1 .. SPDX-License-Identifier: GPL-2.0
2 .. include:: <isonum.txt>
3
4 ==============================================
5 ``intel_pstate`` CPU Performance Scaling Drive
6 ==============================================
7
8 :Copyright: |copy| 2017 Intel Corporation
9
10 :Author: Rafael J. Wysocki <rafael.j.wysocki@in
11
12
13 General Information
14 ===================
15
16 ``intel_pstate`` is a part of the
17 :doc:`CPU performance scaling subsystem <cpufr
18 (``CPUFreq``). It is a scaling driver for the
19 generations of Intel processors. Note, howeve
20 may not be supported. [To understand ``intel_
21 how ``CPUFreq`` works in general, so this is t
22 Documentation/admin-guide/pm/cpufreq.rst if yo
23
24 For the processors supported by ``intel_pstate
25 than just an operating frequency or an operati
26 LinuxCon Europe 2015 presentation by Kristen A
27 information about that). For this reason, the
28 by ``intel_pstate`` internally follows the har
29 refer to Intel Software Developer’s Manual [
30 uses frequencies for identifying operating per
31 frequencies are involved in the user space int
32 ``intel_pstate`` maps its internal representat
33 (fortunately, that mapping is unambiguous). A
34 practical for ``intel_pstate`` to supply the `
35 available frequencies due to the possible size
36 that. Some functionality of the core is limit
37
38 Since the hardware P-state selection interface
39 available at the logical CPU level, the driver
40 CPUs. Consequently, if ``intel_pstate`` is in
41 object corresponds to one logical CPU and ``CP
42 equivalent to CPUs. In particular, this means
43 time the corresponding CPU is taken offline an
44 it goes back online.
45
46 ``intel_pstate`` is not modular, so it cannot
47 only way to pass early-configuration-time para
48 command line. However, its configuration can
49 great extent. In some configurations it even
50 ``sysfs`` which allows another ``CPUFreq`` sca
51 registered (see `below <status_attr_>`_).
52
53
54 Operation Modes
55 ===============
56
57 ``intel_pstate`` can operate in two different
58 active mode, it uses its own internal performa
59 allows the hardware to do performance scaling
60 mode it responds to requests made by a generic
61 a certain performance scaling algorithm. Whic
62 depends on what kernel command line options ar
63 the processor.
64
65 Active Mode
66 -----------
67
68 This is the default operation mode of ``intel_
69 hardware-managed P-states (HWP) support. If i
70 ``scaling_driver`` policy attribute in ``sysfs
71 contains the string "intel_pstate".
72
73 In this mode the driver bypasses the scaling g
74 provides its own scaling algorithms for P-stat
75 can be applied to ``CPUFreq`` policies in the
76 governors (that is, through the ``scaling_gove
77 ``sysfs``). [Note that different P-state sele
78 different policies, but that is not recommende
79
80 They are not generic scaling governors, but th
81 names of some of those governors. Moreover, c
82 do not work in the same way as the generic gov
83 For example, the ``powersave`` P-state selecti
84 ``intel_pstate`` is not a counterpart of the g
85 (roughly, it corresponds to the ``schedutil``
86
87 There are two P-state selection algorithms pro
88 active mode: ``powersave`` and ``performance``
89 depends on whether or not the hardware-managed
90 enabled in the processor and possibly on the p
91
92 Which of the P-state selection algorithms is u
93 :c:macro:`CONFIG_CPU_FREQ_DEFAULT_GOV_PERFORMA
94 Namely, if that option is set, the ``performan
95 default, and the other one will be used by def
96
97 Active Mode With HWP
98 ~~~~~~~~~~~~~~~~~~~~
99
100 If the processor supports the HWP feature, it
101 processor initialization and cannot be disable
102 to avoid enabling it by passing the ``intel_ps
103 kernel in the command line.
104
105 If the HWP feature has been enabled, ``intel_p
106 select P-states by itself, but still it can gi
107 internal P-state selection logic. What those
108 selection algorithm has been applied to the gi
109 corresponds to).
110
111 Even though the P-state selection is carried o
112 ``intel_pstate`` registers utilization update
113 in this mode. However, they are not used for
114 algorithm, but for periodic updates of the cur
115 be made available from the ``scaling_cur_freq`
116
117 HWP + ``performance``
118 .....................
119
120 In this configuration ``intel_pstate`` will wr
121 Energy-Performance Preference (EPP) knob (if s
122 Energy-Performance Bias (EPB) knob (otherwise)
123 internal P-state selection logic is expected t
124
125 This will override the EPP/EPB setting coming
126 (see `Energy vs Performance Hints`_ below). M
127 the EPP/EPB to a value different from 0 ("perf
128 configuration will be rejected.
129
130 Also, in this configuration the range of P-sta
131 internal P-state selection logic is always res
132 (that is, the maximum P-state that the driver
133
134 HWP + ``powersave``
135 ...................
136
137 In this configuration ``intel_pstate`` will se
138 Energy-Performance Preference (EPP) knob (if s
139 Energy-Performance Bias (EPB) knob (otherwise)
140 previously set to via ``sysfs`` (or whatever d
141 set to by the platform firmware). This usuall
142 internal P-state selection logic to be less pe
143
144 Active Mode Without HWP
145 ~~~~~~~~~~~~~~~~~~~~~~~
146
147 This operation mode is optional for processors
148 feature or when the ``intel_pstate=no_hwp`` ar
149 the command line. The active mode is used in
150 ``intel_pstate=active`` argument is passed to
151 In this mode ``intel_pstate`` may refuse to wo
152 recognized by it. [Note that ``intel_pstate``
153 any processor with the HWP feature enabled.]
154
155 In this mode ``intel_pstate`` registers utiliz
156 CPU scheduler in order to run a P-state select
157 ``powersave`` or ``performance``, depending on
158 setting in ``sysfs``. The current CPU frequen
159 available from the ``scaling_cur_freq`` policy
160 periodically updated by those utilization upda
161
162 ``performance``
163 ...............
164
165 Without HWP, this P-state selection algorithm
166 the processor model and platform configuration
167
168 It selects the maximum P-state it is allowed t
169 ``sysfs``, every time the driver configuration
170 (e.g. via ``sysfs``).
171
172 This is the default P-state selection algorith
173 :c:macro:`CONFIG_CPU_FREQ_DEFAULT_GOV_PERFORMA
174 is set.
175
176 ``powersave``
177 .............
178
179 Without HWP, this P-state selection algorithm
180 implemented by the generic ``schedutil`` scali
181 utilization metric used by it is based on numb
182 registers of the CPU. It generally selects P-
183 current CPU utilization.
184
185 This algorithm is run by the driver's utilizat
186 given CPU when it is invoked by the CPU schedu
187 every 10 ms. Like in the ``performance`` case
188 is not touched if the new P-state turns out to
189 one.
190
191 This is the default P-state selection algorith
192 :c:macro:`CONFIG_CPU_FREQ_DEFAULT_GOV_PERFORMA
193 is not set.
194
195 Passive Mode
196 ------------
197
198 This is the default operation mode of ``intel_
199 hardware-managed P-states (HWP) support. It i
200 ``intel_pstate=passive`` argument is passed to
201 regardless of whether or not the given process
202 ``intel_pstate=no_hwp`` setting causes the dri
203 if it is not combined with ``intel_pstate=acti
204 without HWP support, in this mode ``intel_psta
205 processors that are not recognized by it if HW
206 through the kernel command line.
207
208 If the driver works in this mode, the ``scalin
209 ``sysfs`` for all ``CPUFreq`` policies contain
210 Then, the driver behaves like a regular ``CPUF
211 it is invoked by generic scaling governors whe
212 hardware in order to change the P-state of a C
213 ``schedutil`` governor can invoke it directly
214
215 While in this mode, ``intel_pstate`` can be us
216 scaling governors listed by the ``scaling_avai
217 in ``sysfs`` (and the P-state selection algori
218 used). Then, it is responsible for the config
219 corresponding to CPUs and provides the ``CPUFr
220 governors attached to the policy objects) with
221 maximum and minimum operating frequencies supp
222 the so-called "turbo" frequency ranges). In o
223 the entire range of available P-states is expo
224 ``CPUFreq`` core. However, in this mode the d
225 utilization update callbacks with the CPU sche
226 information comes from the ``CPUFreq`` core (a
227 by the current scaling governor for the given
228
229
230 .. _turbo:
231
232 Turbo P-states Support
233 ======================
234
235 In the majority of cases, the entire range of
236 ``intel_pstate`` can be divided into two sub-r
237 different types of processor behavior, above a
238 will be referred to as the "turbo threshold" i
239
240 The P-states above the turbo threshold are ref
241 the whole sub-range of P-states they belong to
242 range". These names are related to the Turbo
243 multicore processor to opportunistically incre
244 cores if there is enough power to do that and
245 thermal envelope of the processor package to b
246
247 Specifically, if software sets the P-state of
248 (that is, above the turbo threshold), the proc
249 performance scaling control for that core and
250 choice going forward. However, that permissio
251 different processor generations. Namely, the
252 processors will never use any P-states above t
253 the given core, even if it is within the turbo
254 processor generations will take it as a licens
255 turbo range, even above the one set by softwar
256 processors setting any P-state from the turbo
257 to put the given core into all turbo P-states
258 supported one as it sees fit.
259
260 One important property of turbo P-states is th
261 precisely, there is no guarantee that any CPUs
262 those states indefinitely, because the power d
263 package may change over time or the thermal e
264 be exceeded if a turbo P-state was used for to
265
266 In turn, the P-states below the turbo threshol
267 fact, if one of them is set by software, the p
268 it to a lower one unless in a thermal stress o
269 situation (a higher P-state may still be used
270 the same package at the same time, for example
271
272 Some processors allow multiple cores to be in
273 but the maximum P-state that can be set for th
274 of cores running concurrently. The maximum tu
275 cores at the same time usually is lower than t
276 2 cores, which in turn usually is lower than t
277 be set for 1 core. The one-core maximum turbo
278 supported one overall.
279
280 The maximum supported turbo P-state, the turbo
281 non-turbo P-state) and the minimum supported P
282 processor model and can be determined by readi
283 registers (MSRs). Moreover, some processors s
284 (Thermal Design Power) feature and, when that
285 threshold effectively becomes a configurable v
286 platform firmware.
287
288 Unlike ``_PSS`` objects in the ACPI tables, ``
289 the entire range of available P-states, includ
290 ``CPUFreq`` core and (in the passive mode) to
291 generally causes turbo P-states to be set more
292 used relative to ACPI-based CPU performance sc
293 for more information).
294
295 Moreover, since ``intel_pstate`` always knows
296 (even if the Configurable TDP feature is enabl
297 ``no_turbo`` attribute in ``sysfs`` (described
298 work as expected in all cases (that is, if set
299 always should prevent ``intel_pstate`` from us
300
301
302 Processor Support
303 =================
304
305 To handle a given processor ``intel_pstate`` r
306 pieces of information on it to be known, inclu
307
308 * The minimum supported P-state.
309
310 * The maximum supported `non-turbo P-state <t
311
312 * Whether or not turbo P-states are supported
313
314 * The maximum supported `one-core turbo P-sta
315 are supported).
316
317 * The scaling formula to translate the driver
318 of P-states into frequencies and the other
319
320 Generally, ways to obtain that information are
321 or family. Although it often is possible to o
322 itself (using model-specific registers), there
323 manuals need to be consulted to get to it too.
324
325 For this reason, there is a list of supported
326 the driver initialization will fail if the det
327 list, unless it supports the HWP feature. [Th
328 information listed above is the same for all o
329 HWP feature, which is why ``intel_pstate`` wor
330
331
332 User Space Interface in ``sysfs``
333 =================================
334
335 Global Attributes
336 -----------------
337
338 ``intel_pstate`` exposes several global attrib
339 control its functionality at the system level.
340 ``/sys/devices/system/cpu/intel_pstate/`` dire
341
342 Some of them are not present if the ``intel_ps
343 argument is passed to the kernel in the comman
344
345 ``max_perf_pct``
346 Maximum P-state the driver is allowed
347 maximum supported performance level (t
348 P-state <turbo_>`_).
349
350 This attribute will not be exposed if
351 ``intel_pstate=per_cpu_perf_limits`` a
352 command line.
353
354 ``min_perf_pct``
355 Minimum P-state the driver is allowed
356 maximum supported performance level (t
357 P-state <turbo_>`_).
358
359 This attribute will not be exposed if
360 ``intel_pstate=per_cpu_perf_limits`` a
361 command line.
362
363 ``num_pstates``
364 Number of P-states supported by the pr
365 inclusive) including both turbo and no
366 `Turbo P-states Support`_).
367
368 This attribute is present only if the
369 for all of the CPUs in the system.
370
371 The value of this attribute is not aff
372 setting described `below <no_turbo_att
373
374 This attribute is read-only.
375
376 ``turbo_pct``
377 Ratio of the `turbo range <turbo_>`_ s
378 range of supported P-states, in percen
379
380 This attribute is present only if the
381 for all of the CPUs in the system.
382
383 This attribute is read-only.
384
385 .. _no_turbo_attr:
386
387 ``no_turbo``
388 If set (equal to 1), the driver is not
389 (see `Turbo P-states Support`_). If u
390 default), turbo P-states can be set by
391 [Note that ``intel_pstate`` does not s
392 attribute (supported by some other sca
393 by this one.]
394
395 This attribute does not affect the max
396 supplied to the ``CPUFreq`` core and e
397 but it affects the maximum possible va
398 (see `Interpretation of Policy Attribu
399
400 ``hwp_dynamic_boost``
401 This attribute is only present if ``in
402 `active mode with the HWP feature enab
403 the processor. If set (equal to 1), i
404 to be increased dynamically for a shor
405 waiting on I/O is selected to run on a
406 of this mechanism is to improve perfor
407
408 This setting has no effect on logical
409 is directly set to the highest non-tur
410
411 .. _status_attr:
412
413 ``status``
414 Operation mode of the driver: "active"
415
416 "active"
417 The driver is functional and i
418 <Active Mode_>`_.
419
420 "passive"
421 The driver is functional and i
422 <Passive Mode_>`_.
423
424 "off"
425 The driver is not functional (
426 driver with the ``CPUFreq`` co
427
428 This attribute can be written to in or
429 operation mode or to unregister it. T
430 one of the possible values of it and,
431 cause the driver to switch over to the
432 that string - or to be unregistered in
433 switching over from the active mode to
434 way around causes the driver to be unr
435 with a different set of callbacks, so
436 as well as the per-policy ones) are th
437 values, possibly depending on the targ
438
439 ``energy_efficiency``
440 This attribute is only present on plat
441 Lake or Coffee Lake desktop CPU model.
442 optimizations are disabled on these CP
443 Enabling energy-efficiency optimizatio
444 frequency with or without the HWP feat
445 optimizations are done only in the tur
446 they are done in the entire available
447 attribute to "1" enables the energy-ef
448 to "0" disables them.
449
450 Interpretation of Policy Attributes
451 -----------------------------------
452
453 The interpretation of some ``CPUFreq`` policy
454 Documentation/admin-guide/pm/cpufreq.rst is sp
455 as the current scaling driver and it generally
456 `operation mode <Operation Modes_>`_.
457
458 First of all, the values of the ``cpuinfo_max_
459 ``scaling_cur_freq`` attributes are produced b
460 multiplier to the internal P-state representat
461 Also, the values of the ``scaling_max_freq`` a
462 attributes are capped by the frequency corresp
463 the driver is allowed to set.
464
465 If the ``no_turbo`` `global attribute <no_turb
466 not allowed to use turbo P-states, so the maxi
467 and ``scaling_min_freq`` is limited to the max
468 Accordingly, setting ``no_turbo`` causes ``sca
469 ``scaling_min_freq`` to go down to that value
470 However, the old values of ``scaling_max_freq`
471 restored after unsetting ``no_turbo``, unless
472 to after ``no_turbo`` was set.
473
474 If ``no_turbo`` is not set, the maximum possib
475 and ``scaling_min_freq`` corresponds to the ma
476 which also is the value of ``cpuinfo_max_freq`
477
478 Next, the following policy attributes have spe
479 ``intel_pstate`` works in the `active mode <Ac
480
481 ``scaling_available_governors``
482 List of P-state selection algorithms p
483
484 ``scaling_governor``
485 P-state selection algorithm provided b
486 use with the given policy.
487
488 ``scaling_cur_freq``
489 Frequency of the average P-state of th
490 policy for the time interval between t
491 driver's utilization update callback b
492
493 One more policy attribute is present if the HW
494 processor:
495
496 ``base_frequency``
497 Shows the base frequency of the CPU. A
498 in the turbo frequency range.
499
500 The meaning of these attributes in the `passiv
501 same as for other scaling drivers.
502
503 Additionally, the value of the ``scaling_drive
504 depends on the operation mode of the driver.
505 "intel_pstate" (in the `active mode <Active Mo
506 `passive mode <Passive Mode_>`_).
507
508 Coordination of P-State Limits
509 ------------------------------
510
511 ``intel_pstate`` allows P-state limits to be s
512 the ``max_perf_pct`` and ``min_perf_pct`` `glo
513 <Global Attributes_>`_ or via the ``scaling_ma
514 ``CPUFreq`` policy attributes. The coordinati
515 on the following rules, regardless of the curr
516
517 1. All CPUs are affected by the global limits
518 requested to run faster than the global ma
519 requested to run slower than the global mi
520
521 2. Each individual CPU is affected by its own
522 cannot be requested to run faster than its
523 cannot be requested to run slower than its
524 effective performance depends on whether t
525 P-states, hyper-threading is enabled and o
526 from other CPUs. When platform doesn't sup
527 effective performance can be more than the
528 other CPUs are requesting higher performan
529 core P-states support, when hyper-threadin
530 is requesting higher performance, the othe
531 performance than their policy limits.
532
533 3. The global and per-policy limits can be se
534
535 In the `active mode with the HWP feature enabl
536 resulting effective values are written into ha
537 limits change in order to request its internal
538 set P-states within these limits. Otherwise,
539 by scaling governors (in the `passive mode <Pa
540 every time before setting a new P-state for a
541
542 Additionally, if the ``intel_pstate=per_cpu_pe
543 is passed to the kernel, ``max_perf_pct`` and
544 at all and the only way to set the limits is b
545
546
547 Energy vs Performance Hints
548 ---------------------------
549
550 If the hardware-managed P-states (HWP) is enab
551 attributes, intended to allow user space to he
552 processor's internal P-state selection logic b
553 energy-efficiency, or somewhere between the tw
554 ``CPUFreq`` policy directory in ``sysfs``. Th
555
556 ``energy_performance_preference``
557 Current value of the energy vs perform
558 (or the CPU represented by it).
559
560 The hint can be changed by writing to
561
562 ``energy_performance_available_preferences``
563 List of strings that can be written to
564 ``energy_performance_preference`` attr
565
566 They represent different energy vs per
567 self-explanatory, except that ``defaul
568 value was set by the platform firmware
569
570 Strings written to the ``energy_performance_pr
571 internally translated to integer values writte
572 Energy-Performance Preference (EPP) knob (if s
573 Energy-Performance Bias (EPB) knob. It is also
574 integer value between 0 to 255, if the EPP fea
575 feature is not present, writing integer value
576 supported. In this case, user can use the
577 "/sys/devices/system/cpu/cpu*/power/energy_per
578
579 [Note that tasks may by migrated from one CPU
580 load-balancing algorithm and if different ener
581 set for those CPUs, that may lead to undesirab
582 issues it is better to set the same energy vs
583 or to pin every task potentially sensitive to
584
585 .. _acpi-cpufreq:
586
587 ``intel_pstate`` vs ``acpi-cpufreq``
588 ====================================
589
590 On the majority of systems supported by ``inte
591 provided by the platform firmware contain ``_P
592 that can be used for CPU performance scaling (
593 [3]_ for details on the ``_PSS`` objects and t
594 returned by them).
595
596 The information returned by the ACPI ``_PSS``
597 ``acpi-cpufreq`` scaling driver. On systems s
598 the ``acpi-cpufreq`` driver uses the same hard
599 interface, but the set of P-states it can use
600 output.
601
602 On those systems each ``_PSS`` object returns
603 the corresponding CPU which basically is a sub
604 be used by ``intel_pstate`` on the same system
605 `turbo range <turbo_>`_ is represented by one
606 convention, the frequency returned by ``_PSS``
607 than the frequency of the highest non-turbo P-
608 corresponding P-state representation (followin
609 returned for it matches the maximum supported
610 special value 255 meaning essentially "go as h
611
612 The list of P-states returned by ``_PSS`` is r
613 available frequencies supplied by ``acpi-cpufr
614 scaling governors and the minimum and maximum
615 it come from that list as well. In particular
616 of the turbo range described above, this means
617 frequency reported by ``acpi-cpufreq`` is high
618 of the highest supported non-turbo P-state lis
619 affects decisions made by the scaling governor
620 ``performance``.
621
622 For example, if a given governor attempts to s
623 estimated CPU load and maps the load of 100% t
624 (possibly multiplied by a constant), then it w
625 the turbo threshold if ``acpi-cpufreq`` is use
626 in that case the turbo range corresponds to a
627 band it can use (1 MHz vs 1 GHz or more). In
628 the turbo range for the highest loads and the
629 benefit from running at turbo frequencies will
630 instead.
631
632 One more issue related to that may appear on s
633 `Configurable TDP feature <turbo_>`_ allowing
634 turbo threshold. Namely, if that is not coord
635 returned by ``_PSS`` properly, there may be mo
636 a turbo P-state in those lists and there may b
637 turbo range (if desirable or necessary). Usua
638 P-states overall, ``acpi-cpufreq`` simply avoi
639 by ``_PSS``, but that is not sufficient when t
640 the list returned by it.
641
642 Apart from the above, ``acpi-cpufreq`` works l
643 `passive mode <Passive Mode_>`_, except that t
644 is limited to the ones listed by the ACPI ``_P
645
646
647 Kernel Command Line Options for ``intel_pstate
648 ==============================================
649
650 Several kernel command line options can be use
651 parameters to ``intel_pstate`` in order to enf
652 of them have to be prepended with the ``intel_
653
654 ``disable``
655 Do not register ``intel_pstate`` as th
656 processor is supported by it.
657
658 ``active``
659 Register ``intel_pstate`` in the `acti
660 with.
661
662 ``passive``
663 Register ``intel_pstate`` in the `pass
664 start with.
665
666 ``force``
667 Register ``intel_pstate`` as the scali
668 ``acpi-cpufreq`` even if the latter is
669
670 This may prevent some platform feature
671 power capping) that rely on the availa
672 information from functioning as expect
673 caution.
674
675 This option does not work with process
676 ``intel_pstate`` and on platforms wher
677 driver is used instead of ``acpi-cpufr
678
679 ``no_hwp``
680 Do not enable the hardware-managed P-s
681 supported by the processor.
682
683 ``hwp_only``
684 Register ``intel_pstate`` as the scali
685 hardware-managed P-states (HWP) featur
686
687 ``support_acpi_ppc``
688 Take ACPI ``_PPC`` performance limits
689
690 If the preferred power management prof
691 Description Table) is set to "Enterpri
692 Server", the ACPI ``_PPC`` limits are
693 and this option has no effect.
694
695 ``per_cpu_perf_limits``
696 Use per-logical-CPU P-State limits (se
697 Limits`_ for details).
698
699
700 Diagnostics and Tuning
701 ======================
702
703 Trace Events
704 ------------
705
706 There are two static trace events that can be
707 diagnostics. One of them is the ``cpu_frequen
708 by ``CPUFreq``, and the other one is the ``pst
709 to ``intel_pstate``. Both of them are trigger
710 it works in the `active mode <Active Mode_>`_.
711
712 The following sequence of shell commands can b
713 their output (if the kernel is generally confi
714
715 # cd /sys/kernel/tracing/
716 # echo 1 > events/power/pstate_sample/enable
717 # echo 1 > events/power/cpu_frequency/enable
718 # cat trace
719 gnome-terminal--4510 [001] ..s. 1177.680733
720 cat-5235 [002] ..s. 1177.681723: cpu_freque
721
722 If ``intel_pstate`` works in the `passive mode
723 ``cpu_frequency`` trace event will be triggere
724 scaling governor (for the policies it is attac
725 core (for the policies with other scaling gove
726
727 ``ftrace``
728 ----------
729
730 The ``ftrace`` interface can be used for low-l
731 ``intel_pstate``. For example, to check how o
732 P-state is called, the ``ftrace`` filter can b
733 :c:func:`intel_pstate_set_pstate`::
734
735 # cd /sys/kernel/tracing/
736 # cat available_filter_functions | grep -i ps
737 intel_pstate_set_pstate
738 intel_pstate_cpu_init
739 ...
740 # echo intel_pstate_set_pstate > set_ftrace_f
741 # echo function > current_tracer
742 # cat trace | head -15
743 # tracer: function
744 #
745 # entries-in-buffer/entries-written: 80/80
746 #
747 # _-----=> irqs-
748 # / _----=> need-
749 # | / _---=> hardi
750 # || / _--=> preem
751 # ||| / delay
752 # TASK-PID CPU# |||| TIMESTAM
753 # | | | |||| |
754 Xorg-3129 [000] ..s. 2537.64484
755 gnome-terminal--4510 [002] ..s. 2537.64984
756 gnome-shell-3409 [001] ..s. 2537.65085
757 <idle>-0 [000] ..s. 2537.65484
758
759
760 References
761 ==========
762
763 .. [1] Kristen Accardi, *Balancing Power and P
764 https://events.static.linuxfound.org/si
765
766 .. [2] *Intel® 64 and IA-32 Architectures Sof
767 https://www.intel.com/content/www/us/en
768
769 .. [3] *Advanced Configuration and Power Inter
770 https://uefi.org/sites/default/files/re
Linux® is a registered trademark of Linus Torvalds in the United States and other countries.
TOMOYO® is a registered trademark of NTT DATA CORPORATION.