1 .. SPDX-License-Identifier: GPL-2.0 1 .. SPDX-License-Identifier: GPL-2.0
2 .. include:: <isonum.txt> 2 .. include:: <isonum.txt>
3 3
>> 4 .. |struct cpufreq_policy| replace:: :c:type:`struct cpufreq_policy <cpufreq_policy>`
4 .. |intel_pstate| replace:: :doc:`intel_pstate 5 .. |intel_pstate| replace:: :doc:`intel_pstate <intel_pstate>`
5 6
6 ======================= 7 =======================
7 CPU Performance Scaling 8 CPU Performance Scaling
8 ======================= 9 =======================
9 10
10 :Copyright: |copy| 2017 Intel Corporation 11 :Copyright: |copy| 2017 Intel Corporation
11 12
12 :Author: Rafael J. Wysocki <rafael.j.wysocki@in 13 :Author: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
13 14
14 15
15 The Concept of CPU Performance Scaling 16 The Concept of CPU Performance Scaling
16 ====================================== 17 ======================================
17 18
18 The majority of modern processors are capable 19 The majority of modern processors are capable of operating in a number of
19 different clock frequency and voltage configur 20 different clock frequency and voltage configurations, often referred to as
20 Operating Performance Points or P-states (in A 21 Operating Performance Points or P-states (in ACPI terminology). As a rule,
21 the higher the clock frequency and the higher 22 the higher the clock frequency and the higher the voltage, the more instructions
22 can be retired by the CPU over a unit of time, 23 can be retired by the CPU over a unit of time, but also the higher the clock
23 frequency and the higher the voltage, the more 24 frequency and the higher the voltage, the more energy is consumed over a unit of
24 time (or the more power is drawn) by the CPU i 25 time (or the more power is drawn) by the CPU in the given P-state. Therefore
25 there is a natural tradeoff between the CPU ca 26 there is a natural tradeoff between the CPU capacity (the number of instructions
26 that can be executed over a unit of time) and 27 that can be executed over a unit of time) and the power drawn by the CPU.
27 28
28 In some situations it is desirable or even nec 29 In some situations it is desirable or even necessary to run the program as fast
29 as possible and then there is no reason to use 30 as possible and then there is no reason to use any P-states different from the
30 highest one (i.e. the highest-performance freq 31 highest one (i.e. the highest-performance frequency/voltage configuration
31 available). In some other cases, however, it 32 available). In some other cases, however, it may not be necessary to execute
32 instructions so quickly and maintaining the hi 33 instructions so quickly and maintaining the highest available CPU capacity for a
33 relatively long time without utilizing it enti 34 relatively long time without utilizing it entirely may be regarded as wasteful.
34 It also may not be physically possible to main 35 It also may not be physically possible to maintain maximum CPU capacity for too
35 long for thermal or power supply capacity reas 36 long for thermal or power supply capacity reasons or similar. To cover those
36 cases, there are hardware interfaces allowing 37 cases, there are hardware interfaces allowing CPUs to be switched between
37 different frequency/voltage configurations or 38 different frequency/voltage configurations or (in the ACPI terminology) to be
38 put into different P-states. 39 put into different P-states.
39 40
40 Typically, they are used along with algorithms 41 Typically, they are used along with algorithms to estimate the required CPU
41 capacity, so as to decide which P-states to pu 42 capacity, so as to decide which P-states to put the CPUs into. Of course, since
42 the utilization of the system generally change 43 the utilization of the system generally changes over time, that has to be done
43 repeatedly on a regular basis. The activity b 44 repeatedly on a regular basis. The activity by which this happens is referred
44 to as CPU performance scaling or CPU frequency 45 to as CPU performance scaling or CPU frequency scaling (because it involves
45 adjusting the CPU clock frequency). 46 adjusting the CPU clock frequency).
46 47
47 48
48 CPU Performance Scaling in Linux 49 CPU Performance Scaling in Linux
49 ================================ 50 ================================
50 51
51 The Linux kernel supports CPU performance scal 52 The Linux kernel supports CPU performance scaling by means of the ``CPUFreq``
52 (CPU Frequency scaling) subsystem that consist 53 (CPU Frequency scaling) subsystem that consists of three layers of code: the
53 core, scaling governors and scaling drivers. 54 core, scaling governors and scaling drivers.
54 55
55 The ``CPUFreq`` core provides the common code 56 The ``CPUFreq`` core provides the common code infrastructure and user space
56 interfaces for all platforms that support CPU 57 interfaces for all platforms that support CPU performance scaling. It defines
57 the basic framework in which the other compone 58 the basic framework in which the other components operate.
58 59
59 Scaling governors implement algorithms to esti 60 Scaling governors implement algorithms to estimate the required CPU capacity.
60 As a rule, each governor implements one, possi 61 As a rule, each governor implements one, possibly parametrized, scaling
61 algorithm. 62 algorithm.
62 63
63 Scaling drivers talk to the hardware. They pr 64 Scaling drivers talk to the hardware. They provide scaling governors with
64 information on the available P-states (or P-st 65 information on the available P-states (or P-state ranges in some cases) and
65 access platform-specific hardware interfaces t 66 access platform-specific hardware interfaces to change CPU P-states as requested
66 by scaling governors. 67 by scaling governors.
67 68
68 In principle, all available scaling governors 69 In principle, all available scaling governors can be used with every scaling
69 driver. That design is based on the observati 70 driver. That design is based on the observation that the information used by
70 performance scaling algorithms for P-state sel 71 performance scaling algorithms for P-state selection can be represented in a
71 platform-independent form in the majority of c 72 platform-independent form in the majority of cases, so it should be possible
72 to use the same performance scaling algorithm 73 to use the same performance scaling algorithm implemented in exactly the same
73 way regardless of which scaling driver is used 74 way regardless of which scaling driver is used. Consequently, the same set of
74 scaling governors should be suitable for every 75 scaling governors should be suitable for every supported platform.
75 76
76 However, that observation may not hold for per 77 However, that observation may not hold for performance scaling algorithms
77 based on information provided by the hardware 78 based on information provided by the hardware itself, for example through
78 feedback registers, as that information is typ 79 feedback registers, as that information is typically specific to the hardware
79 interface it comes from and may not be easily 80 interface it comes from and may not be easily represented in an abstract,
80 platform-independent way. For this reason, `` 81 platform-independent way. For this reason, ``CPUFreq`` allows scaling drivers
81 to bypass the governor layer and implement the 82 to bypass the governor layer and implement their own performance scaling
82 algorithms. That is done by the |intel_pstate 83 algorithms. That is done by the |intel_pstate| scaling driver.
83 84
84 85
85 ``CPUFreq`` Policy Objects 86 ``CPUFreq`` Policy Objects
86 ========================== 87 ==========================
87 88
88 In some cases the hardware interface for P-sta 89 In some cases the hardware interface for P-state control is shared by multiple
89 CPUs. That is, for example, the same register 90 CPUs. That is, for example, the same register (or set of registers) is used to
90 control the P-state of multiple CPUs at the sa 91 control the P-state of multiple CPUs at the same time and writing to it affects
91 all of those CPUs simultaneously. 92 all of those CPUs simultaneously.
92 93
93 Sets of CPUs sharing hardware P-state control 94 Sets of CPUs sharing hardware P-state control interfaces are represented by
94 ``CPUFreq`` as struct cpufreq_policy objects. !! 95 ``CPUFreq`` as |struct cpufreq_policy| objects. For consistency,
95 struct cpufreq_policy is also used when there !! 96 |struct cpufreq_policy| is also used when there is only one CPU in the given
96 set. 97 set.
97 98
98 The ``CPUFreq`` core maintains a pointer to a !! 99 The ``CPUFreq`` core maintains a pointer to a |struct cpufreq_policy| object for
99 every CPU in the system, including CPUs that a 100 every CPU in the system, including CPUs that are currently offline. If multiple
100 CPUs share the same hardware P-state control i 101 CPUs share the same hardware P-state control interface, all of the pointers
101 corresponding to them point to the same struct !! 102 corresponding to them point to the same |struct cpufreq_policy| object.
102 103
103 ``CPUFreq`` uses struct cpufreq_policy as its !! 104 ``CPUFreq`` uses |struct cpufreq_policy| as its basic data type and the design
104 of its user space interface is based on the po 105 of its user space interface is based on the policy concept.
105 106
106 107
107 CPU Initialization 108 CPU Initialization
108 ================== 109 ==================
109 110
110 First of all, a scaling driver has to be regis 111 First of all, a scaling driver has to be registered for ``CPUFreq`` to work.
111 It is only possible to register one scaling dr 112 It is only possible to register one scaling driver at a time, so the scaling
112 driver is expected to be able to handle all CP 113 driver is expected to be able to handle all CPUs in the system.
113 114
114 The scaling driver may be registered before or 115 The scaling driver may be registered before or after CPU registration. If
115 CPUs are registered earlier, the driver core i 116 CPUs are registered earlier, the driver core invokes the ``CPUFreq`` core to
116 take a note of all of the already registered C 117 take a note of all of the already registered CPUs during the registration of the
117 scaling driver. In turn, if any CPUs are regi 118 scaling driver. In turn, if any CPUs are registered after the registration of
118 the scaling driver, the ``CPUFreq`` core will 119 the scaling driver, the ``CPUFreq`` core will be invoked to take note of them
119 at their registration time. 120 at their registration time.
120 121
121 In any case, the ``CPUFreq`` core is invoked t 122 In any case, the ``CPUFreq`` core is invoked to take note of any logical CPU it
122 has not seen so far as soon as it is ready to 123 has not seen so far as soon as it is ready to handle that CPU. [Note that the
123 logical CPU may be a physical single-core proc 124 logical CPU may be a physical single-core processor, or a single core in a
124 multicore processor, or a hardware thread in a 125 multicore processor, or a hardware thread in a physical processor or processor
125 core. In what follows "CPU" always means "log 126 core. In what follows "CPU" always means "logical CPU" unless explicitly stated
126 otherwise and the word "processor" is used to 127 otherwise and the word "processor" is used to refer to the physical part
127 possibly including multiple logical CPUs.] 128 possibly including multiple logical CPUs.]
128 129
129 Once invoked, the ``CPUFreq`` core checks if t 130 Once invoked, the ``CPUFreq`` core checks if the policy pointer is already set
130 for the given CPU and if so, it skips the poli 131 for the given CPU and if so, it skips the policy object creation. Otherwise,
131 a new policy object is created and initialized 132 a new policy object is created and initialized, which involves the creation of
132 a new policy directory in ``sysfs``, and the p 133 a new policy directory in ``sysfs``, and the policy pointer corresponding to
133 the given CPU is set to the new policy object' 134 the given CPU is set to the new policy object's address in memory.
134 135
135 Next, the scaling driver's ``->init()`` callba 136 Next, the scaling driver's ``->init()`` callback is invoked with the policy
136 pointer of the new CPU passed to it as the arg 137 pointer of the new CPU passed to it as the argument. That callback is expected
137 to initialize the performance scaling hardware 138 to initialize the performance scaling hardware interface for the given CPU (or,
138 more precisely, for the set of CPUs sharing th 139 more precisely, for the set of CPUs sharing the hardware interface it belongs
139 to, represented by its policy object) and, if 140 to, represented by its policy object) and, if the policy object it has been
140 called for is new, to set parameters of the po 141 called for is new, to set parameters of the policy, like the minimum and maximum
141 frequencies supported by the hardware, the tab 142 frequencies supported by the hardware, the table of available frequencies (if
142 the set of supported P-states is not a continu 143 the set of supported P-states is not a continuous range), and the mask of CPUs
143 that belong to the same policy (including both 144 that belong to the same policy (including both online and offline CPUs). That
144 mask is then used by the core to populate the 145 mask is then used by the core to populate the policy pointers for all of the
145 CPUs in it. 146 CPUs in it.
146 147
147 The next major initialization step for a new p 148 The next major initialization step for a new policy object is to attach a
148 scaling governor to it (to begin with, that is 149 scaling governor to it (to begin with, that is the default scaling governor
149 determined by the kernel command line or confi !! 150 determined by the kernel configuration, but it may be changed later
150 later via ``sysfs``). First, a pointer to the !! 151 via ``sysfs``). First, a pointer to the new policy object is passed to the
151 the governor's ``->init()`` callback which is !! 152 governor's ``->init()`` callback which is expected to initialize all of the
152 data structures necessary to handle the given 153 data structures necessary to handle the given policy and, possibly, to add
153 a governor ``sysfs`` interface to it. Next, t 154 a governor ``sysfs`` interface to it. Next, the governor is started by
154 invoking its ``->start()`` callback. 155 invoking its ``->start()`` callback.
155 156
156 That callback is expected to register per-CPU 157 That callback is expected to register per-CPU utilization update callbacks for
157 all of the online CPUs belonging to the given 158 all of the online CPUs belonging to the given policy with the CPU scheduler.
158 The utilization update callbacks will be invok 159 The utilization update callbacks will be invoked by the CPU scheduler on
159 important events, like task enqueue and dequeu 160 important events, like task enqueue and dequeue, on every iteration of the
160 scheduler tick or generally whenever the CPU u 161 scheduler tick or generally whenever the CPU utilization may change (from the
161 scheduler's perspective). They are expected t 162 scheduler's perspective). They are expected to carry out computations needed
162 to determine the P-state to use for the given 163 to determine the P-state to use for the given policy going forward and to
163 invoke the scaling driver to make changes to t 164 invoke the scaling driver to make changes to the hardware in accordance with
164 the P-state selection. The scaling driver may 165 the P-state selection. The scaling driver may be invoked directly from
165 scheduler context or asynchronously, via a ker 166 scheduler context or asynchronously, via a kernel thread or workqueue, depending
166 on the configuration and capabilities of the s 167 on the configuration and capabilities of the scaling driver and the governor.
167 168
168 Similar steps are taken for policy objects tha 169 Similar steps are taken for policy objects that are not new, but were "inactive"
169 previously, meaning that all of the CPUs belon 170 previously, meaning that all of the CPUs belonging to them were offline. The
170 only practical difference in that case is that 171 only practical difference in that case is that the ``CPUFreq`` core will attempt
171 to use the scaling governor previously used wi 172 to use the scaling governor previously used with the policy that became
172 "inactive" (and is re-initialized now) instead 173 "inactive" (and is re-initialized now) instead of the default governor.
173 174
174 In turn, if a previously offline CPU is being 175 In turn, if a previously offline CPU is being brought back online, but some
175 other CPUs sharing the policy object with it a 176 other CPUs sharing the policy object with it are online already, there is no
176 need to re-initialize the policy object at all 177 need to re-initialize the policy object at all. In that case, it only is
177 necessary to restart the scaling governor so t 178 necessary to restart the scaling governor so that it can take the new online CPU
178 into account. That is achieved by invoking th 179 into account. That is achieved by invoking the governor's ``->stop`` and
179 ``->start()`` callbacks, in this order, for th 180 ``->start()`` callbacks, in this order, for the entire policy.
180 181
181 As mentioned before, the |intel_pstate| scalin 182 As mentioned before, the |intel_pstate| scaling driver bypasses the scaling
182 governor layer of ``CPUFreq`` and provides its 183 governor layer of ``CPUFreq`` and provides its own P-state selection algorithms.
183 Consequently, if |intel_pstate| is used, scali 184 Consequently, if |intel_pstate| is used, scaling governors are not attached to
184 new policy objects. Instead, the driver's ``- 185 new policy objects. Instead, the driver's ``->setpolicy()`` callback is invoked
185 to register per-CPU utilization update callbac 186 to register per-CPU utilization update callbacks for each policy. These
186 callbacks are invoked by the CPU scheduler in 187 callbacks are invoked by the CPU scheduler in the same way as for scaling
187 governors, but in the |intel_pstate| case they 188 governors, but in the |intel_pstate| case they both determine the P-state to
188 use and change the hardware configuration acco 189 use and change the hardware configuration accordingly in one go from scheduler
189 context. 190 context.
190 191
191 The policy objects created during CPU initiali 192 The policy objects created during CPU initialization and other data structures
192 associated with them are torn down when the sc 193 associated with them are torn down when the scaling driver is unregistered
193 (which happens when the kernel module containi 194 (which happens when the kernel module containing it is unloaded, for example) or
194 when the last CPU belonging to the given polic 195 when the last CPU belonging to the given policy in unregistered.
195 196
196 197
197 Policy Interface in ``sysfs`` 198 Policy Interface in ``sysfs``
198 ============================= 199 =============================
199 200
200 During the initialization of the kernel, the ` 201 During the initialization of the kernel, the ``CPUFreq`` core creates a
201 ``sysfs`` directory (kobject) called ``cpufreq 202 ``sysfs`` directory (kobject) called ``cpufreq`` under
202 :file:`/sys/devices/system/cpu/`. 203 :file:`/sys/devices/system/cpu/`.
203 204
204 That directory contains a ``policyX`` subdirec 205 That directory contains a ``policyX`` subdirectory (where ``X`` represents an
205 integer number) for every policy object mainta 206 integer number) for every policy object maintained by the ``CPUFreq`` core.
206 Each ``policyX`` directory is pointed to by `` 207 Each ``policyX`` directory is pointed to by ``cpufreq`` symbolic links
207 under :file:`/sys/devices/system/cpu/cpuY/` (w 208 under :file:`/sys/devices/system/cpu/cpuY/` (where ``Y`` represents an integer
208 that may be different from the one represented 209 that may be different from the one represented by ``X``) for all of the CPUs
209 associated with (or belonging to) the given po 210 associated with (or belonging to) the given policy. The ``policyX`` directories
210 in :file:`/sys/devices/system/cpu/cpufreq` eac 211 in :file:`/sys/devices/system/cpu/cpufreq` each contain policy-specific
211 attributes (files) to control ``CPUFreq`` beha 212 attributes (files) to control ``CPUFreq`` behavior for the corresponding policy
212 objects (that is, for all of the CPUs associat 213 objects (that is, for all of the CPUs associated with them).
213 214
214 Some of those attributes are generic. They ar 215 Some of those attributes are generic. They are created by the ``CPUFreq`` core
215 and their behavior generally does not depend o 216 and their behavior generally does not depend on what scaling driver is in use
216 and what scaling governor is attached to the g 217 and what scaling governor is attached to the given policy. Some scaling drivers
217 also add driver-specific attributes to the pol 218 also add driver-specific attributes to the policy directories in ``sysfs`` to
218 control policy-specific aspects of driver beha 219 control policy-specific aspects of driver behavior.
219 220
220 The generic attributes under :file:`/sys/devic 221 The generic attributes under :file:`/sys/devices/system/cpu/cpufreq/policyX/`
221 are the following: 222 are the following:
222 223
223 ``affected_cpus`` 224 ``affected_cpus``
224 List of online CPUs belonging to this 225 List of online CPUs belonging to this policy (i.e. sharing the hardware
225 performance scaling interface represen 226 performance scaling interface represented by the ``policyX`` policy
226 object). 227 object).
227 228
228 ``bios_limit`` 229 ``bios_limit``
229 If the platform firmware (BIOS) tells 230 If the platform firmware (BIOS) tells the OS to apply an upper limit to
230 CPU frequencies, that limit will be re 231 CPU frequencies, that limit will be reported through this attribute (if
231 present). 232 present).
232 233
233 The existence of the limit may be a re 234 The existence of the limit may be a result of some (often unintentional)
234 BIOS settings, restrictions coming fro 235 BIOS settings, restrictions coming from a service processor or another
235 BIOS/HW-based mechanisms. 236 BIOS/HW-based mechanisms.
236 237
237 This does not cover ACPI thermal limit 238 This does not cover ACPI thermal limitations which can be discovered
238 through a generic thermal driver. 239 through a generic thermal driver.
239 240
240 This attribute is not present if the s 241 This attribute is not present if the scaling driver in use does not
241 support it. 242 support it.
242 243
243 ``cpuinfo_cur_freq`` 244 ``cpuinfo_cur_freq``
244 Current frequency of the CPUs belongin 245 Current frequency of the CPUs belonging to this policy as obtained from
245 the hardware (in KHz). 246 the hardware (in KHz).
246 247
247 This is expected to be the frequency t 248 This is expected to be the frequency the hardware actually runs at.
248 If that frequency cannot be determined 249 If that frequency cannot be determined, this attribute should not
249 be present. 250 be present.
250 251
251 ``cpuinfo_max_freq`` 252 ``cpuinfo_max_freq``
252 Maximum possible operating frequency t 253 Maximum possible operating frequency the CPUs belonging to this policy
253 can run at (in kHz). 254 can run at (in kHz).
254 255
255 ``cpuinfo_min_freq`` 256 ``cpuinfo_min_freq``
256 Minimum possible operating frequency t 257 Minimum possible operating frequency the CPUs belonging to this policy
257 can run at (in kHz). 258 can run at (in kHz).
258 259
259 ``cpuinfo_transition_latency`` 260 ``cpuinfo_transition_latency``
260 The time it takes to switch the CPUs b 261 The time it takes to switch the CPUs belonging to this policy from one
261 P-state to another, in nanoseconds. 262 P-state to another, in nanoseconds.
262 263
263 If unknown or if known to be so high t 264 If unknown or if known to be so high that the scaling driver does not
264 work with the `ondemand`_ governor, -1 265 work with the `ondemand`_ governor, -1 (:c:macro:`CPUFREQ_ETERNAL`)
265 will be returned by reads from this at 266 will be returned by reads from this attribute.
266 267
267 ``related_cpus`` 268 ``related_cpus``
268 List of all (online and offline) CPUs 269 List of all (online and offline) CPUs belonging to this policy.
269 270
270 ``scaling_available_frequencies`` <<
271 List of available frequencies of the C <<
272 (in kHz). <<
273 <<
274 ``scaling_available_governors`` 271 ``scaling_available_governors``
275 List of ``CPUFreq`` scaling governors 272 List of ``CPUFreq`` scaling governors present in the kernel that can
276 be attached to this policy or (if the 273 be attached to this policy or (if the |intel_pstate| scaling driver is
277 in use) list of scaling algorithms pro 274 in use) list of scaling algorithms provided by the driver that can be
278 applied to this policy. 275 applied to this policy.
279 276
280 [Note that some governors are modular 277 [Note that some governors are modular and it may be necessary to load a
281 kernel module for the governor held by 278 kernel module for the governor held by it to become available and be
282 listed by this attribute.] 279 listed by this attribute.]
283 280
284 ``scaling_cur_freq`` 281 ``scaling_cur_freq``
285 Current frequency of all of the CPUs b 282 Current frequency of all of the CPUs belonging to this policy (in kHz).
286 283
287 In the majority of cases, this is the 284 In the majority of cases, this is the frequency of the last P-state
288 requested by the scaling driver from t 285 requested by the scaling driver from the hardware using the scaling
289 interface provided by it, which may or 286 interface provided by it, which may or may not reflect the frequency
290 the CPU is actually running at (due to 287 the CPU is actually running at (due to hardware design and other
291 limitations). 288 limitations).
292 289
293 Some architectures (e.g. ``x86``) may 290 Some architectures (e.g. ``x86``) may attempt to provide information
294 more precisely reflecting the current 291 more precisely reflecting the current CPU frequency through this
295 attribute, but that still may not be t 292 attribute, but that still may not be the exact current CPU frequency as
296 seen by the hardware at the moment. 293 seen by the hardware at the moment.
297 294
298 ``scaling_driver`` 295 ``scaling_driver``
299 The scaling driver currently in use. 296 The scaling driver currently in use.
300 297
301 ``scaling_governor`` 298 ``scaling_governor``
302 The scaling governor currently attache 299 The scaling governor currently attached to this policy or (if the
303 |intel_pstate| scaling driver is in us 300 |intel_pstate| scaling driver is in use) the scaling algorithm
304 provided by the driver that is current 301 provided by the driver that is currently applied to this policy.
305 302
306 This attribute is read-write and writi 303 This attribute is read-write and writing to it will cause a new scaling
307 governor to be attached to this policy 304 governor to be attached to this policy or a new scaling algorithm
308 provided by the scaling driver to be a 305 provided by the scaling driver to be applied to it (in the
309 |intel_pstate| case), as indicated by 306 |intel_pstate| case), as indicated by the string written to this
310 attribute (which must be one of the na 307 attribute (which must be one of the names listed by the
311 ``scaling_available_governors`` attrib 308 ``scaling_available_governors`` attribute described above).
312 309
313 ``scaling_max_freq`` 310 ``scaling_max_freq``
314 Maximum frequency the CPUs belonging t 311 Maximum frequency the CPUs belonging to this policy are allowed to be
315 running at (in kHz). 312 running at (in kHz).
316 313
317 This attribute is read-write and writi 314 This attribute is read-write and writing a string representing an
318 integer to it will cause a new limit t 315 integer to it will cause a new limit to be set (it must not be lower
319 than the value of the ``scaling_min_fr 316 than the value of the ``scaling_min_freq`` attribute).
320 317
321 ``scaling_min_freq`` 318 ``scaling_min_freq``
322 Minimum frequency the CPUs belonging t 319 Minimum frequency the CPUs belonging to this policy are allowed to be
323 running at (in kHz). 320 running at (in kHz).
324 321
325 This attribute is read-write and writi 322 This attribute is read-write and writing a string representing a
326 non-negative integer to it will cause 323 non-negative integer to it will cause a new limit to be set (it must not
327 be higher than the value of the ``scal 324 be higher than the value of the ``scaling_max_freq`` attribute).
328 325
329 ``scaling_setspeed`` 326 ``scaling_setspeed``
330 This attribute is functional only if t 327 This attribute is functional only if the `userspace`_ scaling governor
331 is attached to the given policy. 328 is attached to the given policy.
332 329
333 It returns the last frequency requeste 330 It returns the last frequency requested by the governor (in kHz) or can
334 be written to in order to set a new fr 331 be written to in order to set a new frequency for the policy.
335 332
336 333
337 Generic Scaling Governors 334 Generic Scaling Governors
338 ========================= 335 =========================
339 336
340 ``CPUFreq`` provides generic scaling governors 337 ``CPUFreq`` provides generic scaling governors that can be used with all
341 scaling drivers. As stated before, each of th 338 scaling drivers. As stated before, each of them implements a single, possibly
342 parametrized, performance scaling algorithm. 339 parametrized, performance scaling algorithm.
343 340
344 Scaling governors are attached to policy objec 341 Scaling governors are attached to policy objects and different policy objects
345 can be handled by different scaling governors 342 can be handled by different scaling governors at the same time (although that
346 may lead to suboptimal results in some cases). 343 may lead to suboptimal results in some cases).
347 344
348 The scaling governor for a given policy object 345 The scaling governor for a given policy object can be changed at any time with
349 the help of the ``scaling_governor`` policy at 346 the help of the ``scaling_governor`` policy attribute in ``sysfs``.
350 347
351 Some governors expose ``sysfs`` attributes to 348 Some governors expose ``sysfs`` attributes to control or fine-tune the scaling
352 algorithms implemented by them. Those attribu 349 algorithms implemented by them. Those attributes, referred to as governor
353 tunables, can be either global (system-wide) o 350 tunables, can be either global (system-wide) or per-policy, depending on the
354 scaling driver in use. If the driver requires 351 scaling driver in use. If the driver requires governor tunables to be
355 per-policy, they are located in a subdirectory 352 per-policy, they are located in a subdirectory of each policy directory.
356 Otherwise, they are located in a subdirectory 353 Otherwise, they are located in a subdirectory under
357 :file:`/sys/devices/system/cpu/cpufreq/`. In 354 :file:`/sys/devices/system/cpu/cpufreq/`. In either case the name of the
358 subdirectory containing the governor tunables 355 subdirectory containing the governor tunables is the name of the governor
359 providing them. 356 providing them.
360 357
361 ``performance`` 358 ``performance``
362 --------------- 359 ---------------
363 360
364 When attached to a policy object, this governo 361 When attached to a policy object, this governor causes the highest frequency,
365 within the ``scaling_max_freq`` policy limit, 362 within the ``scaling_max_freq`` policy limit, to be requested for that policy.
366 363
367 The request is made once at that time the gove 364 The request is made once at that time the governor for the policy is set to
368 ``performance`` and whenever the ``scaling_max 365 ``performance`` and whenever the ``scaling_max_freq`` or ``scaling_min_freq``
369 policy limits change after that. 366 policy limits change after that.
370 367
371 ``powersave`` 368 ``powersave``
372 ------------- 369 -------------
373 370
374 When attached to a policy object, this governo 371 When attached to a policy object, this governor causes the lowest frequency,
375 within the ``scaling_min_freq`` policy limit, 372 within the ``scaling_min_freq`` policy limit, to be requested for that policy.
376 373
377 The request is made once at that time the gove 374 The request is made once at that time the governor for the policy is set to
378 ``powersave`` and whenever the ``scaling_max_f 375 ``powersave`` and whenever the ``scaling_max_freq`` or ``scaling_min_freq``
379 policy limits change after that. 376 policy limits change after that.
380 377
381 ``userspace`` 378 ``userspace``
382 ------------- 379 -------------
383 380
384 This governor does not do anything by itself. 381 This governor does not do anything by itself. Instead, it allows user space
385 to set the CPU frequency for the policy it is 382 to set the CPU frequency for the policy it is attached to by writing to the
386 ``scaling_setspeed`` attribute of that policy. 383 ``scaling_setspeed`` attribute of that policy.
387 384
388 ``schedutil`` 385 ``schedutil``
389 ------------- 386 -------------
390 387
391 This governor uses CPU utilization data availa 388 This governor uses CPU utilization data available from the CPU scheduler. It
392 generally is regarded as a part of the CPU sch 389 generally is regarded as a part of the CPU scheduler, so it can access the
393 scheduler's internal data structures directly. 390 scheduler's internal data structures directly.
394 391
395 It runs entirely in scheduler context, althoug 392 It runs entirely in scheduler context, although in some cases it may need to
396 invoke the scaling driver asynchronously when 393 invoke the scaling driver asynchronously when it decides that the CPU frequency
397 should be changed for a given policy (that dep 394 should be changed for a given policy (that depends on whether or not the driver
398 is capable of changing the CPU frequency from 395 is capable of changing the CPU frequency from scheduler context).
399 396
400 The actions of this governor for a particular 397 The actions of this governor for a particular CPU depend on the scheduling class
401 invoking its utilization update callback for t 398 invoking its utilization update callback for that CPU. If it is invoked by the
402 RT or deadline scheduling classes, the governo 399 RT or deadline scheduling classes, the governor will increase the frequency to
403 the allowed maximum (that is, the ``scaling_ma 400 the allowed maximum (that is, the ``scaling_max_freq`` policy limit). In turn,
404 if it is invoked by the CFS scheduling class, 401 if it is invoked by the CFS scheduling class, the governor will use the
405 Per-Entity Load Tracking (PELT) metric for the 402 Per-Entity Load Tracking (PELT) metric for the root control group of the
406 given CPU as the CPU utilization estimate (see 403 given CPU as the CPU utilization estimate (see the *Per-entity load tracking*
407 LWN.net article [1]_ for a description of the 404 LWN.net article [1]_ for a description of the PELT mechanism). Then, the new
408 CPU frequency to apply is computed in accordan 405 CPU frequency to apply is computed in accordance with the formula
409 406
410 f = 1.25 * ``f_0`` * ``util`` / ``max` 407 f = 1.25 * ``f_0`` * ``util`` / ``max``
411 408
412 where ``util`` is the PELT number, ``max`` is 409 where ``util`` is the PELT number, ``max`` is the theoretical maximum of
413 ``util``, and ``f_0`` is either the maximum po 410 ``util``, and ``f_0`` is either the maximum possible CPU frequency for the given
414 policy (if the PELT number is frequency-invari 411 policy (if the PELT number is frequency-invariant), or the current CPU frequency
415 (otherwise). 412 (otherwise).
416 413
417 This governor also employs a mechanism allowin 414 This governor also employs a mechanism allowing it to temporarily bump up the
418 CPU frequency for tasks that have been waiting 415 CPU frequency for tasks that have been waiting on I/O most recently, called
419 "IO-wait boosting". That happens when the :c: 416 "IO-wait boosting". That happens when the :c:macro:`SCHED_CPUFREQ_IOWAIT` flag
420 is passed by the scheduler to the governor cal 417 is passed by the scheduler to the governor callback which causes the frequency
421 to go up to the allowed maximum immediately an 418 to go up to the allowed maximum immediately and then draw back to the value
422 returned by the above formula over time. 419 returned by the above formula over time.
423 420
424 This governor exposes only one tunable: 421 This governor exposes only one tunable:
425 422
426 ``rate_limit_us`` 423 ``rate_limit_us``
427 Minimum time (in microseconds) that ha 424 Minimum time (in microseconds) that has to pass between two consecutive
428 runs of governor computations (default !! 425 runs of governor computations (default: 1000 times the scaling driver's
429 transition latency or the maximum 2ms) !! 426 transition latency).
430 427
431 The purpose of this tunable is to redu 428 The purpose of this tunable is to reduce the scheduler context overhead
432 of the governor which might be excessi 429 of the governor which might be excessive without it.
433 430
434 This governor generally is regarded as a repla 431 This governor generally is regarded as a replacement for the older `ondemand`_
435 and `conservative`_ governors (described below 432 and `conservative`_ governors (described below), as it is simpler and more
436 tightly integrated with the CPU scheduler, its 433 tightly integrated with the CPU scheduler, its overhead in terms of CPU context
437 switches and similar is less significant, and 434 switches and similar is less significant, and it uses the scheduler's own CPU
438 utilization metric, so in principle its decisi 435 utilization metric, so in principle its decisions should not contradict the
439 decisions made by the other parts of the sched 436 decisions made by the other parts of the scheduler.
440 437
441 ``ondemand`` 438 ``ondemand``
442 ------------ 439 ------------
443 440
444 This governor uses CPU load as a CPU frequency 441 This governor uses CPU load as a CPU frequency selection metric.
445 442
446 In order to estimate the current CPU load, it 443 In order to estimate the current CPU load, it measures the time elapsed between
447 consecutive invocations of its worker routine 444 consecutive invocations of its worker routine and computes the fraction of that
448 time in which the given CPU was not idle. The 445 time in which the given CPU was not idle. The ratio of the non-idle (active)
449 time to the total CPU time is taken as an esti 446 time to the total CPU time is taken as an estimate of the load.
450 447
451 If this governor is attached to a policy share 448 If this governor is attached to a policy shared by multiple CPUs, the load is
452 estimated for all of them and the greatest res 449 estimated for all of them and the greatest result is taken as the load estimate
453 for the entire policy. 450 for the entire policy.
454 451
455 The worker routine of this governor has to run 452 The worker routine of this governor has to run in process context, so it is
456 invoked asynchronously (via a workqueue) and C 453 invoked asynchronously (via a workqueue) and CPU P-states are updated from
457 there if necessary. As a result, the schedule 454 there if necessary. As a result, the scheduler context overhead from this
458 governor is minimum, but it causes additional 455 governor is minimum, but it causes additional CPU context switches to happen
459 relatively often and the CPU P-state updates t 456 relatively often and the CPU P-state updates triggered by it can be relatively
460 irregular. Also, it affects its own CPU load 457 irregular. Also, it affects its own CPU load metric by running code that
461 reduces the CPU idle time (even though the CPU 458 reduces the CPU idle time (even though the CPU idle time is only reduced very
462 slightly by it). 459 slightly by it).
463 460
464 It generally selects CPU frequencies proportio 461 It generally selects CPU frequencies proportional to the estimated load, so that
465 the value of the ``cpuinfo_max_freq`` policy a 462 the value of the ``cpuinfo_max_freq`` policy attribute corresponds to the load of
466 1 (or 100%), and the value of the ``cpuinfo_mi 463 1 (or 100%), and the value of the ``cpuinfo_min_freq`` policy attribute
467 corresponds to the load of 0, unless when the 464 corresponds to the load of 0, unless when the load exceeds a (configurable)
468 speedup threshold, in which case it will go st 465 speedup threshold, in which case it will go straight for the highest frequency
469 it is allowed to use (the ``scaling_max_freq`` 466 it is allowed to use (the ``scaling_max_freq`` policy limit).
470 467
471 This governor exposes the following tunables: 468 This governor exposes the following tunables:
472 469
473 ``sampling_rate`` 470 ``sampling_rate``
474 This is how often the governor's worke 471 This is how often the governor's worker routine should run, in
475 microseconds. 472 microseconds.
476 473
477 Typically, it is set to values of the !! 474 Typically, it is set to values of the order of 10000 (10 ms). Its
478 default value is to add a 50% breathin !! 475 default value is equal to the value of ``cpuinfo_transition_latency``
479 to ``cpuinfo_transition_latency`` on e !! 476 for each policy this governor is attached to (but since the unit here
480 attached to. The minimum is typically !! 477 is greater by 1000, this means that the time represented by
481 ticks. !! 478 ``sampling_rate`` is 1000 times greater than the transition latency by
>> 479 default).
482 480
483 If this tunable is per-policy, the fol 481 If this tunable is per-policy, the following shell command sets the time
484 represented by it to be 1.5 times as h !! 482 represented by it to be 750 times as high as the transition latency::
485 (the default):: <<
486 483
487 # echo `$(($(cat cpuinfo_transition_la !! 484 # echo `$(($(cat cpuinfo_transition_latency) * 750 / 1000)) > ondemand/sampling_rate
488 485
489 ``up_threshold`` 486 ``up_threshold``
490 If the estimated CPU load is above thi 487 If the estimated CPU load is above this value (in percent), the governor
491 will set the frequency to the maximum 488 will set the frequency to the maximum value allowed for the policy.
492 Otherwise, the selected frequency will 489 Otherwise, the selected frequency will be proportional to the estimated
493 CPU load. 490 CPU load.
494 491
495 ``ignore_nice_load`` 492 ``ignore_nice_load``
496 If set to 1 (default 0), it will cause 493 If set to 1 (default 0), it will cause the CPU load estimation code to
497 treat the CPU time spent on executing 494 treat the CPU time spent on executing tasks with "nice" levels greater
498 than 0 as CPU idle time. 495 than 0 as CPU idle time.
499 496
500 This may be useful if there are tasks 497 This may be useful if there are tasks in the system that should not be
501 taken into account when deciding what 498 taken into account when deciding what frequency to run the CPUs at.
502 Then, to make that happen it is suffic 499 Then, to make that happen it is sufficient to increase the "nice" level
503 of those tasks above 0 and set this at 500 of those tasks above 0 and set this attribute to 1.
504 501
505 ``sampling_down_factor`` 502 ``sampling_down_factor``
506 Temporary multiplier, between 1 (defau 503 Temporary multiplier, between 1 (default) and 100 inclusive, to apply to
507 the ``sampling_rate`` value if the CPU 504 the ``sampling_rate`` value if the CPU load goes above ``up_threshold``.
508 505
509 This causes the next execution of the 506 This causes the next execution of the governor's worker routine (after
510 setting the frequency to the allowed m 507 setting the frequency to the allowed maximum) to be delayed, so the
511 frequency stays at the maximum level f 508 frequency stays at the maximum level for a longer time.
512 509
513 Frequency fluctuations in some bursty 510 Frequency fluctuations in some bursty workloads may be avoided this way
514 at the cost of additional energy spent 511 at the cost of additional energy spent on maintaining the maximum CPU
515 capacity. 512 capacity.
516 513
517 ``powersave_bias`` 514 ``powersave_bias``
518 Reduction factor to apply to the origi 515 Reduction factor to apply to the original frequency target of the
519 governor (including the maximum value 516 governor (including the maximum value used when the ``up_threshold``
520 value is exceeded by the estimated CPU 517 value is exceeded by the estimated CPU load) or sensitivity threshold
521 for the AMD frequency sensitivity powe 518 for the AMD frequency sensitivity powersave bias driver
522 (:file:`drivers/cpufreq/amd_freq_sensi 519 (:file:`drivers/cpufreq/amd_freq_sensitivity.c`), between 0 and 1000
523 inclusive. 520 inclusive.
524 521
525 If the AMD frequency sensitivity power 522 If the AMD frequency sensitivity powersave bias driver is not loaded,
526 the effective frequency to apply is gi 523 the effective frequency to apply is given by
527 524
528 f * (1 - ``powersave_bias`` / 525 f * (1 - ``powersave_bias`` / 1000)
529 526
530 where f is the governor's original fre 527 where f is the governor's original frequency target. The default value
531 of this attribute is 0 in that case. 528 of this attribute is 0 in that case.
532 529
533 If the AMD frequency sensitivity power 530 If the AMD frequency sensitivity powersave bias driver is loaded, the
534 value of this attribute is 400 by defa 531 value of this attribute is 400 by default and it is used in a different
535 way. 532 way.
536 533
537 On Family 16h (and later) AMD processo 534 On Family 16h (and later) AMD processors there is a mechanism to get a
538 measured workload sensitivity, between 535 measured workload sensitivity, between 0 and 100% inclusive, from the
539 hardware. That value can be used to e 536 hardware. That value can be used to estimate how the performance of the
540 workload running on a CPU will change 537 workload running on a CPU will change in response to frequency changes.
541 538
542 The performance of a workload with the 539 The performance of a workload with the sensitivity of 0 (memory-bound or
543 IO-bound) is not expected to increase 540 IO-bound) is not expected to increase at all as a result of increasing
544 the CPU frequency, whereas workloads w 541 the CPU frequency, whereas workloads with the sensitivity of 100%
545 (CPU-bound) are expected to perform mu 542 (CPU-bound) are expected to perform much better if the CPU frequency is
546 increased. 543 increased.
547 544
548 If the workload sensitivity is less th 545 If the workload sensitivity is less than the threshold represented by
549 the ``powersave_bias`` value, the sens 546 the ``powersave_bias`` value, the sensitivity powersave bias driver
550 will cause the governor to select a fr 547 will cause the governor to select a frequency lower than its original
551 target, so as to avoid over-provisioni 548 target, so as to avoid over-provisioning workloads that will not benefit
552 from running at higher CPU frequencies 549 from running at higher CPU frequencies.
553 550
554 ``conservative`` 551 ``conservative``
555 ---------------- 552 ----------------
556 553
557 This governor uses CPU load as a CPU frequency 554 This governor uses CPU load as a CPU frequency selection metric.
558 555
559 It estimates the CPU load in the same way as t 556 It estimates the CPU load in the same way as the `ondemand`_ governor described
560 above, but the CPU frequency selection algorit 557 above, but the CPU frequency selection algorithm implemented by it is different.
561 558
562 Namely, it avoids changing the frequency signi 559 Namely, it avoids changing the frequency significantly over short time intervals
563 which may not be suitable for systems with lim 560 which may not be suitable for systems with limited power supply capacity (e.g.
564 battery-powered). To achieve that, it changes 561 battery-powered). To achieve that, it changes the frequency in relatively
565 small steps, one step at a time, up or down - 562 small steps, one step at a time, up or down - depending on whether or not a
566 (configurable) threshold has been exceeded by 563 (configurable) threshold has been exceeded by the estimated CPU load.
567 564
568 This governor exposes the following tunables: 565 This governor exposes the following tunables:
569 566
570 ``freq_step`` 567 ``freq_step``
571 Frequency step in percent of the maxim 568 Frequency step in percent of the maximum frequency the governor is
572 allowed to set (the ``scaling_max_freq 569 allowed to set (the ``scaling_max_freq`` policy limit), between 0 and
573 100 (5 by default). 570 100 (5 by default).
574 571
575 This is how much the frequency is allo 572 This is how much the frequency is allowed to change in one go. Setting
576 it to 0 will cause the default frequen 573 it to 0 will cause the default frequency step (5 percent) to be used
577 and setting it to 100 effectively caus 574 and setting it to 100 effectively causes the governor to periodically
578 switch the frequency between the ``sca 575 switch the frequency between the ``scaling_min_freq`` and
579 ``scaling_max_freq`` policy limits. 576 ``scaling_max_freq`` policy limits.
580 577
581 ``down_threshold`` 578 ``down_threshold``
582 Threshold value (in percent, 20 by def 579 Threshold value (in percent, 20 by default) used to determine the
583 frequency change direction. 580 frequency change direction.
584 581
585 If the estimated CPU load is greater t 582 If the estimated CPU load is greater than this value, the frequency will
586 go up (by ``freq_step``). If the load 583 go up (by ``freq_step``). If the load is less than this value (and the
587 ``sampling_down_factor`` mechanism is 584 ``sampling_down_factor`` mechanism is not in effect), the frequency will
588 go down. Otherwise, the frequency wil 585 go down. Otherwise, the frequency will not be changed.
589 586
590 ``sampling_down_factor`` 587 ``sampling_down_factor``
591 Frequency decrease deferral factor, be 588 Frequency decrease deferral factor, between 1 (default) and 10
592 inclusive. 589 inclusive.
593 590
594 It effectively causes the frequency to 591 It effectively causes the frequency to go down ``sampling_down_factor``
595 times slower than it ramps up. 592 times slower than it ramps up.
596 593
597 594
598 Frequency Boost Support 595 Frequency Boost Support
599 ======================= 596 =======================
600 597
601 Background 598 Background
602 ---------- 599 ----------
603 600
604 Some processors support a mechanism to raise t 601 Some processors support a mechanism to raise the operating frequency of some
605 cores in a multicore package temporarily (and 602 cores in a multicore package temporarily (and above the sustainable frequency
606 threshold for the whole package) under certain 603 threshold for the whole package) under certain conditions, for example if the
607 whole chip is not fully utilized and below its 604 whole chip is not fully utilized and below its intended thermal or power budget.
608 605
609 Different names are used by different vendors 606 Different names are used by different vendors to refer to this functionality.
610 For Intel processors it is referred to as "Tur 607 For Intel processors it is referred to as "Turbo Boost", AMD calls it
611 "Turbo-Core" or (in technical documentation) " 608 "Turbo-Core" or (in technical documentation) "Core Performance Boost" and so on.
612 As a rule, it also is implemented differently 609 As a rule, it also is implemented differently by different vendors. The simple
613 term "frequency boost" is used here for brevit 610 term "frequency boost" is used here for brevity to refer to all of those
614 implementations. 611 implementations.
615 612
616 The frequency boost mechanism may be either ha 613 The frequency boost mechanism may be either hardware-based or software-based.
617 If it is hardware-based (e.g. on x86), the dec 614 If it is hardware-based (e.g. on x86), the decision to trigger the boosting is
618 made by the hardware (although in general it r 615 made by the hardware (although in general it requires the hardware to be put
619 into a special state in which it can control t 616 into a special state in which it can control the CPU frequency within certain
620 limits). If it is software-based (e.g. on ARM 617 limits). If it is software-based (e.g. on ARM), the scaling driver decides
621 whether or not to trigger boosting and when to 618 whether or not to trigger boosting and when to do that.
622 619
623 The ``boost`` File in ``sysfs`` 620 The ``boost`` File in ``sysfs``
624 ------------------------------- 621 -------------------------------
625 622
626 This file is located under :file:`/sys/devices 623 This file is located under :file:`/sys/devices/system/cpu/cpufreq/` and controls
627 the "boost" setting for the whole system. It 624 the "boost" setting for the whole system. It is not present if the underlying
628 scaling driver does not support the frequency 625 scaling driver does not support the frequency boost mechanism (or supports it,
629 but provides a driver-specific interface for c 626 but provides a driver-specific interface for controlling it, like
630 |intel_pstate|). 627 |intel_pstate|).
631 628
632 If the value in this file is 1, the frequency 629 If the value in this file is 1, the frequency boost mechanism is enabled. This
633 means that either the hardware can be put into 630 means that either the hardware can be put into states in which it is able to
634 trigger boosting (in the hardware-based case), 631 trigger boosting (in the hardware-based case), or the software is allowed to
635 trigger boosting (in the software-based case). 632 trigger boosting (in the software-based case). It does not mean that boosting
636 is actually in use at the moment on any CPUs i 633 is actually in use at the moment on any CPUs in the system. It only means a
637 permission to use the frequency boost mechanis 634 permission to use the frequency boost mechanism (which still may never be used
638 for other reasons). 635 for other reasons).
639 636
640 If the value in this file is 0, the frequency 637 If the value in this file is 0, the frequency boost mechanism is disabled and
641 cannot be used at all. 638 cannot be used at all.
642 639
643 The only values that can be written to this fi 640 The only values that can be written to this file are 0 and 1.
644 641
645 Rationale for Boost Control Knob 642 Rationale for Boost Control Knob
646 -------------------------------- 643 --------------------------------
647 644
648 The frequency boost mechanism is generally int 645 The frequency boost mechanism is generally intended to help to achieve optimum
649 CPU performance on time scales below software 646 CPU performance on time scales below software resolution (e.g. below the
650 scheduler tick interval) and it is demonstrabl 647 scheduler tick interval) and it is demonstrably suitable for many workloads, but
651 it may lead to problems in certain situations. 648 it may lead to problems in certain situations.
652 649
653 For this reason, many systems make it possible 650 For this reason, many systems make it possible to disable the frequency boost
654 mechanism in the platform firmware (BIOS) setu 651 mechanism in the platform firmware (BIOS) setup, but that requires the system to
655 be restarted for the setting to be adjusted as 652 be restarted for the setting to be adjusted as desired, which may not be
656 practical at least in some cases. For example 653 practical at least in some cases. For example:
657 654
658 1. Boosting means overclocking the processor 655 1. Boosting means overclocking the processor, although under controlled
659 conditions. Generally, the processor's e 656 conditions. Generally, the processor's energy consumption increases
660 as a result of increasing its frequency a 657 as a result of increasing its frequency and voltage, even temporarily.
661 That may not be desirable on systems that 658 That may not be desirable on systems that switch to power sources of
662 limited capacity, such as batteries, so t 659 limited capacity, such as batteries, so the ability to disable the boost
663 mechanism while the system is running may 660 mechanism while the system is running may help there (but that depends on
664 the workload too). 661 the workload too).
665 662
666 2. In some situations deterministic behavior 663 2. In some situations deterministic behavior is more important than
667 performance or energy consumption (or bot 664 performance or energy consumption (or both) and the ability to disable
668 boosting while the system is running may 665 boosting while the system is running may be useful then.
669 666
670 3. To examine the impact of the frequency bo 667 3. To examine the impact of the frequency boost mechanism itself, it is useful
671 to be able to run tests with and without 668 to be able to run tests with and without boosting, preferably without
672 restarting the system in the meantime. 669 restarting the system in the meantime.
673 670
674 4. Reproducible results are important when r 671 4. Reproducible results are important when running benchmarks. Since
675 the boosting functionality depends on the 672 the boosting functionality depends on the load of the whole package,
676 single-thread performance may vary becaus 673 single-thread performance may vary because of it which may lead to
677 unreproducible results sometimes. That c 674 unreproducible results sometimes. That can be avoided by disabling the
678 frequency boost mechanism before running 675 frequency boost mechanism before running benchmarks sensitive to that
679 issue. 676 issue.
680 677
681 Legacy AMD ``cpb`` Knob 678 Legacy AMD ``cpb`` Knob
682 ----------------------- 679 -----------------------
683 680
684 The AMD powernow-k8 scaling driver supports a 681 The AMD powernow-k8 scaling driver supports a ``sysfs`` knob very similar to
685 the global ``boost`` one. It is used for disa 682 the global ``boost`` one. It is used for disabling/enabling the "Core
686 Performance Boost" feature of some AMD process 683 Performance Boost" feature of some AMD processors.
687 684
688 If present, that knob is located in every ``CP 685 If present, that knob is located in every ``CPUFreq`` policy directory in
689 ``sysfs`` (:file:`/sys/devices/system/cpu/cpuf 686 ``sysfs`` (:file:`/sys/devices/system/cpu/cpufreq/policyX/`) and is called
690 ``cpb``, which indicates a more fine grained c 687 ``cpb``, which indicates a more fine grained control interface. The actual
691 implementation, however, works on the system-w 688 implementation, however, works on the system-wide basis and setting that knob
692 for one policy causes the same value of it to 689 for one policy causes the same value of it to be set for all of the other
693 policies at the same time. 690 policies at the same time.
694 691
695 That knob is still supported on AMD processors 692 That knob is still supported on AMD processors that support its underlying
696 hardware feature, but it may be configured out 693 hardware feature, but it may be configured out of the kernel (via the
697 :c:macro:`CONFIG_X86_ACPI_CPUFREQ_CPB` configu 694 :c:macro:`CONFIG_X86_ACPI_CPUFREQ_CPB` configuration option) and the global
698 ``boost`` knob is present regardless. Thus it 695 ``boost`` knob is present regardless. Thus it is always possible use the
699 ``boost`` knob instead of the ``cpb`` one whic 696 ``boost`` knob instead of the ``cpb`` one which is highly recommended, as that
700 is more consistent with what all of the other 697 is more consistent with what all of the other systems do (and the ``cpb`` knob
701 may not be supported any more in the future). 698 may not be supported any more in the future).
702 699
703 The ``cpb`` knob is never present for any proc 700 The ``cpb`` knob is never present for any processors without the underlying
704 hardware feature (e.g. all Intel ones), even i 701 hardware feature (e.g. all Intel ones), even if the
705 :c:macro:`CONFIG_X86_ACPI_CPUFREQ_CPB` configu 702 :c:macro:`CONFIG_X86_ACPI_CPUFREQ_CPB` configuration option is set.
706 703
707 704
708 References 705 References
709 ========== 706 ==========
710 707
711 .. [1] Jonathan Corbet, *Per-entity load track 708 .. [1] Jonathan Corbet, *Per-entity load tracking*,
712 https://lwn.net/Articles/531853/ 709 https://lwn.net/Articles/531853/
Linux® is a registered trademark of Linus Torvalds in the United States and other countries.
TOMOYO® is a registered trademark of NTT DATA CORPORATION.