1 .. SPDX-License-Identifier: GPL-2.0
2 .. include:: <isonum.txt>
4 ===============================================
5 ``intel_pstate`` CPU Performance Scaling Driver
6 ===============================================
8 :Copyright: |copy| 2017 Intel Corporation
10 :Author: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
16 ``intel_pstate`` is a part of the
17 :doc:`CPU performance scaling subsystem <cpufreq>` in the Linux kernel
18 (``CPUFreq``). It is a scaling driver for the Sandy Bridge and later
19 generations of Intel processors. Note, however, that some of those processors
20 may not be supported. [To understand ``intel_pstate`` it is necessary to know
21 how ``CPUFreq`` works in general, so this is the time to read :doc:`cpufreq` if
22 you have not done that yet.]
24 For the processors supported by ``intel_pstate``, the P-state concept is broader
25 than just an operating frequency or an operating performance point (see the
26 LinuxCon Europe 2015 presentation by Kristen Accardi [1]_ for more
27 information about that). For this reason, the representation of P-states used
28 by ``intel_pstate`` internally follows the hardware specification (for details
29 refer to Intel Software Developer’s Manual [2]_). However, the ``CPUFreq`` core
30 uses frequencies for identifying operating performance points of CPUs and
31 frequencies are involved in the user space interface exposed by it, so
32 ``intel_pstate`` maps its internal representation of P-states to frequencies too
33 (fortunately, that mapping is unambiguous). At the same time, it would not be
34 practical for ``intel_pstate`` to supply the ``CPUFreq`` core with a table of
35 available frequencies due to the possible size of it, so the driver does not do
36 that. Some functionality of the core is limited by that.
38 Since the hardware P-state selection interface used by ``intel_pstate`` is
39 available at the logical CPU level, the driver always works with individual
40 CPUs. Consequently, if ``intel_pstate`` is in use, every ``CPUFreq`` policy
41 object corresponds to one logical CPU and ``CPUFreq`` policies are effectively
42 equivalent to CPUs. In particular, this means that they become "inactive" every
43 time the corresponding CPU is taken offline and need to be re-initialized when
46 ``intel_pstate`` is not modular, so it cannot be unloaded, which means that the
47 only way to pass early-configuration-time parameters to it is via the kernel
48 command line. However, its configuration can be adjusted via ``sysfs`` to a
49 great extent. In some configurations it even is possible to unregister it via
50 ``sysfs`` which allows another ``CPUFreq`` scaling driver to be loaded and
51 registered (see `below <status_attr_>`_).
57 ``intel_pstate`` can operate in three different modes: in the active mode with
58 or without hardware-managed P-states support and in the passive mode. Which of
59 them will be in effect depends on what kernel command line options are used and
60 on the capabilities of the processor.
65 This is the default operation mode of ``intel_pstate`` for processors with
66 hardware-managed P-states (HWP) support. If it works in this mode, the
67 ``scaling_driver`` policy attribute in ``sysfs`` for all ``CPUFreq`` policies
68 contains the string "intel_pstate".
70 In this mode the driver bypasses the scaling governors layer of ``CPUFreq`` and
71 provides its own scaling algorithms for P-state selection. Those algorithms
72 can be applied to ``CPUFreq`` policies in the same way as generic scaling
73 governors (that is, through the ``scaling_governor`` policy attribute in
74 ``sysfs``). [Note that different P-state selection algorithms may be chosen for
75 different policies, but that is not recommended.]
77 They are not generic scaling governors, but their names are the same as the
78 names of some of those governors. Moreover, confusingly enough, they generally
79 do not work in the same way as the generic governors they share the names with.
80 For example, the ``powersave`` P-state selection algorithm provided by
81 ``intel_pstate`` is not a counterpart of the generic ``powersave`` governor
82 (roughly, it corresponds to the ``schedutil`` and ``ondemand`` governors).
84 There are two P-state selection algorithms provided by ``intel_pstate`` in the
85 active mode: ``powersave`` and ``performance``. The way they both operate
86 depends on whether or not the hardware-managed P-states (HWP) feature has been
87 enabled in the processor and possibly on the processor model.
89 Which of the P-state selection algorithms is used by default depends on the
90 :c:macro:`CONFIG_CPU_FREQ_DEFAULT_GOV_PERFORMANCE` kernel configuration option.
91 Namely, if that option is set, the ``performance`` algorithm will be used by
92 default, and the other one will be used by default if it is not set.
97 If the processor supports the HWP feature, it will be enabled during the
98 processor initialization and cannot be disabled after that. It is possible
99 to avoid enabling it by passing the ``intel_pstate=no_hwp`` argument to the
100 kernel in the command line.
102 If the HWP feature has been enabled, ``intel_pstate`` relies on the processor to
103 select P-states by itself, but still it can give hints to the processor's
104 internal P-state selection logic. What those hints are depends on which P-state
105 selection algorithm has been applied to the given policy (or to the CPU it
108 Even though the P-state selection is carried out by the processor automatically,
109 ``intel_pstate`` registers utilization update callbacks with the CPU scheduler
110 in this mode. However, they are not used for running a P-state selection
111 algorithm, but for periodic updates of the current CPU frequency information to
112 be made available from the ``scaling_cur_freq`` policy attribute in ``sysfs``.
114 HWP + ``performance``
115 .....................
117 In this configuration ``intel_pstate`` will write 0 to the processor's
118 Energy-Performance Preference (EPP) knob (if supported) or its
119 Energy-Performance Bias (EPB) knob (otherwise), which means that the processor's
120 internal P-state selection logic is expected to focus entirely on performance.
122 This will override the EPP/EPB setting coming from the ``sysfs`` interface
123 (see `Energy vs Performance Hints`_ below).
125 Also, in this configuration the range of P-states available to the processor's
126 internal P-state selection logic is always restricted to the upper boundary
127 (that is, the maximum P-state that the driver is allowed to use).
132 In this configuration ``intel_pstate`` will set the processor's
133 Energy-Performance Preference (EPP) knob (if supported) or its
134 Energy-Performance Bias (EPB) knob (otherwise) to whatever value it was
135 previously set to via ``sysfs`` (or whatever default value it was
136 set to by the platform firmware). This usually causes the processor's
137 internal P-state selection logic to be less performance-focused.
139 Active Mode Without HWP
140 ~~~~~~~~~~~~~~~~~~~~~~~
142 This operation mode is optional for processors that do not support the HWP
143 feature or when the ``intel_pstate=no_hwp`` argument is passed to the kernel in
144 the command line. The active mode is used in those cases if the
145 ``intel_pstate=active`` argument is passed to the kernel in the command line.
146 In this mode ``intel_pstate`` may refuse to work with processors that are not
147 recognized by it. [Note that ``intel_pstate`` will never refuse to work with
148 any processor with the HWP feature enabled.]
150 In this mode ``intel_pstate`` registers utilization update callbacks with the
151 CPU scheduler in order to run a P-state selection algorithm, either
152 ``powersave`` or ``performance``, depending on the ``scaling_governor`` policy
153 setting in ``sysfs``. The current CPU frequency information to be made
154 available from the ``scaling_cur_freq`` policy attribute in ``sysfs`` is
155 periodically updated by those utilization update callbacks too.
160 Without HWP, this P-state selection algorithm is always the same regardless of
161 the processor model and platform configuration.
163 It selects the maximum P-state it is allowed to use, subject to limits set via
164 ``sysfs``, every time the driver configuration for the given CPU is updated
165 (e.g. via ``sysfs``).
167 This is the default P-state selection algorithm if the
168 :c:macro:`CONFIG_CPU_FREQ_DEFAULT_GOV_PERFORMANCE` kernel configuration option
174 Without HWP, this P-state selection algorithm is similar to the algorithm
175 implemented by the generic ``schedutil`` scaling governor except that the
176 utilization metric used by it is based on numbers coming from feedback
177 registers of the CPU. It generally selects P-states proportional to the
178 current CPU utilization.
180 This algorithm is run by the driver's utilization update callback for the
181 given CPU when it is invoked by the CPU scheduler, but not more often than
182 every 10 ms. Like in the ``performance`` case, the hardware configuration
183 is not touched if the new P-state turns out to be the same as the current
186 This is the default P-state selection algorithm if the
187 :c:macro:`CONFIG_CPU_FREQ_DEFAULT_GOV_PERFORMANCE` kernel configuration option
193 This is the default operation mode of ``intel_pstate`` for processors without
194 hardware-managed P-states (HWP) support. It is always used if the
195 ``intel_pstate=passive`` argument is passed to the kernel in the command line
196 regardless of whether or not the given processor supports HWP. [Note that the
197 ``intel_pstate=no_hwp`` setting implies ``intel_pstate=passive`` if it is used
198 without ``intel_pstate=active``.] Like in the active mode without HWP support,
199 in this mode ``intel_pstate`` may refuse to work with processors that are not
202 If the driver works in this mode, the ``scaling_driver`` policy attribute in
203 ``sysfs`` for all ``CPUFreq`` policies contains the string "intel_cpufreq".
204 Then, the driver behaves like a regular ``CPUFreq`` scaling driver. That is,
205 it is invoked by generic scaling governors when necessary to talk to the
206 hardware in order to change the P-state of a CPU (in particular, the
207 ``schedutil`` governor can invoke it directly from scheduler context).
209 While in this mode, ``intel_pstate`` can be used with all of the (generic)
210 scaling governors listed by the ``scaling_available_governors`` policy attribute
211 in ``sysfs`` (and the P-state selection algorithms described above are not
212 used). Then, it is responsible for the configuration of policy objects
213 corresponding to CPUs and provides the ``CPUFreq`` core (and the scaling
214 governors attached to the policy objects) with accurate information on the
215 maximum and minimum operating frequencies supported by the hardware (including
216 the so-called "turbo" frequency ranges). In other words, in the passive mode
217 the entire range of available P-states is exposed by ``intel_pstate`` to the
218 ``CPUFreq`` core. However, in this mode the driver does not register
219 utilization update callbacks with the CPU scheduler and the ``scaling_cur_freq``
220 information comes from the ``CPUFreq`` core (and is the last frequency selected
221 by the current scaling governor for the given policy).
226 Turbo P-states Support
227 ======================
229 In the majority of cases, the entire range of P-states available to
230 ``intel_pstate`` can be divided into two sub-ranges that correspond to
231 different types of processor behavior, above and below a boundary that
232 will be referred to as the "turbo threshold" in what follows.
234 The P-states above the turbo threshold are referred to as "turbo P-states" and
235 the whole sub-range of P-states they belong to is referred to as the "turbo
236 range". These names are related to the Turbo Boost technology allowing a
237 multicore processor to opportunistically increase the P-state of one or more
238 cores if there is enough power to do that and if that is not going to cause the
239 thermal envelope of the processor package to be exceeded.
241 Specifically, if software sets the P-state of a CPU core within the turbo range
242 (that is, above the turbo threshold), the processor is permitted to take over
243 performance scaling control for that core and put it into turbo P-states of its
244 choice going forward. However, that permission is interpreted differently by
245 different processor generations. Namely, the Sandy Bridge generation of
246 processors will never use any P-states above the last one set by software for
247 the given core, even if it is within the turbo range, whereas all of the later
248 processor generations will take it as a license to use any P-states from the
249 turbo range, even above the one set by software. In other words, on those
250 processors setting any P-state from the turbo range will enable the processor
251 to put the given core into all turbo P-states up to and including the maximum
252 supported one as it sees fit.
254 One important property of turbo P-states is that they are not sustainable. More
255 precisely, there is no guarantee that any CPUs will be able to stay in any of
256 those states indefinitely, because the power distribution within the processor
257 package may change over time or the thermal envelope it was designed for might
258 be exceeded if a turbo P-state was used for too long.
260 In turn, the P-states below the turbo threshold generally are sustainable. In
261 fact, if one of them is set by software, the processor is not expected to change
262 it to a lower one unless in a thermal stress or a power limit violation
263 situation (a higher P-state may still be used if it is set for another CPU in
264 the same package at the same time, for example).
266 Some processors allow multiple cores to be in turbo P-states at the same time,
267 but the maximum P-state that can be set for them generally depends on the number
268 of cores running concurrently. The maximum turbo P-state that can be set for 3
269 cores at the same time usually is lower than the analogous maximum P-state for
270 2 cores, which in turn usually is lower than the maximum turbo P-state that can
271 be set for 1 core. The one-core maximum turbo P-state is thus the maximum
272 supported one overall.
274 The maximum supported turbo P-state, the turbo threshold (the maximum supported
275 non-turbo P-state) and the minimum supported P-state are specific to the
276 processor model and can be determined by reading the processor's model-specific
277 registers (MSRs). Moreover, some processors support the Configurable TDP
278 (Thermal Design Power) feature and, when that feature is enabled, the turbo
279 threshold effectively becomes a configurable value that can be set by the
282 Unlike ``_PSS`` objects in the ACPI tables, ``intel_pstate`` always exposes
283 the entire range of available P-states, including the whole turbo range, to the
284 ``CPUFreq`` core and (in the passive mode) to generic scaling governors. This
285 generally causes turbo P-states to be set more often when ``intel_pstate`` is
286 used relative to ACPI-based CPU performance scaling (see `below <acpi-cpufreq_>`_
287 for more information).
289 Moreover, since ``intel_pstate`` always knows what the real turbo threshold is
290 (even if the Configurable TDP feature is enabled in the processor), its
291 ``no_turbo`` attribute in ``sysfs`` (described `below <no_turbo_attr_>`_) should
292 work as expected in all cases (that is, if set to disable turbo P-states, it
293 always should prevent ``intel_pstate`` from using them).
299 To handle a given processor ``intel_pstate`` requires a number of different
300 pieces of information on it to be known, including:
302 * The minimum supported P-state.
304 * The maximum supported `non-turbo P-state <turbo_>`_.
306 * Whether or not turbo P-states are supported at all.
308 * The maximum supported `one-core turbo P-state <turbo_>`_ (if turbo P-states
311 * The scaling formula to translate the driver's internal representation
312 of P-states into frequencies and the other way around.
314 Generally, ways to obtain that information are specific to the processor model
315 or family. Although it often is possible to obtain all of it from the processor
316 itself (using model-specific registers), there are cases in which hardware
317 manuals need to be consulted to get to it too.
319 For this reason, there is a list of supported processors in ``intel_pstate`` and
320 the driver initialization will fail if the detected processor is not in that
321 list, unless it supports the `HWP feature <Active Mode_>`_. [The interface to
322 obtain all of the information listed above is the same for all of the processors
323 supporting the HWP feature, which is why they all are supported by
327 User Space Interface in ``sysfs``
328 =================================
333 ``intel_pstate`` exposes several global attributes (files) in ``sysfs`` to
334 control its functionality at the system level. They are located in the
335 ``/sys/devices/system/cpu/intel_pstate/`` directory and affect all CPUs.
337 Some of them are not present if the ``intel_pstate=per_cpu_perf_limits``
338 argument is passed to the kernel in the command line.
341 Maximum P-state the driver is allowed to set in percent of the
342 maximum supported performance level (the highest supported `turbo
345 This attribute will not be exposed if the
346 ``intel_pstate=per_cpu_perf_limits`` argument is present in the kernel
350 Minimum P-state the driver is allowed to set in percent of the
351 maximum supported performance level (the highest supported `turbo
354 This attribute will not be exposed if the
355 ``intel_pstate=per_cpu_perf_limits`` argument is present in the kernel
359 Number of P-states supported by the processor (between 0 and 255
360 inclusive) including both turbo and non-turbo P-states (see
361 `Turbo P-states Support`_).
363 The value of this attribute is not affected by the ``no_turbo``
364 setting described `below <no_turbo_attr_>`_.
366 This attribute is read-only.
369 Ratio of the `turbo range <turbo_>`_ size to the size of the entire
370 range of supported P-states, in percent.
372 This attribute is read-only.
377 If set (equal to 1), the driver is not allowed to set any turbo P-states
378 (see `Turbo P-states Support`_). If unset (equalt to 0, which is the
379 default), turbo P-states can be set by the driver.
380 [Note that ``intel_pstate`` does not support the general ``boost``
381 attribute (supported by some other scaling drivers) which is replaced
384 This attrubute does not affect the maximum supported frequency value
385 supplied to the ``CPUFreq`` core and exposed via the policy interface,
386 but it affects the maximum possible value of per-policy P-state limits
387 (see `Interpretation of Policy Attributes`_ below for details).
389 ``hwp_dynamic_boost``
390 This attribute is only present if ``intel_pstate`` works in the
391 `active mode with the HWP feature enabled <Active Mode With HWP_>`_ in
392 the processor. If set (equal to 1), it causes the minimum P-state limit
393 to be increased dynamically for a short time whenever a task previously
394 waiting on I/O is selected to run on a given logical CPU (the purpose
395 of this mechanism is to improve performance).
397 This setting has no effect on logical CPUs whose minimum P-state limit
398 is directly set to the highest non-turbo P-state or above it.
403 Operation mode of the driver: "active", "passive" or "off".
406 The driver is functional and in the `active mode
410 The driver is functional and in the `passive mode
414 The driver is not functional (it is not registered as a scaling
415 driver with the ``CPUFreq`` core).
417 This attribute can be written to in order to change the driver's
418 operation mode or to unregister it. The string written to it must be
419 one of the possible values of it and, if successful, the write will
420 cause the driver to switch over to the operation mode represented by
421 that string - or to be unregistered in the "off" case. [Actually,
422 switching over from the active mode to the passive mode or the other
423 way around causes the driver to be unregistered and registered again
424 with a different set of callbacks, so all of its settings (the global
425 as well as the per-policy ones) are then reset to their default
426 values, possibly depending on the target operation mode.]
428 That only is supported in some configurations, though (for example, if
429 the `HWP feature is enabled in the processor <Active Mode With HWP_>`_,
430 the operation mode of the driver cannot be changed), and if it is not
431 supported in the current configuration, writes to this attribute will
432 fail with an appropriate error.
434 Interpretation of Policy Attributes
435 -----------------------------------
437 The interpretation of some ``CPUFreq`` policy attributes described in
438 :doc:`cpufreq` is special with ``intel_pstate`` as the current scaling driver
439 and it generally depends on the driver's `operation mode <Operation Modes_>`_.
441 First of all, the values of the ``cpuinfo_max_freq``, ``cpuinfo_min_freq`` and
442 ``scaling_cur_freq`` attributes are produced by applying a processor-specific
443 multiplier to the internal P-state representation used by ``intel_pstate``.
444 Also, the values of the ``scaling_max_freq`` and ``scaling_min_freq``
445 attributes are capped by the frequency corresponding to the maximum P-state that
446 the driver is allowed to set.
448 If the ``no_turbo`` `global attribute <no_turbo_attr_>`_ is set, the driver is
449 not allowed to use turbo P-states, so the maximum value of ``scaling_max_freq``
450 and ``scaling_min_freq`` is limited to the maximum non-turbo P-state frequency.
451 Accordingly, setting ``no_turbo`` causes ``scaling_max_freq`` and
452 ``scaling_min_freq`` to go down to that value if they were above it before.
453 However, the old values of ``scaling_max_freq`` and ``scaling_min_freq`` will be
454 restored after unsetting ``no_turbo``, unless these attributes have been written
455 to after ``no_turbo`` was set.
457 If ``no_turbo`` is not set, the maximum possible value of ``scaling_max_freq``
458 and ``scaling_min_freq`` corresponds to the maximum supported turbo P-state,
459 which also is the value of ``cpuinfo_max_freq`` in either case.
461 Next, the following policy attributes have special meaning if
462 ``intel_pstate`` works in the `active mode <Active Mode_>`_:
464 ``scaling_available_governors``
465 List of P-state selection algorithms provided by ``intel_pstate``.
468 P-state selection algorithm provided by ``intel_pstate`` currently in
469 use with the given policy.
472 Frequency of the average P-state of the CPU represented by the given
473 policy for the time interval between the last two invocations of the
474 driver's utilization update callback by the CPU scheduler for that CPU.
476 One more policy attribute is present if the `HWP feature is enabled in the
477 processor <Active Mode With HWP_>`_:
480 Shows the base frequency of the CPU. Any frequency above this will be
481 in the turbo frequency range.
483 The meaning of these attributes in the `passive mode <Passive Mode_>`_ is the
484 same as for other scaling drivers.
486 Additionally, the value of the ``scaling_driver`` attribute for ``intel_pstate``
487 depends on the operation mode of the driver. Namely, it is either
488 "intel_pstate" (in the `active mode <Active Mode_>`_) or "intel_cpufreq" (in the
489 `passive mode <Passive Mode_>`_).
491 Coordination of P-State Limits
492 ------------------------------
494 ``intel_pstate`` allows P-state limits to be set in two ways: with the help of
495 the ``max_perf_pct`` and ``min_perf_pct`` `global attributes
496 <Global Attributes_>`_ or via the ``scaling_max_freq`` and ``scaling_min_freq``
497 ``CPUFreq`` policy attributes. The coordination between those limits is based
498 on the following rules, regardless of the current operation mode of the driver:
500 1. All CPUs are affected by the global limits (that is, none of them can be
501 requested to run faster than the global maximum and none of them can be
502 requested to run slower than the global minimum).
504 2. Each individual CPU is affected by its own per-policy limits (that is, it
505 cannot be requested to run faster than its own per-policy maximum and it
506 cannot be requested to run slower than its own per-policy minimum). The
507 effective performance depends on whether the platform supports per core
508 P-states, hyper-threading is enabled and on current performance requests
509 from other CPUs. When platform doesn't support per core P-states, the
510 effective performance can be more than the policy limits set on a CPU, if
511 other CPUs are requesting higher performance at that moment. Even with per
512 core P-states support, when hyper-threading is enabled, if the sibling CPU
513 is requesting higher performance, the other siblings will get higher
514 performance than their policy limits.
516 3. The global and per-policy limits can be set independently.
518 If the `HWP feature is enabled in the processor <Active Mode With HWP_>`_, the
519 resulting effective values are written into its registers whenever the limits
520 change in order to request its internal P-state selection logic to always set
521 P-states within these limits. Otherwise, the limits are taken into account by
522 scaling governors (in the `passive mode <Passive Mode_>`_) and by the driver
523 every time before setting a new P-state for a CPU.
525 Additionally, if the ``intel_pstate=per_cpu_perf_limits`` command line argument
526 is passed to the kernel, ``max_perf_pct`` and ``min_perf_pct`` are not exposed
527 at all and the only way to set the limits is by using the policy attributes.
530 Energy vs Performance Hints
531 ---------------------------
533 If ``intel_pstate`` works in the `active mode with the HWP feature enabled
534 <Active Mode With HWP_>`_ in the processor, additional attributes are present
535 in every ``CPUFreq`` policy directory in ``sysfs``. They are intended to allow
536 user space to help ``intel_pstate`` to adjust the processor's internal P-state
537 selection logic by focusing it on performance or on energy-efficiency, or
538 somewhere between the two extremes:
540 ``energy_performance_preference``
541 Current value of the energy vs performance hint for the given policy
542 (or the CPU represented by it).
544 The hint can be changed by writing to this attribute.
546 ``energy_performance_available_preferences``
547 List of strings that can be written to the
548 ``energy_performance_preference`` attribute.
550 They represent different energy vs performance hints and should be
551 self-explanatory, except that ``default`` represents whatever hint
552 value was set by the platform firmware.
554 Strings written to the ``energy_performance_preference`` attribute are
555 internally translated to integer values written to the processor's
556 Energy-Performance Preference (EPP) knob (if supported) or its
557 Energy-Performance Bias (EPB) knob.
559 [Note that tasks may by migrated from one CPU to another by the scheduler's
560 load-balancing algorithm and if different energy vs performance hints are
561 set for those CPUs, that may lead to undesirable outcomes. To avoid such
562 issues it is better to set the same energy vs performance hint for all CPUs
563 or to pin every task potentially sensitive to them to a specific CPU.]
567 ``intel_pstate`` vs ``acpi-cpufreq``
568 ====================================
570 On the majority of systems supported by ``intel_pstate``, the ACPI tables
571 provided by the platform firmware contain ``_PSS`` objects returning information
572 that can be used for CPU performance scaling (refer to the ACPI specification
573 [3]_ for details on the ``_PSS`` objects and the format of the information
576 The information returned by the ACPI ``_PSS`` objects is used by the
577 ``acpi-cpufreq`` scaling driver. On systems supported by ``intel_pstate``
578 the ``acpi-cpufreq`` driver uses the same hardware CPU performance scaling
579 interface, but the set of P-states it can use is limited by the ``_PSS``
582 On those systems each ``_PSS`` object returns a list of P-states supported by
583 the corresponding CPU which basically is a subset of the P-states range that can
584 be used by ``intel_pstate`` on the same system, with one exception: the whole
585 `turbo range <turbo_>`_ is represented by one item in it (the topmost one). By
586 convention, the frequency returned by ``_PSS`` for that item is greater by 1 MHz
587 than the frequency of the highest non-turbo P-state listed by it, but the
588 corresponding P-state representation (following the hardware specification)
589 returned for it matches the maximum supported turbo P-state (or is the
590 special value 255 meaning essentially "go as high as you can get").
592 The list of P-states returned by ``_PSS`` is reflected by the table of
593 available frequencies supplied by ``acpi-cpufreq`` to the ``CPUFreq`` core and
594 scaling governors and the minimum and maximum supported frequencies reported by
595 it come from that list as well. In particular, given the special representation
596 of the turbo range described above, this means that the maximum supported
597 frequency reported by ``acpi-cpufreq`` is higher by 1 MHz than the frequency
598 of the highest supported non-turbo P-state listed by ``_PSS`` which, of course,
599 affects decisions made by the scaling governors, except for ``powersave`` and
602 For example, if a given governor attempts to select a frequency proportional to
603 estimated CPU load and maps the load of 100% to the maximum supported frequency
604 (possibly multiplied by a constant), then it will tend to choose P-states below
605 the turbo threshold if ``acpi-cpufreq`` is used as the scaling driver, because
606 in that case the turbo range corresponds to a small fraction of the frequency
607 band it can use (1 MHz vs 1 GHz or more). In consequence, it will only go to
608 the turbo range for the highest loads and the other loads above 50% that might
609 benefit from running at turbo frequencies will be given non-turbo P-states
612 One more issue related to that may appear on systems supporting the
613 `Configurable TDP feature <turbo_>`_ allowing the platform firmware to set the
614 turbo threshold. Namely, if that is not coordinated with the lists of P-states
615 returned by ``_PSS`` properly, there may be more than one item corresponding to
616 a turbo P-state in those lists and there may be a problem with avoiding the
617 turbo range (if desirable or necessary). Usually, to avoid using turbo
618 P-states overall, ``acpi-cpufreq`` simply avoids using the topmost state listed
619 by ``_PSS``, but that is not sufficient when there are other turbo P-states in
620 the list returned by it.
622 Apart from the above, ``acpi-cpufreq`` works like ``intel_pstate`` in the
623 `passive mode <Passive Mode_>`_, except that the number of P-states it can set
624 is limited to the ones listed by the ACPI ``_PSS`` objects.
627 Kernel Command Line Options for ``intel_pstate``
628 ================================================
630 Several kernel command line options can be used to pass early-configuration-time
631 parameters to ``intel_pstate`` in order to enforce specific behavior of it. All
632 of them have to be prepended with the ``intel_pstate=`` prefix.
635 Do not register ``intel_pstate`` as the scaling driver even if the
636 processor is supported by it.
639 Register ``intel_pstate`` in the `passive mode <Passive Mode_>`_ to
642 This option implies the ``no_hwp`` one described below.
645 Register ``intel_pstate`` as the scaling driver instead of
646 ``acpi-cpufreq`` even if the latter is preferred on the given system.
648 This may prevent some platform features (such as thermal controls and
649 power capping) that rely on the availability of ACPI P-states
650 information from functioning as expected, so it should be used with
653 This option does not work with processors that are not supported by
654 ``intel_pstate`` and on platforms where the ``pcc-cpufreq`` scaling
655 driver is used instead of ``acpi-cpufreq``.
658 Do not enable the `hardware-managed P-states (HWP) feature
659 <Active Mode With HWP_>`_ even if it is supported by the processor.
662 Register ``intel_pstate`` as the scaling driver only if the
663 `hardware-managed P-states (HWP) feature <Active Mode With HWP_>`_ is
664 supported by the processor.
667 Take ACPI ``_PPC`` performance limits into account.
669 If the preferred power management profile in the FADT (Fixed ACPI
670 Description Table) is set to "Enterprise Server" or "Performance
671 Server", the ACPI ``_PPC`` limits are taken into account by default
672 and this option has no effect.
674 ``per_cpu_perf_limits``
675 Use per-logical-CPU P-State limits (see `Coordination of P-state
676 Limits`_ for details).
679 Diagnostics and Tuning
680 ======================
685 There are two static trace events that can be used for ``intel_pstate``
686 diagnostics. One of them is the ``cpu_frequency`` trace event generally used
687 by ``CPUFreq``, and the other one is the ``pstate_sample`` trace event specific
688 to ``intel_pstate``. Both of them are triggered by ``intel_pstate`` only if
689 it works in the `active mode <Active Mode_>`_.
691 The following sequence of shell commands can be used to enable them and see
692 their output (if the kernel is generally configured to support event tracing)::
694 # cd /sys/kernel/debug/tracing/
695 # echo 1 > events/power/pstate_sample/enable
696 # echo 1 > events/power/cpu_frequency/enable
698 gnome-terminal--4510 [001] ..s. 1177.680733: pstate_sample: core_busy=107 scaled=94 from=26 to=26 mperf=1143818 aperf=1230607 tsc=29838618 freq=2474476
699 cat-5235 [002] ..s. 1177.681723: cpu_frequency: state=2900000 cpu_id=2
701 If ``intel_pstate`` works in the `passive mode <Passive Mode_>`_, the
702 ``cpu_frequency`` trace event will be triggered either by the ``schedutil``
703 scaling governor (for the policies it is attached to), or by the ``CPUFreq``
704 core (for the policies with other scaling governors).
709 The ``ftrace`` interface can be used for low-level diagnostics of
710 ``intel_pstate``. For example, to check how often the function to set a
711 P-state is called, the ``ftrace`` filter can be set to to
712 :c:func:`intel_pstate_set_pstate`::
714 # cd /sys/kernel/debug/tracing/
715 # cat available_filter_functions | grep -i pstate
716 intel_pstate_set_pstate
717 intel_pstate_cpu_init
719 # echo intel_pstate_set_pstate > set_ftrace_filter
720 # echo function > current_tracer
721 # cat trace | head -15
724 # entries-in-buffer/entries-written: 80/80 #P:4
727 # / _----=> need-resched
728 # | / _---=> hardirq/softirq
729 # || / _--=> preempt-depth
731 # TASK-PID CPU# |||| TIMESTAMP FUNCTION
733 Xorg-3129 [000] ..s. 2537.644844: intel_pstate_set_pstate <-intel_pstate_timer_func
734 gnome-terminal--4510 [002] ..s. 2537.649844: intel_pstate_set_pstate <-intel_pstate_timer_func
735 gnome-shell-3409 [001] ..s. 2537.650850: intel_pstate_set_pstate <-intel_pstate_timer_func
736 <idle>-0 [000] ..s. 2537.654843: intel_pstate_set_pstate <-intel_pstate_timer_func
742 .. [1] Kristen Accardi, *Balancing Power and Performance in the Linux Kernel*,
743 https://events.static.linuxfound.org/sites/events/files/slides/LinuxConEurope_2015.pdf
745 .. [2] *Intel® 64 and IA-32 Architectures Software Developer’s Manual Volume 3: System Programming Guide*,
746 https://www.intel.com/content/www/us/en/architecture-and-technology/64-ia-32-architectures-software-developer-system-programming-manual-325384.html
748 .. [3] *Advanced Configuration and Power Interface Specification*,
749 https://uefi.org/sites/default/files/resources/ACPI_6_3_final_Jan30.pdf