Documentation/ABI/testing/sysfs-devices-system-cpu

   1 What:           /sys/devices/system/cpu/
   2 Date:           pre-git history
   3 Contact:        Linux kernel mailing list <linux-kernel@vger.kernel.org>
   4 Description:
   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/cpu#/
  11
  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
  17 Date:           December 2008
  18 Contact:        Linux kernel mailing list <linux-kernel@vger.kernel.org>
  19 Description:    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/cputopology.txt for more information.
  38
  39
  40 What:           /sys/devices/system/cpu/probe
  41                 /sys/devices/system/cpu/release
  42 Date:           November 2009
  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
  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 writtento the file to remove CPU's
  54                 is architecture specific.
  55
  56 What:           /sys/devices/system/cpu/cpu#/node
  57 Date:           October 2009
  58 Contact:        Linux memory management mailing list <linux-mm@kvack.org>
  59 Description:    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
  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
  76 Date:           December 2008
  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.
  80
  81                 One cpu# 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_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.
  89
  90                 core_siblings: internal kernel map of cpu#'s hardware threads
  91                 within the same physical_package_id.
  92
  93                 core_siblings_list: human-readable list of the logical CPU
  94                 numbers within the same physical_package_id as cpu#.
  95
  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.
  99
 100                 thread_siblings: internel kernel map of cpu#'s hardware
 101                 threads within the same core as cpu#
 102
 103                 thread_siblings_list: human-readable list of cpu#'s hardware
 104                 threads within the same core as cpu#
 105
 106                 See Documentation/cputopology.txt for more information.
 107
 108
 109 What:           /sys/devices/system/cpu/cpuidle/current_driver
 110                 /sys/devices/system/cpu/cpuidle/current_governer_ro
 111                 /sys/devices/system/cpu/cpuidle/available_governors
 112                 /sys/devices/system/cpu/cpuidle/current_governor
 113 Date:           September 2007
 114 Contact:        Linux kernel mailing list <linux-kernel@vger.kernel.org>
 115 Description:    Discover cpuidle policy and mechanism
 116
 117                 Various CPUs today support multiple idle levels that are
 118                 differentiated by varying exit latencies and power
 119                 consumption during idle.
 120
 121                 Idle policy (governor) is differentiated from idle mechanism
 122                 (driver)
 123
 124                 current_driver: (RO) displays current idle mechanism
 125
 126                 current_governor_ro: (RO) displays current idle policy
 127
 128                 With the cpuidle_sysfs_switch boot option enabled (meant for
 129                 developer testing), the following three attributes are visible
 130                 instead:
 131
 132                 current_driver: same as described above
 133
 134                 available_governors: (RO) displays a space separated list of
 135                 available governors
 136
 137                 current_governor: (RW) displays current idle policy. Users can
 138                 switch the governor at runtime by writing to this file.
 139
 140                 See files in Documentation/cpuidle/ for more information.
 141
 142
 143 What:           /sys/devices/system/cpu/cpuX/cpuidle/stateN/name
 144                 /sys/devices/system/cpu/cpuX/cpuidle/stateN/latency
 145                 /sys/devices/system/cpu/cpuX/cpuidle/stateN/power
 146                 /sys/devices/system/cpu/cpuX/cpuidle/stateN/time
 147                 /sys/devices/system/cpu/cpuX/cpuidle/stateN/usage
 148 Date:           September 2007
 149 KernelVersion:  v2.6.24
 150 Contact:        Linux power management list <linux-pm@vger.kernel.org>
 151 Description:
 152                 The directory /sys/devices/system/cpu/cpuX/cpuidle contains per
 153                 logical CPU specific cpuidle information for each online cpu X.
 154                 The processor idle states which are available for use have the
 155                 following attributes:
 156
 157                 name: (RO) Name of the idle state (string).
 158
 159                 latency: (RO) The latency to exit out of this idle state (in
 160                 microseconds).
 161
 162                 power: (RO) The power consumed while in this idle state (in
 163                 milliwatts).
 164
 165                 time: (RO) The total time spent in this idle state (in microseconds).
 166
 167                 usage: (RO) Number of times this state was entered (a count).
 168
 169
 170 What:           /sys/devices/system/cpu/cpuX/cpuidle/stateN/desc
 171 Date:           February 2008
 172 KernelVersion:  v2.6.25
 173 Contact:        Linux power management list <linux-pm@vger.kernel.org>
 174 Description:
 175                 (RO) A small description about the idle state (string).
 176
 177
 178 What:           /sys/devices/system/cpu/cpuX/cpuidle/stateN/disable
 179 Date:           March 2012
 180 KernelVersion:  v3.10
 181 Contact:        Linux power management list <linux-pm@vger.kernel.org>
 182 Description:
 183                 (RW) Option to disable this idle state (bool). The behavior and
 184                 the effect of the disable variable depends on the implementation
 185                 of a particular governor. In the ladder governor, for example,
 186                 it is not coherent, i.e. if one is disabling a light state, then
 187                 all deeper states are disabled as well, but the disable variable
 188                 does not reflect it. Likewise, if one enables a deep state but a
 189                 lighter state still is disabled, then this has no effect.
 190
 191
 192 What:           /sys/devices/system/cpu/cpuX/cpuidle/stateN/residency
 193 Date:           March 2014
 194 KernelVersion:  v3.15
 195 Contact:        Linux power management list <linux-pm@vger.kernel.org>
 196 Description:
 197                 (RO) Display the target residency i.e. the minimum amount of
 198                 time (in microseconds) this cpu should spend in this idle state
 199                 to make the transition worth the effort.
 200
 201
 202 What:           /sys/devices/system/cpu/cpu#/cpufreq/*
 203 Date:           pre-git history
 204 Contact:        linux-pm@vger.kernel.org
 205 Description:    Discover and change clock speed of CPUs
 206
 207                 Clock scaling allows you to change the clock speed of the
 208                 CPUs on the fly. This is a nice method to save battery
 209                 power, because the lower the clock speed, the less power
 210                 the CPU consumes.
 211
 212                 There are many knobs to tweak in this directory.
 213
 214                 See files in Documentation/cpu-freq/ for more information.
 215
 216                 In particular, read Documentation/cpu-freq/user-guide.txt
 217                 to learn how to control the knobs.
 218
 219
 220 What:           /sys/devices/system/cpu/cpu#/cpufreq/freqdomain_cpus
 221 Date:           June 2013
 222 Contact:        linux-pm@vger.kernel.org
 223 Description:    Discover CPUs in the same CPU frequency coordination domain
 224
 225                 freqdomain_cpus is the list of CPUs (online+offline) that share
 226                 the same clock/freq domain (possibly at the hardware level).
 227                 That information may be hidden from the cpufreq core and the
 228                 value of related_cpus may be different from freqdomain_cpus. This
 229                 attribute is useful for user space DVFS controllers to get better
 230                 power/performance results for platforms using acpi-cpufreq.
 231
 232                 This file is only present if the acpi-cpufreq driver is in use.
 233
 234
 235 What:           /sys/devices/system/cpu/cpu*/cache/index3/cache_disable_{0,1}
 236 Date:           August 2008
 237 KernelVersion:  2.6.27
 238 Contact:        Linux kernel mailing list <linux-kernel@vger.kernel.org>
 239 Description:    Disable L3 cache indices
 240
 241                 These files exist in every CPU's cache/index3 directory. Each
 242                 cache_disable_{0,1} file corresponds to one disable slot which
 243                 can be used to disable a cache index. Reading from these files
 244                 on a processor with this functionality will return the currently
 245                 disabled index for that node. There is one L3 structure per
 246                 node, or per internal node on MCM machines. Writing a valid
 247                 index to one of these files will cause the specificed cache
 248                 index to be disabled.
 249
 250                 All AMD processors with L3 caches provide this functionality.
 251                 For details, see BKDGs at
 252                 http://developer.amd.com/documentation/guides/Pages/default.aspx
 253
 254
 255 What:           /sys/devices/system/cpu/cpufreq/boost
 256 Date:           August 2012
 257 Contact:        Linux kernel mailing list <linux-kernel@vger.kernel.org>
 258 Description:    Processor frequency boosting control
 259
 260                 This switch controls the boost setting for the whole system.
 261                 Boosting allows the CPU and the firmware to run at a frequency
 262                 beyound it's nominal limit.
 263                 More details can be found in
 264                 Documentation/admin-guide/pm/cpufreq.rst
 265
 266
 267 What:           /sys/devices/system/cpu/cpu#/crash_notes
 268                 /sys/devices/system/cpu/cpu#/crash_notes_size
 269 Date:           April 2013
 270 Contact:        kexec@lists.infradead.org
 271 Description:    address and size of the percpu note.
 272
 273                 crash_notes: the physical address of the memory that holds the
 274                 note of cpu#.
 275
 276                 crash_notes_size: size of the note of cpu#.
 277
 278
 279 What:           /sys/devices/system/cpu/intel_pstate/max_perf_pct
 280                 /sys/devices/system/cpu/intel_pstate/min_perf_pct
 281                 /sys/devices/system/cpu/intel_pstate/no_turbo
 282 Date:           February 2013
 283 Contact:        linux-pm@vger.kernel.org
 284 Description:    Parameters for the Intel P-state driver
 285
 286                 Logic for selecting the current P-state in Intel
 287                 Sandybridge+ processors. The three knobs control
 288                 limits for the P-state that will be requested by the
 289                 driver.
 290
 291                 max_perf_pct: limits the maximum P state that will be requested by
 292                 the driver stated as a percentage of the available performance.
 293
 294                 min_perf_pct: limits the minimum P state that will be requested by
 295                 the driver stated as a percentage of the available performance.
 296
 297                 no_turbo: limits the driver to selecting P states below the turbo
 298                 frequency range.
 299
 300                 More details can be found in
 301                 Documentation/admin-guide/pm/intel_pstate.rst
 302
 303 What:           /sys/devices/system/cpu/cpu*/cache/index*/<set_of_attributes_mentioned_below>
 304 Date:           July 2014(documented, existed before August 2008)
 305 Contact:        Sudeep Holla <sudeep.holla@arm.com>
 306                 Linux kernel mailing list <linux-kernel@vger.kernel.org>
 307 Description:    Parameters for the CPU cache attributes
 308
 309                 allocation_policy:
 310                         - WriteAllocate: allocate a memory location to a cache line
 311                                          on a cache miss because of a write
 312                         - ReadAllocate: allocate a memory location to a cache line
 313                                         on a cache miss because of a read
 314                         - ReadWriteAllocate: both writeallocate and readallocate
 315
 316                 attributes: LEGACY used only on IA64 and is same as write_policy
 317
 318                 coherency_line_size: the minimum amount of data in bytes that gets
 319                                      transferred from memory to cache
 320
 321                 level: the cache hierarchy in the multi-level cache configuration
 322
 323                 number_of_sets: total number of sets in the cache, a set is a
 324                                 collection of cache lines with the same cache index
 325
 326                 physical_line_partition: number of physical cache line per cache tag
 327
 328                 shared_cpu_list: the list of logical cpus sharing the cache
 329
 330                 shared_cpu_map: logical cpu mask containing the list of cpus sharing
 331                                 the cache
 332
 333                 size: the total cache size in kB
 334
 335                 type:
 336                         - Instruction: cache that only holds instructions
 337                         - Data: cache that only caches data
 338                         - Unified: cache that holds both data and instructions
 339
 340                 ways_of_associativity: degree of freedom in placing a particular block
 341                                         of memory in the cache
 342
 343                 write_policy:
 344                         - WriteThrough: data is written to both the cache line
 345                                         and to the block in the lower-level memory
 346                         - WriteBack: data is written only to the cache line and
 347                                      the modified cache line is written to main
 348                                      memory only when it is replaced
 349
 350
 351 What:           /sys/devices/system/cpu/cpu*/cache/index*/id
 352 Date:           September 2016
 353 Contact:        Linux kernel mailing list <linux-kernel@vger.kernel.org>
 354 Description:    Cache id
 355
 356                 The id provides a unique number for a specific instance of
 357                 a cache of a particular type. E.g. there may be a level
 358                 3 unified cache on each socket in a server and we may
 359                 assign them ids 0, 1, 2, ...
 360
 361                 Note that id value can be non-contiguous. E.g. level 1
 362                 caches typically exist per core, but there may not be a
 363                 power of two cores on a socket, so these caches may be
 364                 numbered 0, 1, 2, 3, 4, 5, 8, 9, 10, ...
 365
 366 What:           /sys/devices/system/cpu/cpuX/cpufreq/throttle_stats
 367                 /sys/devices/system/cpu/cpuX/cpufreq/throttle_stats/turbo_stat
 368                 /sys/devices/system/cpu/cpuX/cpufreq/throttle_stats/sub_turbo_stat
 369                 /sys/devices/system/cpu/cpuX/cpufreq/throttle_stats/unthrottle
 370                 /sys/devices/system/cpu/cpuX/cpufreq/throttle_stats/powercap
 371                 /sys/devices/system/cpu/cpuX/cpufreq/throttle_stats/overtemp
 372                 /sys/devices/system/cpu/cpuX/cpufreq/throttle_stats/supply_fault
 373                 /sys/devices/system/cpu/cpuX/cpufreq/throttle_stats/overcurrent
 374                 /sys/devices/system/cpu/cpuX/cpufreq/throttle_stats/occ_reset
 375 Date:           March 2016
 376 Contact:        Linux kernel mailing list <linux-kernel@vger.kernel.org>
 377                 Linux for PowerPC mailing list <linuxppc-dev@ozlabs.org>
 378 Description:    POWERNV CPUFreq driver's frequency throttle stats directory and
 379                 attributes
 380
 381                 'cpuX/cpufreq/throttle_stats' directory contains the CPU frequency
 382                 throttle stat attributes for the chip. The throttle stats of a cpu
 383                 is common across all the cpus belonging to a chip. Below are the
 384                 throttle attributes exported in the 'throttle_stats' directory:
 385
 386                 - turbo_stat : This file gives the total number of times the max
 387                 frequency is throttled to lower frequency in turbo (at and above
 388                 nominal frequency) range of frequencies.
 389
 390                 - sub_turbo_stat : This file gives the total number of times the
 391                 max frequency is throttled to lower frequency in sub-turbo(below
 392                 nominal frequency) range of frequencies.
 393
 394                 - unthrottle : This file gives the total number of times the max
 395                 frequency is unthrottled after being throttled.
 396
 397                 - powercap : This file gives the total number of times the max
 398                 frequency is throttled due to 'Power Capping'.
 399
 400                 - overtemp : This file gives the total number of times the max
 401                 frequency is throttled due to 'CPU Over Temperature'.
 402
 403                 - supply_fault : This file gives the total number of times the
 404                 max frequency is throttled due to 'Power Supply Failure'.
 405
 406                 - overcurrent : This file gives the total number of times the
 407                 max frequency is throttled due to 'Overcurrent'.
 408
 409                 - occ_reset : This file gives the total number of times the max
 410                 frequency is throttled due to 'OCC Reset'.
 411
 412                 The sysfs attributes representing different throttle reasons like
 413                 powercap, overtemp, supply_fault, overcurrent and occ_reset map to
 414                 the reasons provided by OCC firmware for throttling the frequency.
 415
 416 What:           /sys/devices/system/cpu/cpufreq/policyX/throttle_stats
 417                 /sys/devices/system/cpu/cpufreq/policyX/throttle_stats/turbo_stat
 418                 /sys/devices/system/cpu/cpufreq/policyX/throttle_stats/sub_turbo_stat
 419                 /sys/devices/system/cpu/cpufreq/policyX/throttle_stats/unthrottle
 420                 /sys/devices/system/cpu/cpufreq/policyX/throttle_stats/powercap
 421                 /sys/devices/system/cpu/cpufreq/policyX/throttle_stats/overtemp
 422                 /sys/devices/system/cpu/cpufreq/policyX/throttle_stats/supply_fault
 423                 /sys/devices/system/cpu/cpufreq/policyX/throttle_stats/overcurrent
 424                 /sys/devices/system/cpu/cpufreq/policyX/throttle_stats/occ_reset
 425 Date:           March 2016
 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
 429                 attributes
 430
 431                 'policyX/throttle_stats' directory and all the attributes are same as
 432                 the /sys/devices/system/cpu/cpuX/cpufreq/throttle_stats directory and
 433                 attributes which give the frequency throttle information of the chip.
 434
 435 What:           /sys/devices/system/cpu/cpuX/regs/
 436                 /sys/devices/system/cpu/cpuX/regs/identification/
 437                 /sys/devices/system/cpu/cpuX/regs/identification/midr_el1
 438                 /sys/devices/system/cpu/cpuX/regs/identification/revidr_el1
 439 Date:           June 2016
 440 Contact:        Linux ARM Kernel Mailing list <linux-arm-kernel@lists.infradead.org>
 441 Description:    AArch64 CPU registers
 442                 'identification' directory exposes the CPU ID registers for
 443                  identifying model and revision of the CPU.
 444
 445 What:           /sys/devices/system/cpu/cpu#/cpu_capacity
 446 Date:           December 2016
 447 Contact:        Linux kernel mailing list <linux-kernel@vger.kernel.org>
 448 Description:    information about CPUs heterogeneity.
 449
 450                 cpu_capacity: capacity of cpu#.
 451
 452 What:           /sys/devices/system/cpu/vulnerabilities
 453                 /sys/devices/system/cpu/vulnerabilities/meltdown
 454                 /sys/devices/system/cpu/vulnerabilities/spectre_v1
 455                 /sys/devices/system/cpu/vulnerabilities/spectre_v2
 456 Date:           January 2018
 457 Contact:        Linux kernel mailing list <linux-kernel@vger.kernel.org>
 458 Description:    Information about CPU vulnerabilities
 459
 460                 The files are named after the code names of CPU
 461                 vulnerabilities. The output of those files reflects the
 462                 state of the CPUs in the system. Possible output values:
 463
 464                 "Not affected"    CPU is not affected by the vulnerability
 465                 "Vulnerable"      CPU is affected and no mitigation in effect
 466                 "Mitigation: $M"  CPU is affected and mitigation $M is in effect