKE_X, rapl_defaults_hsw_server),
|
|
|
|
INTEL_CPU_FAM6(KABYLAKE_MOBILE, rapl_defaults_core),
|
|
|
|
INTEL_CPU_FAM6(KABYLAKE_DESKTOP, rapl_defaults_core),
|
|
|
|
INTEL_CPU_FAM6(CANNONLAKE_MOBILE, rapl_defaults_core),
|
|
|
|
INTEL_CPU_FAM6(ICELAKE_MOBILE, rapl_defaults_core),
|
2019-07-10 07:44:35 -06:00
|
|
|
INTEL_CPU_FAM6(ICELAKE_DESKTOP, rapl_defaults_core),
|
2019-07-10 07:44:38 -06:00
|
|
|
INTEL_CPU_FAM6(ICELAKE_NNPI, rapl_defaults_core),
|
2019-07-10 07:44:36 -06:00
|
|
|
INTEL_CPU_FAM6(ICELAKE_X, rapl_defaults_hsw_server),
|
2019-07-10 07:44:37 -06:00
|
|
|
INTEL_CPU_FAM6(ICELAKE_XEON_D, rapl_defaults_hsw_server),
|
2019-07-10 07:44:30 -06:00
|
|
|
|
|
|
|
INTEL_CPU_FAM6(ATOM_SILVERMONT, rapl_defaults_byt),
|
|
|
|
INTEL_CPU_FAM6(ATOM_AIRMONT, rapl_defaults_cht),
|
|
|
|
INTEL_CPU_FAM6(ATOM_SILVERMONT_MID, rapl_defaults_tng),
|
|
|
|
INTEL_CPU_FAM6(ATOM_AIRMONT_MID, rapl_defaults_ann),
|
|
|
|
INTEL_CPU_FAM6(ATOM_GOLDMONT, rapl_defaults_core),
|
|
|
|
INTEL_CPU_FAM6(ATOM_GOLDMONT_PLUS, rapl_defaults_core),
|
|
|
|
INTEL_CPU_FAM6(ATOM_GOLDMONT_X, rapl_defaults_core),
|
|
|
|
INTEL_CPU_FAM6(ATOM_TREMONT_X, rapl_defaults_core),
|
|
|
|
|
|
|
|
INTEL_CPU_FAM6(XEON_PHI_KNL, rapl_defaults_hsw_server),
|
|
|
|
INTEL_CPU_FAM6(XEON_PHI_KNM, rapl_defaults_hsw_server),
|
PowerCap: Introduce Intel RAPL power capping driver
The Intel Running Average Power Limit (RAPL) technology provides platform
software with the ability to monitor, control, and get notifications on
power usage.
This feature is present in all Sandy Bridge and later Intel processors.
Newer models allow more fine grained controls to be applied. In RAPL,
power control is divided into domains, which include package, DRAM
controller, CPU core (Power Plane 0), graphics uncore (power plane 1), etc.
The purpose of this driver is to expose the RAPL settings to userspace.
Overall, RAPL fits in the new powercap class driver in that platform
level power capping controls are exposed via this generic interface.
This driver is based on an earlier patch from Zhang Rui.
However, while the previous work was mainly focused on thermal monitoring
the focus here is on the usability from user space perspective.
References: https://lkml.org/lkml/2011/5/26/93
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Signed-off-by: Jacob Pan <jacob.jun.pan@linux.intel.com>
Reviewed-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
2013-10-17 11:28:35 -06:00
|
|
|
{}
|
|
|
|
};
|
2019-07-10 07:44:30 -06:00
|
|
|
|
PowerCap: Introduce Intel RAPL power capping driver
The Intel Running Average Power Limit (RAPL) technology provides platform
software with the ability to monitor, control, and get notifications on
power usage.
This feature is present in all Sandy Bridge and later Intel processors.
Newer models allow more fine grained controls to be applied. In RAPL,
power control is divided into domains, which include package, DRAM
controller, CPU core (Power Plane 0), graphics uncore (power plane 1), etc.
The purpose of this driver is to expose the RAPL settings to userspace.
Overall, RAPL fits in the new powercap class driver in that platform
level power capping controls are exposed via this generic interface.
This driver is based on an earlier patch from Zhang Rui.
However, while the previous work was mainly focused on thermal monitoring
the focus here is on the usability from user space perspective.
References: https://lkml.org/lkml/2011/5/26/93
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Signed-off-by: Jacob Pan <jacob.jun.pan@linux.intel.com>
Reviewed-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
2013-10-17 11:28:35 -06:00
|
|
|
MODULE_DEVICE_TABLE(x86cpu, rapl_ids);
|
|
|
|
|
2016-11-22 14:15:58 -07:00
|
|
|
/* Read once for all raw primitive data for domains */
|
|
|
|
static void rapl_update_domain_data(struct rapl_package *rp)
|
PowerCap: Introduce Intel RAPL power capping driver
The Intel Running Average Power Limit (RAPL) technology provides platform
software with the ability to monitor, control, and get notifications on
power usage.
This feature is present in all Sandy Bridge and later Intel processors.
Newer models allow more fine grained controls to be applied. In RAPL,
power control is divided into domains, which include package, DRAM
controller, CPU core (Power Plane 0), graphics uncore (power plane 1), etc.
The purpose of this driver is to expose the RAPL settings to userspace.
Overall, RAPL fits in the new powercap class driver in that platform
level power capping controls are exposed via this generic interface.
This driver is based on an earlier patch from Zhang Rui.
However, while the previous work was mainly focused on thermal monitoring
the focus here is on the usability from user space perspective.
References: https://lkml.org/lkml/2011/5/26/93
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Signed-off-by: Jacob Pan <jacob.jun.pan@linux.intel.com>
Reviewed-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
2013-10-17 11:28:35 -06:00
|
|
|
{
|
|
|
|
int dmn, prim;
|
|
|
|
u64 val;
|
|
|
|
|
2016-11-22 14:15:58 -07:00
|
|
|
for (dmn = 0; dmn < rp->nr_domains; dmn++) {
|
2019-05-13 11:58:53 -06:00
|
|
|
pr_debug("update %s domain %s data\n", rp->name,
|
2016-11-22 14:15:58 -07:00
|
|
|
rp->domains[dmn].name);
|
|
|
|
/* exclude non-raw primitives */
|
|
|
|
for (prim = 0; prim < NR_RAW_PRIMITIVES; prim++) {
|
|
|
|
if (!rapl_read_data_raw(&rp->domains[dmn], prim,
|
|
|
|
rpi[prim].unit, &val))
|
2019-07-10 07:44:30 -06:00
|
|
|
rp->domains[dmn].rdd.primitives[prim] = val;
|
PowerCap: Introduce Intel RAPL power capping driver
The Intel Running Average Power Limit (RAPL) technology provides platform
software with the ability to monitor, control, and get notifications on
power usage.
This feature is present in all Sandy Bridge and later Intel processors.
Newer models allow more fine grained controls to be applied. In RAPL,
power control is divided into domains, which include package, DRAM
controller, CPU core (Power Plane 0), graphics uncore (power plane 1), etc.
The purpose of this driver is to expose the RAPL settings to userspace.
Overall, RAPL fits in the new powercap class driver in that platform
level power capping controls are exposed via this generic interface.
This driver is based on an earlier patch from Zhang Rui.
However, while the previous work was mainly focused on thermal monitoring
the focus here is on the usability from user space perspective.
References: https://lkml.org/lkml/2011/5/26/93
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Signed-off-by: Jacob Pan <jacob.jun.pan@linux.intel.com>
Reviewed-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
2013-10-17 11:28:35 -06:00
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
}
|
|
|
|
|
|
|
|
static int rapl_package_register_powercap(struct rapl_package *rp)
|
|
|
|
{
|
|
|
|
struct rapl_domain *rd;
|
|
|
|
struct powercap_zone *power_zone = NULL;
|
2018-01-17 03:30:34 -07:00
|
|
|
int nr_pl, ret;
|
2016-11-22 14:15:58 -07:00
|
|
|
|
|
|
|
/* Update the domain data of the new package */
|
|
|
|
rapl_update_domain_data(rp);
|
PowerCap: Introduce Intel RAPL power capping driver
The Intel Running Average Power Limit (RAPL) technology provides platform
software with the ability to monitor, control, and get notifications on
power usage.
This feature is present in all Sandy Bridge and later Intel processors.
Newer models allow more fine grained controls to be applied. In RAPL,
power control is divided into domains, which include package, DRAM
controller, CPU core (Power Plane 0), graphics uncore (power plane 1), etc.
The purpose of this driver is to expose the RAPL settings to userspace.
Overall, RAPL fits in the new powercap class driver in that platform
level power capping controls are exposed via this generic interface.
This driver is based on an earlier patch from Zhang Rui.
However, while the previous work was mainly focused on thermal monitoring
the focus here is on the usability from user space perspective.
References: https://lkml.org/lkml/2011/5/26/93
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Signed-off-by: Jacob Pan <jacob.jun.pan@linux.intel.com>
Reviewed-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
2013-10-17 11:28:35 -06:00
|
|
|
|
2019-07-10 07:44:30 -06:00
|
|
|
/* first we register package domain as the parent zone */
|
PowerCap: Introduce Intel RAPL power capping driver
The Intel Running Average Power Limit (RAPL) technology provides platform
software with the ability to monitor, control, and get notifications on
power usage.
This feature is present in all Sandy Bridge and later Intel processors.
Newer models allow more fine grained controls to be applied. In RAPL,
power control is divided into domains, which include package, DRAM
controller, CPU core (Power Plane 0), graphics uncore (power plane 1), etc.
The purpose of this driver is to expose the RAPL settings to userspace.
Overall, RAPL fits in the new powercap class driver in that platform
level power capping controls are exposed via this generic interface.
This driver is based on an earlier patch from Zhang Rui.
However, while the previous work was mainly focused on thermal monitoring
the focus here is on the usability from user space perspective.
References: https://lkml.org/lkml/2011/5/26/93
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Signed-off-by: Jacob Pan <jacob.jun.pan@linux.intel.com>
Reviewed-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
2013-10-17 11:28:35 -06:00
|
|
|
for (rd = rp->domains; rd < rp->domains + rp->nr_domains; rd++) {
|
|
|
|
if (rd->id == RAPL_DOMAIN_PACKAGE) {
|
|
|
|
nr_pl = find_nr_power_limit(rd);
|
2019-05-13 11:58:53 -06:00
|
|
|
pr_debug("register package domain %s\n", rp->name);
|
PowerCap: Introduce Intel RAPL power capping driver
The Intel Running Average Power Limit (RAPL) technology provides platform
software with the ability to monitor, control, and get notifications on
power usage.
This feature is present in all Sandy Bridge and later Intel processors.
Newer models allow more fine grained controls to be applied. In RAPL,
power control is divided into domains, which include package, DRAM
controller, CPU core (Power Plane 0), graphics uncore (power plane 1), etc.
The purpose of this driver is to expose the RAPL settings to userspace.
Overall, RAPL fits in the new powercap class driver in that platform
level power capping controls are exposed via this generic interface.
This driver is based on an earlier patch from Zhang Rui.
However, while the previous work was mainly focused on thermal monitoring
the focus here is on the usability from user space perspective.
References: https://lkml.org/lkml/2011/5/26/93
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Signed-off-by: Jacob Pan <jacob.jun.pan@linux.intel.com>
Reviewed-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
2013-10-17 11:28:35 -06:00
|
|
|
power_zone = powercap_register_zone(&rd->power_zone,
|
2019-07-10 07:44:30 -06:00
|
|
|
rp->priv->control_type, rp->name,
|
|
|
|
NULL, &zone_ops[rd->id], nr_pl,
|
|
|
|
&constraint_ops);
|
PowerCap: Introduce Intel RAPL power capping driver
The Intel Running Average Power Limit (RAPL) technology provides platform
software with the ability to monitor, control, and get notifications on
power usage.
This feature is present in all Sandy Bridge and later Intel processors.
Newer models allow more fine grained controls to be applied. In RAPL,
power control is divided into domains, which include package, DRAM
controller, CPU core (Power Plane 0), graphics uncore (power plane 1), etc.
The purpose of this driver is to expose the RAPL settings to userspace.
Overall, RAPL fits in the new powercap class driver in that platform
level power capping controls are exposed via this generic interface.
This driver is based on an earlier patch from Zhang Rui.
However, while the previous work was mainly focused on thermal monitoring
the focus here is on the usability from user space perspective.
References: https://lkml.org/lkml/2011/5/26/93
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Signed-off-by: Jacob Pan <jacob.jun.pan@linux.intel.com>
Reviewed-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
2013-10-17 11:28:35 -06:00
|
|
|
if (IS_ERR(power_zone)) {
|
2019-05-13 11:58:53 -06:00
|
|
|
pr_debug("failed to register power zone %s\n",
|
2019-07-10 07:44:30 -06:00
|
|
|
rp->name);
|
2016-11-22 14:15:58 -07:00
|
|
|
return PTR_ERR(power_zone);
|
PowerCap: Introduce Intel RAPL power capping driver
The Intel Running Average Power Limit (RAPL) technology provides platform
software with the ability to monitor, control, and get notifications on
power usage.
This feature is present in all Sandy Bridge and later Intel processors.
Newer models allow more fine grained controls to be applied. In RAPL,
power control is divided into domains, which include package, DRAM
controller, CPU core (Power Plane 0), graphics uncore (power plane 1), etc.
The purpose of this driver is to expose the RAPL settings to userspace.
Overall, RAPL fits in the new powercap class driver in that platform
level power capping controls are exposed via this generic interface.
This driver is based on an earlier patch from Zhang Rui.
However, while the previous work was mainly focused on thermal monitoring
the focus here is on the usability from user space perspective.
References: https://lkml.org/lkml/2011/5/26/93
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Signed-off-by: Jacob Pan <jacob.jun.pan@linux.intel.com>
Reviewed-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
2013-10-17 11:28:35 -06:00
|
|
|
}
|
|
|
|
/* track parent zone in per package/socket data */
|
|
|
|
rp->power_zone = power_zone;
|
|
|
|
/* done, only one package domain per socket */
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
if (!power_zone) {
|
|
|
|
pr_err("no package domain found, unknown topology!\n");
|
2016-11-22 14:15:58 -07:00
|
|
|
return -ENODEV;
|
PowerCap: Introduce Intel RAPL power capping driver
The Intel Running Average Power Limit (RAPL) technology provides platform
software with the ability to monitor, control, and get notifications on
power usage.
This feature is present in all Sandy Bridge and later Intel processors.
Newer models allow more fine grained controls to be applied. In RAPL,
power control is divided into domains, which include package, DRAM
controller, CPU core (Power Plane 0), graphics uncore (power plane 1), etc.
The purpose of this driver is to expose the RAPL settings to userspace.
Overall, RAPL fits in the new powercap class driver in that platform
level power capping controls are exposed via this generic interface.
This driver is based on an earlier patch from Zhang Rui.
However, while the previous work was mainly focused on thermal monitoring
the focus here is on the usability from user space perspective.
References: https://lkml.org/lkml/2011/5/26/93
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Signed-off-by: Jacob Pan <jacob.jun.pan@linux.intel.com>
Reviewed-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
2013-10-17 11:28:35 -06:00
|
|
|
}
|
2019-07-10 07:44:30 -06:00
|
|
|
/* now register domains as children of the socket/package */
|
PowerCap: Introduce Intel RAPL power capping driver
The Intel Running Average Power Limit (RAPL) technology provides platform
software with the ability to monitor, control, and get notifications on
power usage.
This feature is present in all Sandy Bridge and later Intel processors.
Newer models allow more fine grained controls to be applied. In RAPL,
power control is divided into domains, which include package, DRAM
controller, CPU core (Power Plane 0), graphics uncore (power plane 1), etc.
The purpose of this driver is to expose the RAPL settings to userspace.
Overall, RAPL fits in the new powercap class driver in that platform
level power capping controls are exposed via this generic interface.
This driver is based on an earlier patch from Zhang Rui.
However, while the previous work was mainly focused on thermal monitoring
the focus here is on the usability from user space perspective.
References: https://lkml.org/lkml/2011/5/26/93
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Signed-off-by: Jacob Pan <jacob.jun.pan@linux.intel.com>
Reviewed-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
2013-10-17 11:28:35 -06:00
|
|
|
for (rd = rp->domains; rd < rp->domains + rp->nr_domains; rd++) {
|
|
|
|
if (rd->id == RAPL_DOMAIN_PACKAGE)
|
|
|
|
continue;
|
|
|
|
/* number of power limits per domain varies */
|
|
|
|
nr_pl = find_nr_power_limit(rd);
|
|
|
|
power_zone = powercap_register_zone(&rd->power_zone,
|
2019-07-10 07:44:30 -06:00
|
|
|
rp->priv->control_type,
|
|
|
|
rd->name, rp->power_zone,
|
|
|
|
&zone_ops[rd->id], nr_pl,
|
|
|
|
&constraint_ops);
|
PowerCap: Introduce Intel RAPL power capping driver
The Intel Running Average Power Limit (RAPL) technology provides platform
software with the ability to monitor, control, and get notifications on
power usage.
This feature is present in all Sandy Bridge and later Intel processors.
Newer models allow more fine grained controls to be applied. In RAPL,
power control is divided into domains, which include package, DRAM
controller, CPU core (Power Plane 0), graphics uncore (power plane 1), etc.
The purpose of this driver is to expose the RAPL settings to userspace.
Overall, RAPL fits in the new powercap class driver in that platform
level power capping controls are exposed via this generic interface.
This driver is based on an earlier patch from Zhang Rui.
However, while the previous work was mainly focused on thermal monitoring
the focus here is on the usability from user space perspective.
References: https://lkml.org/lkml/2011/5/26/93
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Signed-off-by: Jacob Pan <jacob.jun.pan@linux.intel.com>
Reviewed-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
2013-10-17 11:28:35 -06:00
|
|
|
|
|
|
|
if (IS_ERR(power_zone)) {
|
2019-05-13 11:58:53 -06:00
|
|
|
pr_debug("failed to register power_zone, %s:%s\n",
|
2019-07-10 07:44:30 -06:00
|
|
|
rp->name, rd->name);
|
PowerCap: Introduce Intel RAPL power capping driver
The Intel Running Average Power Limit (RAPL) technology provides platform
software with the ability to monitor, control, and get notifications on
power usage.
This feature is present in all Sandy Bridge and later Intel processors.
Newer models allow more fine grained controls to be applied. In RAPL,
power control is divided into domains, which include package, DRAM
controller, CPU core (Power Plane 0), graphics uncore (power plane 1), etc.
The purpose of this driver is to expose the RAPL settings to userspace.
Overall, RAPL fits in the new powercap class driver in that platform
level power capping controls are exposed via this generic interface.
This driver is based on an earlier patch from Zhang Rui.
However, while the previous work was mainly focused on thermal monitoring
the focus here is on the usability from user space perspective.
References: https://lkml.org/lkml/2011/5/26/93
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Signed-off-by: Jacob Pan <jacob.jun.pan@linux.intel.com>
Reviewed-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
2013-10-17 11:28:35 -06:00
|
|
|
ret = PTR_ERR(power_zone);
|
|
|
|
goto err_cleanup;
|
|
|
|
}
|
|
|
|
}
|
2016-11-22 14:15:58 -07:00
|
|
|
return 0;
|
PowerCap: Introduce Intel RAPL power capping driver
The Intel Running Average Power Limit (RAPL) technology provides platform
software with the ability to monitor, control, and get notifications on
power usage.
This feature is present in all Sandy Bridge and later Intel processors.
Newer models allow more fine grained controls to be applied. In RAPL,
power control is divided into domains, which include package, DRAM
controller, CPU core (Power Plane 0), graphics uncore (power plane 1), etc.
The purpose of this driver is to expose the RAPL settings to userspace.
Overall, RAPL fits in the new powercap class driver in that platform
level power capping controls are exposed via this generic interface.
This driver is based on an earlier patch from Zhang Rui.
However, while the previous work was mainly focused on thermal monitoring
the focus here is on the usability from user space perspective.
References: https://lkml.org/lkml/2011/5/26/93
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Signed-off-by: Jacob Pan <jacob.jun.pan@linux.intel.com>
Reviewed-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
2013-10-17 11:28:35 -06:00
|
|
|
|
|
|
|
err_cleanup:
|
2016-11-22 14:16:02 -07:00
|
|
|
/*
|
|
|
|
* Clean up previously initialized domains within the package if we
|
PowerCap: Introduce Intel RAPL power capping driver
The Intel Running Average Power Limit (RAPL) technology provides platform
software with the ability to monitor, control, and get notifications on
power usage.
This feature is present in all Sandy Bridge and later Intel processors.
Newer models allow more fine grained controls to be applied. In RAPL,
power control is divided into domains, which include package, DRAM
controller, CPU core (Power Plane 0), graphics uncore (power plane 1), etc.
The purpose of this driver is to expose the RAPL settings to userspace.
Overall, RAPL fits in the new powercap class driver in that platform
level power capping controls are exposed via this generic interface.
This driver is based on an earlier patch from Zhang Rui.
However, while the previous work was mainly focused on thermal monitoring
the focus here is on the usability from user space perspective.
References: https://lkml.org/lkml/2011/5/26/93
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Signed-off-by: Jacob Pan <jacob.jun.pan@linux.intel.com>
Reviewed-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
2013-10-17 11:28:35 -06:00
|
|
|
* failed after the first domain setup.
|
|
|
|
*/
|
|
|
|
while (--rd >= rp->domains) {
|
2019-05-13 11:58:53 -06:00
|
|
|
pr_debug("unregister %s domain %s\n", rp->name, rd->name);
|
2019-07-10 07:44:30 -06:00
|
|
|
powercap_unregister_zone(rp->priv->control_type,
|
|
|
|
&rd->power_zone);
|
PowerCap: Introduce Intel RAPL power capping driver
The Intel Running Average Power Limit (RAPL) technology provides platform
software with the ability to monitor, control, and get notifications on
power usage.
This feature is present in all Sandy Bridge and later Intel processors.
Newer models allow more fine grained controls to be applied. In RAPL,
power control is divided into domains, which include package, DRAM
controller, CPU core (Power Plane 0), graphics uncore (power plane 1), etc.
The purpose of this driver is to expose the RAPL settings to userspace.
Overall, RAPL fits in the new powercap class driver in that platform
level power capping controls are exposed via this generic interface.
This driver is based on an earlier patch from Zhang Rui.
However, while the previous work was mainly focused on thermal monitoring
the focus here is on the usability from user space perspective.
References: https://lkml.org/lkml/2011/5/26/93
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Signed-off-by: Jacob Pan <jacob.jun.pan@linux.intel.com>
Reviewed-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
2013-10-17 11:28:35 -06:00
|
|
|
}
|
|
|
|
|
|
|
|
return ret;
|
|
|
|
}
|
|
|
|
|
2019-07-10 07:44:30 -06:00
|
|
|
int rapl_add_platform_domain(struct rapl_if_priv *priv)
|
2016-04-17 16:03:01 -06:00
|
|
|
{
|
|
|
|
struct rapl_domain *rd;
|
|
|
|
struct powercap_zone *power_zone;
|
intel_rapl: cleanup some functions
Previously, there are three functions:
rapl_register_psys(), which registers platform rapl domain.
rapl_register_powercap(), which registers powercap control type.
rapl_unregsiter_powercap(), which unregisters platform rapl domain and
powercap control type.
This is confusing as the function name does not describe what it does
clearly.
With this patch, the three functions are removed, and two new functions
rapl_register_platform_domain()/rapl_unregister_platform_domain() are
introduced instead, and they do exactly what their function name describes.
Plus, as part of the common code, hardcoded MSR accesses in these functions
are converted to follow the abstracted register access.
Reviewed-by: Pandruvada, Srinivas <srinivas.pandruvada@intel.com>
Tested-by: Pandruvada, Srinivas <srinivas.pandruvada@intel.com>
Signed-off-by: Zhang Rui <rui.zhang@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
2019-07-10 07:44:28 -06:00
|
|
|
struct reg_action ra;
|
|
|
|
int ret;
|
2016-04-17 16:03:01 -06:00
|
|
|
|
intel_rapl: cleanup some functions
Previously, there are three functions:
rapl_register_psys(), which registers platform rapl domain.
rapl_register_powercap(), which registers powercap control type.
rapl_unregsiter_powercap(), which unregisters platform rapl domain and
powercap control type.
This is confusing as the function name does not describe what it does
clearly.
With this patch, the three functions are removed, and two new functions
rapl_register_platform_domain()/rapl_unregister_platform_domain() are
introduced instead, and they do exactly what their function name describes.
Plus, as part of the common code, hardcoded MSR accesses in these functions
are converted to follow the abstracted register access.
Reviewed-by: Pandruvada, Srinivas <srinivas.pandruvada@intel.com>
Tested-by: Pandruvada, Srinivas <srinivas.pandruvada@intel.com>
Signed-off-by: Zhang Rui <rui.zhang@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
2019-07-10 07:44:28 -06:00
|
|
|
ra.reg = priv->regs[RAPL_DOMAIN_PLATFORM][RAPL_DOMAIN_REG_STATUS];
|
|
|
|
ra.mask = ~0;
|
|
|
|
ret = priv->read_raw(0, &ra);
|
|
|
|
if (ret || !ra.value)
|
2016-04-17 16:03:01 -06:00
|
|
|
return -ENODEV;
|
|
|
|
|
intel_rapl: cleanup some functions
Previously, there are three functions:
rapl_register_psys(), which registers platform rapl domain.
rapl_register_powercap(), which registers powercap control type.
rapl_unregsiter_powercap(), which unregisters platform rapl domain and
powercap control type.
This is confusing as the function name does not describe what it does
clearly.
With this patch, the three functions are removed, and two new functions
rapl_register_platform_domain()/rapl_unregister_platform_domain() are
introduced instead, and they do exactly what their function name describes.
Plus, as part of the common code, hardcoded MSR accesses in these functions
are converted to follow the abstracted register access.
Reviewed-by: Pandruvada, Srinivas <srinivas.pandruvada@intel.com>
Tested-by: Pandruvada, Srinivas <srinivas.pandruvada@intel.com>
Signed-off-by: Zhang Rui <rui.zhang@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
2019-07-10 07:44:28 -06:00
|
|
|
ra.reg = priv->regs[RAPL_DOMAIN_PLATFORM][RAPL_DOMAIN_REG_LIMIT];
|
|
|
|
ra.mask = ~0;
|
|
|
|
ret = priv->read_raw(0, &ra);
|
|
|
|
if (ret || !ra.value)
|
2016-04-17 16:03:01 -06:00
|
|
|
return -ENODEV;
|
|
|
|
|
|
|
|
rd = kzalloc(sizeof(*rd), GFP_KERNEL);
|
|
|
|
if (!rd)
|
|
|
|
return -ENOMEM;
|
|
|
|
|
|
|
|
rd->name = rapl_domain_names[RAPL_DOMAIN_PLATFORM];
|
|
|
|
rd->id = RAPL_DOMAIN_PLATFORM;
|
2019-07-10 07:44:30 -06:00
|
|
|
rd->regs[RAPL_DOMAIN_REG_LIMIT] =
|
|
|
|
priv->regs[RAPL_DOMAIN_PLATFORM][RAPL_DOMAIN_REG_LIMIT];
|
|
|
|
rd->regs[RAPL_DOMAIN_REG_STATUS] =
|
|
|
|
priv->regs[RAPL_DOMAIN_PLATFORM][RAPL_DOMAIN_REG_STATUS];
|
2016-04-17 16:03:01 -06:00
|
|
|
rd->rpl[0].prim_id = PL1_ENABLE;
|
|
|
|
rd->rpl[0].name = pl1_name;
|
|
|
|
rd->rpl[1].prim_id = PL2_ENABLE;
|
|
|
|
rd->rpl[1].name = pl2_name;
|
intel_rapl: cleanup some functions
Previously, there are three functions:
rapl_register_psys(), which registers platform rapl domain.
rapl_register_powercap(), which registers powercap control type.
rapl_unregsiter_powercap(), which unregisters platform rapl domain and
powercap control type.
This is confusing as the function name does not describe what it does
clearly.
With this patch, the three functions are removed, and two new functions
rapl_register_platform_domain()/rapl_unregister_platform_domain() are
introduced instead, and they do exactly what their function name describes.
Plus, as part of the common code, hardcoded MSR accesses in these functions
are converted to follow the abstracted register access.
Reviewed-by: Pandruvada, Srinivas <srinivas.pandruvada@intel.com>
Tested-by: Pandruvada, Srinivas <srinivas.pandruvada@intel.com>
Signed-off-by: Zhang Rui <rui.zhang@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
2019-07-10 07:44:28 -06:00
|
|
|
rd->rp = rapl_find_package_domain(0, priv);
|
2016-04-17 16:03:01 -06:00
|
|
|
|
intel_rapl: cleanup some functions
Previously, there are three functions:
rapl_register_psys(), which registers platform rapl domain.
rapl_register_powercap(), which registers powercap control type.
rapl_unregsiter_powercap(), which unregisters platform rapl domain and
powercap control type.
This is confusing as the function name does not describe what it does
clearly.
With this patch, the three functions are removed, and two new functions
rapl_register_platform_domain()/rapl_unregister_platform_domain() are
introduced instead, and they do exactly what their function name describes.
Plus, as part of the common code, hardcoded MSR accesses in these functions
are converted to follow the abstracted register access.
Reviewed-by: Pandruvada, Srinivas <srinivas.pandruvada@intel.com>
Tested-by: Pandruvada, Srinivas <srinivas.pandruvada@intel.com>
Signed-off-by: Zhang Rui <rui.zhang@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
2019-07-10 07:44:28 -06:00
|
|
|
power_zone = powercap_register_zone(&rd->power_zone, priv->control_type,
|
2016-04-17 16:03:01 -06:00
|
|
|
"psys", NULL,
|
|
|
|
&zone_ops[RAPL_DOMAIN_PLATFORM],
|
|
|
|
2, &constraint_ops);
|
|
|
|
|
|
|
|
if (IS_ERR(power_zone)) {
|
|
|
|
kfree(rd);
|
|
|
|
return PTR_ERR(power_zone);
|
|
|
|
}
|
|
|
|
|
intel_rapl: cleanup some functions
Previously, there are three functions:
rapl_register_psys(), which registers platform rapl domain.
rapl_register_powercap(), which registers powercap control type.
rapl_unregsiter_powercap(), which unregisters platform rapl domain and
powercap control type.
This is confusing as the function name does not describe what it does
clearly.
With this patch, the three functions are removed, and two new functions
rapl_register_platform_domain()/rapl_unregister_platform_domain() are
introduced instead, and they do exactly what their function name describes.
Plus, as part of the common code, hardcoded MSR accesses in these functions
are converted to follow the abstracted register access.
Reviewed-by: Pandruvada, Srinivas <srinivas.pandruvada@intel.com>
Tested-by: Pandruvada, Srinivas <srinivas.pandruvada@intel.com>
Signed-off-by: Zhang Rui <rui.zhang@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
2019-07-10 07:44:28 -06:00
|
|
|
priv->platform_rapl_domain = rd;
|
2016-04-17 16:03:01 -06:00
|
|
|
|
|
|
|
return 0;
|
|
|
|
}
|
2019-07-10 07:44:30 -06:00
|
|
|
EXPORT_SYMBOL_GPL(rapl_add_platform_domain);
|
2016-04-17 16:03:01 -06:00
|
|
|
|
2019-07-10 07:44:30 -06:00
|
|
|
void rapl_remove_platform_domain(struct rapl_if_priv *priv)
|
PowerCap: Introduce Intel RAPL power capping driver
The Intel Running Average Power Limit (RAPL) technology provides platform
software with the ability to monitor, control, and get notifications on
power usage.
This feature is present in all Sandy Bridge and later Intel processors.
Newer models allow more fine grained controls to be applied. In RAPL,
power control is divided into domains, which include package, DRAM
controller, CPU core (Power Plane 0), graphics uncore (power plane 1), etc.
The purpose of this driver is to expose the RAPL settings to userspace.
Overall, RAPL fits in the new powercap class driver in that platform
level power capping controls are exposed via this generic interface.
This driver is based on an earlier patch from Zhang Rui.
However, while the previous work was mainly focused on thermal monitoring
the focus here is on the usability from user space perspective.
References: https://lkml.org/lkml/2011/5/26/93
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Signed-off-by: Jacob Pan <jacob.jun.pan@linux.intel.com>
Reviewed-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
2013-10-17 11:28:35 -06:00
|
|
|
{
|
intel_rapl: cleanup some functions
Previously, there are three functions:
rapl_register_psys(), which registers platform rapl domain.
rapl_register_powercap(), which registers powercap control type.
rapl_unregsiter_powercap(), which unregisters platform rapl domain and
powercap control type.
This is confusing as the function name does not describe what it does
clearly.
With this patch, the three functions are removed, and two new functions
rapl_register_platform_domain()/rapl_unregister_platform_domain() are
introduced instead, and they do exactly what their function name describes.
Plus, as part of the common code, hardcoded MSR accesses in these functions
are converted to follow the abstracted register access.
Reviewed-by: Pandruvada, Srinivas <srinivas.pandruvada@intel.com>
Tested-by: Pandruvada, Srinivas <srinivas.pandruvada@intel.com>
Signed-off-by: Zhang Rui <rui.zhang@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
2019-07-10 07:44:28 -06:00
|
|
|
if (priv->platform_rapl_domain) {
|
|
|
|
powercap_unregister_zone(priv->control_type,
|
2019-07-10 07:44:30 -06:00
|
|
|
&priv->platform_rapl_domain->power_zone);
|
intel_rapl: cleanup some functions
Previously, there are three functions:
rapl_register_psys(), which registers platform rapl domain.
rapl_register_powercap(), which registers powercap control type.
rapl_unregsiter_powercap(), which unregisters platform rapl domain and
powercap control type.
This is confusing as the function name does not describe what it does
clearly.
With this patch, the three functions are removed, and two new functions
rapl_register_platform_domain()/rapl_unregister_platform_domain() are
introduced instead, and they do exactly what their function name describes.
Plus, as part of the common code, hardcoded MSR accesses in these functions
are converted to follow the abstracted register access.
Reviewed-by: Pandruvada, Srinivas <srinivas.pandruvada@intel.com>
Tested-by: Pandruvada, Srinivas <srinivas.pandruvada@intel.com>
Signed-off-by: Zhang Rui <rui.zhang@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
2019-07-10 07:44:28 -06:00
|
|
|
kfree(priv->platform_rapl_domain);
|
PowerCap: Introduce Intel RAPL power capping driver
The Intel Running Average Power Limit (RAPL) technology provides platform
software with the ability to monitor, control, and get notifications on
power usage.
This feature is present in all Sandy Bridge and later Intel processors.
Newer models allow more fine grained controls to be applied. In RAPL,
power control is divided into domains, which include package, DRAM
controller, CPU core (Power Plane 0), graphics uncore (power plane 1), etc.
The purpose of this driver is to expose the RAPL settings to userspace.
Overall, RAPL fits in the new powercap class driver in that platform
level power capping controls are exposed via this generic interface.
This driver is based on an earlier patch from Zhang Rui.
However, while the previous work was mainly focused on thermal monitoring
the focus here is on the usability from user space perspective.
References: https://lkml.org/lkml/2011/5/26/93
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Signed-off-by: Jacob Pan <jacob.jun.pan@linux.intel.com>
Reviewed-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
2013-10-17 11:28:35 -06:00
|
|
|
}
|
|
|
|
}
|
2019-07-10 07:44:30 -06:00
|
|
|
EXPORT_SYMBOL_GPL(rapl_remove_platform_domain);
|
PowerCap: Introduce Intel RAPL power capping driver
The Intel Running Average Power Limit (RAPL) technology provides platform
software with the ability to monitor, control, and get notifications on
power usage.
This feature is present in all Sandy Bridge and later Intel processors.
Newer models allow more fine grained controls to be applied. In RAPL,
power control is divided into domains, which include package, DRAM
controller, CPU core (Power Plane 0), graphics uncore (power plane 1), etc.
The purpose of this driver is to expose the RAPL settings to userspace.
Overall, RAPL fits in the new powercap class driver in that platform
level power capping controls are exposed via this generic interface.
This driver is based on an earlier patch from Zhang Rui.
However, while the previous work was mainly focused on thermal monitoring
the focus here is on the usability from user space perspective.
References: https://lkml.org/lkml/2011/5/26/93
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Signed-off-by: Jacob Pan <jacob.jun.pan@linux.intel.com>
Reviewed-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
2013-10-17 11:28:35 -06:00
|
|
|
|
2019-07-10 07:44:26 -06:00
|
|
|
static int rapl_check_domain(int cpu, int domain, struct rapl_package *rp)
|
PowerCap: Introduce Intel RAPL power capping driver
The Intel Running Average Power Limit (RAPL) technology provides platform
software with the ability to monitor, control, and get notifications on
power usage.
This feature is present in all Sandy Bridge and later Intel processors.
Newer models allow more fine grained controls to be applied. In RAPL,
power control is divided into domains, which include package, DRAM
controller, CPU core (Power Plane 0), graphics uncore (power plane 1), etc.
The purpose of this driver is to expose the RAPL settings to userspace.
Overall, RAPL fits in the new powercap class driver in that platform
level power capping controls are exposed via this generic interface.
This driver is based on an earlier patch from Zhang Rui.
However, while the previous work was mainly focused on thermal monitoring
the focus here is on the usability from user space perspective.
References: https://lkml.org/lkml/2011/5/26/93
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Signed-off-by: Jacob Pan <jacob.jun.pan@linux.intel.com>
Reviewed-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
2013-10-17 11:28:35 -06:00
|
|
|
{
|
2019-07-10 07:44:29 -06:00
|
|
|
struct reg_action ra;
|
PowerCap: Introduce Intel RAPL power capping driver
The Intel Running Average Power Limit (RAPL) technology provides platform
software with the ability to monitor, control, and get notifications on
power usage.
This feature is present in all Sandy Bridge and later Intel processors.
Newer models allow more fine grained controls to be applied. In RAPL,
power control is divided into domains, which include package, DRAM
controller, CPU core (Power Plane 0), graphics uncore (power plane 1), etc.
The purpose of this driver is to expose the RAPL settings to userspace.
Overall, RAPL fits in the new powercap class driver in that platform
level power capping controls are exposed via this generic interface.
This driver is based on an earlier patch from Zhang Rui.
However, while the previous work was mainly focused on thermal monitoring
the focus here is on the usability from user space perspective.
References: https://lkml.org/lkml/2011/5/26/93
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Signed-off-by: Jacob Pan <jacob.jun.pan@linux.intel.com>
Reviewed-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
2013-10-17 11:28:35 -06:00
|
|
|
|
|
|
|
switch (domain) {
|
|
|
|
case RAPL_DOMAIN_PACKAGE:
|
|
|
|
case RAPL_DOMAIN_PP0:
|
|
|
|
case RAPL_DOMAIN_PP1:
|
|
|
|
case RAPL_DOMAIN_DRAM:
|
2019-07-10 07:44:29 -06:00
|
|
|
ra.reg = rp->priv->regs[domain][RAPL_DOMAIN_REG_STATUS];
|
PowerCap: Introduce Intel RAPL power capping driver
The Intel Running Average Power Limit (RAPL) technology provides platform
software with the ability to monitor, control, and get notifications on
power usage.
This feature is present in all Sandy Bridge and later Intel processors.
Newer models allow more fine grained controls to be applied. In RAPL,
power control is divided into domains, which include package, DRAM
controller, CPU core (Power Plane 0), graphics uncore (power plane 1), etc.
The purpose of this driver is to expose the RAPL settings to userspace.
Overall, RAPL fits in the new powercap class driver in that platform
level power capping controls are exposed via this generic interface.
This driver is based on an earlier patch from Zhang Rui.
However, while the previous work was mainly focused on thermal monitoring
the focus here is on the usability from user space perspective.
References: https://lkml.org/lkml/2011/5/26/93
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Signed-off-by: Jacob Pan <jacob.jun.pan@linux.intel.com>
Reviewed-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
2013-10-17 11:28:35 -06:00
|
|
|
break;
|
2016-04-17 16:03:01 -06:00
|
|
|
case RAPL_DOMAIN_PLATFORM:
|
|
|
|
/* PSYS(PLATFORM) is not a CPU domain, so avoid printng error */
|
|
|
|
return -EINVAL;
|
PowerCap: Introduce Intel RAPL power capping driver
The Intel Running Average Power Limit (RAPL) technology provides platform
software with the ability to monitor, control, and get notifications on
power usage.
This feature is present in all Sandy Bridge and later Intel processors.
Newer models allow more fine grained controls to be applied. In RAPL,
power control is divided into domains, which include package, DRAM
controller, CPU core (Power Plane 0), graphics uncore (power plane 1), etc.
The purpose of this driver is to expose the RAPL settings to userspace.
Overall, RAPL fits in the new powercap class driver in that platform
level power capping controls are exposed via this generic interface.
This driver is based on an earlier patch from Zhang Rui.
However, while the previous work was mainly focused on thermal monitoring
the focus here is on the usability from user space perspective.
References: https://lkml.org/lkml/2011/5/26/93
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Signed-off-by: Jacob Pan <jacob.jun.pan@linux.intel.com>
Reviewed-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
2013-10-17 11:28:35 -06:00
|
|
|
default:
|
|
|
|
pr_err("invalid domain id %d\n", domain);
|
|
|
|
return -EINVAL;
|
|
|
|
}
|
2014-04-29 16:33:06 -06:00
|
|
|
/* make sure domain counters are available and contains non-zero
|
|
|
|
* values, otherwise skip it.
|
2014-02-10 07:11:51 -07:00
|
|
|
*/
|
2019-07-10 07:44:29 -06:00
|
|
|
|
|
|
|
ra.mask = ~0;
|
|
|
|
if (rp->priv->read_raw(cpu, &ra) || !ra.value)
|
2014-04-29 16:33:06 -06:00
|
|
|
return -ENODEV;
|
PowerCap: Introduce Intel RAPL power capping driver
The Intel Running Average Power Limit (RAPL) technology provides platform
software with the ability to monitor, control, and get notifications on
power usage.
This feature is present in all Sandy Bridge and later Intel processors.
Newer models allow more fine grained controls to be applied. In RAPL,
power control is divided into domains, which include package, DRAM
controller, CPU core (Power Plane 0), graphics uncore (power plane 1), etc.
The purpose of this driver is to expose the RAPL settings to userspace.
Overall, RAPL fits in the new powercap class driver in that platform
level power capping controls are exposed via this generic interface.
This driver is based on an earlier patch from Zhang Rui.
However, while the previous work was mainly focused on thermal monitoring
the focus here is on the usability from user space perspective.
References: https://lkml.org/lkml/2011/5/26/93
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Signed-off-by: Jacob Pan <jacob.jun.pan@linux.intel.com>
Reviewed-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
2013-10-17 11:28:35 -06:00
|
|
|
|
2014-04-29 16:33:06 -06:00
|
|
|
return 0;
|
PowerCap: Introduce Intel RAPL power capping driver
The Intel Running Average Power Limit (RAPL) technology provides platform
software with the ability to monitor, control, and get notifications on
power usage.
This feature is present in all Sandy Bridge and later Intel processors.
Newer models allow more fine grained controls to be applied. In RAPL,
power control is divided into domains, which include package, DRAM
controller, CPU core (Power Plane 0), graphics uncore (power plane 1), etc.
The purpose of this driver is to expose the RAPL settings to userspace.
Overall, RAPL fits in the new powercap class driver in that platform
level power capping controls are exposed via this generic interface.
This driver is based on an earlier patch from Zhang Rui.
However, while the previous work was mainly focused on thermal monitoring
the focus here is on the usability from user space perspective.
References: https://lkml.org/lkml/2011/5/26/93
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Signed-off-by: Jacob Pan <jacob.jun.pan@linux.intel.com>
Reviewed-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
2013-10-17 11:28:35 -06:00
|
|
|
}
|
|
|
|
|
2016-05-31 14:41:29 -06:00
|
|
|
/*
|
|
|
|
* Check if power limits are available. Two cases when they are not available:
|
|
|
|
* 1. Locked by BIOS, in this case we still provide read-only access so that
|
|
|
|
* users can see what limit is set by the BIOS.
|
|
|
|
* 2. Some CPUs make some domains monitoring only which means PLx MSRs may not
|
2019-07-10 07:44:30 -06:00
|
|
|
* exist at all. In this case, we do not show the constraints in powercap.
|
2016-05-31 14:41:29 -06:00
|
|
|
*
|
|
|
|
* Called after domains are detected and initialized.
|
|
|
|
*/
|
|
|
|
static void rapl_detect_powerlimit(struct rapl_domain *rd)
|
|
|
|
{
|
|
|
|
u64 val64;
|
|
|
|
int i;
|
|
|
|
|
|
|
|
/* check if the domain is locked by BIOS, ignore if MSR doesn't exist */
|
|
|
|
if (!rapl_read_data_raw(rd, FW_LOCK, false, &val64)) {
|
|
|
|
if (val64) {
|
2019-05-13 11:58:53 -06:00
|
|
|
pr_info("RAPL %s domain %s locked by BIOS\n",
|
|
|
|
rd->rp->name, rd->name);
|
2016-05-31 14:41:29 -06:00
|
|
|
rd->state |= DOMAIN_STATE_BIOS_LOCKED;
|
|
|
|
}
|
|
|
|
}
|
2019-07-10 07:44:30 -06:00
|
|
|
/* check if power limit MSR exists, otherwise domain is monitoring only */
|
2016-05-31 14:41:29 -06:00
|
|
|
for (i = 0; i < NR_POWER_LIMITS; i++) {
|
|
|
|
int prim = rd->rpl[i].prim_id;
|
2019-07-10 07:44:30 -06:00
|
|
|
|
2016-05-31 14:41:29 -06:00
|
|
|
if (rapl_read_data_raw(rd, prim, false, &val64))
|
|
|
|
rd->rpl[i].name = NULL;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
PowerCap: Introduce Intel RAPL power capping driver
The Intel Running Average Power Limit (RAPL) technology provides platform
software with the ability to monitor, control, and get notifications on
power usage.
This feature is present in all Sandy Bridge and later Intel processors.
Newer models allow more fine grained controls to be applied. In RAPL,
power control is divided into domains, which include package, DRAM
controller, CPU core (Power Plane 0), graphics uncore (power plane 1), etc.
The purpose of this driver is to expose the RAPL settings to userspace.
Overall, RAPL fits in the new powercap class driver in that platform
level power capping controls are exposed via this generic interface.
This driver is based on an earlier patch from Zhang Rui.
However, while the previous work was mainly focused on thermal monitoring
the focus here is on the usability from user space perspective.
References: https://lkml.org/lkml/2011/5/26/93
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Signed-off-by: Jacob Pan <jacob.jun.pan@linux.intel.com>
Reviewed-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
2013-10-17 11:28:35 -06:00
|
|
|
/* Detect active and valid domains for the given CPU, caller must
|
|
|
|
* ensure the CPU belongs to the targeted package and CPU hotlug is disabled.
|
|
|
|
*/
|
|
|
|
static int rapl_detect_domains(struct rapl_package *rp, int cpu)
|
|
|
|
{
|
|
|
|
struct rapl_domain *rd;
|
2016-11-22 14:16:02 -07:00
|
|
|
int i;
|
PowerCap: Introduce Intel RAPL power capping driver
The Intel Running Average Power Limit (RAPL) technology provides platform
software with the ability to monitor, control, and get notifications on
power usage.
This feature is present in all Sandy Bridge and later Intel processors.
Newer models allow more fine grained controls to be applied. In RAPL,
power control is divided into domains, which include package, DRAM
controller, CPU core (Power Plane 0), graphics uncore (power plane 1), etc.
The purpose of this driver is to expose the RAPL settings to userspace.
Overall, RAPL fits in the new powercap class driver in that platform
level power capping controls are exposed via this generic interface.
This driver is based on an earlier patch from Zhang Rui.
However, while the previous work was mainly focused on thermal monitoring
the focus here is on the usability from user space perspective.
References: https://lkml.org/lkml/2011/5/26/93
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Signed-off-by: Jacob Pan <jacob.jun.pan@linux.intel.com>
Reviewed-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
2013-10-17 11:28:35 -06:00
|
|
|
|
|
|
|
for (i = 0; i < RAPL_DOMAIN_MAX; i++) {
|
|
|
|
/* use physical package id to read counters */
|
2019-07-10 07:44:26 -06:00
|
|
|
if (!rapl_check_domain(cpu, i, rp)) {
|
PowerCap: Introduce Intel RAPL power capping driver
The Intel Running Average Power Limit (RAPL) technology provides platform
software with the ability to monitor, control, and get notifications on
power usage.
This feature is present in all Sandy Bridge and later Intel processors.
Newer models allow more fine grained controls to be applied. In RAPL,
power control is divided into domains, which include package, DRAM
controller, CPU core (Power Plane 0), graphics uncore (power plane 1), etc.
The purpose of this driver is to expose the RAPL settings to userspace.
Overall, RAPL fits in the new powercap class driver in that platform
level power capping controls are exposed via this generic interface.
This driver is based on an earlier patch from Zhang Rui.
However, while the previous work was mainly focused on thermal monitoring
the focus here is on the usability from user space perspective.
References: https://lkml.org/lkml/2011/5/26/93
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Signed-off-by: Jacob Pan <jacob.jun.pan@linux.intel.com>
Reviewed-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
2013-10-17 11:28:35 -06:00
|
|
|
rp->domain_map |= 1 << i;
|
2014-09-02 03:55:21 -06:00
|
|
|
pr_info("Found RAPL domain %s\n", rapl_domain_names[i]);
|
|
|
|
}
|
PowerCap: Introduce Intel RAPL power capping driver
The Intel Running Average Power Limit (RAPL) technology provides platform
software with the ability to monitor, control, and get notifications on
power usage.
This feature is present in all Sandy Bridge and later Intel processors.
Newer models allow more fine grained controls to be applied. In RAPL,
power control is divided into domains, which include package, DRAM
controller, CPU core (Power Plane 0), graphics uncore (power plane 1), etc.
The purpose of this driver is to expose the RAPL settings to userspace.
Overall, RAPL fits in the new powercap class driver in that platform
level power capping controls are exposed via this generic interface.
This driver is based on an earlier patch from Zhang Rui.
However, while the previous work was mainly focused on thermal monitoring
the focus here is on the usability from user space perspective.
References: https://lkml.org/lkml/2011/5/26/93
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Signed-off-by: Jacob Pan <jacob.jun.pan@linux.intel.com>
Reviewed-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
2013-10-17 11:28:35 -06:00
|
|
|
}
|
2019-07-10 07:44:30 -06:00
|
|
|
rp->nr_domains = bitmap_weight(&rp->domain_map, RAPL_DOMAIN_MAX);
|
PowerCap: Introduce Intel RAPL power capping driver
The Intel Running Average Power Limit (RAPL) technology provides platform
software with the ability to monitor, control, and get notifications on
power usage.
This feature is present in all Sandy Bridge and later Intel processors.
Newer models allow more fine grained controls to be applied. In RAPL,
power control is divided into domains, which include package, DRAM
controller, CPU core (Power Plane 0), graphics uncore (power plane 1), etc.
The purpose of this driver is to expose the RAPL settings to userspace.
Overall, RAPL fits in the new powercap class driver in that platform
level power capping controls are exposed via this generic interface.
This driver is based on an earlier patch from Zhang Rui.
However, while the previous work was mainly focused on thermal monitoring
the focus here is on the usability from user space perspective.
References: https://lkml.org/lkml/2011/5/26/93
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Signed-off-by: Jacob Pan <jacob.jun.pan@linux.intel.com>
Reviewed-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
2013-10-17 11:28:35 -06:00
|
|
|
if (!rp->nr_domains) {
|
2019-05-13 11:58:53 -06:00
|
|
|
pr_debug("no valid rapl domains found in %s\n", rp->name);
|
2016-11-22 14:16:02 -07:00
|
|
|
return -ENODEV;
|
PowerCap: Introduce Intel RAPL power capping driver
The Intel Running Average Power Limit (RAPL) technology provides platform
software with the ability to monitor, control, and get notifications on
power usage.
This feature is present in all Sandy Bridge and later Intel processors.
Newer models allow more fine grained controls to be applied. In RAPL,
power control is divided into domains, which include package, DRAM
controller, CPU core (Power Plane 0), graphics uncore (power plane 1), etc.
The purpose of this driver is to expose the RAPL settings to userspace.
Overall, RAPL fits in the new powercap class driver in that platform
level power capping controls are exposed via this generic interface.
This driver is based on an earlier patch from Zhang Rui.
However, while the previous work was mainly focused on thermal monitoring
the focus here is on the usability from user space perspective.
References: https://lkml.org/lkml/2011/5/26/93
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Signed-off-by: Jacob Pan <jacob.jun.pan@linux.intel.com>
Reviewed-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
2013-10-17 11:28:35 -06:00
|
|
|
}
|
2019-05-13 11:58:53 -06:00
|
|
|
pr_debug("found %d domains on %s\n", rp->nr_domains, rp->name);
|
PowerCap: Introduce Intel RAPL power capping driver
The Intel Running Average Power Limit (RAPL) technology provides platform
software with the ability to monitor, control, and get notifications on
power usage.
This feature is present in all Sandy Bridge and later Intel processors.
Newer models allow more fine grained controls to be applied. In RAPL,
power control is divided into domains, which include package, DRAM
controller, CPU core (Power Plane 0), graphics uncore (power plane 1), etc.
The purpose of this driver is to expose the RAPL settings to userspace.
Overall, RAPL fits in the new powercap class driver in that platform
level power capping controls are exposed via this generic interface.
This driver is based on an earlier patch from Zhang Rui.
However, while the previous work was mainly focused on thermal monitoring
the focus here is on the usability from user space perspective.
References: https://lkml.org/lkml/2011/5/26/93
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Signed-off-by: Jacob Pan <jacob.jun.pan@linux.intel.com>
Reviewed-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
2013-10-17 11:28:35 -06:00
|
|
|
|
|
|
|
rp->domains = kcalloc(rp->nr_domains + 1, sizeof(struct rapl_domain),
|
2019-07-10 07:44:30 -06:00
|
|
|
GFP_KERNEL);
|
2016-11-22 14:16:02 -07:00
|
|
|
if (!rp->domains)
|
|
|
|
return -ENOMEM;
|
|
|
|
|
PowerCap: Introduce Intel RAPL power capping driver
The Intel Running Average Power Limit (RAPL) technology provides platform
software with the ability to monitor, control, and get notifications on
power usage.
This feature is present in all Sandy Bridge and later Intel processors.
Newer models allow more fine grained controls to be applied. In RAPL,
power control is divided into domains, which include package, DRAM
controller, CPU core (Power Plane 0), graphics uncore (power plane 1), etc.
The purpose of this driver is to expose the RAPL settings to userspace.
Overall, RAPL fits in the new powercap class driver in that platform
level power capping controls are exposed via this generic interface.
This driver is based on an earlier patch from Zhang Rui.
However, while the previous work was mainly focused on thermal monitoring
the focus here is on the usability from user space perspective.
References: https://lkml.org/lkml/2011/5/26/93
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Signed-off-by: Jacob Pan <jacob.jun.pan@linux.intel.com>
Reviewed-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
2013-10-17 11:28:35 -06:00
|
|
|
rapl_init_domains(rp);
|
|
|
|
|
2016-05-31 14:41:29 -06:00
|
|
|
for (rd = rp->domains; rd < rp->domains + rp->nr_domains; rd++)
|
|
|
|
rapl_detect_powerlimit(rd);
|
|
|
|
|
PowerCap: Introduce Intel RAPL power capping driver
The Intel Running Average Power Limit (RAPL) technology provides platform
software with the ability to monitor, control, and get notifications on
power usage.
This feature is present in all Sandy Bridge and later Intel processors.
Newer models allow more fine grained controls to be applied. In RAPL,
power control is divided into domains, which include package, DRAM
controller, CPU core (Power Plane 0), graphics uncore (power plane 1), etc.
The purpose of this driver is to expose the RAPL settings to userspace.
Overall, RAPL fits in the new powercap class driver in that platform
level power capping controls are exposed via this generic interface.
This driver is based on an earlier patch from Zhang Rui.
However, while the previous work was mainly focused on thermal monitoring
the focus here is on the usability from user space perspective.
References: https://lkml.org/lkml/2011/5/26/93
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Signed-off-by: Jacob Pan <jacob.jun.pan@linux.intel.com>
Reviewed-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
2013-10-17 11:28:35 -06:00
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
/* called from CPU hotplug notifier, hotplug lock held */
|
2019-07-10 07:44:30 -06:00
|
|
|
void rapl_remove_package(struct rapl_package *rp)
|
PowerCap: Introduce Intel RAPL power capping driver
The Intel Running Average Power Limit (RAPL) technology provides platform
software with the ability to monitor, control, and get notifications on
power usage.
This feature is present in all Sandy Bridge and later Intel processors.
Newer models allow more fine grained controls to be applied. In RAPL,
power control is divided into domains, which include package, DRAM
controller, CPU core (Power Plane 0), graphics uncore (power plane 1), etc.
The purpose of this driver is to expose the RAPL settings to userspace.
Overall, RAPL fits in the new powercap class driver in that platform
level power capping controls are exposed via this generic interface.
This driver is based on an earlier patch from Zhang Rui.
However, while the previous work was mainly focused on thermal monitoring
the focus here is on the usability from user space perspective.
References: https://lkml.org/lkml/2011/5/26/93
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Signed-off-by: Jacob Pan <jacob.jun.pan@linux.intel.com>
Reviewed-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
2013-10-17 11:28:35 -06:00
|
|
|
{
|
|
|
|
struct rapl_domain *rd, *rd_package = NULL;
|
|
|
|
|
2016-11-22 14:16:02 -07:00
|
|
|
package_power_limit_irq_restore(rp);
|
|
|
|
|
PowerCap: Introduce Intel RAPL power capping driver
The Intel Running Average Power Limit (RAPL) technology provides platform
software with the ability to monitor, control, and get notifications on
power usage.
This feature is present in all Sandy Bridge and later Intel processors.
Newer models allow more fine grained controls to be applied. In RAPL,
power control is divided into domains, which include package, DRAM
controller, CPU core (Power Plane 0), graphics uncore (power plane 1), etc.
The purpose of this driver is to expose the RAPL settings to userspace.
Overall, RAPL fits in the new powercap class driver in that platform
level power capping controls are exposed via this generic interface.
This driver is based on an earlier patch from Zhang Rui.
However, while the previous work was mainly focused on thermal monitoring
the focus here is on the usability from user space perspective.
References: https://lkml.org/lkml/2011/5/26/93
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Signed-off-by: Jacob Pan <jacob.jun.pan@linux.intel.com>
Reviewed-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
2013-10-17 11:28:35 -06:00
|
|
|
for (rd = rp->domains; rd < rp->domains + rp->nr_domains; rd++) {
|
2016-11-22 14:16:02 -07:00
|
|
|
rapl_write_data_raw(rd, PL1_ENABLE, 0);
|
|
|
|
rapl_write_data_raw(rd, PL1_CLAMP, 0);
|
|
|
|
if (find_nr_power_limit(rd) > 1) {
|
|
|
|
rapl_write_data_raw(rd, PL2_ENABLE, 0);
|
|
|
|
rapl_write_data_raw(rd, PL2_CLAMP, 0);
|
|
|
|
}
|
PowerCap: Introduce Intel RAPL power capping driver
The Intel Running Average Power Limit (RAPL) technology provides platform
software with the ability to monitor, control, and get notifications on
power usage.
This feature is present in all Sandy Bridge and later Intel processors.
Newer models allow more fine grained controls to be applied. In RAPL,
power control is divided into domains, which include package, DRAM
controller, CPU core (Power Plane 0), graphics uncore (power plane 1), etc.
The purpose of this driver is to expose the RAPL settings to userspace.
Overall, RAPL fits in the new powercap class driver in that platform
level power capping controls are exposed via this generic interface.
This driver is based on an earlier patch from Zhang Rui.
However, while the previous work was mainly focused on thermal monitoring
the focus here is on the usability from user space perspective.
References: https://lkml.org/lkml/2011/5/26/93
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Signed-off-by: Jacob Pan <jacob.jun.pan@linux.intel.com>
Reviewed-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
2013-10-17 11:28:35 -06:00
|
|
|
if (rd->id == RAPL_DOMAIN_PACKAGE) {
|
|
|
|
rd_package = rd;
|
|
|
|
continue;
|
|
|
|
}
|
2019-05-13 11:58:53 -06:00
|
|
|
pr_debug("remove package, undo power limit on %s: %s\n",
|
|
|
|
rp->name, rd->name);
|
2019-07-10 07:44:30 -06:00
|
|
|
powercap_unregister_zone(rp->priv->control_type,
|
|
|
|
&rd->power_zone);
|
PowerCap: Introduce Intel RAPL power capping driver
The Intel Running Average Power Limit (RAPL) technology provides platform
software with the ability to monitor, control, and get notifications on
power usage.
This feature is present in all Sandy Bridge and later Intel processors.
Newer models allow more fine grained controls to be applied. In RAPL,
power control is divided into domains, which include package, DRAM
controller, CPU core (Power Plane 0), graphics uncore (power plane 1), etc.
The purpose of this driver is to expose the RAPL settings to userspace.
Overall, RAPL fits in the new powercap class driver in that platform
level power capping controls are exposed via this generic interface.
This driver is based on an earlier patch from Zhang Rui.
However, while the previous work was mainly focused on thermal monitoring
the focus here is on the usability from user space perspective.
References: https://lkml.org/lkml/2011/5/26/93
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Signed-off-by: Jacob Pan <jacob.jun.pan@linux.intel.com>
Reviewed-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
2013-10-17 11:28:35 -06:00
|
|
|
}
|
|
|
|
/* do parent zone last */
|
2019-07-10 07:44:30 -06:00
|
|
|
powercap_unregister_zone(rp->priv->control_type,
|
|
|
|
&rd_package->power_zone);
|
PowerCap: Introduce Intel RAPL power capping driver
The Intel Running Average Power Limit (RAPL) technology provides platform
software with the ability to monitor, control, and get notifications on
power usage.
This feature is present in all Sandy Bridge and later Intel processors.
Newer models allow more fine grained controls to be applied. In RAPL,
power control is divided into domains, which include package, DRAM
controller, CPU core (Power Plane 0), graphics uncore (power plane 1), etc.
The purpose of this driver is to expose the RAPL settings to userspace.
Overall, RAPL fits in the new powercap class driver in that platform
level power capping controls are exposed via this generic interface.
This driver is based on an earlier patch from Zhang Rui.
However, while the previous work was mainly focused on thermal monitoring
the focus here is on the usability from user space perspective.
References: https://lkml.org/lkml/2011/5/26/93
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Signed-off-by: Jacob Pan <jacob.jun.pan@linux.intel.com>
Reviewed-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
2013-10-17 11:28:35 -06:00
|
|
|
list_del(&rp->plist);
|
|
|
|
kfree(rp);
|
|
|
|
}
|
2019-07-10 07:44:30 -06:00
|
|
|
EXPORT_SYMBOL_GPL(rapl_remove_package);
|
|
|
|
|
|
|
|
/* caller to ensure CPU hotplug lock is held */
|
|
|
|
struct rapl_package *rapl_find_package_domain(int cpu, struct rapl_if_priv *priv)
|
|
|
|
{
|
|
|
|
int id = topology_logical_die_id(cpu);
|
|
|
|
struct rapl_package *rp;
|
|
|
|
|
|
|
|
list_for_each_entry(rp, &rapl_packages, plist) {
|
|
|
|
if (rp->id == id
|
|
|
|
&& rp->priv->control_type == priv->control_type)
|
|
|
|
return rp;
|
|
|
|
}
|
|
|
|
|
|
|
|
return NULL;
|
|
|
|
}
|
|
|
|
EXPORT_SYMBOL_GPL(rapl_find_package_domain);
|
PowerCap: Introduce Intel RAPL power capping driver
The Intel Running Average Power Limit (RAPL) technology provides platform
software with the ability to monitor, control, and get notifications on
power usage.
This feature is present in all Sandy Bridge and later Intel processors.
Newer models allow more fine grained controls to be applied. In RAPL,
power control is divided into domains, which include package, DRAM
controller, CPU core (Power Plane 0), graphics uncore (power plane 1), etc.
The purpose of this driver is to expose the RAPL settings to userspace.
Overall, RAPL fits in the new powercap class driver in that platform
level power capping controls are exposed via this generic interface.
This driver is based on an earlier patch from Zhang Rui.
However, while the previous work was mainly focused on thermal monitoring
the focus here is on the usability from user space perspective.
References: https://lkml.org/lkml/2011/5/26/93
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Signed-off-by: Jacob Pan <jacob.jun.pan@linux.intel.com>
Reviewed-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
2013-10-17 11:28:35 -06:00
|
|
|
|
|
|
|
/* called from CPU hotplug notifier, hotplug lock held */
|
2019-07-10 07:44:30 -06:00
|
|
|
struct rapl_package *rapl_add_package(int cpu, struct rapl_if_priv *priv)
|
PowerCap: Introduce Intel RAPL power capping driver
The Intel Running Average Power Limit (RAPL) technology provides platform
software with the ability to monitor, control, and get notifications on
power usage.
This feature is present in all Sandy Bridge and later Intel processors.
Newer models allow more fine grained controls to be applied. In RAPL,
power control is divided into domains, which include package, DRAM
controller, CPU core (Power Plane 0), graphics uncore (power plane 1), etc.
The purpose of this driver is to expose the RAPL settings to userspace.
Overall, RAPL fits in the new powercap class driver in that platform
level power capping controls are exposed via this generic interface.
This driver is based on an earlier patch from Zhang Rui.
However, while the previous work was mainly focused on thermal monitoring
the focus here is on the usability from user space perspective.
References: https://lkml.org/lkml/2011/5/26/93
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Signed-off-by: Jacob Pan <jacob.jun.pan@linux.intel.com>
Reviewed-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
2013-10-17 11:28:35 -06:00
|
|
|
{
|
2019-05-13 11:58:51 -06:00
|
|
|
int id = topology_logical_die_id(cpu);
|
PowerCap: Introduce Intel RAPL power capping driver
The Intel Running Average Power Limit (RAPL) technology provides platform
software with the ability to monitor, control, and get notifications on
power usage.
This feature is present in all Sandy Bridge and later Intel processors.
Newer models allow more fine grained controls to be applied. In RAPL,
power control is divided into domains, which include package, DRAM
controller, CPU core (Power Plane 0), graphics uncore (power plane 1), etc.
The purpose of this driver is to expose the RAPL settings to userspace.
Overall, RAPL fits in the new powercap class driver in that platform
level power capping controls are exposed via this generic interface.
This driver is based on an earlier patch from Zhang Rui.
However, while the previous work was mainly focused on thermal monitoring
the focus here is on the usability from user space perspective.
References: https://lkml.org/lkml/2011/5/26/93
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Signed-off-by: Jacob Pan <jacob.jun.pan@linux.intel.com>
Reviewed-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
2013-10-17 11:28:35 -06:00
|
|
|
struct rapl_package *rp;
|
2019-05-13 11:58:53 -06:00
|
|
|
struct cpuinfo_x86 *c = &cpu_data(cpu);
|
2016-11-22 14:16:05 -07:00
|
|
|
int ret;
|
PowerCap: Introduce Intel RAPL power capping driver
The Intel Running Average Power Limit (RAPL) technology provides platform
software with the ability to monitor, control, and get notifications on
power usage.
This feature is present in all Sandy Bridge and later Intel processors.
Newer models allow more fine grained controls to be applied. In RAPL,
power control is divided into domains, which include package, DRAM
controller, CPU core (Power Plane 0), graphics uncore (power plane 1), etc.
The purpose of this driver is to expose the RAPL settings to userspace.
Overall, RAPL fits in the new powercap class driver in that platform
level power capping controls are exposed via this generic interface.
This driver is based on an earlier patch from Zhang Rui.
However, while the previous work was mainly focused on thermal monitoring
the focus here is on the usability from user space perspective.
References: https://lkml.org/lkml/2011/5/26/93
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Signed-off-by: Jacob Pan <jacob.jun.pan@linux.intel.com>
Reviewed-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
2013-10-17 11:28:35 -06:00
|
|
|
|
|
|
|
rp = kzalloc(sizeof(struct rapl_package), GFP_KERNEL);
|
|
|
|
if (!rp)
|
2016-11-22 14:16:05 -07:00
|
|
|
return ERR_PTR(-ENOMEM);
|
PowerCap: Introduce Intel RAPL power capping driver
The Intel Running Average Power Limit (RAPL) technology provides platform
software with the ability to monitor, control, and get notifications on
power usage.
This feature is present in all Sandy Bridge and later Intel processors.
Newer models allow more fine grained controls to be applied. In RAPL,
power control is divided into domains, which include package, DRAM
controller, CPU core (Power Plane 0), graphics uncore (power plane 1), etc.
The purpose of this driver is to expose the RAPL settings to userspace.
Overall, RAPL fits in the new powercap class driver in that platform
level power capping controls are exposed via this generic interface.
This driver is based on an earlier patch from Zhang Rui.
However, while the previous work was mainly focused on thermal monitoring
the focus here is on the usability from user space perspective.
References: https://lkml.org/lkml/2011/5/26/93
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Signed-off-by: Jacob Pan <jacob.jun.pan@linux.intel.com>
Reviewed-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
2013-10-17 11:28:35 -06:00
|
|
|
|
|
|
|
/* add the new package to the list */
|
2019-05-13 11:58:50 -06:00
|
|
|
rp->id = id;
|
2016-02-24 14:31:38 -07:00
|
|
|
rp->lead_cpu = cpu;
|
2019-07-10 07:44:25 -06:00
|
|
|
rp->priv = priv;
|
2016-02-24 14:31:38 -07:00
|
|
|
|
2019-05-13 11:58:53 -06:00
|
|
|
if (topology_max_die_per_package() > 1)
|
|
|
|
snprintf(rp->name, PACKAGE_DOMAIN_NAME_LENGTH,
|
2019-07-10 07:44:30 -06:00
|
|
|
"package-%d-die-%d", c->phys_proc_id, c->cpu_die_id);
|
2019-05-13 11:58:53 -06:00
|
|
|
else
|
|
|
|
snprintf(rp->name, PACKAGE_DOMAIN_NAME_LENGTH, "package-%d",
|
2019-07-10 07:44:30 -06:00
|
|
|
c->phys_proc_id);
|
2019-05-13 11:58:53 -06:00
|
|
|
|
PowerCap: Introduce Intel RAPL power capping driver
The Intel Running Average Power Limit (RAPL) technology provides platform
software with the ability to monitor, control, and get notifications on
power usage.
This feature is present in all Sandy Bridge and later Intel processors.
Newer models allow more fine grained controls to be applied. In RAPL,
power control is divided into domains, which include package, DRAM
controller, CPU core (Power Plane 0), graphics uncore (power plane 1), etc.
The purpose of this driver is to expose the RAPL settings to userspace.
Overall, RAPL fits in the new powercap class driver in that platform
level power capping controls are exposed via this generic interface.
This driver is based on an earlier patch from Zhang Rui.
However, while the previous work was mainly focused on thermal monitoring
the focus here is on the usability from user space perspective.
References: https://lkml.org/lkml/2011/5/26/93
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Signed-off-by: Jacob Pan <jacob.jun.pan@linux.intel.com>
Reviewed-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
2013-10-17 11:28:35 -06:00
|
|
|
/* check if the package contains valid domains */
|
2019-07-10 07:44:30 -06:00
|
|
|
if (rapl_detect_domains(rp, cpu) || rapl_defaults->check_unit(rp, cpu)) {
|
PowerCap: Introduce Intel RAPL power capping driver
The Intel Running Average Power Limit (RAPL) technology provides platform
software with the ability to monitor, control, and get notifications on
power usage.
This feature is present in all Sandy Bridge and later Intel processors.
Newer models allow more fine grained controls to be applied. In RAPL,
power control is divided into domains, which include package, DRAM
controller, CPU core (Power Plane 0), graphics uncore (power plane 1), etc.
The purpose of this driver is to expose the RAPL settings to userspace.
Overall, RAPL fits in the new powercap class driver in that platform
level power capping controls are exposed via this generic interface.
This driver is based on an earlier patch from Zhang Rui.
However, while the previous work was mainly focused on thermal monitoring
the focus here is on the usability from user space perspective.
References: https://lkml.org/lkml/2011/5/26/93
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Signed-off-by: Jacob Pan <jacob.jun.pan@linux.intel.com>
Reviewed-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
2013-10-17 11:28:35 -06:00
|
|
|
ret = -ENODEV;
|
|
|
|
goto err_free_package;
|
|
|
|
}
|
2016-11-22 14:15:59 -07:00
|
|
|
ret = rapl_package_register_powercap(rp);
|
|
|
|
if (!ret) {
|
PowerCap: Introduce Intel RAPL power capping driver
The Intel Running Average Power Limit (RAPL) technology provides platform
software with the ability to monitor, control, and get notifications on
power usage.
This feature is present in all Sandy Bridge and later Intel processors.
Newer models allow more fine grained controls to be applied. In RAPL,
power control is divided into domains, which include package, DRAM
controller, CPU core (Power Plane 0), graphics uncore (power plane 1), etc.
The purpose of this driver is to expose the RAPL settings to userspace.
Overall, RAPL fits in the new powercap class driver in that platform
level power capping controls are exposed via this generic interface.
This driver is based on an earlier patch from Zhang Rui.
However, while the previous work was mainly focused on thermal monitoring
the focus here is on the usability from user space perspective.
References: https://lkml.org/lkml/2011/5/26/93
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Signed-off-by: Jacob Pan <jacob.jun.pan@linux.intel.com>
Reviewed-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
2013-10-17 11:28:35 -06:00
|
|
|
INIT_LIST_HEAD(&rp->plist);
|
|
|
|
list_add(&rp->plist, &rapl_packages);
|
2016-11-22 14:16:05 -07:00
|
|
|
return rp;
|
PowerCap: Introduce Intel RAPL power capping driver
The Intel Running Average Power Limit (RAPL) technology provides platform
software with the ability to monitor, control, and get notifications on
power usage.
This feature is present in all Sandy Bridge and later Intel processors.
Newer models allow more fine grained controls to be applied. In RAPL,
power control is divided into domains, which include package, DRAM
controller, CPU core (Power Plane 0), graphics uncore (power plane 1), etc.
The purpose of this driver is to expose the RAPL settings to userspace.
Overall, RAPL fits in the new powercap class driver in that platform
level power capping controls are exposed via this generic interface.
This driver is based on an earlier patch from Zhang Rui.
However, while the previous work was mainly focused on thermal monitoring
the focus here is on the usability from user space perspective.
References: https://lkml.org/lkml/2011/5/26/93
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Signed-off-by: Jacob Pan <jacob.jun.pan@linux.intel.com>
Reviewed-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
2013-10-17 11:28:35 -06:00
|
|
|
}
|
|
|
|
|
|
|
|
err_free_package:
|
|
|
|
kfree(rp->domains);
|
|
|
|
kfree(rp);
|
2016-11-22 14:16:05 -07:00
|
|
|
return ERR_PTR(ret);
|
PowerCap: Introduce Intel RAPL power capping driver
The Intel Running Average Power Limit (RAPL) technology provides platform
software with the ability to monitor, control, and get notifications on
power usage.
This feature is present in all Sandy Bridge and later Intel processors.
Newer models allow more fine grained controls to be applied. In RAPL,
power control is divided into domains, which include package, DRAM
controller, CPU core (Power Plane 0), graphics uncore (power plane 1), etc.
The purpose of this driver is to expose the RAPL settings to userspace.
Overall, RAPL fits in the new powercap class driver in that platform
level power capping controls are exposed via this generic interface.
This driver is based on an earlier patch from Zhang Rui.
However, while the previous work was mainly focused on thermal monitoring
the focus here is on the usability from user space perspective.
References: https://lkml.org/lkml/2011/5/26/93
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Signed-off-by: Jacob Pan <jacob.jun.pan@linux.intel.com>
Reviewed-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
2013-10-17 11:28:35 -06:00
|
|
|
}
|
2019-07-10 07:44:30 -06:00
|
|
|
EXPORT_SYMBOL_GPL(rapl_add_package);
|
PowerCap: Introduce Intel RAPL power capping driver
The Intel Running Average Power Limit (RAPL) technology provides platform
software with the ability to monitor, control, and get notifications on
power usage.
This feature is present in all Sandy Bridge and later Intel processors.
Newer models allow more fine grained controls to be applied. In RAPL,
power control is divided into domains, which include package, DRAM
controller, CPU core (Power Plane 0), graphics uncore (power plane 1), etc.
The purpose of this driver is to expose the RAPL settings to userspace.
Overall, RAPL fits in the new powercap class driver in that platform
level power capping controls are exposed via this generic interface.
This driver is based on an earlier patch from Zhang Rui.
However, while the previous work was mainly focused on thermal monitoring
the focus here is on the usability from user space perspective.
References: https://lkml.org/lkml/2011/5/26/93
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Signed-off-by: Jacob Pan <jacob.jun.pan@linux.intel.com>
Reviewed-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
2013-10-17 11:28:35 -06:00
|
|
|
|
2018-01-09 17:38:23 -07:00
|
|
|
static void power_limit_state_save(void)
|
|
|
|
{
|
|
|
|
struct rapl_package *rp;
|
|
|
|
struct rapl_domain *rd;
|
|
|
|
int nr_pl, ret, i;
|
|
|
|
|
|
|
|
get_online_cpus();
|
|
|
|
list_for_each_entry(rp, &rapl_packages, plist) {
|
|
|
|
if (!rp->power_zone)
|
|
|
|
continue;
|
|
|
|
rd = power_zone_to_rapl_domain(rp->power_zone);
|
|
|
|
nr_pl = find_nr_power_limit(rd);
|
|
|
|
for (i = 0; i < nr_pl; i++) {
|
|
|
|
switch (rd->rpl[i].prim_id) {
|
|
|
|
case PL1_ENABLE:
|
|
|
|
ret = rapl_read_data_raw(rd,
|
2019-07-10 07:44:30 -06:00
|
|
|
POWER_LIMIT1, true,
|
|
|
|
&rd->rpl[i].last_power_limit);
|
2018-01-09 17:38:23 -07:00
|
|
|
if (ret)
|
|
|
|
rd->rpl[i].last_power_limit = 0;
|
|
|
|
break;
|
|
|
|
case PL2_ENABLE:
|
|
|
|
ret = rapl_read_data_raw(rd,
|
2019-07-10 07:44:30 -06:00
|
|
|
POWER_LIMIT2, true,
|
|
|
|
&rd->rpl[i].last_power_limit);
|
2018-01-09 17:38:23 -07:00
|
|
|
if (ret)
|
|
|
|
rd->rpl[i].last_power_limit = 0;
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
}
|
|
|
|
put_online_cpus();
|
|
|
|
}
|
|
|
|
|
|
|
|
static void power_limit_state_restore(void)
|
|
|
|
{
|
|
|
|
struct rapl_package *rp;
|
|
|
|
struct rapl_domain *rd;
|
|
|
|
int nr_pl, i;
|
|
|
|
|
|
|
|
get_online_cpus();
|
|
|
|
list_for_each_entry(rp, &rapl_packages, plist) {
|
|
|
|
if (!rp->power_zone)
|
|
|
|
continue;
|
|
|
|
rd = power_zone_to_rapl_domain(rp->power_zone);
|
|
|
|
nr_pl = find_nr_power_limit(rd);
|
|
|
|
for (i = 0; i < nr_pl; i++) {
|
|
|
|
switch (rd->rpl[i].prim_id) {
|
|
|
|
case PL1_ENABLE:
|
|
|
|
if (rd->rpl[i].last_power_limit)
|
2019-07-10 07:44:30 -06:00
|
|
|
rapl_write_data_raw(rd, POWER_LIMIT1,
|
|
|
|
rd->rpl[i].last_power_limit);
|
2018-01-09 17:38:23 -07:00
|
|
|
break;
|
|
|
|
case PL2_ENABLE:
|
|
|
|
if (rd->rpl[i].last_power_limit)
|
2019-07-10 07:44:30 -06:00
|
|
|
rapl_write_data_raw(rd, POWER_LIMIT2,
|
|
|
|
rd->rpl[i].last_power_limit);
|
2018-01-09 17:38:23 -07:00
|
|
|
break;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
}
|
|
|
|
put_online_cpus();
|
|
|
|
}
|
|
|
|
|
|
|
|
static int rapl_pm_callback(struct notifier_block *nb,
|
2019-07-10 07:44:30 -06:00
|
|
|
unsigned long mode, void *_unused)
|
2018-01-09 17:38:23 -07:00
|
|
|
{
|
|
|
|
switch (mode) {
|
|
|
|
case PM_SUSPEND_PREPARE:
|
|
|
|
power_limit_state_save();
|
|
|
|
break;
|
|
|
|
case PM_POST_SUSPEND:
|
|
|
|
power_limit_state_restore();
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
return NOTIFY_OK;
|
|
|
|
}
|
|
|
|
|
|
|
|
static struct notifier_block rapl_pm_notifier = {
|
|
|
|
.notifier_call = rapl_pm_callback,
|
|
|
|
};
|
|
|
|
|
2019-07-10 07:44:34 -06:00
|
|
|
static struct platform_device *rapl_msr_platdev;
|
|
|
|
|
|
|
|
static int __init rapl_init(void)
|
PowerCap: Introduce Intel RAPL power capping driver
The Intel Running Average Power Limit (RAPL) technology provides platform
software with the ability to monitor, control, and get notifications on
power usage.
This feature is present in all Sandy Bridge and later Intel processors.
Newer models allow more fine grained controls to be applied. In RAPL,
power control is divided into domains, which include package, DRAM
controller, CPU core (Power Plane 0), graphics uncore (power plane 1), etc.
The purpose of this driver is to expose the RAPL settings to userspace.
Overall, RAPL fits in the new powercap class driver in that platform
level power capping controls are exposed via this generic interface.
This driver is based on an earlier patch from Zhang Rui.
However, while the previous work was mainly focused on thermal monitoring
the focus here is on the usability from user space perspective.
References: https://lkml.org/lkml/2011/5/26/93
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Signed-off-by: Jacob Pan <jacob.jun.pan@linux.intel.com>
Reviewed-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
2013-10-17 11:28:35 -06:00
|
|
|
{
|
2014-11-07 10:29:25 -07:00
|
|
|
const struct x86_cpu_id *id;
|
2016-11-22 14:16:02 -07:00
|
|
|
int ret;
|
PowerCap: Introduce Intel RAPL power capping driver
The Intel Running Average Power Limit (RAPL) technology provides platform
software with the ability to monitor, control, and get notifications on
power usage.
This feature is present in all Sandy Bridge and later Intel processors.
Newer models allow more fine grained controls to be applied. In RAPL,
power control is divided into domains, which include package, DRAM
controller, CPU core (Power Plane 0), graphics uncore (power plane 1), etc.
The purpose of this driver is to expose the RAPL settings to userspace.
Overall, RAPL fits in the new powercap class driver in that platform
level power capping controls are exposed via this generic interface.
This driver is based on an earlier patch from Zhang Rui.
However, while the previous work was mainly focused on thermal monitoring
the focus here is on the usability from user space perspective.
References: https://lkml.org/lkml/2011/5/26/93
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Signed-off-by: Jacob Pan <jacob.jun.pan@linux.intel.com>
Reviewed-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
2013-10-17 11:28:35 -06:00
|
|
|
|
2014-11-07 10:29:25 -07:00
|
|
|
id = x86_match_cpu(rapl_ids);
|
|
|
|
if (!id) {
|
PowerCap: Introduce Intel RAPL power capping driver
The Intel Running Average Power Limit (RAPL) technology provides platform
software with the ability to monitor, control, and get notifications on
power usage.
This feature is present in all Sandy Bridge and later Intel processors.
Newer models allow more fine grained controls to be applied. In RAPL,
power control is divided into domains, which include package, DRAM
controller, CPU core (Power Plane 0), graphics uncore (power plane 1), etc.
The purpose of this driver is to expose the RAPL settings to userspace.
Overall, RAPL fits in the new powercap class driver in that platform
level power capping controls are exposed via this generic interface.
This driver is based on an earlier patch from Zhang Rui.
However, while the previous work was mainly focused on thermal monitoring
the focus here is on the usability from user space perspective.
References: https://lkml.org/lkml/2011/5/26/93
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Signed-off-by: Jacob Pan <jacob.jun.pan@linux.intel.com>
Reviewed-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
2013-10-17 11:28:35 -06:00
|
|
|
pr_err("driver does not support CPU family %d model %d\n",
|
2019-07-10 07:44:30 -06:00
|
|
|
boot_cpu_data.x86, boot_cpu_data.x86_model);
|
PowerCap: Introduce Intel RAPL power capping driver
The Intel Running Average Power Limit (RAPL) technology provides platform
software with the ability to monitor, control, and get notifications on
power usage.
This feature is present in all Sandy Bridge and later Intel processors.
Newer models allow more fine grained controls to be applied. In RAPL,
power control is divided into domains, which include package, DRAM
controller, CPU core (Power Plane 0), graphics uncore (power plane 1), etc.
The purpose of this driver is to expose the RAPL settings to userspace.
Overall, RAPL fits in the new powercap class driver in that platform
level power capping controls are exposed via this generic interface.
This driver is based on an earlier patch from Zhang Rui.
However, while the previous work was mainly focused on thermal monitoring
the focus here is on the usability from user space perspective.
References: https://lkml.org/lkml/2011/5/26/93
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Signed-off-by: Jacob Pan <jacob.jun.pan@linux.intel.com>
Reviewed-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
2013-10-17 11:28:35 -06:00
|
|
|
|
|
|
|
return -ENODEV;
|
|
|
|
}
|
2014-03-10 14:39:26 -06:00
|
|
|
|
2014-11-07 10:29:25 -07:00
|
|
|
rapl_defaults = (struct rapl_defaults *)id->driver_data;
|
|
|
|
|
2018-01-09 17:38:23 -07:00
|
|
|
ret = register_pm_notifier(&rapl_pm_notifier);
|
2019-07-10 07:44:34 -06:00
|
|
|
if (ret)
|
|
|
|
return ret;
|
2018-01-09 17:38:23 -07:00
|
|
|
|
2019-07-10 07:44:34 -06:00
|
|
|
rapl_msr_platdev = platform_device_alloc("intel_rapl_msr", 0);
|
|
|
|
if (!rapl_msr_platdev) {
|
|
|
|
ret = -ENOMEM;
|
|
|
|
goto end;
|
|
|
|
}
|
|
|
|
|
|
|
|
ret = platform_device_add(rapl_msr_platdev);
|
|
|
|
if (ret)
|
|
|
|
platform_device_put(rapl_msr_platdev);
|
|
|
|
|
|
|
|
end:
|
|
|
|
if (ret)
|
|
|
|
unregister_pm_notifier(&rapl_pm_notifier);
|
|
|
|
|
|
|
|
return ret;
|
PowerCap: Introduce Intel RAPL power capping driver
The Intel Running Average Power Limit (RAPL) technology provides platform
software with the ability to monitor, control, and get notifications on
power usage.
This feature is present in all Sandy Bridge and later Intel processors.
Newer models allow more fine grained controls to be applied. In RAPL,
power control is divided into domains, which include package, DRAM
controller, CPU core (Power Plane 0), graphics uncore (power plane 1), etc.
The purpose of this driver is to expose the RAPL settings to userspace.
Overall, RAPL fits in the new powercap class driver in that platform
level power capping controls are exposed via this generic interface.
This driver is based on an earlier patch from Zhang Rui.
However, while the previous work was mainly focused on thermal monitoring
the focus here is on the usability from user space perspective.
References: https://lkml.org/lkml/2011/5/26/93
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Signed-off-by: Jacob Pan <jacob.jun.pan@linux.intel.com>
Reviewed-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
2013-10-17 11:28:35 -06:00
|
|
|
}
|
|
|
|
|
2019-07-10 07:44:34 -06:00
|
|
|
static void __exit rapl_exit(void)
|
PowerCap: Introduce Intel RAPL power capping driver
The Intel Running Average Power Limit (RAPL) technology provides platform
software with the ability to monitor, control, and get notifications on
power usage.
This feature is present in all Sandy Bridge and later Intel processors.
Newer models allow more fine grained controls to be applied. In RAPL,
power control is divided into domains, which include package, DRAM
controller, CPU core (Power Plane 0), graphics uncore (power plane 1), etc.
The purpose of this driver is to expose the RAPL settings to userspace.
Overall, RAPL fits in the new powercap class driver in that platform
level power capping controls are exposed via this generic interface.
This driver is based on an earlier patch from Zhang Rui.
However, while the previous work was mainly focused on thermal monitoring
the focus here is on the usability from user space perspective.
References: https://lkml.org/lkml/2011/5/26/93
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Signed-off-by: Jacob Pan <jacob.jun.pan@linux.intel.com>
Reviewed-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
2013-10-17 11:28:35 -06:00
|
|
|
{
|
2019-07-10 07:44:34 -06:00
|
|
|
platform_device_unregister(rapl_msr_platdev);
|
2018-01-09 17:38:23 -07:00
|
|
|
unregister_pm_notifier(&rapl_pm_notifier);
|
PowerCap: Introduce Intel RAPL power capping driver
The Intel Running Average Power Limit (RAPL) technology provides platform
software with the ability to monitor, control, and get notifications on
power usage.
This feature is present in all Sandy Bridge and later Intel processors.
Newer models allow more fine grained controls to be applied. In RAPL,
power control is divided into domains, which include package, DRAM
controller, CPU core (Power Plane 0), graphics uncore (power plane 1), etc.
The purpose of this driver is to expose the RAPL settings to userspace.
Overall, RAPL fits in the new powercap class driver in that platform
level power capping controls are exposed via this generic interface.
This driver is based on an earlier patch from Zhang Rui.
However, while the previous work was mainly focused on thermal monitoring
the focus here is on the usability from user space perspective.
References: https://lkml.org/lkml/2011/5/26/93
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Signed-off-by: Jacob Pan <jacob.jun.pan@linux.intel.com>
Reviewed-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
2013-10-17 11:28:35 -06:00
|
|
|
}
|
|
|
|
|
2019-07-10 07:44:34 -06:00
|
|
|
module_init(rapl_init);
|
|
|
|
module_exit(rapl_exit);
|
|
|
|
|
2019-07-10 07:44:30 -06:00
|
|
|
MODULE_DESCRIPTION("Intel Runtime Average Power Limit (RAPL) common code");
|
PowerCap: Introduce Intel RAPL power capping driver
The Intel Running Average Power Limit (RAPL) technology provides platform
software with the ability to monitor, control, and get notifications on
power usage.
This feature is present in all Sandy Bridge and later Intel processors.
Newer models allow more fine grained controls to be applied. In RAPL,
power control is divided into domains, which include package, DRAM
controller, CPU core (Power Plane 0), graphics uncore (power plane 1), etc.
The purpose of this driver is to expose the RAPL settings to userspace.
Overall, RAPL fits in the new powercap class driver in that platform
level power capping controls are exposed via this generic interface.
This driver is based on an earlier patch from Zhang Rui.
However, while the previous work was mainly focused on thermal monitoring
the focus here is on the usability from user space perspective.
References: https://lkml.org/lkml/2011/5/26/93
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Signed-off-by: Jacob Pan <jacob.jun.pan@linux.intel.com>
Reviewed-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
2013-10-17 11:28:35 -06:00
|
|
|
MODULE_AUTHOR("Jacob Pan <jacob.jun.pan@intel.com>");
|
|
|
|
MODULE_LICENSE("GPL v2");
|