xref: /linux/Documentation/ABI/testing/sysfs-devices-system-cpu (revision 856250ba2e810e772dc95b3234ebf0d6393a51d9)
1What:		/sys/devices/system/cpu/
2Date:		pre-git history
3Contact:	Linux kernel mailing list <linux-kernel@vger.kernel.org>
4Description:
5		A collection of both global and individual CPU attributes
6
7		Individual CPU attributes are contained in subdirectories
8		named by the kernel's logical CPU number, e.g.:
9
10		/sys/devices/system/cpu/cpuX/
11
12What:		/sys/devices/system/cpu/kernel_max
13		/sys/devices/system/cpu/offline
14		/sys/devices/system/cpu/online
15		/sys/devices/system/cpu/possible
16		/sys/devices/system/cpu/present
17Date:		December 2008
18Contact:	Linux kernel mailing list <linux-kernel@vger.kernel.org>
19Description:	CPU topology files that describe kernel limits related to
20		hotplug. Briefly:
21
22		kernel_max: the maximum cpu index allowed by the kernel
23		configuration.
24
25		offline: cpus that are not online because they have been
26		HOTPLUGGED off or exceed the limit of cpus allowed by the
27		kernel configuration (kernel_max above).
28
29		online: cpus that are online and being scheduled.
30
31		possible: cpus that have been allocated resources and can be
32		brought online if they are present.
33
34		present: cpus that have been identified as being present in
35		the system.
36
37		See Documentation/admin-guide/cputopology.rst for more information.
38
39
40What:		/sys/devices/system/cpu/probe
41		/sys/devices/system/cpu/release
42Date:		November 2009
43Contact:	Linux kernel mailing list <linux-kernel@vger.kernel.org>
44Description:	Dynamic addition and removal of CPU's.  This is not hotplug
45		removal, this is meant complete removal/addition of the CPU
46		from the system.
47
48		probe: writes to this file will dynamically add a CPU to the
49		system.  Information written to the file to add CPU's is
50		architecture specific.
51
52		release: writes to this file dynamically remove a CPU from
53		the system.  Information written to the file to remove CPU's
54		is architecture specific.
55
56What:		/sys/devices/system/cpu/cpuX/node
57Date:		October 2009
58Contact:	Linux memory management mailing list <linux-mm@kvack.org>
59Description:	Discover NUMA node a CPU belongs to
60
61		When CONFIG_NUMA is enabled, a symbolic link that points
62		to the corresponding NUMA node directory.
63
64		For example, the following symlink is created for cpu42
65		in NUMA node 2:
66
67		/sys/devices/system/cpu/cpu42/node2 -> ../../node/node2
68
69
70What:		/sys/devices/system/cpu/cpuX/topology/core_siblings
71		/sys/devices/system/cpu/cpuX/topology/core_siblings_list
72		/sys/devices/system/cpu/cpuX/topology/physical_package_id
73		/sys/devices/system/cpu/cpuX/topology/thread_siblings
74		/sys/devices/system/cpu/cpuX/topology/thread_siblings_list
75		/sys/devices/system/cpu/cpuX/topology/ppin
76Date:		December 2008
77Contact:	Linux kernel mailing list <linux-kernel@vger.kernel.org>
78Description:	CPU topology files that describe a logical CPU's relationship
79		to other cores and threads in the same physical package.
80
81		One cpuX directory is created per logical CPU in the system,
82		e.g. /sys/devices/system/cpu/cpu42/.
83
84		Briefly, the files above are:
85
86		core_siblings: internal kernel map of cpuX's hardware threads
87		within the same physical_package_id.
88
89		core_siblings_list: human-readable list of the logical CPU
90		numbers within the same physical_package_id as cpuX.
91
92		physical_package_id: physical package id of cpuX. Typically
93		corresponds to a physical socket number, but the actual value
94		is architecture and platform dependent.
95
96		thread_siblings: internal kernel map of cpuX's hardware
97		threads within the same core as cpuX
98
99		thread_siblings_list: human-readable list of cpuX's hardware
100		threads within the same core as cpuX
101
102		ppin: human-readable Protected Processor Identification
103		Number of the socket the cpu# belongs to. There should be
104		one per physical_package_id. File is readable only to
105		admin.
106
107		See Documentation/admin-guide/cputopology.rst for more information.
108
109
110What:		/sys/devices/system/cpu/cpuidle/available_governors
111		/sys/devices/system/cpu/cpuidle/current_driver
112		/sys/devices/system/cpu/cpuidle/current_governor
113		/sys/devices/system/cpu/cpuidle/current_governer_ro
114		/sys/devices/system/cpu/cpuidle/intel_c1_demotion
115Date:		September 2007
116Contact:	Linux kernel mailing list <linux-kernel@vger.kernel.org>
117Description:	Discover cpuidle policy and mechanism
118
119		Various CPUs today support multiple idle levels that are
120		differentiated by varying exit latencies and power
121		consumption during idle.
122
123		Idle policy (governor) is differentiated from idle mechanism
124		(driver).
125
126		available_governors: (RO) displays a space separated list of
127		available governors.
128
129		current_driver: (RO) displays current idle mechanism.
130
131		current_governor: (RW) displays current idle policy. Users can
132		switch the governor at runtime by writing to this file.
133
134		current_governor_ro: (RO) displays current idle policy.
135
136		intel_c1_demotion: (RW) enables/disables the C1 demotion
137		feature on Intel CPUs.
138
139		See Documentation/admin-guide/pm/cpuidle.rst,
140		Documentation/admin-guide/pm/intel_idle.rst, and
141		Documentation/driver-api/pm/cpuidle.rst for more information.
142
143
144What:		/sys/devices/system/cpu/cpuX/cpuidle/state<N>/name
145		/sys/devices/system/cpu/cpuX/cpuidle/stateN/latency
146		/sys/devices/system/cpu/cpuX/cpuidle/stateN/power
147		/sys/devices/system/cpu/cpuX/cpuidle/stateN/time
148		/sys/devices/system/cpu/cpuX/cpuidle/stateN/usage
149		/sys/devices/system/cpu/cpuX/cpuidle/stateN/above
150		/sys/devices/system/cpu/cpuX/cpuidle/stateN/below
151Date:		September 2007
152KernelVersion:	v2.6.24
153Contact:	Linux power management list <linux-pm@vger.kernel.org>
154Description:
155		The directory /sys/devices/system/cpu/cpuX/cpuidle contains per
156		logical CPU specific cpuidle information for each online cpu X.
157		The processor idle states which are available for use have the
158		following attributes:
159
160		======== ==== =================================================
161		name:	 (RO) Name of the idle state (string).
162
163		latency: (RO) The latency to exit out of this idle state (in
164			      microseconds).
165
166		power:   (RO) The power consumed while in this idle state (in
167			      milliwatts).
168
169		time:    (RO) The total time spent in this idle state
170			      (in microseconds).
171
172		usage:	 (RO) Number of times this state was entered (a count).
173
174		above:	 (RO) Number of times this state was entered, but the
175			      observed CPU idle duration was too short for it
176			      (a count).
177
178		below:	 (RO) Number of times this state was entered, but the
179			      observed CPU idle duration was too long for it
180			      (a count).
181		======== ==== =================================================
182
183What:		/sys/devices/system/cpu/cpuX/cpuidle/state<N>/desc
184Date:		February 2008
185KernelVersion:	v2.6.25
186Contact:	Linux power management list <linux-pm@vger.kernel.org>
187Description:
188		(RO) A small description about the idle state (string).
189
190
191What:		/sys/devices/system/cpu/cpuX/cpuidle/state<N>/disable
192Date:		March 2012
193KernelVersion:	v3.10
194Contact:	Linux power management list <linux-pm@vger.kernel.org>
195Description:
196		(RW) Option to disable this idle state (bool). The behavior and
197		the effect of the disable variable depends on the implementation
198		of a particular governor. In the ladder governor, for example,
199		it is not coherent, i.e. if one is disabling a light state, then
200		all deeper states are disabled as well, but the disable variable
201		does not reflect it. Likewise, if one enables a deep state but a
202		lighter state still is disabled, then this has no effect.
203
204What:		/sys/devices/system/cpu/cpuX/cpuidle/state<N>/default_status
205Date:		December 2019
206KernelVersion:	v5.6
207Contact:	Linux power management list <linux-pm@vger.kernel.org>
208Description:
209		(RO) The default status of this state, "enabled" or "disabled".
210
211What:		/sys/devices/system/cpu/cpuX/cpuidle/state<N>/residency
212Date:		March 2014
213KernelVersion:	v3.15
214Contact:	Linux power management list <linux-pm@vger.kernel.org>
215Description:
216		(RO) Display the target residency i.e. the minimum amount of
217		time (in microseconds) this cpu should spend in this idle state
218		to make the transition worth the effort.
219
220What:		/sys/devices/system/cpu/cpuX/cpuidle/state<N>/s2idle/
221Date:		March 2018
222KernelVersion:	v4.17
223Contact:	Linux power management list <linux-pm@vger.kernel.org>
224Description:
225		Idle state usage statistics related to suspend-to-idle.
226
227		This attribute group is only present for states that can be
228		used in suspend-to-idle with suspended timekeeping.
229
230What:		/sys/devices/system/cpu/cpuX/cpuidle/state<N>/s2idle/time
231Date:		March 2018
232KernelVersion:	v4.17
233Contact:	Linux power management list <linux-pm@vger.kernel.org>
234Description:
235		Total time spent by the CPU in suspend-to-idle (with scheduler
236		tick suspended) after requesting this state.
237
238What:		/sys/devices/system/cpu/cpuX/cpuidle/state<N>/s2idle/usage
239Date:		March 2018
240KernelVersion:	v4.17
241Contact:	Linux power management list <linux-pm@vger.kernel.org>
242Description:
243		Total number of times this state has been requested by the CPU
244		while entering suspend-to-idle.
245
246What:		/sys/devices/system/cpu/cpuX/cpufreq/*
247Date:		pre-git history
248Contact:	linux-pm@vger.kernel.org
249Description:	Discover and change clock speed of CPUs
250
251		Clock scaling allows you to change the clock speed of the
252		CPUs on the fly. This is a nice method to save battery
253		power, because the lower the clock speed, the less power
254		the CPU consumes.
255
256		There are many knobs to tweak in this directory.
257
258		See files in Documentation/cpu-freq/ for more information.
259
260
261What:		/sys/devices/system/cpu/cpuX/cpufreq/freqdomain_cpus
262Date:		June 2013
263Contact:	linux-pm@vger.kernel.org
264Description:	Discover CPUs in the same CPU frequency coordination domain
265
266		freqdomain_cpus is the list of CPUs (online+offline) that share
267		the same clock/freq domain (possibly at the hardware level).
268		That information may be hidden from the cpufreq core and the
269		value of related_cpus may be different from freqdomain_cpus. This
270		attribute is useful for user space DVFS controllers to get better
271		power/performance results for platforms using acpi-cpufreq.
272
273		This file is only present if the acpi-cpufreq or the cppc-cpufreq
274		drivers are in use.
275
276What:		/sys/devices/system/cpu/cpuX/cpufreq/auto_select
277Date:		May 2025
278Contact:	linux-pm@vger.kernel.org
279Description:	Autonomous selection enable
280
281		Read/write interface to control autonomous selection enable
282			Read returns autonomous selection status:
283				0: autonomous selection is disabled
284				1: autonomous selection is enabled
285
286			Write 'y' or '1' or 'on' to enable autonomous selection.
287			Write 'n' or '0' or 'off' to disable autonomous selection.
288
289		This file is only present if the cppc-cpufreq driver is in use.
290
291What:		/sys/devices/system/cpu/cpuX/cpufreq/auto_act_window
292Date:		May 2025
293Contact:	linux-pm@vger.kernel.org
294Description:	Autonomous activity window
295
296		This file indicates a moving utilization sensitivity window to
297		the platform's autonomous selection policy.
298
299		Read/write an integer represents autonomous activity window (in
300		microseconds) from/to this file. The max value to write is
301		1270000000 but the max significand is 127. This means that if 128
302		is written to this file, 127 will be stored. If the value is
303		greater than 130, only the first two digits will be saved as
304		significand.
305
306		Writing a zero value to this file enable the platform to
307		determine an appropriate Activity Window depending on the workload.
308
309		Writing to this file only has meaning when Autonomous Selection is
310		enabled.
311
312		This file is only present if the cppc-cpufreq driver is in use.
313
314What:		/sys/devices/system/cpu/cpuX/cpufreq/energy_performance_preference_val
315Date:		May 2025
316Contact:	linux-pm@vger.kernel.org
317Description:	Energy performance preference
318
319		Read/write an 8-bit integer from/to this file. This file
320		represents a range of values from 0 (performance preference) to
321		0xFF (energy efficiency preference) that influences the rate of
322		performance increase/decrease and the result of the hardware's
323		energy efficiency and performance optimization policies.
324
325		Writing to this file only has meaning when Autonomous Selection is
326		enabled.
327
328		This file is only present if the cppc-cpufreq driver is in use.
329
330What:		/sys/devices/system/cpu/cpuX/cpufreq/perf_limited
331Date:		February 2026
332Contact:	linux-pm@vger.kernel.org
333Description:	Performance Limited
334
335		Read to check if platform throttling (thermal/power/current
336		limits) caused delivered performance to fall below the
337		requested level. A non-zero value indicates throttling occurred.
338
339		Write the bitmask of bits to clear:
340
341		  - 0x1 = clear bit 0 (desired performance excursion)
342		  - 0x2 = clear bit 1 (minimum performance excursion)
343		  - 0x3 = clear both bits
344
345		The platform sets these bits; OSPM can only clear them.
346
347		This file is only present if the cppc-cpufreq driver is in use.
348
349What:		/sys/devices/system/cpu/cpu*/cache/index3/cache_disable_{0,1}
350Date:		August 2008
351KernelVersion:	2.6.27
352Contact:	Linux kernel mailing list <linux-kernel@vger.kernel.org>
353Description:	Disable L3 cache indices
354
355		These files exist in every CPU's cache/index3 directory. Each
356		cache_disable_{0,1} file corresponds to one disable slot which
357		can be used to disable a cache index. Reading from these files
358		on a processor with this functionality will return the currently
359		disabled index for that node. There is one L3 structure per
360		node, or per internal node on MCM machines. Writing a valid
361		index to one of these files will cause the specified cache
362		index to be disabled.
363
364		All AMD processors with L3 caches provide this functionality.
365		For details, see BKDGs at
366                https://www.amd.com/en/support/tech-docs?keyword=bios+kernel
367
368
369What:		/sys/devices/system/cpu/cpufreq/boost
370Date:		August 2012
371Contact:	Linux kernel mailing list <linux-kernel@vger.kernel.org>
372Description:	Processor frequency boosting control
373
374		This switch controls the boost setting for the whole system.
375		Boosting allows the CPU and the firmware to run at a frequency
376		beyond its nominal limit.
377
378		More details can be found in
379		Documentation/admin-guide/pm/cpufreq.rst
380
381
382What:		/sys/devices/system/cpu/cpuX/crash_notes
383		/sys/devices/system/cpu/cpuX/crash_notes_size
384Date:		April 2013
385Contact:	kexec@lists.infradead.org
386Description:	address and size of the percpu note.
387
388		crash_notes: the physical address of the memory that holds the
389		note of cpuX.
390
391		crash_notes_size: size of the note of cpuX.
392
393
394What:		/sys/devices/system/cpu/intel_pstate/max_perf_pct
395		/sys/devices/system/cpu/intel_pstate/min_perf_pct
396		/sys/devices/system/cpu/intel_pstate/no_turbo
397Date:		February 2013
398Contact:	linux-pm@vger.kernel.org
399Description:	Parameters for the Intel P-state driver
400
401		Logic for selecting the current P-state in Intel
402		Sandybridge+ processors. The three knobs control
403		limits for the P-state that will be requested by the
404		driver.
405
406		max_perf_pct: limits the maximum P state that will be requested by
407		the driver stated as a percentage of the available performance.
408
409		min_perf_pct: limits the minimum P state that will be requested by
410		the driver stated as a percentage of the available performance.
411
412		no_turbo: limits the driver to selecting P states below the turbo
413		frequency range.
414
415		More details can be found in
416		Documentation/admin-guide/pm/intel_pstate.rst
417
418What:		/sys/devices/system/cpu/cpu*/cache/index*/<set_of_attributes_mentioned_below>
419Date:		July 2014(documented, existed before August 2008)
420Contact:	Sudeep Holla <sudeep.holla@arm.com>
421		Linux kernel mailing list <linux-kernel@vger.kernel.org>
422Description:	Parameters for the CPU cache attributes
423
424		allocation_policy:
425			- WriteAllocate:
426					allocate a memory location to a cache line
427					on a cache miss because of a write
428			- ReadAllocate:
429					allocate a memory location to a cache line
430					on a cache miss because of a read
431			- ReadWriteAllocate:
432					both writeallocate and readallocate
433
434		coherency_line_size:
435				     the minimum amount of data in bytes that gets
436				     transferred from memory to cache
437
438		level:
439			the cache hierarchy in the multi-level cache configuration
440
441		number_of_sets:
442				total number of sets in the cache, a set is a
443				collection of cache lines with the same cache index
444
445		physical_line_partition:
446				number of physical cache line per cache tag
447
448		shared_cpu_list:
449				the list of logical cpus sharing the cache
450
451		shared_cpu_map:
452				logical cpu mask containing the list of cpus sharing
453				the cache
454
455		size:
456			the total cache size in kB
457
458		type:
459			- Instruction: cache that only holds instructions
460			- Data: cache that only caches data
461			- Unified: cache that holds both data and instructions
462
463		ways_of_associativity:
464			degree of freedom in placing a particular block
465			of memory in the cache
466
467		write_policy:
468			- WriteThrough:
469					data is written to both the cache line
470					and to the block in the lower-level memory
471			- WriteBack:
472				     data is written only to the cache line and
473				     the modified cache line is written to main
474				     memory only when it is replaced
475
476
477What:		/sys/devices/system/cpu/cpu*/cache/index*/id
478Date:		September 2016
479Contact:	Linux kernel mailing list <linux-kernel@vger.kernel.org>
480Description:	Cache id
481
482		The id provides a unique number for a specific instance of
483		a cache of a particular type. E.g. there may be a level
484		3 unified cache on each socket in a server and we may
485		assign them ids 0, 1, 2, ...
486
487		Note that id value can be non-contiguous. E.g. level 1
488		caches typically exist per core, but there may not be a
489		power of two cores on a socket, so these caches may be
490		numbered 0, 1, 2, 3, 4, 5, 8, 9, 10, ...
491
492What:		/sys/devices/system/cpu/cpuX/cpufreq/throttle_stats
493		/sys/devices/system/cpu/cpuX/cpufreq/throttle_stats/turbo_stat
494		/sys/devices/system/cpu/cpuX/cpufreq/throttle_stats/sub_turbo_stat
495		/sys/devices/system/cpu/cpuX/cpufreq/throttle_stats/unthrottle
496		/sys/devices/system/cpu/cpuX/cpufreq/throttle_stats/powercap
497		/sys/devices/system/cpu/cpuX/cpufreq/throttle_stats/overtemp
498		/sys/devices/system/cpu/cpuX/cpufreq/throttle_stats/supply_fault
499		/sys/devices/system/cpu/cpuX/cpufreq/throttle_stats/overcurrent
500		/sys/devices/system/cpu/cpuX/cpufreq/throttle_stats/occ_reset
501Date:		March 2016
502Contact:	Linux kernel mailing list <linux-kernel@vger.kernel.org>
503		Linux for PowerPC mailing list <linuxppc-dev@lists.ozlabs.org>
504Description:	POWERNV CPUFreq driver's frequency throttle stats directory and
505		attributes
506
507		'cpuX/cpufreq/throttle_stats' directory contains the CPU frequency
508		throttle stat attributes for the chip. The throttle stats of a cpu
509		is common across all the cpus belonging to a chip. Below are the
510		throttle attributes exported in the 'throttle_stats' directory:
511
512		- turbo_stat : This file gives the total number of times the max
513		  frequency is throttled to lower frequency in turbo (at and above
514		  nominal frequency) range of frequencies.
515
516		- sub_turbo_stat : This file gives the total number of times the
517		  max frequency is throttled to lower frequency in sub-turbo(below
518		  nominal frequency) range of frequencies.
519
520		- unthrottle : This file gives the total number of times the max
521		  frequency is unthrottled after being throttled.
522
523		- powercap : This file gives the total number of times the max
524		  frequency is throttled due to 'Power Capping'.
525
526		- overtemp : This file gives the total number of times the max
527		  frequency is throttled due to 'CPU Over Temperature'.
528
529		- supply_fault : This file gives the total number of times the
530		  max frequency is throttled due to 'Power Supply Failure'.
531
532		- overcurrent : This file gives the total number of times the
533		  max frequency is throttled due to 'Overcurrent'.
534
535		- occ_reset : This file gives the total number of times the max
536		  frequency is throttled due to 'OCC Reset'.
537
538		The sysfs attributes representing different throttle reasons like
539		powercap, overtemp, supply_fault, overcurrent and occ_reset map to
540		the reasons provided by OCC firmware for throttling the frequency.
541
542What:		/sys/devices/system/cpu/cpufreq/policyX/throttle_stats
543		/sys/devices/system/cpu/cpufreq/policyX/throttle_stats/turbo_stat
544		/sys/devices/system/cpu/cpufreq/policyX/throttle_stats/sub_turbo_stat
545		/sys/devices/system/cpu/cpufreq/policyX/throttle_stats/unthrottle
546		/sys/devices/system/cpu/cpufreq/policyX/throttle_stats/powercap
547		/sys/devices/system/cpu/cpufreq/policyX/throttle_stats/overtemp
548		/sys/devices/system/cpu/cpufreq/policyX/throttle_stats/supply_fault
549		/sys/devices/system/cpu/cpufreq/policyX/throttle_stats/overcurrent
550		/sys/devices/system/cpu/cpufreq/policyX/throttle_stats/occ_reset
551Date:		March 2016
552Contact:	Linux kernel mailing list <linux-kernel@vger.kernel.org>
553		Linux for PowerPC mailing list <linuxppc-dev@lists.ozlabs.org>
554Description:	POWERNV CPUFreq driver's frequency throttle stats directory and
555		attributes
556
557		'policyX/throttle_stats' directory and all the attributes are same as
558		the /sys/devices/system/cpu/cpuX/cpufreq/throttle_stats directory and
559		attributes which give the frequency throttle information of the chip.
560
561What:		/sys/devices/system/cpu/cpuX/regs/
562		/sys/devices/system/cpu/cpuX/regs/identification/
563		/sys/devices/system/cpu/cpuX/regs/identification/midr_el1
564		/sys/devices/system/cpu/cpuX/regs/identification/revidr_el1
565		/sys/devices/system/cpu/cpuX/regs/identification/aidr_el1
566		/sys/devices/system/cpu/cpuX/regs/identification/smidr_el1
567Date:		June 2016
568Contact:	Linux ARM Kernel Mailing list <linux-arm-kernel@lists.infradead.org>
569Description:	AArch64 CPU registers
570
571		'identification' directory exposes the CPU ID registers for
572		identifying model and revision of the CPU and SMCU.
573
574What:		/sys/devices/system/cpu/aarch32_el0
575Date:		May 2021
576Contact:	Linux ARM Kernel Mailing list <linux-arm-kernel@lists.infradead.org>
577Description:	Identifies the subset of CPUs in the system that can execute
578		AArch32 (32-bit ARM) applications. If present, the same format as
579		/sys/devices/system/cpu/{offline,online,possible,present} is used.
580		If absent, then all or none of the CPUs can execute AArch32
581		applications and execve() will behave accordingly.
582
583What:		/sys/devices/system/cpu/cpuX/cpu_capacity
584Date:		December 2016
585Contact:	Linux kernel mailing list <linux-kernel@vger.kernel.org>
586Description:	information about CPUs heterogeneity.
587
588		cpu_capacity: capacity of cpuX.
589
590What:		/sys/devices/system/cpu/vulnerabilities
591		/sys/devices/system/cpu/vulnerabilities/gather_data_sampling
592		/sys/devices/system/cpu/vulnerabilities/indirect_target_selection
593		/sys/devices/system/cpu/vulnerabilities/itlb_multihit
594		/sys/devices/system/cpu/vulnerabilities/l1tf
595		/sys/devices/system/cpu/vulnerabilities/mds
596		/sys/devices/system/cpu/vulnerabilities/meltdown
597		/sys/devices/system/cpu/vulnerabilities/mmio_stale_data
598		/sys/devices/system/cpu/vulnerabilities/old_microcode
599		/sys/devices/system/cpu/vulnerabilities/reg_file_data_sampling
600		/sys/devices/system/cpu/vulnerabilities/retbleed
601		/sys/devices/system/cpu/vulnerabilities/spec_store_bypass
602		/sys/devices/system/cpu/vulnerabilities/spectre_v1
603		/sys/devices/system/cpu/vulnerabilities/spectre_v2
604		/sys/devices/system/cpu/vulnerabilities/srbds
605		/sys/devices/system/cpu/vulnerabilities/tsa
606		/sys/devices/system/cpu/vulnerabilities/tsx_async_abort
607		/sys/devices/system/cpu/vulnerabilities/vmscape
608Date:		January 2018
609Contact:	Linux kernel mailing list <linux-kernel@vger.kernel.org>
610Description:	Information about CPU vulnerabilities
611
612		The files are named after the code names of CPU
613		vulnerabilities. The output of those files reflects the
614		state of the CPUs in the system. Possible output values:
615
616		================  ==============================================
617		"Not affected"	  CPU is not affected by the vulnerability
618		"Vulnerable"	  CPU is affected and no mitigation in effect
619		"Mitigation: $M"  CPU is affected and mitigation $M is in effect
620		================  ==============================================
621
622		See also: Documentation/admin-guide/hw-vuln/index.rst
623
624What:		/sys/devices/system/cpu/smt
625		/sys/devices/system/cpu/smt/active
626		/sys/devices/system/cpu/smt/control
627Date:		June 2018
628Contact:	Linux kernel mailing list <linux-kernel@vger.kernel.org>
629Description:	Control Symmetric Multi Threading (SMT)
630
631		active:  Tells whether SMT is active (enabled and siblings online)
632
633		control: Read/write interface to control SMT. Possible
634			 values:
635
636			 ================ =========================================
637			 "on"		  SMT is enabled
638			 "off"		  SMT is disabled
639			 "<N>"		  SMT is enabled with N threads per core.
640			 "forceoff"	  SMT is force disabled. Cannot be changed.
641			 "notsupported"   SMT is not supported by the CPU
642			 "notimplemented" SMT runtime toggling is not
643					  implemented for the architecture
644			 ================ =========================================
645
646			 If control status is "forceoff" or "notsupported" writes
647			 are rejected. Note that enabling SMT on PowerPC skips
648			 offline cores.
649
650What:		/sys/devices/system/cpu/cpuX/power/energy_perf_bias
651Date:		March 2019
652Contact:	linux-pm@vger.kernel.org
653Description:	Intel Energy and Performance Bias Hint (EPB)
654
655		EPB for the given CPU in a sliding scale 0 - 15, where a value
656		of 0 corresponds to a hint preference for highest performance
657		and a value of 15 corresponds to the maximum energy savings.
658
659		In order to change the EPB value for the CPU, write either
660		a number in the 0 - 15 sliding scale above, or one of the
661		strings: "performance", "balance-performance", "normal",
662		"balance-power", "power" (that represent values reflected by
663		their meaning), to this attribute.
664
665		This attribute is present for all online CPUs supporting the
666		Intel EPB feature.
667
668What:		/sys/devices/system/cpu/umwait_control
669		/sys/devices/system/cpu/umwait_control/enable_c02
670		/sys/devices/system/cpu/umwait_control/max_time
671Date:		May 2019
672Contact:	Linux kernel mailing list <linux-kernel@vger.kernel.org>
673Description:	Umwait control
674
675		enable_c02: Read/write interface to control umwait C0.2 state
676			Read returns C0.2 state status:
677				0: C0.2 is disabled
678				1: C0.2 is enabled
679
680			Write 'y' or '1'  or 'on' to enable C0.2 state.
681			Write 'n' or '0'  or 'off' to disable C0.2 state.
682
683			The interface is case insensitive.
684
685		max_time: Read/write interface to control umwait maximum time
686			  in TSC-quanta that the CPU can reside in either C0.1
687			  or C0.2 state. The time is an unsigned 32-bit number.
688			  Note that a value of zero means there is no limit.
689			  Low order two bits must be zero.
690
691What:		/sys/devices/system/cpu/sev
692		/sys/devices/system/cpu/sev/vmpl
693Date:		May 2024
694Contact:	Linux kernel mailing list <linux-kernel@vger.kernel.org>
695Description:	Secure Encrypted Virtualization (SEV) information
696
697		This directory is only present when running as an SEV-SNP guest.
698
699		vmpl: Reports the Virtual Machine Privilege Level (VMPL) at which
700		      the SEV-SNP guest is running.
701
702
703What:		/sys/devices/system/cpu/svm
704Date:		August 2019
705Contact:	Linux kernel mailing list <linux-kernel@vger.kernel.org>
706		Linux for PowerPC mailing list <linuxppc-dev@lists.ozlabs.org>
707Description:	Secure Virtual Machine
708
709		If 1, it means the system is using the Protected Execution
710		Facility in POWER9 and newer processors. i.e., it is a Secure
711		Virtual Machine.
712
713What:		/sys/devices/system/cpu/cpuX/purr
714Date:		Apr 2005
715Contact:	Linux for PowerPC mailing list <linuxppc-dev@lists.ozlabs.org>
716Description:	PURR ticks for this CPU since the system boot.
717
718		The Processor Utilization Resources Register (PURR) is
719		a 64-bit counter which provides an estimate of the
720		resources used by the CPU thread. The contents of this
721		register increases monotonically. This sysfs interface
722		exposes the number of PURR ticks for cpuX.
723
724What: 		/sys/devices/system/cpu/cpuX/spurr
725Date:		Dec 2006
726Contact:	Linux for PowerPC mailing list <linuxppc-dev@lists.ozlabs.org>
727Description:	SPURR ticks for this CPU since the system boot.
728
729		The Scaled Processor Utilization Resources Register
730		(SPURR) is a 64-bit counter that provides a frequency
731		invariant estimate of the resources used by the CPU
732		thread. The contents of this register increases
733		monotonically. This sysfs interface exposes the number
734		of SPURR ticks for cpuX.
735
736What: 		/sys/devices/system/cpu/cpuX/idle_purr
737Date:		Apr 2020
738Contact:	Linux for PowerPC mailing list <linuxppc-dev@lists.ozlabs.org>
739Description:	PURR ticks for cpuX when it was idle.
740
741		This sysfs interface exposes the number of PURR ticks
742		for cpuX when it was idle.
743
744What: 		/sys/devices/system/cpu/cpuX/idle_spurr
745Date:		Apr 2020
746Contact:	Linux for PowerPC mailing list <linuxppc-dev@lists.ozlabs.org>
747Description:	SPURR ticks for cpuX when it was idle.
748
749		This sysfs interface exposes the number of SPURR ticks
750		for cpuX when it was idle.
751
752What: 		/sys/devices/system/cpu/cpuX/mte_tcf_preferred
753Date:		July 2021
754Contact:	Linux ARM Kernel Mailing list <linux-arm-kernel@lists.infradead.org>
755Description:	Preferred MTE tag checking mode
756
757		When a user program specifies more than one MTE tag checking
758		mode, this sysfs node is used to specify which mode should
759		be preferred when scheduling a task on that CPU. Possible
760		values:
761
762		================  ==============================================
763		"sync"	  	  Prefer synchronous mode
764		"asymm"	  	  Prefer asymmetric mode
765		"async"	  	  Prefer asynchronous mode
766		================  ==============================================
767
768		See also: Documentation/arch/arm64/memory-tagging-extension.rst
769
770What:		/sys/devices/system/cpu/nohz_full
771Date:		Apr 2015
772Contact:	Linux kernel mailing list <linux-kernel@vger.kernel.org>
773Description:
774		(RO) the list of CPUs that are in nohz_full mode.
775		These CPUs are set by boot parameter "nohz_full=".
776
777What:		/sys/devices/system/cpu/isolated
778Date:		Apr 2015
779Contact:	Linux kernel mailing list <linux-kernel@vger.kernel.org>
780Description:
781		(RO) the list of CPUs that are isolated and don't
782		participate in load balancing. These CPUs are set by
783		boot parameter "isolcpus=".
784
785What:		/sys/devices/system/cpu/housekeeping
786Date:		Oct 2025
787Contact:	Linux kernel mailing list <linux-kernel@vger.kernel.org>
788Description:
789		(RO) the list of logical CPUs that are designated by the kernel as
790		"housekeeping". Each CPU are responsible for handling essential
791		system-wide background tasks, including RCU callbacks, delayed
792		timer callbacks, and unbound workqueues, minimizing scheduling
793		jitter on low-latency, isolated CPUs. These CPUs are set when boot
794		parameter "isolcpus=nohz" or "nohz_full=" is specified.
795
796What:		/sys/devices/system/cpu/crash_hotplug
797Date:		Aug 2023
798Contact:	Linux kernel mailing list <linux-kernel@vger.kernel.org>
799Description:
800		(RO) indicates whether or not the kernel updates relevant kexec
801		segments on memory hot un/plug and/or on/offline events, avoiding the
802		need to reload kdump kernel.
803
804What:		/sys/devices/system/cpu/enabled
805Date:		Nov 2022
806Contact:	Linux kernel mailing list <linux-kernel@vger.kernel.org>
807Description:
808		(RO) the list of CPUs that can be brought online.
809