1 What: /sys/devices/system/cpu/
3 Contact: Linux kernel mailing list <linux-kernel@vger.kernel.org>
5 A collection of both global and individual CPU attributes
7 Individual CPU attributes are contained in subdirectories
8 named by the kernel's logical CPU number, e.g.:
10 /sys/devices/system/cpu/cpu#/
12 What: /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
18 Contact: Linux kernel mailing list <linux-kernel@vger.kernel.org>
19 Description: CPU topology files that describe kernel limits related to
22 kernel_max: the maximum cpu index allowed by the kernel
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).
29 online: cpus that are online and being scheduled.
31 possible: cpus that have been allocated resources and can be
32 brought online if they are present.
34 present: cpus that have been identified as being present in
37 See Documentation/admin-guide/cputopology.rst for more information.
40 What: /sys/devices/system/cpu/probe
41 /sys/devices/system/cpu/release
43 Contact: Linux kernel mailing list <linux-kernel@vger.kernel.org>
44 Description: Dynamic addition and removal of CPU's. This is not hotplug
45 removal, this is meant complete removal/addition of the CPU
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.
52 release: writes to this file dynamically remove a CPU from
53 the system. Information writtento the file to remove CPU's
54 is architecture specific.
56 What: /sys/devices/system/cpu/cpu#/node
58 Contact: Linux memory management mailing list <linux-mm@kvack.org>
59 Description: Discover NUMA node a CPU belongs to
61 When CONFIG_NUMA is enabled, a symbolic link that points
62 to the corresponding NUMA node directory.
64 For example, the following symlink is created for cpu42
67 /sys/devices/system/cpu/cpu42/node2 -> ../../node/node2
70 What: /sys/devices/system/cpu/cpu#/topology/core_id
71 /sys/devices/system/cpu/cpu#/topology/core_siblings
72 /sys/devices/system/cpu/cpu#/topology/core_siblings_list
73 /sys/devices/system/cpu/cpu#/topology/physical_package_id
74 /sys/devices/system/cpu/cpu#/topology/thread_siblings
75 /sys/devices/system/cpu/cpu#/topology/thread_siblings_list
77 Contact: Linux kernel mailing list <linux-kernel@vger.kernel.org>
78 Description: CPU topology files that describe a logical CPU's relationship
79 to other cores and threads in the same physical package.
81 One cpu# directory is created per logical CPU in the system,
82 e.g. /sys/devices/system/cpu/cpu42/.
84 Briefly, the files above are:
86 core_id: the CPU core ID of cpu#. Typically it is the
87 hardware platform's identifier (rather than the kernel's).
88 The actual value is architecture and platform dependent.
90 core_siblings: internal kernel map of cpu#'s hardware threads
91 within the same physical_package_id.
93 core_siblings_list: human-readable list of the logical CPU
94 numbers within the same physical_package_id as cpu#.
96 physical_package_id: physical package id of cpu#. Typically
97 corresponds to a physical socket number, but the actual value
98 is architecture and platform dependent.
100 thread_siblings: internel kernel map of cpu#'s hardware
101 threads within the same core as cpu#
103 thread_siblings_list: human-readable list of cpu#'s hardware
104 threads within the same core as cpu#
106 See Documentation/admin-guide/cputopology.rst for more information.
109 What: /sys/devices/system/cpu/cpuidle/available_governors
110 /sys/devices/system/cpu/cpuidle/current_driver
111 /sys/devices/system/cpu/cpuidle/current_governor
112 /sys/devices/system/cpu/cpuidle/current_governer_ro
114 Contact: Linux kernel mailing list <linux-kernel@vger.kernel.org>
115 Description: Discover cpuidle policy and mechanism
117 Various CPUs today support multiple idle levels that are
118 differentiated by varying exit latencies and power
119 consumption during idle.
121 Idle policy (governor) is differentiated from idle mechanism
124 available_governors: (RO) displays a space separated list of
127 current_driver: (RO) displays current idle mechanism.
129 current_governor: (RW) displays current idle policy. Users can
130 switch the governor at runtime by writing to this file.
132 current_governor_ro: (RO) displays current idle policy.
134 See Documentation/admin-guide/pm/cpuidle.rst and
135 Documentation/driver-api/pm/cpuidle.rst for more information.
138 What: /sys/devices/system/cpu/cpuX/cpuidle/stateN/name
139 /sys/devices/system/cpu/cpuX/cpuidle/stateN/latency
140 /sys/devices/system/cpu/cpuX/cpuidle/stateN/power
141 /sys/devices/system/cpu/cpuX/cpuidle/stateN/time
142 /sys/devices/system/cpu/cpuX/cpuidle/stateN/usage
143 /sys/devices/system/cpu/cpuX/cpuidle/stateN/above
144 /sys/devices/system/cpu/cpuX/cpuidle/stateN/below
146 KernelVersion: v2.6.24
147 Contact: Linux power management list <linux-pm@vger.kernel.org>
149 The directory /sys/devices/system/cpu/cpuX/cpuidle contains per
150 logical CPU specific cpuidle information for each online cpu X.
151 The processor idle states which are available for use have the
152 following attributes:
154 ======== ==== =================================================
155 name: (RO) Name of the idle state (string).
157 latency: (RO) The latency to exit out of this idle state (in
160 power: (RO) The power consumed while in this idle state (in
163 time: (RO) The total time spent in this idle state
166 usage: (RO) Number of times this state was entered (a count).
168 above: (RO) Number of times this state was entered, but the
169 observed CPU idle duration was too short for it
172 below: (RO) Number of times this state was entered, but the
173 observed CPU idle duration was too long for it
175 ======== ==== =================================================
177 What: /sys/devices/system/cpu/cpuX/cpuidle/stateN/desc
179 KernelVersion: v2.6.25
180 Contact: Linux power management list <linux-pm@vger.kernel.org>
182 (RO) A small description about the idle state (string).
185 What: /sys/devices/system/cpu/cpuX/cpuidle/stateN/disable
188 Contact: Linux power management list <linux-pm@vger.kernel.org>
190 (RW) Option to disable this idle state (bool). The behavior and
191 the effect of the disable variable depends on the implementation
192 of a particular governor. In the ladder governor, for example,
193 it is not coherent, i.e. if one is disabling a light state, then
194 all deeper states are disabled as well, but the disable variable
195 does not reflect it. Likewise, if one enables a deep state but a
196 lighter state still is disabled, then this has no effect.
198 What: /sys/devices/system/cpu/cpuX/cpuidle/stateN/default_status
201 Contact: Linux power management list <linux-pm@vger.kernel.org>
203 (RO) The default status of this state, "enabled" or "disabled".
205 What: /sys/devices/system/cpu/cpuX/cpuidle/stateN/residency
208 Contact: Linux power management list <linux-pm@vger.kernel.org>
210 (RO) Display the target residency i.e. the minimum amount of
211 time (in microseconds) this cpu should spend in this idle state
212 to make the transition worth the effort.
214 What: /sys/devices/system/cpu/cpuX/cpuidle/stateN/s2idle/
217 Contact: Linux power management list <linux-pm@vger.kernel.org>
219 Idle state usage statistics related to suspend-to-idle.
221 This attribute group is only present for states that can be
222 used in suspend-to-idle with suspended timekeeping.
224 What: /sys/devices/system/cpu/cpuX/cpuidle/stateN/s2idle/time
227 Contact: Linux power management list <linux-pm@vger.kernel.org>
229 Total time spent by the CPU in suspend-to-idle (with scheduler
230 tick suspended) after requesting this state.
232 What: /sys/devices/system/cpu/cpuX/cpuidle/stateN/s2idle/usage
235 Contact: Linux power management list <linux-pm@vger.kernel.org>
237 Total number of times this state has been requested by the CPU
238 while entering suspend-to-idle.
240 What: /sys/devices/system/cpu/cpu#/cpufreq/*
241 Date: pre-git history
242 Contact: linux-pm@vger.kernel.org
243 Description: Discover and change clock speed of CPUs
245 Clock scaling allows you to change the clock speed of the
246 CPUs on the fly. This is a nice method to save battery
247 power, because the lower the clock speed, the less power
250 There are many knobs to tweak in this directory.
252 See files in Documentation/cpu-freq/ for more information.
255 What: /sys/devices/system/cpu/cpu#/cpufreq/freqdomain_cpus
257 Contact: linux-pm@vger.kernel.org
258 Description: Discover CPUs in the same CPU frequency coordination domain
260 freqdomain_cpus is the list of CPUs (online+offline) that share
261 the same clock/freq domain (possibly at the hardware level).
262 That information may be hidden from the cpufreq core and the
263 value of related_cpus may be different from freqdomain_cpus. This
264 attribute is useful for user space DVFS controllers to get better
265 power/performance results for platforms using acpi-cpufreq.
267 This file is only present if the acpi-cpufreq driver is in use.
270 What: /sys/devices/system/cpu/cpu*/cache/index3/cache_disable_{0,1}
272 KernelVersion: 2.6.27
273 Contact: Linux kernel mailing list <linux-kernel@vger.kernel.org>
274 Description: Disable L3 cache indices
276 These files exist in every CPU's cache/index3 directory. Each
277 cache_disable_{0,1} file corresponds to one disable slot which
278 can be used to disable a cache index. Reading from these files
279 on a processor with this functionality will return the currently
280 disabled index for that node. There is one L3 structure per
281 node, or per internal node on MCM machines. Writing a valid
282 index to one of these files will cause the specificed cache
283 index to be disabled.
285 All AMD processors with L3 caches provide this functionality.
286 For details, see BKDGs at
287 http://developer.amd.com/documentation/guides/Pages/default.aspx
290 What: /sys/devices/system/cpu/cpufreq/boost
292 Contact: Linux kernel mailing list <linux-kernel@vger.kernel.org>
293 Description: Processor frequency boosting control
295 This switch controls the boost setting for the whole system.
296 Boosting allows the CPU and the firmware to run at a frequency
297 beyound it's nominal limit.
299 More details can be found in
300 Documentation/admin-guide/pm/cpufreq.rst
303 What: /sys/devices/system/cpu/cpu#/crash_notes
304 /sys/devices/system/cpu/cpu#/crash_notes_size
306 Contact: kexec@lists.infradead.org
307 Description: address and size of the percpu note.
309 crash_notes: the physical address of the memory that holds the
312 crash_notes_size: size of the note of cpu#.
315 What: /sys/devices/system/cpu/intel_pstate/max_perf_pct
316 /sys/devices/system/cpu/intel_pstate/min_perf_pct
317 /sys/devices/system/cpu/intel_pstate/no_turbo
319 Contact: linux-pm@vger.kernel.org
320 Description: Parameters for the Intel P-state driver
322 Logic for selecting the current P-state in Intel
323 Sandybridge+ processors. The three knobs control
324 limits for the P-state that will be requested by the
327 max_perf_pct: limits the maximum P state that will be requested by
328 the driver stated as a percentage of the available performance.
330 min_perf_pct: limits the minimum P state that will be requested by
331 the driver stated as a percentage of the available performance.
333 no_turbo: limits the driver to selecting P states below the turbo
336 More details can be found in
337 Documentation/admin-guide/pm/intel_pstate.rst
339 What: /sys/devices/system/cpu/cpu*/cache/index*/<set_of_attributes_mentioned_below>
340 Date: July 2014(documented, existed before August 2008)
341 Contact: Sudeep Holla <sudeep.holla@arm.com>
342 Linux kernel mailing list <linux-kernel@vger.kernel.org>
343 Description: Parameters for the CPU cache attributes
347 allocate a memory location to a cache line
348 on a cache miss because of a write
350 allocate a memory location to a cache line
351 on a cache miss because of a read
353 both writeallocate and readallocate
356 LEGACY used only on IA64 and is same as write_policy
359 the minimum amount of data in bytes that gets
360 transferred from memory to cache
363 the cache hierarchy in the multi-level cache configuration
366 total number of sets in the cache, a set is a
367 collection of cache lines with the same cache index
369 physical_line_partition:
370 number of physical cache line per cache tag
373 the list of logical cpus sharing the cache
376 logical cpu mask containing the list of cpus sharing
380 the total cache size in kB
383 - Instruction: cache that only holds instructions
384 - Data: cache that only caches data
385 - Unified: cache that holds both data and instructions
387 ways_of_associativity:
388 degree of freedom in placing a particular block
389 of memory in the cache
393 data is written to both the cache line
394 and to the block in the lower-level memory
396 data is written only to the cache line and
397 the modified cache line is written to main
398 memory only when it is replaced
401 What: /sys/devices/system/cpu/cpu*/cache/index*/id
403 Contact: Linux kernel mailing list <linux-kernel@vger.kernel.org>
404 Description: Cache id
406 The id provides a unique number for a specific instance of
407 a cache of a particular type. E.g. there may be a level
408 3 unified cache on each socket in a server and we may
409 assign them ids 0, 1, 2, ...
411 Note that id value can be non-contiguous. E.g. level 1
412 caches typically exist per core, but there may not be a
413 power of two cores on a socket, so these caches may be
414 numbered 0, 1, 2, 3, 4, 5, 8, 9, 10, ...
416 What: /sys/devices/system/cpu/cpuX/cpufreq/throttle_stats
417 /sys/devices/system/cpu/cpuX/cpufreq/throttle_stats/turbo_stat
418 /sys/devices/system/cpu/cpuX/cpufreq/throttle_stats/sub_turbo_stat
419 /sys/devices/system/cpu/cpuX/cpufreq/throttle_stats/unthrottle
420 /sys/devices/system/cpu/cpuX/cpufreq/throttle_stats/powercap
421 /sys/devices/system/cpu/cpuX/cpufreq/throttle_stats/overtemp
422 /sys/devices/system/cpu/cpuX/cpufreq/throttle_stats/supply_fault
423 /sys/devices/system/cpu/cpuX/cpufreq/throttle_stats/overcurrent
424 /sys/devices/system/cpu/cpuX/cpufreq/throttle_stats/occ_reset
426 Contact: Linux kernel mailing list <linux-kernel@vger.kernel.org>
427 Linux for PowerPC mailing list <linuxppc-dev@ozlabs.org>
428 Description: POWERNV CPUFreq driver's frequency throttle stats directory and
431 'cpuX/cpufreq/throttle_stats' directory contains the CPU frequency
432 throttle stat attributes for the chip. The throttle stats of a cpu
433 is common across all the cpus belonging to a chip. Below are the
434 throttle attributes exported in the 'throttle_stats' directory:
436 - turbo_stat : This file gives the total number of times the max
437 frequency is throttled to lower frequency in turbo (at and above
438 nominal frequency) range of frequencies.
440 - sub_turbo_stat : This file gives the total number of times the
441 max frequency is throttled to lower frequency in sub-turbo(below
442 nominal frequency) range of frequencies.
444 - unthrottle : This file gives the total number of times the max
445 frequency is unthrottled after being throttled.
447 - powercap : This file gives the total number of times the max
448 frequency is throttled due to 'Power Capping'.
450 - overtemp : This file gives the total number of times the max
451 frequency is throttled due to 'CPU Over Temperature'.
453 - supply_fault : This file gives the total number of times the
454 max frequency is throttled due to 'Power Supply Failure'.
456 - overcurrent : This file gives the total number of times the
457 max frequency is throttled due to 'Overcurrent'.
459 - occ_reset : This file gives the total number of times the max
460 frequency is throttled due to 'OCC Reset'.
462 The sysfs attributes representing different throttle reasons like
463 powercap, overtemp, supply_fault, overcurrent and occ_reset map to
464 the reasons provided by OCC firmware for throttling the frequency.
466 What: /sys/devices/system/cpu/cpufreq/policyX/throttle_stats
467 /sys/devices/system/cpu/cpufreq/policyX/throttle_stats/turbo_stat
468 /sys/devices/system/cpu/cpufreq/policyX/throttle_stats/sub_turbo_stat
469 /sys/devices/system/cpu/cpufreq/policyX/throttle_stats/unthrottle
470 /sys/devices/system/cpu/cpufreq/policyX/throttle_stats/powercap
471 /sys/devices/system/cpu/cpufreq/policyX/throttle_stats/overtemp
472 /sys/devices/system/cpu/cpufreq/policyX/throttle_stats/supply_fault
473 /sys/devices/system/cpu/cpufreq/policyX/throttle_stats/overcurrent
474 /sys/devices/system/cpu/cpufreq/policyX/throttle_stats/occ_reset
476 Contact: Linux kernel mailing list <linux-kernel@vger.kernel.org>
477 Linux for PowerPC mailing list <linuxppc-dev@ozlabs.org>
478 Description: POWERNV CPUFreq driver's frequency throttle stats directory and
481 'policyX/throttle_stats' directory and all the attributes are same as
482 the /sys/devices/system/cpu/cpuX/cpufreq/throttle_stats directory and
483 attributes which give the frequency throttle information of the chip.
485 What: /sys/devices/system/cpu/cpuX/regs/
486 /sys/devices/system/cpu/cpuX/regs/identification/
487 /sys/devices/system/cpu/cpuX/regs/identification/midr_el1
488 /sys/devices/system/cpu/cpuX/regs/identification/revidr_el1
490 Contact: Linux ARM Kernel Mailing list <linux-arm-kernel@lists.infradead.org>
491 Description: AArch64 CPU registers
493 'identification' directory exposes the CPU ID registers for
494 identifying model and revision of the CPU.
496 What: /sys/devices/system/cpu/cpu#/cpu_capacity
498 Contact: Linux kernel mailing list <linux-kernel@vger.kernel.org>
499 Description: information about CPUs heterogeneity.
501 cpu_capacity: capacity of cpu#.
503 What: /sys/devices/system/cpu/vulnerabilities
504 /sys/devices/system/cpu/vulnerabilities/meltdown
505 /sys/devices/system/cpu/vulnerabilities/spectre_v1
506 /sys/devices/system/cpu/vulnerabilities/spectre_v2
507 /sys/devices/system/cpu/vulnerabilities/spec_store_bypass
508 /sys/devices/system/cpu/vulnerabilities/l1tf
509 /sys/devices/system/cpu/vulnerabilities/mds
510 /sys/devices/system/cpu/vulnerabilities/srbds
511 /sys/devices/system/cpu/vulnerabilities/tsx_async_abort
512 /sys/devices/system/cpu/vulnerabilities/itlb_multihit
514 Contact: Linux kernel mailing list <linux-kernel@vger.kernel.org>
515 Description: Information about CPU vulnerabilities
517 The files are named after the code names of CPU
518 vulnerabilities. The output of those files reflects the
519 state of the CPUs in the system. Possible output values:
521 ================ ==============================================
522 "Not affected" CPU is not affected by the vulnerability
523 "Vulnerable" CPU is affected and no mitigation in effect
524 "Mitigation: $M" CPU is affected and mitigation $M is in effect
525 ================ ==============================================
527 See also: Documentation/admin-guide/hw-vuln/index.rst
529 What: /sys/devices/system/cpu/smt
530 /sys/devices/system/cpu/smt/active
531 /sys/devices/system/cpu/smt/control
533 Contact: Linux kernel mailing list <linux-kernel@vger.kernel.org>
534 Description: Control Symetric Multi Threading (SMT)
536 active: Tells whether SMT is active (enabled and siblings online)
538 control: Read/write interface to control SMT. Possible
541 ================ =========================================
543 "off" SMT is disabled
544 "forceoff" SMT is force disabled. Cannot be changed.
545 "notsupported" SMT is not supported by the CPU
546 "notimplemented" SMT runtime toggling is not
547 implemented for the architecture
548 ================ =========================================
550 If control status is "forceoff" or "notsupported" writes
553 What: /sys/devices/system/cpu/cpu#/power/energy_perf_bias
555 Contact: linux-pm@vger.kernel.org
556 Description: Intel Energy and Performance Bias Hint (EPB)
558 EPB for the given CPU in a sliding scale 0 - 15, where a value
559 of 0 corresponds to a hint preference for highest performance
560 and a value of 15 corresponds to the maximum energy savings.
562 In order to change the EPB value for the CPU, write either
563 a number in the 0 - 15 sliding scale above, or one of the
564 strings: "performance", "balance-performance", "normal",
565 "balance-power", "power" (that represent values reflected by
566 their meaning), to this attribute.
568 This attribute is present for all online CPUs supporting the
571 What: /sys/devices/system/cpu/umwait_control
572 /sys/devices/system/cpu/umwait_control/enable_c02
573 /sys/devices/system/cpu/umwait_control/max_time
575 Contact: Linux kernel mailing list <linux-kernel@vger.kernel.org>
576 Description: Umwait control
578 enable_c02: Read/write interface to control umwait C0.2 state
579 Read returns C0.2 state status:
583 Write 'y' or '1' or 'on' to enable C0.2 state.
584 Write 'n' or '0' or 'off' to disable C0.2 state.
586 The interface is case insensitive.
588 max_time: Read/write interface to control umwait maximum time
589 in TSC-quanta that the CPU can reside in either C0.1
590 or C0.2 state. The time is an unsigned 32-bit number.
591 Note that a value of zero means there is no limit.
592 Low order two bits must be zero.
594 What: /sys/devices/system/cpu/svm
596 Contact: Linux kernel mailing list <linux-kernel@vger.kernel.org>
597 Linux for PowerPC mailing list <linuxppc-dev@ozlabs.org>
598 Description: Secure Virtual Machine
600 If 1, it means the system is using the Protected Execution
601 Facility in POWER9 and newer processors. i.e., it is a Secure
604 What: /sys/devices/system/cpu/cpuX/purr
606 Contact: Linux for PowerPC mailing list <linuxppc-dev@ozlabs.org>
607 Description: PURR ticks for this CPU since the system boot.
609 The Processor Utilization Resources Register (PURR) is
610 a 64-bit counter which provides an estimate of the
611 resources used by the CPU thread. The contents of this
612 register increases monotonically. This sysfs interface
613 exposes the number of PURR ticks for cpuX.
615 What: /sys/devices/system/cpu/cpuX/spurr
617 Contact: Linux for PowerPC mailing list <linuxppc-dev@ozlabs.org>
618 Description: SPURR ticks for this CPU since the system boot.
620 The Scaled Processor Utilization Resources Register
621 (SPURR) is a 64-bit counter that provides a frequency
622 invariant estimate of the resources used by the CPU
623 thread. The contents of this register increases
624 monotonically. This sysfs interface exposes the number
625 of SPURR ticks for cpuX.
627 What: /sys/devices/system/cpu/cpuX/idle_purr
629 Contact: Linux for PowerPC mailing list <linuxppc-dev@ozlabs.org>
630 Description: PURR ticks for cpuX when it was idle.
632 This sysfs interface exposes the number of PURR ticks
633 for cpuX when it was idle.
635 What: /sys/devices/system/cpu/cpuX/idle_spurr
637 Contact: Linux for PowerPC mailing list <linuxppc-dev@ozlabs.org>
638 Description: SPURR ticks for cpuX when it was idle.
640 This sysfs interface exposes the number of SPURR ticks
641 for cpuX when it was idle.